University of Southern California Dissertations and Theses

Eicosanoid regulation of spermatogonial stem cell development

(USC Thesis Other)

Eicosanoid regulation of spermatogonial stem cell development

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
Copyright 2022 Amy Tran-Guzman

EICOSANOID REGULATION OF SPERMATOGONIAL STEM CELL
DEVELOPMENT

by

AMY TRAN-GUZMAN

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(MOLECULAR PHARMACOLOGY AND TOXICOLOGY)

August 2022

ii
DEDICATION

This dissertation is dedicated to my husband, Adrian Guzman, who has always been my
biggest supporter and advocate throughout this journey. For my family, who wholeheartedly
believed that I could attain a PhD degree even before I was accepted into a graduate program.
For my close friends who were extremely patient with me, especially for all the times they had to
be flexible with my work commitments. And for my late grandmother, whose lost battle with
pancreatic cancer inspired my lifelong mission to become a scientist and help develop lifesaving
treatments for future cancer patients.

iii
ACKNOWLEDGEMENTS

I would like to acknowledge and express my sincere gratitude towards my PhD advisor,
Dr. Martine Culty who welcomed me into the field of toxicology with open arms. She taught me
how to design and interpret scientific studies, reviewed every piece of writing I submitted, and
enthusiastically supported all the extracurricular activities I participated in throughout my duration
in the PhD program. I would not have been the graduate student I am today if it weren’t for the
effort she put into my growth, not only as a scientist but also as a well-rounded member of
toxicology community.
I would like to thank Dr. Vassilios Papadopoulos for his support in my scientific
development. Dr. Papadopoulos taught me the importance of tackling a scientific question using
multiple different methods and had always given me perspective on my project. He often
encouraged me to draw ideas from studies outside my field and to think about the impact of my
research beyond the scope of male reproductive health.
I would like to thank Dr. Kimberly Brannen who mentored me during my 2021 summer
internship with Merck Pharmaceuticals. She supported my career development in pharma
toxicology and taught me the importance of being able to communicate my findings to a broad
scientific audience.
I would like to thank all my senior lab members in both the Culty and Papadopoulos lab
who were always willing to help me whenever I had questions. Chantal Sottas has been incredibly
supportive in helping me design my animal studies and a great sounding board for all my ideas.
Dr.’s Vanessa Brouard, Lu Li, and Yuchang Li have been instrumental in training me and helping
me troubleshoot my experiments over the years. I would also like to thank our collaborator, Dr.
Wen, and her lab for giving me an opportunity to work with an incredibly interesting side project
that is outside my specialty, one that I wouldn’t have been exposed to otherwise. As well as Dr.

iv
Gopalakrishna for participating as a contributing member of my thesis committee and supporting
my development as a graduate student since my qualification exam.
Lastly, I would like to thank my lab mates for their daily support throughout these past 5
years, whether it’d be troubleshooting my experiments, getting me through mental hurdles, or
giving me feedback on my work. I would not have been able to get to this point in my career
without help of incredibly smart and hardworking graduate students like Christina Lin, Melanie
Galano, Casandra Walker, Maia Corpuz, Sam Garza, and Garett Cheung. I would also like to
acknowledge the direct contributions of my mentees Amina Khan and Renita Moradian on my
projects who I have no doubt will go on and make great contributions to the scientific field in the
future.

v

TABLE OF CONTENTS
DEDICATION......................................................................................................................... ii
ACKNOWLEDGEMENTS..................................................................................................... iii
LIST OF TABLES.................................................................................................................. ix
LIST OF FIGURES................................................................................................................ x
ABSTRACT.......................................................................................................................... xii
Chapter 1: Introduction.......................................................................................................... 1
1.1 Abstract................................................................................................................ 1
1.2 Introduction.......................................................................................................... 2
1.3 Fetal and neonatal male reproductive development in rodents and humans....... 3
1.3.1 Main cell types and timepoints in testis development......................... 3
1.3.2 Testicular steroidogenesis.................................................................. 6
1.3.3 Disruption of testis development......................................................... 7
1.4 Eicosanoid biosynthesis and COX inhibitors....................................................... 9
1.4.1 Enzymes of the eicosanoid pathway................................................... 9
1.4.2 Prostaglandins................................................................................. 11
1.4.3 Pharmacology of NSAIDs and analgesic drugs................................ 12
1.5 Eicosanoid system in male reproductive development...................................... 14
1.5.1 Expression and role of cyclooxygenases......................................... 14
1.5.2 Expression of PG synthases and prostanoid receptors in the testis. 17
1.5.2.1 Prostaglandin synthases...................................................... 22
1.5.2.2 Prostaglandin receptors....................................................... 20
1.5.2.3 Roles of PGD2 and 15-PGJ2............................................... 23
1.5.2.4 Roles of PGE2 and PGF2a.................................................. 24
1.5.3 Effects of NSAIDs and analgesic drugs on male reproductive
development.................................................................................... 25
1.5.3.1 Effects of NSAIDs in testis.................................................... 25
1.5.3.2 Effects of NSAIDs on fetal and neonatal germ cells.............. 27
1.5.3.3 Effects of NSAIDs on Leydig cells and steroidogenesis........ 29
1.5.3.4 Effects of NSAIDs on Sertoli cells......................................... 33
1.5.3.5 Effects of NSAIDs on testicular macrophages...................... 34
1.6 Epidemiological Evidence................................................................................. 34
1.6.1 Cryptorchidism................................................................................ 34
1.6.2 Hypospadias.................................................................................... 36
1.6.3 Reduced AGD................................................................................. 36
1.6.4 Long-term male reproductive effects................................................ 37
1.7 Non-rodent studies............................................................................................ 38
1.8 Conclusion........................................................................................................ 41
1.9 Acknowledgements.......................................................................................... 43

vi
Chapter 2: Characterizing the eicosanoid pathway in SSCs and its involvement in modulating
EDC effects........................................................................................................ 44
2.1 Abstract............................................................................................................. 44
2.2 Introduction....................................................................................................... 46
2.3 Materials and methods....................................................................................... 50
2.3.1 Chemicals........................................................................................ 50
2.3.2 C18-4 spermatogonial cell line culture.............................................. 50
2.3.3 Primary spermatogonia isolation...................................................... 51
2.3.4 Spermatogonia treatments............................................................... 51
2.3.5 Gene array analysis.......................................................................... 52
2.3.6 Immunofluorescence........................................................................ 52
2.3.7 MTT viability assay........................................................................... 53
2.3.8 RNA extraction, cDNA synthesis, and RT-qPCR analysis................ 53
2.3.9 LC-MS analysis of PGs PGE2 and PGF2a....................................... 54
2.3.10 Prostaglandin D2 ELISA assays....................................................... 55
2.3.11 Immunoblot protein analysis............................................................. 56
2.3.12 Statistics........................................................................................... 58
2.4 Results............................................................................................................... 58
2.4.1 Characterization of the eicosanoid pathway in mouse SSC model C18-
4 cell line and in primary rat spermatogonia...................................... 58
2.4.2 GEN, MEHP, and GEN+MEHP treatments alter differently the
expression of Cox1 and Cox2 eicosanoid pathway enzymes............ 62
2.4.3 Treatments with GEN, MEHP, and GEN+MEHP mixture disrupt
prostaglandin synthesis.................................................................... 63
2.4.3.1 Effects on Cbr1 expression and PGF2a synthesis................ 65
2.4.3.2 Effects on Ptges expression and PGE2 synthesis................. 67
2.4.3.3 Effects on Ptgds expression and PGD2 synthesis................ 69
2.4.3.4 GEN and GEN+MEHP alter SSC differentiation markers
concomitant to changes in prostaglandins............................ 71
2.5 Discussion......................................................................................................... 73
2.5.1 SSCs express major eicosanoid pathway enzymes and produce PGs..
......................................................................................................... 74
2.5.2 GEN and MEHP exert differential effects on the eicosanoid pathway
in SSCs............................................................................................ 75
2.5.3 GEN emerges as modulator of SSC differentiation concurrently to
increasing PGs................................................................................. 77
2.5.4 Distinctive effects of MEHP alone on PGD2 synthesis and SSC
differentiation.................................................................................... 78
2.5.5 Synergistic effects of GEN and MEHP on PGD2 synthesis in
SSCs................................................................................................ 79
2.6 Summary........................................................................................................... 81
2.7 Acknowledgements............................................................................................ 81
Chapter 3: Differential roles of cyclooxygenase enzymes in the regulation of murine
spermatogonial stem cells.................................................................................. 82
3.1 Abstract.............................................................................................................. 82
3.2 Introduction........................................................................................................ 84
3.3 Materials and methods....................................................................................... 86
3.3.1 Cell lines and treatments.................................................................. 86
3.3.2 Cox1 shRNA mediated silencing...................................................... 87

vii
3.3.3 LC-MS of prostaglandins................................................................. 87
3.3.4 Prostaglandin D2 ELISA................................................................... 88
3.3.5 EDU proliferation assay.................................................................... 89
3.3.6 RNA extraction, cDNA synthesis, and qPCR................................... 89
3.3.7 Western blotting............................................................................... 90
3.3.8 Immunofluorescence........................................................................ 91
3.3.9 Total RNA sequencing...................................................................... 91
3.3.10 Bioinformatics analysis..................................................................... 92
3.3.11 Human testicular tissue collection and gene array........................... 92
3.4 Results............................................................................................................... 93
3.4.1 Cox inhibition decrease prostaglandin production in the C18-4 cell
line.................................................................................................... 93
3.4.2 Effects of pharmacological Cox inhibitors on cell proliferation and
differentiation.................................................................................... 95
3.4.3 Morphological and transcriptomic characterization of C18-4
Cox1-KD1
.. 98
3.4.4 Ontology and pathway analysis of differentially expressed genes in
C18-4
Cox1-KD1
................................................................................... 100
3.4.5 Activation of Notch3 signaling pathway in C18-4
Cox1-KD1
.................. 104
3.4.6 Differential effects of pharmacological Cox inhibitors on Notch
pathway
expression...................................................................................... 106
3.4.7 PGD2 negatively regulates Notch signaling in C18-4 spermatogonial
cells................................................................................................ 109
3.4.8 Notch signaling is disrupted in testicular cancer.............................. 110
3.5 Discussion....................................................................................................... 111
3.6 Summary......................................................................................................... 117
3.7 Acknowledgements.......................................................................................... 117
Chapter 4: Characterizing the effects of early postnatal acetaminophen and ibuprofen
administration on immediate and long-term male reproductive
development.................................................................................................... 118
4.1 Abstract............................................................................................................ 118
4.2 Introduction...................................................................................................... 120
4.3 Materials and methods..................................................................................... 122
4.3.1 Animals and treatments.................................................................. 122
4.3.2 Single cell RNA sequencing............................................................ 122
4.3.3 Bioinformatics analysis................................................................... 123
4.3.4 Tissue histology.............................................................................. 124
4.3.5 Testosterone and luteinizing hormone ELISAs............................... 126
4.3.6 Statistics......................................................................................... 126
4.4 Results............................................................................................................. 126
4.4.1 No effects of treatment on animal weights or AGD......................... 126
4.4.2 Ace and ibu induces pronounced vacuolation in PND4 and PND8 pup
testes.............................................................................................. 130
4.4.3 Characterization of molecular effects of ace and ibu on testicular
cell populations in PND8 pups using scRNAseq............................ 131
4.4.4 Characterization of molecular changes in spermatogonia……...... 131
4.4.5 Characterization of molecular changes in Sertoli cells................... 138
4.4.6 Characterization of molecular changes in Leydig cells.................. 141

viii
4.4.7 Histological characterizations of adverse morphological alterations in
treated adult rats............................................................................ 144
4.4.8 High dose ace and ibu increase testosterone and LH levels in adult
rats................................................................................................ 148
4.5 Discussion....................................................................................................... 149
4.6 Summary......................................................................................................... 154
4.7 Acknowledgements.......................................................................................... 155
Chapter 5: Summary, Challenges, and Future Perspectives.............................................. 156
5.1 Summary......................................................................................................... 156
5.1.1 Summary of Chapter 2.................................................................... 156
5.1.2 Summary of Chapter 3.................................................................... 158
5.1.3 Summary of Chapter 4.................................................................... 160
5.2 Challenges....................................................................................................... 163
5.2.1 Translatability between rodent models to human............................ 163
5.2.2 Cox2 knockdown C18-4 model for comparison............................... 163
5.2.3 Defining a role of Notch3 to the malignant transformation of
SSCs.............................................................................................. 164
5.3 Future Perspectives......................................................................................... 164
5.3.1 Contribution to the field of reproductive biology and toxicology....... 164
5.3.2 Role of Notch signaling in spermatogonial development................. 165
5.3.3 Interaction between Notch and eicosanoid signaling...................... 166
5.3.4 Further exploration of pathways revealed by scRNAseq................. 167
5.4 Conclusion....................................................................................................... 168
References........................................................................................................................ 169
Appendices........................................................................................................................ 189
Appendix 1: Supplemental Table 1......................................................................... 189
Appendix 2: Supplemental Table 2......................................................................... 190
Appendix 3: Supplemental Table 3......................................................................... 191
Appendix 4: Supplemental Figure 1....................................................................... 211

ix
LIST OF TABLES

Table 1. Reports of the expression of prostaglandins (PGs), PG synthases and receptors in the
male reproductive system.
Table 2. List of qPCR Primer sets.
Table 3. Table showing the proportions of cells identified in each population of the testes using
scRNAseq.
Table 4. List of top differentially expressed genes upregulated and downregulated by high ace
and high ibu treatment in the spermatogonial cell cluster and their respective p-values, false
discovery rate, fold change, and functions.
Table 5. List of shared differentially expressed genes altered in the same direction with high ace
and high ibu treatment and their respective fold changes and functions.
Table 6. List of top differentially expressed genes upregulated and downregulated by high ace
and high ibu treatment in the Sertoli cell cluster and their respective p-values, false discovery
rate, fold change, and functions.
Table 7. List of top differentially expressed genes upregulated and downregulated by high ace
and high ibu treatment in the Leydig cell cluster and their respective p-values, false discovery
rate, fold change, and functions.
Supplemental Table 1. List of primers used for qPCR.
Supplemental Table 2. List of differentially expressed genes involved in signaling pathways
predicted to be altered by Cox1 silencing.
Supplemental Table 3. Complete list of 1,265 differentially expressed genes altered by Cox1
silencing.

x
LIST OF FIGURES

Figure 1. Schematic representation of the eicosanoid pathway.
Figure 2. Schematic illustration of the production and roles of prostaglandins (PGs) in
mammalian testis from development to adulthood.
Figure 3. Characterization of eicosanoid pathway in rodent SSC models.
Figure 4. C18-4 spermatogonia produce prostaglandins PGE2 and PGF2ɑ.
Figure 5. Effect of retinoic acid treatment on the expression spermatogonial gene markers and
Cox genes in C18-4 cell line.
Figure 6. Effects of GEN and MEHP, alone or mixed, on cell viability and on the expression of
upstream eicosanoid pathway enzymes.
Figure 7. Effects of GEN and MEHP, alone or mixed, on Cbr1 gene expression and on
Prostaglandin F2a Production.
Figure 8. Effects of GEN and MEHP, alone or mixed, on Ptges gene expression and on
Prostaglandin E2 Production.
Figure 9. Effects of GEN and MEHP, alone or mixed, on Prostaglandin D2 production.
Figure 10. Effects of GEN and MEHP, alone or mixed, on undifferentiated and differentiated
spermatogonial gene markers.
Figure 11. Diagram summarizing the effects of GEN and MEHP, alone or mixed, on the
eicosanoid biosynthetic pathway and on SSC differentiation.
Figure 12. Measurements of prostaglandin levels with Cox inhibitor treatment.
Figure 13. Evaluation of pharmacological Cox inhibitors on C18-4 proliferation and
differentiation.
Figure 14. Generation of based Cox1 knockdown cellular model C18-4
Cox1-KD1
.
Figure 15. RNAseq characterization of C18-4
Cox1-KD1
.
Figure 16. Notch3 pathway is upregulated in C18-4
Cox1-KD1
.

xi
Figure 17. Effect of 24-hr treatment of Cox inhibitors on expression of Notch pathway genes.
Figure 18. Supplemental PGD2 negatively regulates Notch3 and Notch pathway components.
Figure 19. Expression of NOTCH3 and APH1B in human testicular cancer.
Figure 20. Schematic of crosstalk between eicosanoid and Notch pathways and predicted
downstream effects.
Figure 21. Schematic of animal study design.
Figure 22. Effects of treatments on animal weights.
Figure 23. Effects of treatments on anogenital distances.
Figure 24. Effect of treatments on testicular morphology in pups.
Figure 25. scRNAseq characterization of treated PND8 testicular cell suspensions.
Figure 26. scRNAseq characterization of the spermatogonial cell cluster.
Figure 27. Volcano plot illustrating up and downregulated differentially expressed genes in the
Sertoli cell cluster with high ace and high ibu treatment relative to the vehicle control.
Figure 28. Volcano plot illustrating up and downregulated differentially expressed genes in the
Leydig cell cluster with high ace and high ibu treatment relative to the vehicle control.
Figure 29. Morphological and histological evaluation of treated adult testes.
Figure 30. Effect of treatment on serum testosterone and luteinizing hormone levels in PND90
rats.
Supplemental Figure 1. Mycoplasma results for C18-4 cell line passage 29 (p29).

xii
ABSTRACT

The objective of this thesis is to investigate the role of the eicosanoid biosynthetic pathway
in regulating neonatal and juvenile stages of spermatogenesis, and the potential impact of
compounds targeting this pathway on male reproductive health. These include prevalent
environmental toxicants such as plasticizer metabolite Mono(2-ethylhexyl) Phthalate (MEHP) and
soy-derived isoflavone genistein as well as commonly used drugs acetaminophen (ace) and
ibuprofen (ibu). We hypothesize that the eicosanoid pathway plays a role in spermatogonial stem
cell (SSC) fate and that chemicals disrupting eicosanoid synthesis can disturb germ cell growth,
of which Cox enzymes and prostaglandin effectors can play important roles in regulating.
Maintaining a balance between self-renewal and differentiation in the SSC pool is critical for
sustaining male fertility. Starting at puberty, some SSCs differentiate to form spermatozoa,
whereas others retain their stem cell identity to maintain an adequate SSC pool. Disruption to
proper germ cell development has been shown to be a source of infertility and germ cell derived
seminomas, emphasizing the relevance of this study for male reproductive health.
Three specific aims have been developed to address the objective of this project and our
hypothesis. Firstly, we have characterized the eicosanoid pathway in the C18-4 cell line, an SSC
model, and established that SSCs express eicosanoid pathway components and can produce
measurable amounts of PGs. We also discovered that MEHP and GEN can modulate eicosanoid
pathway expression and alter PG levels. Next, using a combination of treatments with
pharmacological Cox inhibitors and gene silencing methods, we unveiled a mechanism of action
behind Cox inhibition on SSCs, which involves Notch3 and the Notch signaling pathway. Lastly,
7-day postnatal treatment with human relevant doses of ace and ibu had potential to alter testis
morphology and the transcriptome and functional pathways of several testicular cell types,
including spermatogonia, with long-term consequences on testicular morphology and hormonal

xiii
regulation in adult rats. Taken together these results support our hypothesis, and thus greater
caution should be taken when exposing drugs that target eicosanoid biosynthesis to infants.

1
CHAPTER 1: INTRODUCTION
1

1.1 ABSTRACT

Increasing rates of infertility associated with declining sperm counts and quality, as well as
increasing rates of testicular cancer are contemporary issues in the United States and abroad.
These conditions are part of the Testicular Dysgenesis Syndrome, which includes a variety of
male reproductive disorders hypothesized to share a common origin based on disrupted testicular
development during fetal and neonatal stages of life. Male reproductive development is a highly
regulated and complex process that relies on an intricate coordination between germ, Leydig, and
Sertoli cells as well as other supporting cell types, to ensure proper spermatogenesis, testicular
immune privilege, and endocrine function. The eicosanoid system has been reported to be
involved in the regulation of fetal and neonatal germ cell development as well as overall testicular
homeostasis. Moreover, non-steroidal anti-inflammatory drugs (NSAIDs) and analgesics with
abilities to block eicosanoid synthesis by targeting either or both isoforms of cyclooxygenase
enzymes, have been found to adversely affect male reproductive development. This review will
explore the current body of knowledge on the involvement of the eicosanoid system in male
reproductive development, as well as discuss adverse effects of NSAIDs and analgesic drugs
administered perinatally, focusing on toxicities reported in the testis and on major testicular cell
types. Rodent and epidemiological studies will be corroborated by findings in invertebrate models
for a comprehensive report of the state of the field, and to add to our understanding of the potential
long-term effects of NSAID and analgesic drug administration in infants.

1
This chapter is derived from a manuscript entitled “Eicosanoid biosynthesis in male reproductive development: Effects
of perinatal exposure to NSAIDs and analgesic drugs” by Tran-Guzman, A. and Culty, M.

2
1.2 INTRODUCTION

The recent decades have seen concerning increases in male reproductive disorders and
diseases. Total fertility rates have been trending downward in places such as Europe, Japan, and
United States, and this decline has been correlated with increases in testicular dysgenesis
syndrome (TDS). TDS includes a variety of male reproductive disorders such as infertility, poor
semen quality, cryptorchidism, and hypospadias, which are hypothesized to share a common
origin of disrupted testicular development during fetal and neonatal stages of life [1]. A study
evaluating cancer rates of a Danish registry found that testicular cancer rates have more than
tripled between 1944-2009 [2]. Furthermore, cryptorchidism, or undescended testes, is one of the
most commonly diagnosed birth defects [3]. Overall, negative trends in global sperm counts and
decreases in fertility rates [4, 5] suggest that there is much more to learn about the biological
factors that contribute to reductions in male reproductive capacity. One hypothesis is that lifestyle
factors, such as greater exposure to environmental toxicants and pharmaceuticals in the modern
world, may be contributing to the significant rates of TDS diagnoses in contemporary times [1].
Over-the-counter pharmaceuticals have been drawing greater attention recently as potential
endocrine disrupting compounds (EDCs) to be concerned about, due to their widespread use and
presumed safety for administration throughout life. One such class of drugs include
cyclooxygenase (COX) inhibitors or non-steroidal anti-inflammatory drugs (NSAIDs), which are
widely prescribed for pain relief and fever reduction. Acetaminophen (paracetamol), a non-specific
COX inhibitor used for its analgesic and antipyretic properties, was reported to be used during
pregnancy in 62% of woman surveyed [6], and ibuprofen, an NSAID, was also found to be used
commonly during pregnancy [7]. However, a growing body of evidence suggests that in utero
exposures to such drugs have the potential of altering reproductive development during fetal
stages and subsequent reproductive function in the offspring. Maternal exposure to common
NSAIDs and analgesic drugs during pregnancy was found to be significantly positively associated
with cryptorchidism in infants [8, 9], specifically in the second trimester [7]. Currently, the US Food

3
and Drug Administration is recommending against taking NSAIDs at 20 weeks or later during
pregnancy, due to their ability to cause kidney problems in unborn babies [10], but no
recommendation has been made against the use of analgesic drugs such as acetaminophen
during pregnancy or the use of NSAIDs in infants. In addition to in utero exposure, babies can be
exposed to these drugs either in the form of direct administration or passed through breast milk
from maternal ingestion after birth [11, 12]. The first postnatal months represent a complex period
during male reproductive development in which the population of spermatogonial stem cells
(SSCs) is established to support lifelong spermatogenesis. Therefore, an assessment of how
NSAID and analgesic drugs are impacting male reproductive development during the early stages
of life is necessary to obtain a comprehensive understanding of their contributions to male
reproductive diseases.

1.3 Fetal and neonatal male reproductive development in rodents and human
1.3.1 Main cell types and timepoints in testis development

Gonadogenesis is a dynamic and finely tuned system, requiring the regulation of multiple
molecular processes, many occurring simultaneously. The first sign of gonadogenesis is the
thickening of the coelomic epithelium, forming the genital ridge. Germ cell development initiates
with the commitment of embryonic stem cells to the germ cell lineage, forming primordial germ
cells (PGCs) expressing characteristic germ cell signature genes [13]. From gestational day
(GD)7.5 to 13 in rodents, sexually undifferentiated PGCs, upon receiving Kit and CXCR4 signals,
migrate and become resident at the genital ridge. PGCs undergo a genomic erasure process in
which parental DNA methylation patterns are removed [14]. Sex determination occurs on GD12.5,
driven by Sry expression in male somatic cells which differentiate into fetal Sertoli cells, whose
main function is to assist germ cell development and spermatogenesis. At this stage, fetal Sertoli
cells committed to epithelial differentiation organize intimately around the germ cells, which are

4
now considered fetal gonocytes (pro/pre-spermatogonia). Sertoli cells drive the organization of
the seminiferous cords as well as the recruitment and differentiation of mesonephros-derived cells
to peritubular myoid cells located at the outer periphery of the cords, and fetal Leydig cells, the
steroidogenic cells found in the interstitium surrounding the cords. Cells derived from the
mesonephros give rise to endothelial cells that contribute to the formation of blood vessels and
support the structure of the cords. Both Sertoli and peritubular myoid cells contribute to the
formation of the basement membrane surrounding the seminiferous cords. Leydig cells are
primarily responsible for the production of male sex hormones, mainly testosterone and
dihydrotestosterone, that regulate male sex differentiation, the development of internal and
external genitalia, as well as secondary sex characteristics. Male germ cells from fetal to juvenile
phases of development are prevented from entering meiosis through the degradation of retinoic
acid (RA) by cytochrome CYP26B1 and NANOS2 expression [14, 15].
Fetal Sertoli cells in humans double every two weeks from GW7-19 and continue to
amplify until cell numbers peak at birth. After birth, Sertoli cell numbers remain stable until puberty
where the cells reach a final count of approximately 1800x10
6
cells/testis. Leydig cells are first
detected in humans at GW 8 and cell numbers increase exponentially until GW18. Thereafter, a
dedifferentiation process occurs and Leydig cell numbers decline in the fetus until after birth when
numbers increase up to three months postnatal, likely due to increases in LH levels. The cells
then decline at four months and are scarce until age six to eight years, at which point they increase
to reach adult levels of approximately 800x10
6
cells/testis. Germ cell numbers in humans
exponentially increase from the initiation of testis differentiation to the end of the second trimester,
and is greatest from GW6-10, at which point it decreases until the second trimester.
Asynchronous development of undifferentiated gonocytes occurs throughout fetal and neonatal
stages of development, followed by a marked increase of undifferentiated spermatogonial
including SSCs between birth and puberty [16].

5
In rodents and humans, early male germ cell development comprises phases of
quiescence, proliferation, migration and differentiation [17]. Apoptotic processes also play integral
roles by ensuring removal of defective cells and maintaining proper germ-Sertoli cell ratios. Sertoli
cells assist with the maturation of neonatal gonocytes into spermatogonial stem cells (SSCs),
which will support lifelong spermatogenesis. At postnatal day (PND)3 in rats, gonocytes re-enter
mitosis and migrate towards the basement membrane of the seminiferous cords, where they
undergo differentiation into SSCs or Type A spermatogonia of the first spermatogenic wave
around PND6-8. In human, testis cords begin to form between the 7
th
and 9
th
gestational
week (GW). Formation of the essential structure of the testis takes approximately eight weeks,
leading to establishment of major testicular cell types [18]. The cords give rise to seminiferous
tubules and the differentiation of immature Sertoli cells into androgen-responsive cells leads to
the formation of an intratubular lumen. Tight junction proteins organize between Sertoli cells into
the testis-blood-barrier, a crucial structure to maintaining testicular immune privilege [19].
The rodent spermatogenic cycle involves distinct phases along the seminiferous tubules
taking place simultaneously: 1) mitosis in SSCs to type B differentiated spermatogonia; 2)
spermatogonial differentiation in which SSCs mature from undifferentiated progenitors to
differentiated spermatogonia; 3) the differentiation of successive types of spermatocytes and a
lengthy meiotic phase; 4) the formation of haploid spermatids during spermiogenesis, a
metamorphosis phase ultimately forming spermatozoa released into the lumen during
spermiation. Several steps of postnatal germ cell differentiation and the entry to meiosis are
regulated by Sertoli cell-produced RA [20]. The human spermatogenic cycle is similar. Type A
spermatogonia, comprising of dark type A and pale type A spermatogonia, undergo mitotic
proliferation and differentiate to Type B spermatogonia. Type B spermatogonia further
differentiate to spermatocytes (leptotene, zygotene, pachytene, and diplotene), which are present
upon puberty in humans. Diplotene spermatocytes undergo meiosis I to form secondary haploid

6
spermatocytes and further undergo morphological transformations from round spermatids and to
spermatozoa [21].
Another type of cell that plays multiple roles in testis development and adaptation to its
environment are the testicular macrophages, known to exert not only immune functions, but also
to modulate Leydig and germ cell functions. While rodent primitive macrophages are first seen in
gonadal primordium, as early as E10.5 in mice, fetal rat macrophages have been observed in
testis starting at GD16 to 19. Rodent fetal testicular macrophages express mainly M2 polarization-
type markers and are believed to be important for maternal-fetal tolerance [22, 23]. The numbers
of macrophages increase slowly until the onset of puberty, where macrophages rapidly increase
and decline thereafter. During puberty, testicular macrophages interact intimately with Leydig cells
through digitations or microvillus-like processes that are inserted within coated pits of the
macrophages [16, 22, 24, 25].
Animal studies, including ours, have established the susceptibility of the fetal and neonatal
testis to environmental insults, leading to short- and long-term adverse effects in testicular cells,
including germ cells [26-29]. Several groups have theorized that testicular germ cell tumors
(TGCTs) may arise from the disruption of SSC formation resulting from the vulnerability of PGC
or gonocyte populations to environmental exposures [30, 31]. Therefore, there is great interest in
understanding the molecular mechanisms driving the establishment of the SSC pool, with the goal
of determining the origins of disrupted spermatogenesis and/or TGCT formation. Furthermore, a
greater understanding of the etiology of TDS can lead to developing interventions to rescue the
SSC population [32-34].

1.3.2 Testicular steroidogenesis

The steroidogenic pathway is generally well-conserved between rodents and humans, but
there are several differences. In early fetal Leydig cells, androgen production occurs

7
independently of the pituitary, and it is only in late fetal age that testosterone production becomes
LH-dependent [35]. In the mouse, studies have shown that LH is not required for fetal Leydig cell
to produce testosterone, but rather fetal Leydig cells produce androstenedione that is converted
to testosterone by fetal Sertoli cells, a fact recently supported by Single-cell RNA-seq analysis of
fetal and adult mouse Leydig cells [36, 37]. In rats, fetal Leydig cell development is largely
independent of LH, but become LH-dependent few days before birth [35, 38, 39]. In humans, fetal
Leydig cells are critically dependent upon stimulation of LH throughout most of fetal development.
Testosterone synthesis also differs between rodent and humans. Mice and rats synthesize
testosterone through the D
4
pathway, whereas humans predominantly use the D
5
pathway. In the
D
4
pathway, pregnenolone is converted to all D
4
steroids by HSD17B and the final step in
testosterone synthesis, the reduction of androstenedione, is also catalyzed by HSD17B in fetal
Leydig cells. However, this process does not completely rely on Leydig cells, as androstenedione
can also be converted by Sertoli cells in mice lacking HSD17B [40, 41]. Human Leydig cells
synthesize testosterone using the D
5
pathway, in which the first reaction is catalyzed by CYP17A1
[42]. Often more than one HSD enzyme can catalyze the same reaction. For example, the
reduction of androstenedione to testosterone can involve both ARK1C3 and HSD17B [16].

1.3.3 Disruption of testis development

As described in previous sections, the development of the male reproductive system relies
on the formation of a functional fetal testis, the sequential production of hormones from fetal
Sertoli and Leydig cells, and the establishment of a pool of spermatogonial stem cells from
perinatal germ cell precursors in early postnatal ages. The disruption of these processes leads to
conditions that define the Testicular Dysgenesis Syndrome (TDS). In addition to testis cancer,
TDS includes cryptorchidism, hypospadias, altered testosterone levels, poor semen quality, and
reduced anogenital distance (AGD). Cryptorchidism is one of the most common birth defects,

8
corresponding to the failure of one or both testes to descend to the scrotum before birth. If not
corrected, the retention of testes in the abdomen results in the loss of spermatogenic cells in early
childhood, and increased risks of infertility or TGCTs, whereas corrective surgery via early
orchidopexy can reduce chances of developing such reproductive disorders. Hypospadias, which
is due to androgen deficiency in the fetus, is diagnosed when the urethra opens on the ventral
side of the penis instead of the tip, which may require surgical reconstruction of the penile urethra
in severe cases [1]. Postnatally, testosterone is important for sustaining secondary male sex
characteristics and spermatogenesis. The trend towards lower testosterone levels observed in
the US and Europe is believed to be related to the decline in male reproductive health [43]. As
testosterone is the major paracrine messenger involved in crosstalk between testicular cells,
compromised testosterone levels can affect testicular function. Semen quality has served as a
measurement for male fertility by the World Health Organization [44]. Low sperm count and quality
(motility and morphology) have been associated with decreasing rates of conception on a
population-wide level [45]. In males, the AGD is 50-100% longer than in females [46]. Shorter
AGDs and smaller penis size reflect decreased fetal androgen levels [1, 47]. Reduced AGD has
been found in men in association with cryptorchidism, low sperm counts, and hypospadias, further
supporting the common origin of these disorders. Epidemiological studies have also reported
strong correlations between shorter AGD in children and the levels of phthalate plasticizers
measured in maternal urine [48].
The testis is a complex organ that requires coordinated crosstalk between several major
cell types for normal function. Fetal development and infancy represent periods of dynamic growth
and maturation of various cell types within the testes and disrupting these processes may lead to
long-term reproductive diseases and syndromes. Due to the similarities in male reproductive
development and the regulation of steroidogenic pathways between human and rodents, rodent
models are useful tools in studying specific mechanistic effects of NSAID and analgesic drugs on
the male reproductive system. However, to ultimately translate these findings to real-life

9
applications, human-derived models and epidemiological studies must be considered. Therefore,
this review will weigh evidence from both rodent and human models, to provide a comprehensive
overview of the field.

1.4 Eicosanoid Biosynthesis and COX inhibitors
1.4.1 Enzymes of the eicosanoid pathway

Cyclooxygenases are enzymes that catalyze the first two steps in the biosynthesis of
prostaglandins (PGs) from arachidonic acid (AA). The main steps of the pathway are illustrated
in Figure 1. Two isoforms of COXs have been characterized, identified as COX1 and COX2.
COX1 is constitutively expressed, and COX2 is inducible upon an external stimulus. Both
isozymes are 71 kDa in molecular weight and are almost identical in length, with over 63%
similarity in amino acid sequence. AA is the major prostanoid precursor and both isoforms of COX
enzymes oxygenate this substrate through identical enzymatic processes [49, 50]. The release of
AA from membrane-bound phospholipids is catalyzed by phospholipase A2 (PLA2), of which
there are several isoforms that may be expressed constitutively or can appear after inflammatory
insult [51]. The biosynthesis of prostanoids requires a 3-step enzymatic process: 1) Stimulus-
initiated hydrolysis of AA from glycerophospholipids involving secretory or cytoplasmic PLA2
(sPLA2, cPLA2), 2) oxygenation of AA by the COXs to yield PGG2 and PGH2, and 3) conversion
of PGH2 to biologically active prostaglandins PGs (PGD2, PGE2, PGF2a, PGI2) and
thromboxane (TXA2). Thereafter, the prostanoid end products exit the cells to activate G-protein-
coupled prostanoid receptors, or in some cases interact directly with nuclear receptors [49].

10

Figure 1. Schematic representation of the eicosanoid pathway. Abbreviations: PG -E2, -I2, -F2a, -D2: Prostaglandins
-E2, I2, F2a, D2. CBR1: Carbonyl reductase 1. PTGES: Prostaglandin E Synthase, PTGDS: Prostaglandin D Synthase,
PTGIS: Prostaglandin I Synthase.

PG synthases are responsible for the enzymatic conversion of PGH2 to the major
prostanoids, which are isoform specific. PGD synthase (PGDS/PTGDS) catalyzes the conversion
of PGH2 to PGD2, PGE synthase (PGES) to PGE2, PGFa synthase to PGF2a, and PGI synthase
(PGIS) to PGI2, or prostacyclin [52]. Two forms of PGDSs have been characterized, glutathione-
dependent and -independent. Glutathione-dependent hematopoietic PGDS (H-PGDS) exhibits
glutathione-S-transferase activity, whereas lipocalin PGDS (L-PGDS) conversion of PGH2 to
PGD2 does not require this transferase activity. Three PGESs have been studied and all require
reduced glutathione as a cofactor- cytosolic PGES (cPGES), an inducible, microsomal PGES
(mPGES-1), and a non-inducible microsomal PGES (mPGES-2) [53]. Formation of PGF2a

11
involves first a reduction of PGH2 and a PGF2S catalyzing NADPH, and PGF2a prostanoid that
requires an NADPH-dependent synthase. PGF2a can also be synthesized from PGE2 by
carbonyl reductase 1. PGI synthases are believed to be localized to the cytosolic side of the
endoplasmic reticulum (ER), therefore PGH2 formed in the lumen of the ER can diffuse across
the membrane to be converted to PGI2 on the cytosolic side of the membrane.
Though similar in structure and function, and often co-expressed in the same cell, the
differences between the two COX isozymes lie in their unique downstream signaling properties
[49]. COX2 will accept a wider range of fatty acids than COX1, and has the ability to oxygenate
other substrates in addition to AA, such as eicosapentaenoic acid, y-linolenic acid, a-linolenic acid
and linoleic acid [50]. Stimuli known to induce COX2 expression are those associated with
inflammation such as pro-inflammatory cytokines (IL-2, IL-1, and TNFa) and lipopolysaccharide
(LPS), whereas anti-inflammatory cytokines (IL-4, IL-10, and IL-13) can decrease the induction of
COX2 [50]. As a constitutively expressed enzyme, COX1 is important for maintaining the
homeostasis or “housekeeping” roles of the body, such as providing PGs to the stomach and
intestine to maintain the mucosal epithelium. COX1 also plays a crucial role during development
by synthesizing PGs that are responsible for the survival of fetuses, such that fetuses born from
homozygous COX1 knockout mice did not survive past birth [54]. It is hypothesized that their
unique properties are derived from their expression: COX1 being involved in producing PGs that
act extracellularly to mediate responses to external signals, while the ability of COX2 to be
induced suggests also a role in augmenting the function of COX1 upon an inflammatory response
[49].

1.4.2 Prostaglandins

PGs are important lipid mediators that have widespread roles based on their unique
chemical structures, but are mainly involved in inflammatory responses [55]. They bind and act

12
via specific G-coupled protein receptors (GPCRs) [56]. PGI2 bind to IP receptors that lead to
cAMP elevation, generally leading to an inhibition of many cellular processes. PGE2 can bind on
one of four isoforms of EP receptors (EP1-4) that are each linked to different signal transduction
pathways that can either activate or inhibit cellular responses [57]. PGD2 and PGF2a bind to DP
and FP receptors, respectively, which are linked to activation of inositolphosphates and
subsequent increase in intracellular calcium, and usually result in activation of cellular processes.
Besides mediating varied local cellular responses, the action of PGD2 and PGF2a on calcium
signaling is critical in females for ovulation and during gestation and parturition. Indeed, the
eicosanoid system plays an important role during pregnancy. The COX enzymes are expressed
in the uterine epithelium at different times during pregnancy and are involved in regulating
feedback mechanisms associated with embryogenesis [58]. COX1 is expressed in the uterine
epithelium prior to implantation, whereby thereafter it gets downregulated. Subsequently, COX2
gets expressed, suggesting a possible role of COX2 in maintaining blastocyst attachment. During
labor and delivery, PGs (primarily PGE2 and PGF2a) are involved in smooth muscle contraction
of the uterus and cervical ripening [59]. Due to the significant involvement of PGs to induce
contractions during labor, NSAIDs have been prescribed to delay premature labor by inhibiting
the production of PGs [60]. These cellular responses are dependent on the target tissues and
cells of which the PG receptors are expressed.

1.4.3 Pharmacology of NSAIDs and analgesic drugs

The main pharmacology of NSAIDs is to inhibit COX activity and subsequent PG
production, thereby exerting anti-inflammatory and pain relief properties [56]. The unique
mechanisms of COX inhibition and drug preference to specific COX isozyme are dependent on
the drug classes. Classical NSAIDs developed prior to 1995 have ability to inhibit both isozymes,
but in general bind more tightly to COX1 [61]. The inhibition of COX by aspirin is due to irreversible

13
acetylation of the active site on the enzyme, leaving the peroxidase site unaffected [50]. Other
NSAIDs, such as ibuprofen or indomethacin, can produce either reversible or irreversible
inhibition by competition with AA for the active site of the enzyme. More modern NSAIDs were
developed to target specifically COX2 by taking advantage of a hydrophobic side pocket in its
main channel, which is inaccessible in COX1 due to steric hindrance of the presence of an
isoleucine rather than a valine at position 523 [62, 63]. The main concerns of non-selective COX
inhibitors is due to their ability to also target COX1, which is highly expressed in the stomach and
intestine [49].
Acetaminophen, for example, has a greater selectivity for COX2 than for COX1, and share
many properties of COX2 selective inhibitors such as minimal gastric side effects [64]. However,
in contrast to other NSAIDs, COX inhibition by acetaminophen is dependent on levels of cellular
peroxides, and only at low cellular peroxide levels does acetaminophen inhibit the production of
PGs. In contrast, selective COX2 inhibitors like celecoxib are effective anti-inflammatory agents
in diseases like rheumatoid arthritis where peroxide levels are high as they are not peroxide
dependent [65, 66].
As little is currently known about the contribution of perinatal exposure of pharmacological
COX inhibitors to later life testicular disorders, this review will aim to take a deep dive into the
involvement of the eicosanoid system in the regulation of fetal and neonatal germ cell
development, as well as its role in overall testicular homeostasis. We will provide an overview of
the eicosanoid biosynthetic pathway, explore its expression and function in developing and
healthy mammalian testes, and evaluate the impact of NSAID and analgesics exposure on cells
and processes related to male reproductive development and function. The current body of
knowledge on the role of COXs and PGs in male reproductive systems, in conjunction with the
toxicities reported with exposure of pharmacological COX inhibitors during early male
reproductive development will be discussed in detail. These rodent and epidemiological studies
will be corroborated by findings in non-rodent models for a comprehensive report of the state of

14
the field, and to add to our understanding of long-term effects of NSAID and analgesic drug
administration to infants.

1.5 Eicosanoid System in Male Reproductive Development
1.5.1 Expression and role of cyclooxygenases

While it is well established that the eicosanoid cascade plays a large role in the female
reproductive system, particularly in labor and childbirth, less is known about the involvement of
this pathway in the male reproductive system. Early efforts to establish roles of COX enzymes in
male fertility using knockout mouse models determined that male fertility was not affected in
COX1 or COX2 null mice, whereas COX2 null female mice were infertile [67]. Thus, it was
presumed early on that PGs do not play a role in male reproduction, and there was little interest
in evaluating this further. More recently, however, interest in the involvement of the eicosanoid
pathway in male reproduction peaked again and studies started to contradict or refine this
conclusion. In 2003, Hase et al. found that while there was no expression of COX1 and COX2 in
normal human testes, these enzymes were highly expressed in men with testicular cancer. Upon
investigation of COX inhibitors on human testicular cancer cell lines (NEC-8), they found weak
inhibition of cell viability with both COX1 and COX2 selective inhibitors, suggesting a role in of
COX enzymes in promoting growth of testicular cancer cells [68]. Another study reported
expression of COX2 in interstitial cells in human testes biopsies of infertile male patients with
Sertoli-cell-only syndrome (SCO-Syndrome) and patients with germ arrest (GA) syndrome [69].
In 2010, Matzkin et al. reported that a reason for the upregulation of COX enzymes in infertile
patients may be in response to action of proinflammatory cytokine IL-1. They also reported higher
levels of PGD2 and PGF2a in the testes of these men [70]. It appears that pathological human
testes and those with altered morphology are ones that express COX enzymes and can produce
PGs, suggesting a role of COX enzymes in the pathogenesis of testicular diseases.

15
The dysregulation of COX enzyme expression in pathological conditions can suggest that
these enzymes may have roles in maintaining the normal functioning of the testes, and therefore
efforts have been made to determine baseline levels of expression of COX enzymes in different
cell types of the organ. In 2007, COX1 and COX2 mRNA was detected in both normal and LPS
conditions of rat testes in somatic cells (Leydig and Sertoli cells) and testicular macrophages, as
well as measurable amounts of PGE2 [71]. They reported that COX2 expression was greater than
that of COX1 and treatment with a COX2 specific inhibitor, NS398, suppressed testicular PGE2
production that was not observed with a COX1 specific inhibitor. Detectable COX2 expression
was also reported in Leydig cells in the human [72] and the Syrian hamster [73]. COX2 expression
and PGs production were reported in hamster and rat Sertoli cells where its regulation by FSH
and testosterone is proposed to play a role in glucose uptake and spermatogenic efficiency [74,
75]. COX1 is expressed in human testicular peritubular cells (HTPCs), which were also found to
secrete PGE2 and express EP receptors [76]. Moreover, our lab recently reported the presence
of COX enzymes and detectable levels of PGs in neonatal rat gonocytes [77] and in mouse and
rat spermatogonia [78]. Detectable baseline levels of expression of COX enzymes and PGs in
several models support a maintenance role of the eicosanoid system in male reproduction,
however the exact mechanism behind this role remains elusive.
Several roles of COX enzymes have been considered to explain the positive correlation
between COX expression and/or function with pathological testicular conditions. Reports have
suggested COX involvement with steroid biosynthesis or feedback induction when
spermatogenesis is impaired [79]. More recently, Kubota et al examined the potential role of
COX2 in the pathogenesis of testicular cryptorchidism. In an experimental cryptorchidism mouse
model, COX2 expression was induced in cryptorchid testes with 4-fold greater mRNA expression
compared to controls, whereas COX1 expression remained unchanged [80]. Disturbance of the
arrangement of cells in the seminiferous tubules suggested that spermatogenesis was disrupted
by experimental cryptorchidism, leading to the proposition that COX2 may play a role in protecting

16
germ cells against injury, a proposal which has yet to be confirmed by other studies. However,
we cannot disregard the involvement of other processes contributing to the pathogenesis of
cryptorchidism that may be targeted by COX inhibitors. The leading hypothesis is that COX
inhibitors can act as endocrine disruptors and interfere with steroidogenesis, which will be
elaborated in section 1.5.3.3. Furthermore, the exposure of these drugs during the critical male
programming window (GD15.5-18.5 in rats, GW8-14 in humans) may be involved in diminishing
normal androgen production. Insulin like 3 (INSL3) expression in human fetal Leydig cells was
also shown to be downregulated by ibuprofen, thereby interfering with testicular descent [7, 9, 81,
82].
Though not directly involved with the testes, it is worth briefly mentioning that neuro-
morphological and neurochemical differentiation involved in characteristic male sexual behavior
involve interactions between the eicosanoid pathway and steroid hormones and can be targets
for NSAIDs and analgesic drugs as well. For example, induction of dendritic spines in the first few
days of postnatal life in rats involves an estrogen-dependent increase in COX2 and PGE2, which
maintains the density responsible for the masculinization of the male rat brain [83, 84]. While the
analgesic effect of acetaminophen is due to its ability to inhibit COX activity in the brain, it has
been recently shown that its metabolite, p-aminophenol, can also cross the blood-brain-barrier
and act on vanilloid and cannabinoid receptors. Therefore, it is quite possible that NSAIDs and
analgesic drugs can adversely impact male reproductive development outside the testes [85, 86].
Overall, there have been few studies on the role of COX enzymes in spermatogenesis.
Nonetheless, the accumulation of evidence to support involvement of COX enzymes in the
pathogenesis of testicular disease would suggest a role of the eicosanoid pathway that is
significant to proper male reproductive development but the mechanism behind their involvement
is relatively still unknown. Recent efforts have turned towards understanding potential effects
induced by prostanoid end products due to their well-characterized roles as vital chemical

17
messengers within the body and their involvement in inflammatory responses, the female
reproduction, and embryonic development.

1.5.2 Expression of PG synthases and prostanoid receptors in the testis

Whereas studies would suggest a protective role of COX enzymes against testicular insult,
reports on whether PGs have been beneficial to the testes have been controversial in the past.
Some groups have determined that exogenous PGs can be harmful to the testes [87, 88], but
others have found that PGs are important mediators involved in inflammatory processes behind
testicular damage [89]. Since the 70s, efforts to better understand the function of the eicosanoid
pathway in maintaining testicular homeostasis have been focused on understanding how different
PGs signal on a cellular basis. So next, we will explore the expression of PGs in various cell types
in the testes, and then dive deeper in the known roles of specific PGs in testicular development
as outlined in Table 1.

Table 1. Reports of the expression of prostaglandins (PGs), PG synthases and receptors in the male reproductive system. Addition of species type and main findings aid
to briefly summarize what is currently known about involvement of these compounds in the field of male reproduction.

PGs
PG
Synthase
PG
Receptor
Cell type /
Tissue
Species Finding(s) Reference
PGD2 L-PGDS DP1
Testis
NT2/D1
testicular
cancer cell
line
Human
PIG1 enhances PGD2 production, further increasing
cAMP levels and SOX9 activation. RIG1-PGD2
signaling might play an role in cancer cell supression
in the testis
Wu CC et al. 2012
PGD2 L-PGDS
Leydig Cells,
epididymis
Bovine
and
human
Positive correlation between concentrations of PGDS
and PGD2. L-PGDS was found in Leydig cells; may
play a role as retinoid-binding protein to maintain
spermatogenesis
Tokugawa Y et al.
1998
PGD2 L-PGDS
Sertoli cells
and germ
cells
Mouse
L-PGDS-/- XY gonads showed abnormal SOX9
cellular localization pattern at GD11.5. SOX9
activates and maintains L-Pgds transcription. PGD2
involved in maintaining Sox9 expression during
Sertoli cell differentation. PGD2 is weakly expressed
in germ cells
Moniot B et al. 2009
PGD2 L/H-PGDS DP
Interstitial,
Leydig cells,
Mast cells
Human,
Hamster
DP is expressed in interstitial cell clusters. L-PGDS
is expressed and PDG2 is secreted by hamster
Leydig cells. H-PGDS is expressed in mast cells
Schell C et al. 2007
PGD2 L/H-PGDS DP2
Embryonic
germ cells
Mouse
PGD2 acts through DP2 and is involved in germ cell
differentiation. Nanos2 downregulated in L-PGDS-/-
testes
Moniot B et al. 2014

PGD2 L/H-PGDS
Testes
(postnatal)
Mouse
Spermatogonia apoptosis in L-PGDS-/- mice. 24% of
L-PGDS-/- mice presented phenotype of unilateral
cryptorchidism
Philibert P et al.
2013
PGD2 PGDS
Sertoli cell
lineage
Mouse
PGDS is expressed in a similar dynamic spatio-
temporal expression pattern in developing mouse
gonads. Activating effect of PGD2 on Sox9
transcription
Wilhelm D et al.
2007
PGD2 PGDS
Sertoli cells
and
gonocytes
(fetal)
Mouse
PGDS is expressed at GD11.5 and GD12.5 in male
genital ridge. PGD2 is a response to somatic
masculinizing environment
Adams IR and
McLaren 2002
PGD2
PGES,
PGIS,
L-PGDS
DP1, DP2,
EP2, EP4
Sertoli cells Mouse
Expression of prostanoid receptors (DP1/2, EP2/4),
and PGES, PGIS, and L-PGDS was found in
GD10.5, 11.5, and 12.5 Sertoli cells and Sertoli-like
NT2 cells. PGDS/PGD2 pathway induces Sox9
nuclear translocation in Sertoli cells
Malki S et al. 2005
PGD2
Juvenile
Sertoli cells
(SC5 cell line)
Mouse
PGD2 can be produced in SC5 cell line and
compounds that inhibit COX activity are able to
reduce PGD2 synthesis when PGD2 levels were
measured with an ELISA
Kugathas S et al.
2016
15-
dPGJ2

Peritubular
cells
Human
15-dPGJ2 influences expression of differentiation
markers and contractability of pertibular cells, and is
involved in the generation of fibrosis that occurs in
tubule walls of infertile patients
Schell C et al. 2010
PGD2/
15-
dPGJ2

Pertibular
cells
Human
15d-PGJ2 is detected in infertile patients and acts
via reactive oxygen species to alter the phenotype of
peritubular cells
Kampfer C et al.
2012

L-PGDS
Testicular
Sertoli and
interstitial
cells
Mouse
L-PGDS was expressed in Sertoli cells at stages VI-
VIII of spermatogenic cycle (late spermatids) of the
seminiferous epithelium
Gerena RL et al.
2000
L-PGDS
Seminiferous
tubules (fetal,
postnatal),
Leydig Cells
(Adult)
Mouse
In adult testis, L-PGDS is largely confined to Leydig
cells. In neonatal testis is within seminiferous
tubules. Expression decreases after birth and
increases up to 10-fold between days 30-40
Baker PJ and
O ́Shaughnessy PJ
2001

L-PGDS
(as B-tr)
Leydig Cells Mouse
L-PGDS (as B-tr) is expressed specifically in the
interstitial space of the testis and in epithelia of the
epididymis
Hoffman A et al.
1996

L-PGDS
(mRNA)

Sertoli and
germ cells
Rat
PGDS is highest in the epididymis and is
predominantly accumulated and expressed in the
caput epididymis, likely being involved in sperm
maturation
Sorrentino C et al.
1998
L/H-PGDS DP Epididymis Rat
DP is weakly expressed in the epididymis. H-PGDS
and L-PGDS is expressed weakly in the epididymis
Gerashchenko D et
al. 1998
PGDS
Sertoli-
Sertoli/Sertoli-
germ cell
junctions
Rat
PGDS increases significantly at onset of puberty in
relation to assembly and disassembly of cell
junctions. It is likely to act as a carrier protein in
Sertoli cells to transport molecules that are important
for spermatogenesis (retinoic acid, retinal, T3)
Samy ET et al. 2000
PGE2
EP1, EP2,
EP4
Pertibular
cells
Human
PGE2 is expressed in human testicular peritbular
cells. EP3 was found in spermatogonia, and was the
only EP not expressed in pertibular cells. PGE2
elevated GDNF levels around 2-fold after 3 hours,
and increased mRNA levels for calponin
Rey-Ares V et al.
2018

PGE2,
PGF2a

EP1-4, IP,
FP
Sertoli cells Rat
Il-1B induces Sertoli cell PGE2 and PGF2a
production, PGs can induce IL-1b in dose- and time-
dependent but COX2 independent manner
Ishikawa T and
Morris PL 2006
PGE2,
PGI2
Sertoli cells Rat
Sertoli cells synthesize PGE2 and PGI2 and FSH
potentiates PG production and cAMP production.
Sertoli cell production of PGF2a may regulate germ-
cell development
Cooper DR and
Carpenter MP 1987

EP and
FP
Stem and
Adult Leydig
cells
Rat
Rat Leydig progenitors express FP, EP1, EP2, and
EP4. EP1 and EP4 are expressed in adult Leydig
cells
Walch L et al. 2003
PGF2a FP Leydig Cells Hamster
Inhibitory effect of PGF2a on expression of StAR and
17b-HSD as well as synthesis of testosterone
induced by HcG and LH
Frungieri MB et al.
2006
PGF2a Leydig Cells Hamster
Stimulatory effect on testosterone on PGF2a
production via non-classical mechanism that involves
phosphorylation of ERK1/2. PGF2a may play a role
acting as a brake on testicular steroidogenesis
Matzkin ME et al.
2009

22
1.5.2.1 Prostaglandin Synthases
Out of the several major PG synthases, PGDS has been one of the most characterized in
the testis. PGDS was discovered in the mouse testis in the 1990s expressed in Leydig cells and
in the epithelial cells of the epididymis [90]. This finding was corroborated by several studies
reporting the presence of L-PGDS in the Sertoli cells during stages VI-VIII of the spermatogenic
cycle, and in the interstitial cells of the mouse testes [91], as well as measurable levels of PGD2
produced in juvenile mouse Sertoli cell line SC5 [92]. When tracing L-PGDS expression during
early murine reproductive development, Baker & O’Shaughnessy found L-PGDS localized in the
tubules during fetal and postnatal testes. The synthase begins to be expressed between GD11.5-
12.5 in the male genital ridge, and expression decreases after birth and upregulates
approximately 10-fold between PND30-40, and in the adult, L-PGDS is found in Leydig cells [93,
94]. In the rat, both H- and L-PGDS were found to be expressed in the epididymis, and in Sertoli
and germ cells, where they are involved in sperm maturation [95, 96]. Like in the mouse, PGDS
in the rat increases in expression from birth to adulthood, with clear increases starting in
prepubertal rats when the formation of the blood-testis-barrier takes place. The role of PGDS in
the assembly and disassembly of cell-cell junctions between Sertoli and Sertoli/germ cells
suggests that PGDS may mediate the transport of molecules involved in spermatogenesis [97].
Thus, in the mouse and rat models, PGDS may play important roles in supporting
spermatogenesis.
As for the other synthases, PGES and PGIS were reported to be expressed in a mouse
Sertoli-like cell line [98], and detectable levels of PGI2, PGE2, and PGF2a were found in rat Sertoli
cells, which would suggest functional synthases [99, 100]. PGE2 is produced in human peritubular
cells [76], and PGF2a is produced in hamster Leydig cells [73, 101], which would indicate
expression of PGES and PGF2a in non-rodent models as well.

23

1.5.2.2 Prostaglandin Receptors
Expression of PG receptors have also been reported in several cell types in the male
reproductive system. In the human, DP1 is expressed in a testicular cell line, NT2, suggesting
that PGD2/DP1 may play a role in the suppression of carcinogenesis in the testis [102]. DP2 is
expressed in embryonic germ cells in mice [103], and both DP1 and DP2 were expressed in fetal
mouse Sertoli cells, as well as EP2 and EP4 [98]. Rat epididymis express DP receptors [95], as
well as other PG receptors: EP1-4, IP, and FP in Sertoli cells; EP, EP1-2, EP4 in Leydig cell
progenitors; and EP1 and EP4 in adult Leydig cells [99, 104]. Hamster Leydig cells express DP
and FP receptors [105, 106]. Human peritubular cells express EP1, EP2, and EP4 [76].

1.5.2.3 Roles of PGD2 and 15d-PGJ2
PGD2 has also been studied widely to decipher its role in the testis. A series of studies
have attempted to uncover its role in Sertoli cell maturation through the regulation of Sox9
expression. In the human-derived NT2 Sertoli-like cell line, PGD2 was shown to be involved in
Sox9 nuclear translocation during embryonic development [98]. Later findings in mice uncovered
the ability of PGD2 to activate Sox9 transcription in cells fated to the Sertoli cell lineage [107]. In
L-PGDS
-/-
gonads of mice at GD11.5, Sox9 cellular localization was found to be altered,
definitively confirming a significant role of PGD2 to regulate the differentiation of Sertoli cells
during fetal development [108].
As for mouse germ cells, PGD2 was found to be weakly expressed, but was still crucial in
regulating germ cell differentiation [108]. Its production by mouse PGCs and gonocytes suggests
PGD2 can be a factor released in response to a somatic masculinizing environment [94]. In
mechanistic studies utilizing mouse PGDS knockout models, Nanos2, which promotes fetal germ
cell differentiation, was downregulated in a L/H-PGDS
-/-
embryonic model [103]. Moreover, a
study following the development of a postnatally derived L-PGDS
-/-
mouse model showed that

24

spermatogenesis was disrupted, consistent with the development of unilateral cryptorchidism in
adult animals [109].
In humans, PGDS is expressed in Leydig and peritubular cells and PGD2 levels have
been detected. The PGDS/PGD2 system in human Leydig cells was hypothesized to play a role
in retinoid-binding in regulation of spermatogenesis [110]. Expression of PGD2 in peritubular cells
may be associated with greater infertility, as infertile patients were found to express changed
levels of 15d-PGJ2 (metabolite of PGD2). 15d-PGJ2 is believed to be involved in altering the
phenotype of human peritubular cells such that differentiation markers and contractability were
affected [111, 112]. Thus, PGD2 may contribute to the generation of fibrosis that occurs in tubule
walls of infertile patients.

1.5.2.4 Roles of PGE2 and PGF2a
Though not as widely characterized as PGD2, some groups have also attempted to
decipher roles of PGE2 and PGF2a in male reproductive development. In the rat, Sertoli cells
were able to synthesize PGE2 and PGF2a in response to IL-1b and Follicle Stimulating Hormone
(FSH), and in response PGE2 and PGF2a were able to induce IL-1b in positive feedback loop
[99, 100]. In hamsters, PGF2a was found to stimulate testosterone production in Leydig cells
when induced by human chorionic gonadotropin (HCG) and Luteinizing hormone (LH) via a non-
classical mechanism that involves the phosphorylation of ERK 1/2. PGF2a also has a inhibitory
effect on StAR and 17b-HSD, perhaps acting as a brake on testicular steroidogenesis [73, 101].
PGE2 is expressed in human testicular peritubular cells, and was shown to be able to
elevate GDNF levels and increase mRNA levels of calponin, supporting a role in spermatogenesis
and maintenance of the testicular interstitium [76].

25

1.5.3 Effects of NSAIDs and analgesic drugs on male reproductive development
The accumulation of literature suggests that members of the eicosanoid system are widely
expressed in several cell types of the male reproductive system and are involved in fundamental
developmental processes such as regulating germ and somatic cell differentiation,
spermatogenesis, testosterone production, and more. While the specific signaling mechanisms
behind these effects are still being interrogated, there is ample evidence to propose that altering
normal eicosanoid synthesis with COX modulatory drugs can result in disruptions in
developmental processes. This next section will discuss reported effects of NSAID and analgesic
drugs when administered during the dynamic reproductive maturation that takes place during
perinatal periods and review the weight of evidence in support of reported immediate and long-
term reproductive effects.

1.5.3.1 Effects of NSAIDs in Testis
Several studies have assessed the role of NSAIDs and analgesic drugs on the testis and
their ability to induce characteristic reproductive birth defects. An early study conducted in mice
found that aspirin (150 mg/kg) or indomethacin (1 mg/kg) inhibited the masculinization of male
genitalia in GD18 embryos when exposed in utero from GD11-14. While administration of AA (25
or 100 mg/kg) was able to correct anti-masculinization effects such as hypospadias and
shortening of the AGD, indomethacin (1 mg/kg) exposure completely blocked those protective
effects when administered with AA and a positive control [113]. Other studies focusing on fetal
exposure found similar effects. In utero exposure of acetaminophen (150, 250, 350 mg/kg/day) in
rats from GD13 to GD21 resulted in significantly reduced AGD index (AGDi) in all dose groups,
while aspirin (150, 200, 250 mg/kg/day) exposure resulted in growth retardation of the fetuses.
Although there were no differences in AGD, trends towards decreases in testosterone production
were observed with aspirin exposure [7]. In a subsequent study also conducted in rats,
acetaminophen at 350 mg/kg/day was dosed to dams from GD7 to GD19 and then again from

26

PND14 to PN22. There were no differences in the AGDi in offspring, only a reduction of 1.6%, but
the difference was not significant [114]. This study contrasts with what was reported in Kristensen
et al. 2011, but this difference may be attributed to the different ages of exposure, as trends
towards reduction in AGDi was also observed by Alexstad et al who reported a significant increase
in the rate of retained nipples in the male offspring, suggesting an antiandrogenic activity of
acetaminophen. This further highlights the importance of considering the age of exposure when
assessing the risk of chemicals, due to potential differences in the reproductive susceptibility
windows of the fetal versus pre-pubertal age groups.
While Gupta and Goldman saw effects of indomethacin in GD18 mice fetuses, fetal
exposure to a similar dose of indomethacin (0.8 mg/kg) from GD15.5 to GD18.5 in rats resulted
in very little adverse effects in the later stage fetuses or in offspring. Despite causing a significant
decrease in testes weight at GD21.5, there were no alterations in testosterone levels or AGDi.
The offspring did not exhibit hypospadias or cryptorchidisms and AGDi was not affected in puberty
or adulthood. Interestingly, penile length was significantly decreased in the treatment group at
PND25 [115]. In a similar study evaluating long term effects of in utero exposure of ibuprofen,
adult mice exposed to ibuprofen (5.6 mg/kg/day) from GD5-18 had no differences in sperm
motility, viability, or response to hypoosmotic shock. There were also no differences in fertilization
index or acrosomal integrity when compared to controls [116].
Pre-pubertal exposure of rats (PND23 to PND43) to ibuprofen (2.4-14.3 mg/kg)
compromised sperm parameters, the number of sperm, and daily sperm production. Daily sperm
production was significantly decreased when rats were exposed to the highest dose of ibuprofen
(14.3 mg/kg), but sperm transit time or morphology was not affected by the treatment.
Furthermore, fertility was also reduced in highest dose group, including greater pre-implantation
loss rates of the offspring [117].
Studies suggest that NSAIDs and analgesic drugs can affect testicular development in
rodents. For example, several studies reported that early fetal exposure of pharmacological COX

27

inhibitors induced an immediate reduction of AGDi in rodent models [7, 113]. Other studies did
not observe such an effect on the AGDi of rat offspring exposed in utero to these drugs, which
may suggest an ability of the testes to recover by birth and subsequently, resume normal
reproductive potential in puberty and adulthood [114, 115]. While fetal exposure of ibuprofen did
not have long-term effects on sperm motility or viability [116], pre-pubertal exposure did lead to
decreased sperm production and fertility [117]. However, no report has evaluated the immediate
and long-term effects of neonatal exposure of NSAIDs, which is of interest to us due the significant
plasticity of the testes during this stage of development.

1.5.3.2 Effects of NSAIDs on Fetal and neonatal germ cells
As fetal and postnatal gonocytes support lifelong spermatogenesis, studies have
evaluated effects of NSAIDs and analgesic drugs directly on germ cells, particularly to determine
whether they affected viability or had the ability to affect immediate and long-term
spermatogenesis. Intrauterine exposure of indomethacin (0.8 mg/kg) and acetaminophen (350
mg/kg) in rats induced decreases in the expression of Oct4, a fetal germ cell differentiation marker
at ages between GD15.5 and GD17.5 [118]. In GD13.5 mouse gonadal sections, in utero
exposure of COX inhibitors (30 mg/kg/day acetaminophen, 50 mg/kg/d aspirin, 15 mg/kg/d
ibuprofen) from GD10-13.5 significantly reduced the percentage of S-phase proportion of germ
cells and resulted in significantly lower proliferative index. Intergenerational effects of in utero
exposure were also observed, with the first generation exhibiting decreased sperm counts [119].
In vitro studies found similar abilities of NSAID and analgesic drugs to reduce fetal germ
cell number. First and second semester human fetal testis fragments either cultured in hanging
drops or used as xenografts in mice were exposed for 7 days to acetaminophen or ibuprofen, at
10μM in organ culture, or 10 and 20 mg/kg 3 times daily respectively in mouse xenografts. In both
cases, acetaminophen and ibuprofen significantly decreased fetal gonocyte numbers and
proliferation compared to controls. In the same study, treatment of the human embryonal

28

carcinoma cell line NTera2 with 10 μM acetaminophen or ibuprofen reduced the expression of
pluripotency markers Pou5f1 for both drugs, and Tfap2c with acetaminophen. Both drugs
increased the expression of the epigenetic gene TET1, as well as the relative levels of epigenetic
repressive regulator H3H27me3, while only acetaminophen significantly decreased the
expression of DNMT3b [120]. Using a fetal testis explant gonad assay (FEGA) system, the
expression of five germ cell markers were downregulated upon 48 hrs of ibuprofen exposure:
Pouf51, Tfap2c, Lin28a, Alpp, and Kit. However, neither the morphology nor the density of germ
cells were altered up to 72 hrs of 10
-4
and 10
-5
M treatment of ibuprofen [82].
A study evaluating the in utero effect of the NSAID diclofenac, a derivative of phenyl acetic
acid, at 0.2, 1, and 5 mg/kg/day, did not find alteration in the testes of rat offspring by
histopathological analyses, nor in plasma testosterone concentrations [121]. However, treatment
of adult rats with diclofenac sodium at 0.25, 0.50 and 1.0 mg/kg reduced total sperm count, total
number of motile sperm, and sperm density in the epididymis. Upon histological testis
examination, treatments with 0.50 and 1.0 mg/kg resulted in the complete arrest of
spermatogenesis and shrinkage of the seminiferous tubules, whereas 0.25 mg/kg diclofenac
sodium treatment primarily affected secondary spermatocytes and spermatids [122].
In neonatal germ cells, the treatment of PND3 rat gonocytes with acetaminophen and
ibuprofen at a high dose of 20 μg/ml (132 μM and 97 μM respectively), induced significant
increases in proliferation, whereas ibuprofen at low (5 μg/ml; 24 μM) and high doses inhibited in
a dose-dependent manner the RA-induced expression of Stra8, a gonocyte differentiation marker.
Moreover, acetaminophen and ibuprofen at high doses decreased Cox2 expression, alone or
combined with RA [77].
Overall, findings indicate that pharmacological COX inhibition alter differently germ cell
development depending on the age at which exposure occurs. When administered during fetal
development, these drugs either decrease proliferative ability [115], directly reduce germ cell
number [120], or decrease the expression of germ cell markers [82]. In contrast, neonatal

29

exposure to these drugs increases gonocyte proliferation and decrease the ability of these cells
to differentiate [77], which could lead to subsequent delays on the formation of foundational
spermatogonial stem cell pool and long-term spermatogenesis.

1.5.3.3 Effects of NSAIDs on Leydig cells and steroidogenesis
Most studies evaluating the effect of NSAIDs and analgesic drugs in early male
reproductive development have focused on the ability of these drugs to target Leydig cells and
alter androgen levels. The initial report that NSAIDs had the ability to target Leydig cells and affect
hormone secretion was published 2003 when it was discovered that 1 μM indomethacin treatment
significantly increased StAR protein levels in the mouse MA-10 Leydig cell line stimulated with
0.05 mM Bt2cAMP. This alteration was concomitant with an increase in progesterone production,
though this was not observed with indomethacin treatment alone. Treatment with a COX2 specific
inhibitor, NS398, also significantly increased StAR protein expression in addition to progesterone
production in MA-10 cells stimulated with 0.05 mM Bt2cAMP, whereas no alterations were
observed in StAR levels or steroid production when treated with a COX1 specific inhibitor [79].

Fetal Leydig Cells
In utero exposure of 350 mg/kg acetaminophen in rats starting at GD13.5 resulted in
reduced AGDi in GD21.5 fetuses, and mRNA expression of key enzymes of the steroidogenic
pathway (Cyp11a1, Cyp17a1) were significantly decreased, though there was no change in
Leydig cell number or size [123]. In a separate study, in utero exposure of rats to 10-60 mg/kg of
ibuprofen from GD15 to PND21 was found to lower volumes of Leydig cell nuclei by PND90, with
a concomitant decrease in testosterone levels in the highest dose group. LH and FSH levels were
similar between groups, and no alterations in spermatogenesis or Sertoli cell dynamics were
observed [124]. When fetal rat testes at GD14.5 were exposed to acetaminophen (1 μM) or aspirin
(1 and 10 μM) for 24 to 72 hours ex vivo, 1 uM acetaminophen reduced testosterone production

30

and dose dependent reductions were observed with aspirin as well, with a significant reduction at
the 10 μM dose [7]. This study was corroborated with doses up to 100 μM, showing similar effects.
Indomethacin at 10 μM decreased fetal rat testosterone production, but no alterations in Insulin-
like hormone 3 (INSL3) were observed [125]. However, when pregnant mouse dams were
gavaged with 50 mg/kg/day or 150 mg/kg/day of acetaminophen, there were no abnormalities in
the gonads of the male offspring when they examined germ cell markers, Leydig cell morphology,
or spermatogenesis, despite finding a reduction in AGDi at 10 weeks old after in utero exposure
of acetaminophen [126].

Juvenile and Peripubertal Leydig Cells
In rats exposed to ibuprofen during juvenile and peripubertal stages of development, 2.4
mg/kg/day and 7.2 mg/kg/day doses resulted in reduced testosterone levels, while the 7.2
mg/kg/day dose presented increased FSH and altered LH levels. Furthermore, all treatment
groups exhibited decreased Leydig cell number and significantly reduced Leydig cell volume in
the 2.4 mg/kg/day treatment group [117]. In human adult testis explants exposed to 10
-4
M aspirin,
10
-5
M indomethacin, and 10
-5
M and 10
-4
M acetaminophen, testosterone secretion was
decreased, while INSL3 expression was reduced by both aspirin concentrations and 10
-4
M
indomethacin after 24 hours, but the number of Leydig cells was increased by 10
-4
M aspirin and
10
-5
M indomethacin. In addition, no effects on gross morphology of the testes were observed
[127]. In a study of testes from young men xenografted in mice hosts, testosterone was inhibited
by treating the host mice with ibuprofen at 10
-4
M at 24 hours, and 10
-5
M at 48 hrs. Furthermore,
ibuprofen inhibited all steroids from pregnenolone down to testosterone, which corresponded with
decreases in gene expression involved in testicular steroidogenesis, except Cyp19a1 [128].

31

Studies from human-derived models
Using the FEGA system, dose-dependent decreases in testosterone were observed in
fetal human testis cultures after 24 hrs of exposure to 10
-4
and 10
-5
M ibuprofen, at GW 8 to 9, but
not in younger or older fetal testes. Human fetal testis (GW 10-12) xenografted mice exposed for
48 hours to ibuprofen presented decreased expression of steroidogenic genes Cyp11a1,
Cyp17a1, and Hsd17b3, and decreased INSL3 levels after 72 hours of exposure. However the
treatments had no effect on testosterone production in these conditions [82]. These findings are
in contrast to that of another study, in which three times daily of 20 mg/kg acetaminophen
exposure in human fetal testes (GW 14-20) xenograft mice resulted in significantly reduced
testosterone levels compared to vehicle-exposed mice [123]. This was confirmed in a subsequent
experiment in which 7-day ex vivo exposure of 350 mg/kg acetaminophen on human fetal explants
reduced overall testosterone levels.
A study evaluating mixtures of ketoconazole, BPA, valproic acid, and theophylline found
a dose-dependent pattern of declining testosterone levels in human fetal testes explants (GW10-
12). However, despite alterations in testosterone levels there were no morphological signs of
impairments or significant changes in cleaved caspase 3 expression [129]. The effects of several
analgesic drugs (acetaminophen, aspirin, indomethacin, and ketoconazole at doses 10
-4
to 10
-7

M) on ex vivo human fetal testes (GW 7-12) organ cultures resulted in unique alterations.
Ketoconazole treatment significantly decreased testosterone and INSL3 levels were significantly
decreased, whereas indomethacin and aspirin stimulated testosterone levels. While no alteration
in testosterone level was found with acetaminophen or its metabolite, decreases in INSL3 levels
were observed [130].
Human fetus explant studies suggest decreases in testosterone levels with exposure to
NSAIDs and analgesic drugs, particularly acetaminophen, when dosed between GW10-20,
seemingly when targeting the latter duration of masculinizing programming window in humans
(GW8-14). However, the complexity lies when considering the impact of ages of fetuses, doses,

32

exposure levels, and the type of NSAID or analgesic used in the study on steroidogenic effects.
Thus, future studies aimed at evaluating a wide range of compounds at several age groups are
important in gaining an overall better picture of the effects of these drugs on steroidogenesis.
In an in vitro model of steroidogenic cells, the NCI-H295R human adrenocortical cell line,
INSL3 levels were decreased with exposure to both doses of aspirin and 10
-4
M indomethacin,
but no effects were observed with ketoconazole or acetaminophen treatment on INSL3 production
[127]. Interestingly, in another study, exposure of acetaminophen to up 1000 μM increased
pregnenolone and decreased hormone levels downstream from progesterone. Treatment of
dipyrone (100, 314, and 1000 uM) on human derived H295R cells reduced concentrations of
testosterone and was concomitant with an increase in progesterone and induction of CYP21
activity. However, intrauterine dipyrone (50, 100, or 200 mg/kg/day) did not result in testosterone
reduction in contrast to DEHP at 750 mg/kg/day, which was used as a positive control [131].
Overall, these studies strongly suggest that NSAIDs and analgesic drugs can target
Leydig cells and alter their ability to regulate hormone levels, particularly testosterone, as reported
in ex vivo [82, 123, 125, 127, 129], in utero [124], and in vitro [131] studies. In support of these
effects, several studies reported alterations of steroidogenic enzymes [79, 82, 123, 128, 131].
There is ample evidence to support the conclusion that COX inhibitors can affect normal Leydig
cell function and impact steroidogenesis. Despite the findings presented here, there is a gap in
the understanding of the link between prostaglandins and steroidogenesis. Use of NSAIDs and
analgesic drugs may only partially involve prostaglandin inhibition in the Leydig cell and thereby
may be disrupting steroidogenesis through additional mechanisms such as by inhibiting
steroidogenic enzymes or inducing oxidative stress [7, 115, 125, 132, 133]. Cell-type specific
effects of these drugs on other testicular cell types must be considered as well, as the normal
function of the testes relies on the coordinated efforts of all the cell types involved.

33

1.5.3.4 Effects of NSAIDs on Sertoli cells
While there have been many reports on detrimental the effects of NSAID and analgesic
drugs on the development of germ and Leydig cells, significantly less findings were reported on
Sertoli cells. In utero exposure of either low (3.6 mg/kg/day), medium (9 mg/kg/day), or high (18
mg/kg/day) diclofenac sodium from GD15-21, rat pups at PND7 showed significant decreases in
Sertoli cell at the medium and high drug doses. Interestingly, seminiferous tubules did not exhibit
lumens, and the tubules consisted only of spermatogonia and Sertoli cells. Spermatogonia were
also significantly reduced with the medium and high doses, with many cells exhibiting pycnotic
nuclei and remnants of dead cells. However, no significant differences were observed in AGD or
nipple retention [134].
Ex vivo ibuprofen treatment of 10
-5
and 10
-4
M in human fetal testes (GW 8-12) explants
dose dependently inhibited Anti-Müllerian hormone (AMH) levels. This is supported by decreases
in gene expression of Amh and Sox9 with the 10
-5
M ibuprofen dose at 24 and 48 hours [82].
Aspirin treatment of human fetal testes (GW 7-12) cultures induced AMH production in Sertoli
cells, whereas trends in increases were observed with acetaminophen and ibuprofen treatment,
and ketoconazole inhibited AMH production. However, none of the treatments altered Sertoli or
germ cell ratio [130].
Despite some effects observed in rats with in-utero exposure of NSAID diclofenac sodium
on Sertoli cell number and the morphology of the tubules, there were no alterations in Sertoli cell
ratio observed in human testes explants. Other studies that evaluated effect of NSAIDs on the
testes found no changes in Sertoli cell number [82, 119, 123, 124], despite some decreases in
AMH protein and mRNA levels and Sox9 levels [82, 128]. Therefore, further studies will need to
be conducted on Sertoli cell models to gain a more complete understanding of the effect of
NSAIDs and analgesic drugs on early male reproductive development.

34

1.5.3.5 Effects of NSAIDs on Testicular Macrophages
No studies on the effects of NSAIDs and analgesics drugs on testicular macrophages
have been reported in the literature. However, it is well established that macrophages are targets
of pharmaceutical COX inhibitors and these drugs have been found to affect TNFa release, nitric
oxide production, and proliferation [16]. In the testis, macrophages are intimately involved in cord
formation, germ cell development, Leydig cell function, and the maintenance of testicular privilege
[16, 23, 135]. Studies have reported that prostaglandins, in particular PGE2 and PGI2, can affect
the polarization status of testicular macrophages, from M1 to M2 phenotypes. Their presence in
testis, together with PGF2a and PGD2, and changes in testicular Cox2 levels in testis during
inflammation suggest that exposure to acetaminophen or NSAIDs could alter testicular
macrophages, suggesting a potential effect of these drugs on testicular macrophages functions
[23]. Therefore, it is likely the immunomodulatory effects of NSAID and analgesic drugs can
impact the abilities of testicular macrophages to carry out such functions related to maintaining
testicular homeostasis, highlighting the need for more studies in human.

1.6 Epidemiological Evidence
In support of experimental findings, epidemiological studies in humans have also reported
adverse effects in early male reproductive development with exposure to NSAID and analgesic
drugs. TDS conditions such as cryptorchidism, hypospadias, and reduced AGD were positively
associated with maternal use of COX inhibitors at varying stages of pregnancy.

1.6.1 Cryptorchidism

A 2010 study evaluating infant sons of mothers that were using acetaminophen, ibuprofen,
or acetylsalicylic acid during pregnancy found that drug use during the first and second trimesters
were associated with increased cryptorchidism and orchiopexy, and exposure throughout all 3

35

trimesters were associated with increased risk as well. Use during a single trimester or during the
male programming window at gestational week (GW)8-14 was only weakly associated with
cryptorchidism in infants. Exposure to ibuprofen or acetylsalicylic acid alone was not associated
with cryptorchidism, neither was exposure of to a combination of drugs. Interestingly, when drugs
were used consistently for longer durations (5-8 weeks) or when they were used during and
beyond the 4-week male programming window, this resulted in cryptorchidism in offspring [9].
This study is supported by a 2012 prospective cohort study which followed the growth of infants
to early adulthood in the Netherlands. Mothers were asked to report their use of mild analgesics
(NSAIDs, acetaminophen, painkillers like aspirin) during pregnancy at 12 gestational weeks, 20
gestational weeks, and 30 gestational weeks. Amongst the mothers with cryptorchid sons, 33.3%
reported use of mild analgesics during pregnancy, and amongst those with sons with
hypospadias, 31.8% reported use of mild analgesics. However, there were still 29.9% of women
who report use of mild analgesics with healthy sons. The study did find that the use of mild
analgesics during periconception was not associated with cryptorchidism or hypospadias, but that
use between GW14-22 was associated with increased risk of cryptorchidism, even after adjusting
for individual compound use. Use during GW20-32 was not associated with cryptorchism or
hypospadias [136]. Lastly, a prospective birth cohort study conducted in Denmark found that
mothers who use mild analgesics during pregnancy had a significant association with giving birth
to a cryptorchid baby boy. This was significant for ibuprofen and acetylsalicylic acid and trended
in the same direction but was not significant for acetaminophen. For mothers using multiple
analgesics or using them for more than two weeks during pregnancy, the risk was even higher.
The highest risk for cryptorchid babies was found in mothers who were using multiple compounds
for more than two weeks [7]. A study conducted between 2003 and 2006 assessed women who
were exposed to analgesics (aspirin, acetaminophen, ibuprofen) within the first 2 trimesters of
pregnancy or throughout the entire pregnancy. They reported a 4.6% frequency of undescended
testes in offspring compared with 2.9% of unexposed offspring, as well as an odds ratio of 1.2 for

36

women who used analgesics during the first two trimesters, which was similar to those who were
exposed throughout the pregnancy (odds ratio =1.5) [137].

1.6.2 Hypospadias

Using data from the National Birth Defects Prevention Study (2007-2011), Interrante et al.
evaluated associations between NSAID use and birth defects including hypospadias. They found
that approximately 80% of women reported using analgesics during pregnancy, and among the
COX inhibitor user group, acetaminophen and ibuprofen use during pregnancy was significantly
associated with hypospadias, in addition to a variety of other birth defects including gastroschisis
and spina bifida [138]. However, the ability of acetaminophen to cause hypospadias is somewhat
controversial as another assessment of the Danish National Birth Cohort of mothers who had
been exposed to acetaminophen use during the first trimester of pregnancy did not find any
association between acetaminophen use and higher prevalence of congenital abnormalities [139].
However, a positive correlation between ibuprofen use and the development of hypospadias is
supported by another study that evaluated the associations between maternal use of common
medications and herbal remedies during periconceptual period and early pregnancy. The authors
reported an association with analgesic use (aspirin, meperidine HCl, and ibuprofen), but only
ibuprofen use was significantly associated with hypospadias, after adjusting for confounding
factors in the population, such as maternal age, ethnicity, education, BMI, etc. [140].

1.6.3 Reduced AGD

In 2017, a prospective birth study was conducted in which pregnant women were asked
about their medication use of acetaminophen and NSAIDs during GW10-27 and at GW28. Lind
et al. found no association between acetaminophen and reduced AGD in boys 3 months after
birth, but exposure to both acetaminophen and other NSAIDs was associated with a shorter AGD.

37

There was no statistically significant association between acetaminophen use and reduced AGD
in boys, but a strong tendency was observed when acetaminophen was used during the second
trimester of pregnancy. Simultaneous exposure to multiple pain killers did result in a significant
association. Interestingly, in mothers who reported analgesic use but did not specify when during
pregnancy, analgesic use was positively associated with boys presenting smaller penile widths
compared to non-exposed boys [141]. In a prospective cohort study conducted in the United
Kingdom in 2016, mothers self-reported medication uses during pregnancy, and infants’ AGD,
penile length, and testicular descent were assessed at 0, 3, 12, 18, and 24 months of age.
Acetaminophen exposure during GW8-14 was associated with a reduced AGD, and this reduction
is consistent with those measured at birth through 24 months of age. Penile length or testicular
descent were not significantly associated with acetaminophen exposure during pregnancy, nor
was there an association between maternal exposure and cryptorchidism in this study [142].

1.6.4 Long-term Male Reproductive effects

While NSAID and analgesic exposure have been found to cause adverse effects in infants,
studies are starting to evaluate long-term effects in pubertal youth and in adults. A study using
data derived from the Danish National Birth Cohort evaluated the association between pre- and
perinatal exposure of maternal use of acetaminophen and pubertal effects. While some
associations between intrauterine exposure of greater than 12 weeks and a shift towards later
Tanner pubic hair stages were observed, there were no strong associations with male pubertal
development [143]. In adults, a prospective study of couple’s fecundity found that men with high
urinary acetaminophen had an increased time to pregnancy [144]. Regular NSAID users also had
lower serum testosterone (17% compared to non-users) and albumin levels compared to non-
users, but had no difference in serum AMH, sex-hormone binding globulin, or inhibin B [145].
Another study correlating regular NSAID use of adult men throughout life and male reproductive

38

parameters found no associations between drug use and semen quality or male reproductive
hormones. Interestingly, in this study, use of NSAIDs was correlated with significantly higher
levels of testosterone as well as use of combination drugs (acetaminophen, NSAIDs, and
antihistamines) [146].

1.7 Non-Rodent studies
Non-rodent models have been studied to assess whether eicosanoid pathway genes are
expressed and whether their responses to COX inhibitors are similar to that of humans. In the
bull, PGDS activity was detected in bull seminal plasma, as well as the presence of PGD2. PGDS
was also present in the luminal fluids collected from the testis and the epididymis. L-PGDS is
believed to be involved in the rapid conversion of PGD2 to 15-d-PGJ2 in the bull, and likely
functions as a carrier protein for transport across the blood-testis-barrier into the seminal plasma
[147].
Zebrafish models have many advantages for the study of developmental processes due
to their low cost, small size, and similarity of developmental processes to mammalian models. In
a study evaluating the characterization of the zebrafish and chicken homologues of mammalian
L-PGDS, Grozzer et al. found that exon/intron junctions were conserved and that the enzyme
bound lipophilic molecules like lipocalin gene family proteins [148]. Zebrafish gonads at 42- and
90- days post fertilization (dpf) were found to express ptgs1 and two ptgs2 genes, ptgs2a and
ptgs2b, and their expressions were higher in testes than ovaries. Pgds expression was higher in
90-dpf testes and pges expression was high in both 42- and 90-dpf ovaries. Testes produced
higher PGE2 and PGD2 levels, and higher expression of PGE2 and PGD2 receptors as well. 20-
dpf zebrafish exposed to 30 uM meloxicam for 6 days had increased expression of male-specific
genes Sox9a and dmrt1a, while female specific genes were downregulated. Exposure of a PGD2
analog, BW-245C, induced expression of Sox9a after 24 and 48 hours in testes explants and in
juvenile zebrafish [149]. Exposure of BW-245C at 70 dpf resulted in male-biased sex ratios,

39

suggesting that PGD2 influences male sexual development, similar to what was reported in the
mammalian studies [98, 150]. Furthermore, it was discovered that expression of Pgds in the
gonads is localized to Sertoli cells. Initiation of Pgds expression seems to be initiated between 1
and 2 dpf. Pges, on the other hand, was expressed both in male and female gonads [151]. When
zebrafish embryos were exposed to 0.1, 1, 10 and 50 mg/L NSAIDs (acetylsalicylic acid,
ketoprofen, indomethacin, naproxen, ibuprofen, nimesulide, celecoxib) from 0 to 6 dpf, genes
from the PG pathway were downregulated following to exposure from all NSAIDs (ptgs1, ptgs2a,
and ptgs2b), despite being Cox1 or Cox2 selective. Pges was not altered, but Pgds was
upregulated for only the Cox-2 selective inhibitors (nimesulide and celecoxib). Ibuprofen and
naproxen resulted in 22% and 18% increase in the male population, respectively, while Cox-2
selective inhibitors resulted in 31.6% and 26.6% increase in the male population [152]. This is
consistent with the 2014 study which found that meloxicam, another Cox2 selective inhibitor, was
able to induce male-biased populations in zebrafish [149].
In a study conducted on another fish species, the juvenile rainbow trout, who were fed
once a day with salicylate (100 mg/kg) incorporated into the feed, acute ACTH-mediated cortisol
production was significantly inhibited, reflecting alteration of adrenal steroidogenesis. Gene
expression of StAR and PBR, which are involved in cholesterol transport, were also
downregulated in salicylate exposed fish. No significant difference was observed in 11b-
hydroxylase and P450scc expression. Similar results were obtained from in vitro studies in cells
exposed to ibuprofen and acetaminophen in addition to salicylate [153]. Effect of AA on goldfish
testes was evaluated in the presence and absence of hormonal stimulation, which caused a
stimulation of testosterone production in both conditions of hormonal stimulation over the course
of 20-hour incubation. Incubation of indomethacin and ibuprofen resulted in reductions of AA-
induced testosterone production, while PGE2, PGI2, and PGF2a caused an increase in
testosterone production, whereas PGD2 had no effect [154]. This work was corroborated with a
later study that continued to investigate whether the steroidogenic activity of fatty acids was

40

mediated through conversion to cyclooxygenase products. AA at 400 uM stimulated testosterone
production, and this effect was blocked by indomethacin. Testis pieces were incubated with
PGE1, PGE2, and PGE3, which all induced testosterone production. Studies in both the trout and
goldfish suggest that the eicosanoid pathway can modulate steroidogenesis in fish testes [154].
In the chicken, cPGDS was found to be expressed in embryos starting on day 6.5 (stage
30) in the male genital ridge only. By day 8.5, both cPGDS and cSOX9 are co-expressed by
Sertoli cells within the testicular cords. Some germ cells (cSOX9-) were found to also express
cPGDS, similarly to what has been shown in mammalian studies. Expression of cPGDS increases
in a linear fashion from day 6.0 to day 8.0 [155]. In male chicken gonadal explants, PGD2 was
found to activate cSOX9 expression when visualized with in situ hybridization, similar to what was
shown in the mouse and zebrafish [98, 149, 150].
In insects, the Chironomus ripaius (harlequin fly) were treated with 0, 0.1, 1, 10 and 100
μg/L of ibuprofen for 24, 48, 72, and 96 hours. Met, a receptor for Juvenile hormone (JH) which
regulates of metamorphosis and reproductive processes, was downregulated by ibuprofen
exposure, despite JH expression not being altered by the drug [156].
Studies in non-rodent models have shown presence of eicosanoid pathway activity in
species such as the bull, fish, chicken, and insects. In fish, the PGDS and PGD2 axis is important
for sexual differentiation, even though PGE2 is the major prostanoid found in both the zebrafish
and the rainbow trout. COX inhibitors are able to target the eicosanoid pathway in these models,
and alter sex ratios [149, 152] or hormone synthesis [153, 154]. The ability of NSAIDs to affect
development processes have been observed with the chicken [155] and the fly as well [156].
Overall, studies in non-rodent species have found similar results as what has been reported in
rodent models, with the eicosanoid pathway having wide-ranging roles in reproductive
development, steroidogenesis, and maturation of male gonads.

41

1.8 Conclusion
Overall, multiple studies suggest that the eicosanoid pathway is expressed throughout the
testes and mechanistic studies have shown substantial PG involvement in processes of germ cell
development and steroidogenesis, such that when PG synthesis is disrupted with
pharmacological COX inhibitors, these processes are the ones primarily affected. What is
currently known about the involvement of PGs in the regulation of testicular cell types from the
embryonic stage of development to adulthood in various species is summarized in Figure 2.
Laboratory-based studies reporting adverse effects such as cryptorchidism, hypospadias, and
reduced AGD were corroborated by epidemiological studies of infants exposed to NSAIDs and
analgesic drugs in utero. Similar findings were observed in non-rodent studies as well, which
further supports a critical role of the eicosanoid pathway in early male reproductive development.
However, more research is necessary to correlate these mechanistic findings to adverse
reproductive disorders observed in epidemiological studies. These limitations may be due to a
lack of innovative tools to measure direct effects of exposures in offspring [157] and the lack of
studies conducted to evaluate exposures beyond the pregnancy and throughout development.
Furthermore, these effects are particularly important to explore in premature babies and
neonates, who are often long-term consumers of NSAIDs and analgesic drugs [132]. Despite
these limitations, sufficient evidence supports that the eicosanoid pathway is an important
regulatory pathway involved in male reproductive development, and scientists are just beginning
to unveil the immediate and long-term adverse effects that exposure to pharmacological COX
inhibitors can have on the developing male reproductive system.

Figure 2. Schematic illustration of the production and roles of prostaglandins (PGs) in mammalian testis from development to adulthood. PGD2 induces mouse germ cell
differentiation via Nanos2 and activates Sox9 produced by immature mouse Sertoli cells. PGE2 and PGF2a are produced by neonatal rat gonocytes and levels are inhibited
by NSAIDs. PGE2 and PGI2 synthesized by mature Sertoli cells can activate FSH and Il1b in rats. In turn, Il1b can stimulate Sertoli cells to release more PGs. In hamsters,
hCG- and LH- stimulated PGF2a has an inhibitory effect on StaR and 17b-HSD synthesis. In human PTMCs, PGE2 can elevate GDNF and Calponin levels.

Abbreviations: PGC: Primordial Germ Cell, G: Gonocyte, LC: Leydig Cell, Fetal SC: Fetal Sertoli Cell, Immature SC: Immature Sertoli Cell, Mature SC: Mature Sertoli Cell,
SPG: Spermatogonia, PTMC: Peritubular Myoid Cell. T: Testosterone. Dotted arrow: factors produced by specific cell types; (↑): positive effects; (T): negative effect.

43

1.9 Acknowledgements
This work was supported by funds from the University of Southern California School of Pharmacy
to MC; and did not receive any grant from funding agencies. The authors are thankful to the USC
School of Pharmacy for its financial support.

44

CHAPTER 2: CHARACTERIZING THE EICOSANOID PATHWAY IN SSCs AND ITS
INVOLVEMENT IN MODULATING EDC EFFECTS
1

OBJECTIVE (AIM 1): To characterize the eicosanoid pathway in the C18-4 cell line, a
spermatogonial stem cell model, and determine if the cells express eicosanoid pathway
components and produce measurable amounts of prostaglandins.

2.1 ABSTRACT
Perinatal exposure to endocrine disrupting chemicals (EDCs) alters the male reproductive
system. Infants are exposed to genistein (GEN) through soy-based formula, and to Mono(2-
ethylhexyl) Phthalate (MEHP), metabolite of the plasticizer DEHP. Spermatogonial stem cells
(SSCs) are formed in infancy and their integrity is essential for spermatogenesis. Thus,
understanding the impact of EDCs on SSCs is critical. Prostaglandins (PGs) are inflammatory
mediators synthesized via the eicosanoid pathway starting with cyclooxygenases (Coxs), that
regulate physiological and pathological processes. Our goal was to study the eicosanoid pathway
in SSCs and examine whether it was disrupted by GEN and MEHP, potentially contributing to
their adverse effects. The mouse C18-4 cell line used as SSC model expressed high levels of
Cox1 and Cox2 genes and proteins, and eicosanoid pathway genes similarly to levels measured
in primary rat spermatogonia. Treatments with GEN and MEHP at 10 and 100 μM decreased
Cox1 gene and protein expression, whereas Cox2, phospholipase A2, prostaglandin synthases
transcripts, PGE2, PGF2a and PGD2 were upregulated. Simultaneously, the transcript levels of
spermatogonia progenitor markers Foxo1 and Mcam and differentiated spermatogonial markers
cKit and Stra8 were increased. Foxo1 was also increased by EDCs in primary rat spermatogonia.
This study shows that the eicosanoid pathway is altered during SSC differentiation and that
exposure to GEN and MEHP disrupts this process, mainly driven by GEN effects on Cox2
pathway, while MEHP acts through an alternative mechanism. Thus, understanding the role of

45

Cox enzymes in SSCs and how GEN and MEHP exposures alter their differentiation warrants
further studies.

1
This chapter is derived from a published manuscript entitled “In vitro impact of genistein and mono(2-ethylhexyl)
phthalate (MEHP) on the eicosanoid pathway in spermatogonial stem cells” by Tran-Guzman, et al.

46

2.2. INTRODUCTION
Testicular cancer incidences and developmental male reproductive tract abnormalities
have been steadily increasing, in part due to the prevalence and widespread use of endocrine
disrupting chemicals (EDCs) [158, 159]. The discovery of EDCs, chemicals that alter the hormonal
systems which enable an organism to interact with its environment, dates back to more than 50
years ago, when plant compounds were first reported to induce estrogenic effects in sheep [160,
161]. 25 years later, plasticizers were shown to be carcinogenic in rodents after high lifetime
exposures [162]. The molecules that are identified as EDCs form a highly heterogeneous group,
including both synthetic and natural chemicals that range from byproducts of industrial solvents
to those found in foods consumed by humans. Generally, EDCs have phenolic structures that
mimic natural steroid hormones, allowing them to interact with hormonal receptors as analogs or
antagonists. Exposures of EDCs during sensitive windows of development can interfere with
reproductive function, which can be linked to infertility or other reproductive pathologies later in
life.
The effects of EDC exposure vary across an organism’s life cycle and are especially potent
when the exposure occurs during fetal or neonatal development. Infancy represents a critical
developmental period when male germ cells differentiate into spermatogonial stem cells (SSCs),
the lifelong germline stem cells. Male germ cells proceed through proliferation and differentiation
steps, followed by meiosis, to produce spermatozoa from a reservoir of SSCs. SSCs either self-
renew to maintain the SSC reserve, or differentiate, entering the spermatogenic cycle, for sperm
formation throughout a male’s life. SSCs are established early in infancy from a finite and transient
pool of neonatal precursors, the gonocytes [163]. Disrupting gonocyte and SSC development can
have long-term impacts on male reproduction. The etiology of testicular germ cell tumors (TGCTs)
such as seminomas and embryonal carcinomas is believed to reside in the failure of gonocytes
to properly differentiate, whereas the later onset of spermatocytic carcinomas has been proposed

47

to originate from abnormal SSCs [164, 165]. In addition to cancer, improper gonocyte and SSC
development could also contribute to increasing rates of infertility presently affecting over one-
third of men in the United States [166].
The National Health and Nutrition Examination Survey reported that most children have
detectable levels of multiple EDCs across various chemical classes, and children that have higher
concentrations of one chemical are at risk for having higher exposures to other chemicals [167].
Sources of exposure to the phytoestrogen genistein in infants can be found in soy-based infant
formulas, with urinary concentrations of genistein found to be 500-fold greater in infants fed soy
formulas than in infants who were fed formulas derived from cow’s milk [168]. Moreover, the
plasticizer DEHP is widely used in medical devices, which currently represents one of the major
routes of exposures to this plasticizer and its major bioactive metabolite MEHP [169]. Food is
another route of exposure to DEHP, as infants can be exposed to DEHP from ingestion of breast
milk or infant formulas in the range of 1-10 μg/kg/day. Neonates who undergo serious medical
procedures can be exposed up to 10- to 100-fold higher levels.
As humans are simultaneously exposed to multiple types of EDCs, assessing the effects
of individual EDCs does not provide a comprehensive understanding of EDC toxicity. More
recently, significant efforts have been made to evaluate the effects of combined exposures in
comparison to individual exposures [170, 171]. The Endocrine Society has supported this feat by
recognizing the importance of both individual and EDC mixtures in the contribution to adult-onset
disorders [172]. However, a majority of studies that have been conducted on phthalates have
evaluated only their individual effects. Only recently have scientists started studying the effect
mixtures of EDCs on mammalian reproductive development [173, 174]. A study found that
phytoestrogens previously reported to have low estrogenic potencies as individual compounds
can, as a mixture, induce synergistic transcriptional activity resulting in additive effects [175, 176].
Our own team also reported synergistic effects of the estrogenic compounds genistein and BPA
on neonatal gonocyte proliferation [177]. Furthermore, different modes of actions from multiple

48

compounds, such as those interacting with estrogen receptors or ones interfering with steroid
hormone synthesis, can also result in synergistic responses in vivo [178]. Despite these novel
findings, risk to human health associated with exposure to multiple environmental toxicants still
has not been formally monitored by regulatory agencies, highlighting a significant knowledge gap
in the field [179].
Studies suggest that the perinatal testes, and gonocytes in particular, are very sensitive
to environmental exposure to EDCs genistein, DEHP/MEHP and their combination [180-183].
Rats who were exposed in utero to dietary levels of genistein that mimic normal human intake
were shown to have significant reductions in luteinizing hormone levels in adulthood [184].
Perinatal exposure of DEHP at 10 mg/kg/day was reported to induce anti-androgenic effects such
as reduced anogenital distance, whereas higher doses induced histopathological effects in the
testes and reduced expression of androgen-regulated genes in the prostate [185]. We previously
showed that genistein reduced DEHP-induced oxidative stress responses and normalize the
expression of testicular cell markers by increasing antioxidant genes in the neonatal testis of rats
exposed in utero to a combination of genistein and DEHP at 10 mg/kg/day, a dose relevant to
human exposure [182]. Similarly, the prepubertal exposure of male rats to 50 mg/kg/day of
genistein combined with DEHP at 50 to 450 mg/kg/day showed that coadministration of genistein
could alleviate DEHP-induced oxidative stress damage to male reproduction, via enhancement
of testicular antioxidative enzyme activities [186]. Moreover, our lab reported that fetal exposure
to 10 mg/kg/day of genistein and DEHP resulted long-term alterations in somatic and germ cell
markers as well as increased infertility and abnormal testis phenotypes [181]. Macrophages were
also altered, suggesting that inflammatory events had taken place in the affected testes [181,
187]. Interestingly, testicular mast cells and macrophages have also been linked to infertility in
men [188, 189].
Human studies have established a link between testicular inflammatory processes and
infertility in relation to the actions of major inflammatory compounds, prostaglandins (PGs) [190,

49

191]. PG synthesis involves a cascade on enzymatic reactions starting by the metabolism of
arachidonic acid by the rate-limiting cyclooxygenases (Prostaglandin G/H Synthases) Cox 1 and
2 (PTGS 1 and 2), and several PG synthases. As downstream products of eicosanoid synthesis,
PGs are fatty acid based signaling molecules that can induce both physiological and pathological
effects. They regulate physiological systems such as pain perception, cell growth, pregnancy, and
childbirth in females, and play significant roles in mediating inflammation [192]. PGs may play a
role in the regulation of fetal testicular germ cell development in rats and humans [193-195]. We
recently reported that neonatal rat gonocytes express COX2 and produce two prostaglandins,
PGE2 and PGF2a, and that treating gonocytes with acetaminophen and ibuprofen decreased the
production of the PGs, while altering the proliferation and differentiation processes of the cells,
suggesting that PGs might play a role in the regulation of neonatal germ cell development [77].
Studies have shown that genistein decreases the expression of Cox2, the rate-limiting enzyme of
the eicosanoid pathway, in human prostate cancer cell line and primary prostate epithelial cells
resulting in decrease in downstream PGE2 secretion, which is often elevated in cancers. These
results suggest a pro-inflammatory role of genistein which may be in part contributed by inhibition
of tyrosine kinase signaling [196]. On the other hand, Onorato et al. showed that 200 μM MEHP
treatment over 24h inhibited proliferation and increased cytosolic Cox2 expression in mouse
spermatocyte-derived cell line [197]. Furthermore, the interaction between EDCs and the
eicosanoid pathway may be more intimate than previously thought, due to the structural
similarities between phthalates and commercial COX inhibitors [198]. Therefore, EDCs may play
a significant role in disrupting COX activity and altering downstream prostaglandin signaling that
regulate important physiological systems, including reproductive development.
Taken together, few studies have evaluated the impact of EDCs and their mixtures on
early postnatal phases of male germ cell development, despite strong evidence that they
represent sensitive windows of neonatal development [199, 200]. In the present study, we report
that undifferentiated spermatogonia express the eicosanoid synthesis pathway, using the C18-4

50

mouse spermatogonial cell line as a model for SSCs, suggesting a novel role of prostaglandins
in regulating self-renewal and differentiation processes of neonatal spermatogenesis. Moreover,
we report here that mixtures of genistein and MEHP can disrupt eicosanoid biosynthesis in
undifferentiated spermatogonia.

2.3. MATERIALS AND METHODS
2.3.1. Chemicals
Genistein (4′,5,7-Trihydroxyisoflavone) (GEN), mono (2-ethylhexyl) phthalate (MEHP), the
bioactive metabolite of di-(2-ethylhexyl) phthalate (DEHP), and all-trans retinoic acid (RA) were
purchased from Sigma-Aldrich (St. Louis, MO, USA). Stock solutions of 10
−1
M Genistein were
prepared in ultra-pure grade dimethyl sulfoxide (DMSO from VWR International, Radnor, PA,
USA), whereas as stock solutions of MEHP at 10
-1
M and RA at 10
-2
M were prepared in 100%
ethanol (EtOH) (Commercial Alcohols, USA). Stock solutions of chemicals were stored in -20c.
Treatments were subsequently diluted in culture media to achieved desired concentrations of 10
and 100 μM for GEN and MEHP (10
-5
and 10
-4
M abbreviated as 5 and 4 for brevity in figures),
and 1 μM for RA.

2.3.2. C18-4 Spermatogonial Cell Line Culture
The LTAg-immortalized mouse type A spermatogonia cell line was a gift from Marie-Claude
Hofmann (The University of Texas MD Anderson Cancer Center, Houston, Texas, USA). Cells
were cultured in Gibco™ DMEM containing 4.5 g/L d-Glucose, L-Glutamine and 110 mg/L of
sodium pyruvate (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% heat-
inactivated FBS (Sigma-Aldrich, St. Louis, MO, USA) and 1% Penicillin-Streptomycin Solution
100X (Corning
™
) at 35 °C, 7% CO2. The cells were grown in 75cm
2
Corning
TM
cell culture-treated
flasks and passaged every other day.

51

2.3.3. Primary spermatogonia Isolation
Spermatogonia were isolated from postnatal day (PND) 8 rat testes as previously described [201-
203]. Testes were isolated from 10 rat pups per experiment and decapsulated. Spermatogonia
were isolated by sequential enzymatic digestion and differential plating in RMPI 1640 medium
(Invitrogen, Burlington, ON, CA) with 5% heat-inactivated FBS (Sigma-Aldrich, St. Louis, MO,
USA), 2% Penicillin-Streptomycin Solution 100X (Corning
™
), and 1% amphotericin B (CellGro).
Cells were cultured overnight at 37
O
C, 3.5% CO2 to allow for differential plating. The next morning,
non-adherent cells were collected and further separated using a 2-4% bovine serum albumin
(BSA) (Equitech-Bio, Inc., Kerville Texas, USA) gradient. Spermatogonia cells were identified by
morphology and size compared to Sertoli/myoid cells. Fractions containing spermatogonia were
selected, pooled and centrifuged for a final cell purity of at least 85%. The cells were either treated
and/or directly frozen for mRNA analysis. Experiments were performed using a minimum of three
spermatogonia preparations.

2.3.4. Spermatogonia Treatments
Spermatogonia were plated in 6, 24, or 96 well Corning
TM
culture-treated microplates at 10,000-
1,000,000 cells/well for western blots, qPCR, and viability assays. The cells were grown overnight
(80% confluency) in culture media containing regular FBS. Next day, the cells were incubated for
another 24h with filtered vehicle control (0.2% DMSO + ETOH) (VC), GEN, MEHP or GEN+MEHP
mixtures at 10 μM and 100 μM concentrations. Treatments were diluted in cell culture growth
media containing either 10% heat-inactivated Regular FBS (Reg-FBS) or 10% charcoal-stripped
FBS (CS-FBS) (Sigma-Aldrich, St. Louis, MO, USA). All treatments were filtered with 0.2 µm filters
prior to addition to cells to prevent bacterial contamination. Studies measuring the expression of
undifferentiated and differentiated spermatogonial gene markers in C18-4 cells in response to the
differentiation factor RA (1 μM) followed the same protocol, while primary rat spermatogonia were
treated with vehicle or EDCs diluted in culture medium supplemented with Reg-FBS.

52

2.3.5. Gene Array Analysis
Rat Illumina microarray analysis was performed by McGill University’s Genome Quebec facility.
Total RNA was extracted from spermatogonial cells isolated from 10 PND8 rat pups per cell
preparation using the PicoPure RNA isolation kit (Arcturus, Mountain View, CA, USA) and
digested with DNase I (Qiagen, Valencia, CA, USA) as previously described [202]. For reverse-
transcriptase (RT)-PCR analysis, cDNA was synthesized from the extracted RNA by using the
single strand cDNA transcription synthesis kit (Roche Diagnostics). The samples were analyzed
using the RatRef-12 Expression BeadChip for genome-wide expression analysis (Illumina, San
Diego, CA, USA), containing 22523 probes selected primarily from the NCBI RefSeq database.
Data normalization was performed by Dr. Jaroslav Novak, using quantile normalization corrected
for background signal [204].

2.3.6. Immunofluorescence (IF)
C18-4 cells were plated at 50,000 cells in 8 chamber tissue culture slides (Corning Inc., NY, USA)
and allowed to reach 80% confluency overnight. The next day, the cells were fixed with cold 4%
paraformaldehyde in PBS (Santa Cruz Biotechnology, TX, USA) for 10 minutes at room
temperature and washed with PBS. The cells were then incubated with 1% Triton X-100
(Promega, Madison, WI, USA) in PBS for 10 minutes at room temperature, washed, and
incubated in 5% donkey serum (Sigma-Aldrich, St. Louis, MO, USA) in PBS for another 30
minutes. Cells were incubated with primary antibodies for Cox1 (Cat #4841 Cell Signaling
Technology, Davers, MA, USA), Cox2 (Ab15191 Abcam, Cambridge, MA, USA), PGE2 (bs-
2639R, Bioss Antibodies, Woburn, MA, USA) or PGF2a (bs-2341R, Bioss Antibodies, Woburn,
MA, USA) at 1:300 diluted in 5% donkey serum overnight at 4c. Next day, the wells were washed
and corresponding Alexafluor
®
anti-mouse/ anti-rabbit secondary antibodies (Thermo Fisher,
Eugene, OR, USA) were added at 1:400 in 5% donkey serum for 30 minutes at room temperature.
The cells were washed one last time, and Vectashield
®
DAPI mounting medium (Vector

53

Laboratories, Burlingame, CA, USA) was added to the slides. Images were taken using the
Revolve-G130 fluorescent microscope after adjusting for background signal.

2.3.7. MTT Viability Assay
C18-4 cells were plated in 96 well Corning
TM
culture-treated microplates at 10,000 cells/well for
the MTT viability assay following the recommendations of the manufacturer (Roche Cell
proliferation kit I MTT, Sigma-Aldrich, St. Louis, MO, USA) as previously described (Boisvert et
al. 2016). After 24h treatment, 10 ul MTT labeling reagent was added to each well and incubated
for an additional 4 hours at 35 °C. Then 100 ul solubilization solution was added to dissolve purple
formazan crystals. The conversion of MTT (dimethylthiazol-diphenyltetrazolium bromide) to
formazan was measured with the VICTOR
™
X5 Multilabel Plate Reader (PerkinElmer, Inc.,
Waltham, MA, USA). Data was expressed as fold change relative to VC, calculated from three
independent experiments conducted in triplicates.

2.3.8. RNA extraction, cDNA Synthesis and Quantitative real time PCR (qPCR) analysis
Isolated PND8 spermatogonia and C18-4 cells were plated at 100-200,000 cells/well and treated
as previously described. After treatments, cells were stored in RNA lysis buffer, and samples were
extracted using the RNAqueous™-Micro Total RNA Isolation Kit (Invitrogen, Carlsbad, CA, USA).
The cDNA was synthesized from extracted RNA using PrimeScript™ RT Master Mix (Takara Bio,
Mountain View, CA, USA) according to manufacturer’s instructions. QPCR was performed using
a LightCycler 480 with a SYBR Green PCR Master Mix kit (Roche Diagnostics). Primers were
designed using Primer-BLAST from the NCBI-NIH gene database. Primers are listed in Table 2.
QPCR cycling conditions were as follows: initial step at 95°C followed by 40 cycles at 95°C for 15
sec, 60°C for 1 minute. This was followed by both melting curves and cooling cycles. Direct
detection of PCR products was monitored by measuring the increase in fluorescence caused by
the binding of SYBR Green dye to double-stranded DNA, and the comparative threshold cycle

54

(Ct) method was used to analyze the data. 18S rRNA was used for data normalization for isolated
spermatogonia. GAPDH was used for data normalization for C18-4 cells. Assays were performed
in triplicates. All experiments were performed using a minimum of three independent sample
preparations and the mean ± SEM are shown.
Table 2. List of qPCR Primer sets.
Mouse

Forward Primer Reverse Primer
Cbr1 CCGGTCTTGAACCACTCTCC TTTCCGACACAGGTCACGAG
Cox1 CCTCTTTCCAGGAGCTCACA TCGATGTCACCGTACAGCTC
Cox2 CAGGACTCTGCTCACGAAGG ATCCAGTCCGGGTACAGTCA
Foxo1 CTTCAAGGATAAGGGCGACA GACAGATTGTGGCGAATTGA
Gapdh AAGGTCATCCCAGAGCTGAA CTGCTTCACCACCTTCTTGA
Gfra-1 GCGTGTGAAGCACTGAAGTC GGTTCAGTTCCGACCCAAC
Id4 CAGGGTGACAGCATTCTCTG CCGGTGGCTTGTTTCTCTTA
Kit AGCAAATGTCACAACAACCT CCTCGTATTCAACAACCAAA
Mcam CAAACTGGTGTGCGTCTTCTT CTTTTCCTCTCCTGGCACAC
Pla2g12
a
GGGCAGGAACAGGACCAGACCA
CCG
GGTTTATATCCATAGCGTGGAACAGGC
TTCG
Pla2g4a GCTTAAGGCAGGAGCTAACCT GAGTGTCCAGCATATCGCCA
Pla2g6 GGATTGGTGGAGAGCCGTAG GAGGACATGCCTCAGAGTCG
Ptgds GGCTCCTGGACACTACACCT CTGGGTTCTGCTGTAGAGGGT
Ptges GGCTCCTCCAAAGACGGAAA TGGCACAGCATGGGTCTTAG
Ptgis ATGCAGTGTCAAAAACCGCC TGGGACCCATATTCCCCTGT
Stra8 GCCTCAAAGTGGCAGGTACTG CTTATCCAGGCTTTCTTCCTGTTC
Thy1 GGAGTCCAGAATCCAAGTCGG TATTCTCATGGCGGCAGTCC
Rat
18s
CGGGTGCTCTTAGCTGAGTGTCC
CG
CTCGGGCCTGCTTTGAACAC
Foxo1 TAGGAGTTAGTGAGCAGGCAAC TGCTGCCAAGTCTGACGAAA

2.3.9. LC-MS analysis of Prostaglandins PGE2 and PGF2a
C18-4 cells were plated at 1,000,000 cells/well in 6-well Corning
TM
culture-treated microplates
overnight and treated with EDCs as described above. CS-FBS culture medium (DMEM medium
supplemented with 10% CS-FBS and 1% Penicillin-Streptomycin) without cells and conditioned

55

medium (1 mL) from control and EDC-treated cells were collected and stored in -80c for LC-MS
analysis. The LC-MS protocol was performed as previously described [77]. In short, solutions of
PGE2 (P5640) and PGF2a tris salt (P0424) (Sigma-Aldrich) were used to make calibration curves.
Aliquots (200 uL) of CS-FBS culture medium without cells and conditioned medium from treated
C18-4 cells were extracted using solid-phase-extraction (SPE) by dilution with 2% formic acid in
water (200 uL) and addition to pre-conditioned Phenomenex Strata-X 10 mg cartridges. SPE
eluents were reconstituted in 50 uL 25:75 (v/v) acetonitrile: methanol (0.1% acetic acid total)
solution. Calibration curves were prepared into the matrix (DMEM medium supplemented with
10% CS-FBS and 1% Penicillin-Streptomycin) for an analytical range from 50-5000 pg/ml.
Extracts were analyzed using a Shimadzu Nexera ultra high-performance liquid chromatograph
(HPLC) coupled with a Sciex TripleTOF quadrupole time of flight mass spectrometer (MS). PGs
were detected using electrospray negative mode ionization followed by MS/MS fragmentation.
Sciex Analyst
®
v1.7 software was used for data acquisition. Sciex MultiQuant
TM
v3.02 software
was used to select peak area measurements from selected product ions and to perform calibration
curve regression analysis and sample quantifications. PG concentrations were normalized to total
protein/well quantified with using BCA quantification as described below. The data represent the
means ± SEM of three independent experiments, each conducted in duplicates.

2.3.10. Prostaglandin D2 ELISA assays
Cells were plated at 200,000 cells/well in 24-well Corning
TM
culture-treated microplates overnight
and treated with EDCs. CS-FBS culture medium (DMEM medium supplemented with 10% CS-
FBS and 1% Penicillin-Streptomycin) or Reg-FBS culture medium (DMEM medium supplemented
with 10% Reg-FBS and 1% Penicillin-Streptomycin) without cells and conditioned media (500 μL)
from control and treated C18-4 cells were collected and stored in -80c for ELISA analysis. PGD2
levels were measured using the Prostaglandin D2 ELISA Kit (Cayman Chemical, Ann Arbor, MI,
USA) according to manufacturer’s instructions. This assay is based on competition between

56

PGD2 and a PGD2 tracer (PGD2-acetylcholinesterase conjugate) introduced into wells coated
with goat polyclonal anti-mouse IgG. ELISA standards detecting PGD2 at a range from 19.5 pg/ml
to 2,500 pg/ml were prepared in CS-FBS culture medium (DMEM medium supplemented with
10% CS-FBS and 1% Penicillin-Streptomycin) or culture medium containing regular FBS (DMEM
medium supplemented with 10% FBS and 1% Penicillin-Streptomycin) without cells. VICTOR
™

X5 Multilabel Plate Reader (PerkinElmer, Inc., Waltham, MA, USA) was used to detect OD
measurement of Ellman’s reagent at wavelength 405 nm. Data analysis was evaluated using the
Cayman PGD2 computer spreadsheet, and %B/B0 from standards S1-S8 versus PGD2
concentration was plotted in Prism version 7.0 (GraphPad Software, San Diego, CA) software
using a 4-parameter logistic fit. Assays were performed in triplicate. PGD2 concentrations were
normalized to total protein/well quantified with Bradford reagent according to the manufacturer’s
protocol (VWR Life Science, Solon, OH, USA). Cell layers were solubilized using 0.1 N NaOH
and stored in -80c for total protein quantification. All experiments were performed using a
minimum of three independent experiments, with data normalized to VC and the mean ± SEM are
shown.

2.3.11. Immunoblot Protein Analysis
Cells were plated at 1,000,000 cells/well in 6-well Corning
TM
culture-treated microplates overnight
and treated with EDCs as described above. The cells were scraped from the plates using chilled
PBS (Santa Cruz Biotechnology, TX, USA), centrifuged, collected and cell pellets were stored in
-80c for analysis. Proteins were extracted in 30 μL from cell pellets using RIPA lysis buffer (Santa
Cruz Biotechnology, TX, USA) containing Pierce
TM
protease and phosphatase inhibitors (Thermo
Scientific, Rockford, IL, USA). Contents were vortexed to lyse cells and protein extracts were
collected after centrifugation for 30 minutes at 4c. Protein was quantified using the Pierce BCA
Protein Assay Kit following manufacturer’s instructions. Protein levels were calculated against a
BSA standard curve diluted in RIPA lysis buffer in a working range of 20-2,000 ug/mL BSA. 16

57

ug/sample of protein were mixed at 1:1 ratio with 2x Laemmli Sample Buffer + beta
mercaptoethanol (Bio-Rad Laboratories, Hercules, CA, USA), and boiled for 10 minutes at 95c to
allow for protein denaturation. Samples were frozen at -20c until western blotting. Samples were
loaded onto a 4-20% Mini-Protean TGX precast gel (Bio-Rad Laboratories, Hercules, CA, USA)
at 18 uL per sample. Dual color protein standard (Bio-Rad Laboratories, Hercules, CA, USA) was
loaded at 10 uL. Gels ran at 60V for 10 minutes, then increased to 100V for approximately one
hour with running buffer (10x Tris/glycine/SDS) (Bio-Rad Laboratories, Hercules, CA, USA). Then
the samples were transferred onto nitrocellulose membranes or PVDF membranes activated with
methanol at 4c for 2 hours at 80V with transfer buffer (Efficient Western 20X Transfer Buffer +
20% methanol) (G-Biosciences, St. Louis, MO, USA). After transfer, the membranes were stained
for total protein following manufacturer’s instructions using Ponceau S solution (Sigma-Aldrich,
St. Louis, MO, USA) or TotalStain Q Staining kit (Azure Biosystems, Dublin, CA, USA).
Membranes were visualized on the Azure c600 imaging system and total protein was quantified
using the ImageJ software. Membrane stain was rinsed off with deionized water and membranes
were blocked with 5% BSA in PBST for 1 hour at room temperature. Then the membranes were
incubated with 1:1000 primary Cox1 antibody Cat #4841 (Cell Signaling Technology, MA, USA)
in 5% BSA in PBST at 4c overnight. In some experiments, alpha tubulin band was visualized (anti-
alpha tubulin Cat # T9026, Sigma Aldrich) and used for normalization. Membranes were washed
three times with PBST for 10 minutes each and incubated with WesternSure
®
Anti-Rabbit IgG
secondary antibody (Li-COR biosciences, Lincoln, NE, USA) at 1:3000 for 1 hour at room
temperature. Membranes were washed another three times with PBST and visualized using a
Clarity
TM
Western ECL Substrate Solution (Bio-Rad Laboratories, Hercules, CA, USA) on Azure
c600 imaging system. Band density was quantified using the ImageJ software and normalized
against total protein to determine relative protein quantity. The data represent the means ± SEM
of three independent experiments, each conducted in duplicates normalized against VC.

58

2.3.12. Statistics
Statistical analysis was conducted by Student’s t-test or one-way ANOVA with post-hoc multiple
comparison analysis using Prism version 7.0 (GraphPad Software, San Diego, CA). Statistical
significance was considered at p ≤ 0.05. Results are shown as the mean ± SEM. Values of * p ≤
0.05, ** p ≤ 0.01 and *** p ≤ 0.001: Significant difference of treatments compared to control
samples. Values of # (p≤0.05), ## (p<0.01), ### (p<0.001) indicate significant difference between
an individual treatment and the EDC mixture at the same concentration.

2.4. RESULTS
2.4.1. Characterization of the eicosanoid pathway in mouse SSC model C18-4 cell line and
in primary rat spermatogonia
To elucidate the role of Coxs and PGs in spermatogonial development, we used an in vitro
mouse cell line model that encompasses the cellular properties of undifferentiated spermatogonia,
including SSCs. The C18-4 cell line, established by the Hofmann lab, is an immortalized mouse
spermatogonia cell line generated from PND6 juvenile spermatogonia, which expresses markers
of SSCs and undifferentiated germ cells, such as GNDF family receptor alpha-1 (GFRa1) and
Ret, and proliferates in response to 100 ng/ml glial cell line-derived neurotrophic factor (GDNF)
(Hofmann, 2008). We found that cyclooxygenases Cox1 and Cox2 transcripts and proteins are
highly expressed in the C18-4 cell line (Fig. 2A, B), as well as other intermediates of the
eicosanoid biosynthetic pathway including phospholipase A2 (Pla2), carbonyl reductase 1 (Cbr1),
an enzyme capable of converting prostaglandin E2 to prostaglandin F2α, prostaglandin E
synthase (Ptges), prostaglandin I2 (prostacyclin) synthase (Ptgis), and prostaglandin D2 synthase
(Ptgds). qPCR analysis showed that the expression of Ptgis, Cox1, Pla2 and Cox2 enzymes were
amongst the most highly expressed genes, with Ptgds, Cbr1 and Ptges exhibiting lower
expression levels in C18-4 cells. Similarly, in PND8 rat spermatogonia, Cox2 and Pla2 were highly
expressed in addition to downstream enzymes in the pathway, including Ptgds, Ptgis, Ptges, and

59

Cbr1 (Fig 1C), while Cox1, which was not represented in the rat gene array platform, was found
to be expressed in the same range but at slightly lower levels than Cox2 when measured by qPCR
(data not shown).
Immunocytochemical staining using prostaglandin-specific antibodies demonstrated the
presence of PGs -E2 (PGE2) and -F2a (PGF2a) in C18-4 cells (Fig. 4). The identity and levels of
PGE2 and PGF2α were further confirmed using LC-MS, and detectable levels of PGD2 were
measured using a commercial ELISA kit. Baseline levels of PGs in the C18-4 cell line are 50.23
± 30.93 pg/ml of PGD2, 419.65 ± 94.59 pg/ml of PGE2, and 222.59 ± 37.92 pg/ml of PGF2α. The
presence of Cox enzymes and PG synthases, as well as detectable levels of PGs indicate that
the eicosanoid pathway is biologically active in this SSC model.

Figure 3. Characterization of eicosanoid pathway in rodent SSC models. (A, B) Expression of eicosanoid-related genes
and proteins in C18-4 cells. (A) Relative expression of eicosanoid related transcripts expressed in untreated C18-4
cells, measured by qPCR. Results are expressed as mean ± SEM of 2
ΔCT
values normalized to Gapdh x 1000 and
were obtained from 2-3 independent experiments conducted in triplicates. (B) Representative protein expression of
Cox1 and Cox2. Scale in µm. (C) Expression of eicosanoid-related genes in PND8 rat primary spermatogonia. The
expression profile of Cox2 and related genes in PND8 rat spermatogonia was analyzed using Illumina gene arrays,
measuring transcript levels in 3 independent cell preparations, each isolated from the testes of 10 rats. Results are
expressed as mean ± SD. Abbreviations: Cox (cyclooxygenase), Cbr1 (carbonyl reductase 1), Pla2g12a

60

(phospholipase A2 group 12a), Pla2g6 (phospholipase A2 group 6), Ptgds (prostaglandin D synthase), Ptges
(prostaglandin E synthase), Ptgis (prostaglandin I synthase).

Figure 4. C18-4 spermatogonia produce prostaglandins PGE2 and PGF2ɑ. PGs were visualized by ICC in untreated
C18-4 cell line. A non-specific IgG was used for negative control. Blue: DAPI nuclear signal. Scale in µm.
Representative images are shown.

Next, we examined whether exposing C18-4 spermatogonia for 24h to 1 µM RA, a known
inducer of spermatogonial differentiation, would affect Coxs gene expression, in parallel to
measuring the gene expression of a panel of spermatogenic gene markers. First, the data given
in relative unit for the Cox genes confirmed that Cox1 expression is 10-fold higher than Cox2
expression in C18-4 cells. Moreover, RA treatments induced a 30% significant decrease in the
transcript level of Cox1, whereas it had no effect on Cox2 mRNA (Fig. 5A). The transcripts of the
SSC markers Thy1, Id4 and Gfra1 were decreased by 50-60%, while the undifferentiated
spermatogonial marker Foxa1 was decreased by 33%, in RA-treated compared to control cells
(Fig. 5B). However, the transcript levels of Melanoma cell adhesion molecule (Mcam), a gene
identified in adult SSC and in undifferentiated progenitor spermatogonia, were increased by 40%
in RA-treated cells, suggesting that it acted as a progenitor rather than SSC marker in these cells
(Fig. 5B). The expression of the differentiated spermatogonia markers Kit and Stra8 was
significantly increased by RA treatment (Fig. 5C). The only exception was Sohlh1, a differentiated
spermatogonia and spermatocyte marker that was downregulated by RA treatment (Fig. 5C).

61

Nonetheless, the overall gene expression changes observed in markers of SSC, undifferentiated
progenitors and differentiated spermatogonia in RA-treated cells indicated a progression of the
cells toward differentiation, validating the use of C18-4 cell line as an adequate in vitro model for
studying eicosanoids in SSC development. Furthermore, the downregulation of Cox1 gene by RA
suggested that the decrease of this cyclooxygenase might be part of SSC differentiation process.

Figure 5. Effect of retinoic acid treatment on the expression spermatogonial gene markers and Cox genes in C18-4
cell line. Cells were treated with 1 µM all-trans retinoic acid (RA) for 24h. (A) Results for Cox1 and 2 expression are
presented in relative gene expression (2^dt). Gene expression levels of (B) SSC/undifferentiated spermatogonia
markers, and (C) differentiated spermatogonia markers, presented as fold change of vehicle control (VC). mRNA
expression data were normalized to Gapdh. Results are shown as mean ± SEM from 2-3 independent experiments
conducted in triplicates. Significant differences relative to VC determined with unpaired t-test: * (p≤0.05), ** (p<0.01),
*** (p<0.001).

62

2.4.2. GEN, MEHP, and GEN+MEHP treatments alters differently the expression of Cox1
and Cox2 eicosanoid pathway enzymes
We then examined whether GEN, MEHP and their mixtures altered the eicosanoid pathway
in C18-4 cells by evaluating their impact on the expression of major eicosanoid pathway enzymes
after 24h of treatment. Because our goal was to measure prostaglandins in the conditioned media
from the cells and regular FBS contains measurable levels of prostaglandins that would increase
their baseline levels, we performed parallel experiments using regular FBS (Reg-FBS) and
charcoal-stripped FBS (CS-FBS), after verifying that CS-FBS contained lower levels of
prostaglandins than Reg-FBS (data not shown). Overall, gene expression data obtained in
experiments using the two types of FBS were similar, except for Ptges and Ptgds which presented
different expression levels in control samples depending of the type of FBS used (see below).
Thus, gene expression was measured in parallel for genes of the eicosanoid pathway and data
presented for both serum conditions. GEN and MEHP were used either at the environmentally
relevant concentration of 10 μM (abbreviated as G5, M5, GM5) or a 10-fold higher concentration
of 100 μM (G4, M4, GM4) to evaluate their effects, alone or mixed.
The overall viability of the cells was minimally affected by the EDC treatments, showing 6
to 10% changes in medium containing Reg-FBS (Fig. 6A), and small but significant increases by
20 % and 35% above vehicle control (VC) levels in cells treated with 10 μM GEN+MEHP mixture
and 100 μM MEHP in CS-FBS supplemented medium respectively, suggesting a small increase
in cell proliferation (Fig. 6B). qPCR analysis showed that Pla2 mRNA levels were altered by GEN
and GEN+MEHP at 100 μM, with increasing trends in Reg-FBS medium and significant 2-fold
increases in CS-FBS medium (Fig. 6C, D). However, treatments with GEN+MEHP mixtures and
GEN alone at 10 μM and 100 μM resulted in dose-dependent decreases of Cox1 gene expression
by 30 to 70% compared to controls, as well as 100 μM MEHP, whereas 10 μM MEHP had no
effect, with similar patterns in both FBS conditions (Fig. 4E, F). The protein levels of Cox1 were
measured by immunoblot analysis and confirmed that GEN+MEHP mixtures at 10 and 100 μM

63

reduced the protein levels of Cox1, as well as GEN alone at 100 μM, while 100 μM MEHP
decreased Cox1 protein in CS-FBS medium but not Reg-FBS medium (Fig. 6G, H). In contrast,
dose-dependent increases of 45 to 90% over control levels were observed in the mRNA
expression of Cox2 gene in cells treated with 100 μM GEN and GEN+MEHP in both FBS
conditions (Fig. 6I, J). These results indicate that GEN and GEN+MEHP treatments at a
concentration of 10 μM corresponding to levels measured in human blood alter Cox1 expression,
but not Cox2 levels. At the protein level, GEN+MEHP mixture of 10 μM induced Cox1 decreases
not found with individual compounds at the same concentration, suggesting synergistic effects.
The changes observed with GEN+MEHP mixtures might be the result of unopposed GEN-driven
effects, while MEHP alone at 100 μM decreased Cox1 mRNA and protein but did not affect Cox2
or Pla2 expression. Therefore, Cox1 may exhibit a greater sensitivity to GEN and MEHP
treatments than Cox2.

2.4.3. Treatments with GEN, MEHP and GEN+MEHP mixture disrupt prostaglandin
synthesis
To further explore EDC-induced changes in prostaglandin synthesis, we next quantified the
levels of PG synthases and PGs in C18-4 cells exposed to the same treatments. Secreted PGD2
levels were measured by ELISA using both FBS types. PGE2 and PGF2a secreted levels were
determined using LC-MS in the condition media of cells grown using CS-FBS conditions, to
prevent/reduce the addition of exogenous PGs from FBS. The intracellular levels of PGE2 and
PGF2a in cells grown with Reg-FBS were determined by IF after removing the cell supernatants.

64

Figure 6. Effects of GEN and MEHP, alone or mixed, on cell viability and on the expression of upstream eicosanoid
pathway enzymes. Effects of 24h treatment with GEN, MEHP, or GEN+MEHP mixtures at either 10 μM (10
-5
M
abbreviated as 5) or 100 μM (10
-4
M abbreviated as 4) in media containing either regular FBS (Reg-FBS) (A, C, E, G,
I) or charcoal-stripped FBS (CS-FBS) (B, D, F, H, J). (A, B) Cell viability measured with MTT assays. mRNA expression
of Pla2 (C, D); Cox1 (E,F); Cox2 (I, J); and protein expression of Cox1 (G, H). mRNA expression data were normalized
to Gapdh, and protein loading was normalized to total protein content /lane. (G) Representative immunoblots. Results
are shown as mean ± SEM from 3-4 independent experiments conducted in triplicates. Significant difference relative
to VC with One-way ANOVA test and multiple comparisons using Fisher’s LSD Test: * (p≤0.05), ** (p<0.01), ***
(p<0.001), #: Significant difference of individual treatment to dose-corresponding mixture.

65

2.4.3.1. Effects on Cbr1 expression and PGF2a synthesis. The relative expression levels of
Cbr1, gene coding for an enzyme metabolizing PGE2 to PGF2a, was increased significantly by
100 μM GEN and GEN+MEHP, in a manner similar to the effects on Cox2 and Pla2, with 2-fold
increases in both Reg-FBS and CS-FBS conditions (Fig. 7A-B). Small increases of 30 and 20 %
in Cbr1 levels were also observed in cells treated with 10 μM GEN and GEN+MEHP mixture in
CS-FBS, respectively, but not in medium with Reg-FBS. However, there was no difference in
Cbr1 levels in control samples between the two types of FBS (Fig. 7C), supporting the similar
effects found with GEN and MEHP independent of the serum used.
In agreement with the increases found in Cbr1 gene expression with 100 μM GEN and
GEN+MEHP, LC-MS analysis of PGF2a showed 1.6-fold and 1.4-fold significant increases in
secreted PGF2a with 100 μM GEN and GEN+MEHP treatments respectively, compared to VC,
whereas other treatments had no effect on PGF2a (Fig. 7D). The increase in PGF2a induced by
100 μM GEN and GEN-MEHP was also visible as strong immunofluorescent signal in the nucleus
and in the cytoplasm of cells cultured with Reg-FBS, indicating that part of PGF2a was retained
inside the cells (Fig. 7E). Thus, these treatments upregulated both secreted and intracellular
PGF2a levels independent of serum type (Fig. 7D and E). PGF2a signal was also observed in the
nuclei and perinuclear Golgi apparatus of control cells and other treatments (Fig. 7E). The
changes in PGF2a production and Cbr1 transcript levels induced by 100 μM GEN and the EDC
mixture suggest a good correlation between Cbr1 transcript levels and its enzyme activity in these
cells.

66

Figure 7. Effects of GEN and MEHP, alone or mixed, on Cbr1 gene expression and on Prostaglandin F2a Production.
Effects of 24h treatments of C18-4 cells with GEN, MEHP, or GEN+MEHP at 10 μM (5) or 100 μM (4) in media
containing Reg-FBS (A, C, E) or CS-FBS (B, C, D). mRNA levels of PG synthase Cbr1 were measured by qPCR. (A,
B) EDC-induced changes in Cbr1. (C) Effect of FBS type on basal Cbr1 mRNA levels expressed in relative units. (D)
Secreted PGF2a levels were measured by LC-MS. (E) Representative pictures of PGF2a immunofluorescence in cells
grown in reg-FBS. NC: negative control. Open arrowhead: PGF2a nuclear signal; Plain arrow: PGF2a signal in Golgi
apparatus; Dotted arrow: PGF2a cytoplasmic signal. Scale in µm. qPCR (A, B) and LC-MS results (D) are presented
as fold change over vehicle control (VC). mRNA expression data were normalized to Gapdh. Results are shown as
mean ± SEM from 3 independent experiments conducted in triplicates. Significant difference with One-way ANOVA
test and multiple comparisons using Fisher’s LSD Test: * (p≤0.05), ** (p<0.01), *** (p<0.001), #: Significant difference
of individual treatment to dose-corresponding mixture.

67

2.4.3.2. Effects on Ptges expression and PGE2 synthesis. The mRNA levels of Ptges were
significantly downregulated by GEN and GEN+MEHP at 10 μM and all treatments at 100 μM in
Reg-FBS treatment conditions (Fig. 8A), and similar significant decreases in Ptges were induced
by 10 μM GEN+MEHP and 100 μM MEHP in cells incubated with CS-FBS, with decreasing trends
in response to 10 μM GEN (Fig. 8B). However, Ptges mRNA levels in cells treated with 100 μM
GEN and GEN+MEHP in CS-FBS condition were increased by 20- to 30-fold (Fig. 8B), contrasting
with the decreases found with Reg-FBS. To understand the possible origin of this difference, we
compared the mRNA levels of Ptges in control cells incubated either with Reg-FBS or CS-FBS,
and found that Ptges expression was significantly decreased in control cells incubated with 10%
CS-FBS compared to Reg-FBS (Fig. 8C). This indicates that growing C18-4 cells in CS-FBS
rather than Reg-FBS alters the basal levels of Ptges in these cells. Since levels were expressed
as percent of control levels, it is not surprising that the decreases observed for most treatments
remained similar independently of the serum used. However, the large Ptges increases observed
with 100 μM GEN and GEN+MEHP in cell grown with CS-FBS, which occurred despite
downregulated Ptges basal levels, suggest that Reg-FBS contains a molecule(s) removed by
charcoal treatment, such as a prostaglandin or steroid, which prevented the upregulation of Ptges
by high dose GEN and GEN+MEHP treatments. Further studies will be required to identify the
molecule(s) mediating these effects on Ptges transcript levels.
LC-MS analysis of PGE2 secretion in the conditioned medium of cells cultured with CS-FBS
showed decrease PGE2 levels with GEN and GEN+MEHP at 10 μM and 100 μM MEHP that were
consistent with the decreased observed in Ptges gene expression (Fig 8D). While 100 μM
GEN+MEHP did not increase PGE2 despite increasing Ptges mRNA levels, treatment with 100
μM GEN induced a significant increase of secreted PGE2, in agreement with the increase in Ptges
gene expression, but to a lesser extend (Fig 8D). Thus, changes in PGE2 levels agreed with those
observed with Ptges expression for most treatments in CS-FBS conditions, except for 100 μM

68

GEN+MEHP where secreted PGE2 was lower than expected. This discrepancy, as well as the
difference in the amplitude of the Ptges and PGE2 responses for 100 μM GEN and GEN+MEHP
could be due in part to the retention of PGE2 in the cells. This is supported by the finding of a
strong increase in PGE2 IF signal inside cells treated with 100 μM GEN and GEN+MEHP, but not
for other treatments where PGE2 signal was less or similar to that of control cells, in Reg-FBS
medium (Fig. 8E). These data suggest that the secreted and intracellular fractions of PGE2 are
similarly altered by most treatments, except for 100 μM GEN and GEN+MEHP treatments which
induce a large increase of PGE2 inside the cells that is not secreted.

69

Figure 8. Effects of GEN and MEHP, alone or mixed, on Ptges gene expression and on Prostaglandin E2 Production.
Effects of 24h treatments of C18-4 cells with GEN, MEHP, or GEN+MEHP at 10 μM (5) or 100 μM (4) in media
containing Reg-FBS (A, C, E) or CS-FBS (B, C, D). mRNA levels of the PG synthase Ptges were measured by qPCR.
(A, B) EDC-induced changes in Ptges. (C) Effect of FBS type on basal Ptges mRNA levels expressed in relative units.
(D) Secreted PGE2 levels were measured by LC-MS. (E) Representative pictures of PGE2 immunofluorescence in
cells grown in reg-FBS. NC: negative control. Open arrowhead: examples of PGE2 nuclear signal. Scale in µm. qPCR
(A, B) and LC-MS results (D) are presented as fold change over vehicle control (VC). mRNA expression data were
normalized to Gapdh. Results are shown as mean ± SEM from 3 independent experiments conducted in triplicates.
Significant difference with One-way ANOVA test and multiple comparisons using Fisher’s LSD Test: * (p≤0.05), **
(p<0.01), *** (p<0.001), #: Significant difference of individual treatment to dose-corresponding mixture.

2.4.3.3. Effects on Ptgds expression and PGD2 synthesis. While 10 and 100 μM GEN and
100 μM GEN+MEHP upregulated Ptgds expression in Reg-FBS conditions between 1.5-fold to
more than 3-fold (Fig 9A), GEN and GEN+MEHP lead to the decrease of Ptgds mRNA by 20 to
30 % of control cells in CS-FBS conditions with 10 μM GEN+MEHP and all treatments at 100 μM
(Fig 9B). The difference in Ptgds expression changes between the two serum conditions could be
due to the fact that the cells grown in CS-FBS expressed 5 times more Ptgds than the cells grown
in Reg-FBS (Fig 9C), which could mask or prevent further increases of the transcript. Indeed,
PGD2 secretion was increased by GEN, MEHP and their mixture at 100 μM in both serum
conditions. The upregulation of Ptgds mRNA levels in Reg FBS conditions were in agreement
with the upregulation of PGD2 production, which was increased by 4-fold in cells treated with
MEHP and GEN+MEHP at 100 μM, and a 2-fold increasing trend with 100 μM GEN and 10 μM
MEHP in Reg-FBS conditions (Fig 79). However, in CS-FBS conditions, PGD2 production was
increased by 56-fold in cells treated with 100 μM GEN+MEHP and 22-fold by 100 μM MEHP,
while showing a 3-fold increasing trend both with 10 μM GEN+MEHP and 100 μM GEN, and a 2-
fold increase with 10 μM MEHP (Fig 9E). This further suggests robust effects of the EDCs on the
enzyme activities independent of their mRNA levels in cells grown with 10% CS-FBS. Moreover,
there were clear synergistic effects of 100 μM GEN and MEHP, their mixture inducing much higher
levels of PGD2 than the individual compounds. Overall, EDCs induced dose-dependent increases
of PGD2 production independently of serum conditions, but the responses to 100 μM EDCs were
much stronger in medium containing CS-FBS than regular FBS. Since charcoal treatment
decreases the serum contents of prostaglandins and steroids, the difference in the amplitude of

70

PGD2 increases suggests that Reg-FBS contains a substance that antagonizes PGD2 synthesis.
The fact that MEHP significantly increased PGD2 levels but not the other PGs suggests that
MEHP has a unique ability to target PGD2 synthesis but not that of PGE2 or PGF2a.

Figure 9. Effects of GEN and MEHP, alone or mixed, on Prostaglandin D2 production. Effects of 24h treatments of
C18-4 cells with GEN, MEHP, or GEN+MEHP at 10 μM (5) or 100 μM (4) cultured in media containing Reg-FBS (A, C,
D) or CS-FBS (B, C, E). (A, B) EDC-induced changes in Ptgds mRNA expression. (C) Effect of FBS type on basal
Ptgds mRNA levels expressed in relative units. (D, E) Secreted PGD2 levels measured by ELISA. qPCR and ELISA
results are presented as fold change over vehicle control (VC). mRNA expression data were normalized to Gapdh.
Results are shown as mean ± SEM from 3 independent experiments conducted in triplicates. Significant difference
relative to VC with One-way ANOVA test and multiple comparisons using Fisher’s LSD Test: * (p≤0.05), ** (p<0.01),
*** (p<0.001), #: Significant difference of individual treatment to dose-corresponding mixture.

71

2.4.3.4. GEN and GEN+MEHP alter SSC differentiation markers concomitant to changes in
prostaglandins
To further elucidate whether the changes in prostaglandin synthesis induced by the two EDCs
might be associated with changes in SSC development, we evaluated the impact of GEN, MEHP,
and their combination on genes expressed in progenitor and differentiated spermatogonia, as
indication of SSC differentiation. Treatments with 100 μM GEN and GEN+MEHP upregulated the
expression of undifferentiated progenitor spermatogonial markers Mcam (Fig. 10A, B) and
forkhead box O1 (Foxo1) independently of the type of FBS used (Fig. 10C, D), although the
inductions were greater in medium containing CS-FBS than Reg-FBS. Foxo1 mRNA expression
was also upregulated in primary spermatogonia treated with 100 μM GEN+MEHP, while 100 μM
MEHP significantly reduced Foxo1 expression (Fig. 10E). These data suggest the induction of a
switch towards the progression to progenitor spermatogonia by 100 μM GEN and GEN+MEHP.
The commitment to differentiation was further supported by the upregulation of the
spermatogonial differentiation markers Kit and stimulated by retinoic acid (Stra8). Stra8
expression was increased by 40-fold in cells treated with 100 μM GEN and GEN+MEHP after 24h
of treatment in CS-FBS conditions (Fig 10F). Kit expression was upregulated by 100 μM GEN
and GEN+MEHP, and to a greater extent by 100 μM MEHP in CS-FBS conditions (Fig 10G).
These changes were concomitant with the GEN-driven induction of Pla2 and Cox2 mRNA
expression, the expression of several isoforms of PG synthases, and cellular PG secretion (Figs
6-9). The minimal effects of MEHP treatments on the eicosanoid pathway, but its unique ability to
upregulate Kit expression at 100 μM suggest a separate and unique mode of action of MEHP.

72

Figure 10. Effects of GEN and MEHP, alone or mixed, on undifferentiated and differentiated spermatogonial gene
markers. Effects of 24h treatments of C18-4 cells or PND8 primary spermatogonia with GEN, MEHP, or GEN+MEHP
at 10 μM (5) or 100 μM (4) in media containing Reg-FBS (A, C, D) or CS-FBS (B, C, E). (A-D) EDC-induced changes
in the mRNA expression of the undifferentiated spermatogonial markers Mcam (A, B) and Foxo1 in C18-4 cells (C, D)
and in primary spermatogonia (E). (F-G) EDC-induced changes in the mRNA expression of differentiation markers
Stra8 (A) and Kit (B) in cells grown in CS-FBS. Results are presented as fold change of vehicle control (VC). Results
are presented as fold change of vehicle control (VC). C18-4 mRNA expression data were normalized to Gapdh. Results
are shown as mean ± SEM from 3 independent experiments conducted in triplicates.

Primary spermatogonia mRNA
expression data were normalized to 18S ribosomal RNA, and the results are shown as mean ± SEM from 3 independent
experiments with 10 rats used per cell isolation, n=5-6 per condition. Significant difference relative to VC with One-way
ANOVA test and multiple comparisons using Fisher’s LSD Test: * (p≤0.05), ** (p<0.01), *** (p<0.001), #: Significant
difference of individual treatment to dose-corresponding mixture.

73

2.5. DISCUSSION
Overall, these data support the remodeling of the spermatogonial transcriptome and
disruption of SSC differentiation processes by GEN and MEHP, alone or in combination, in
association to their ability to regulate PGs levels in SSCs. Because most of the effects with
GEN+MEHP were similar to those of GEN alone, we postulate that GEN drove the induction of
differentiation processes in C18-4 cells, and that this may be due to its ability to induce PG release
via upregulation of several genes in the Cox2 pathway (Fig. 11). Furthermore, the fact that
switching to CS-FBS, which contains lower levels of PGs that Reg-FBS, altered the basal
expression of two PG synthases suggests that serum provides nutrient(s) essential to sustain a
functional SSC and hints at a role of PG regulation in SSC development, which will be further
elucidated in future studies.

Figure 11. Diagram summarizing the effects of GEN and MEHP, alone or mixed, on the eicosanoid biosynthetic
pathway and on SSC differentiation. GEN (G); MEHP (M); GEN+MEHP mixture (GM); 10 μM (5); 100 μM (4). ↑
upregulation; ↓ downregulation. As shown in the diagram, the lower concentration of 10 μM EDCs exerted mainly
inhibitory effects on Cox1 and stimulatory on Cbr1 expression, whereas 10 μM concentration had multiple effects,
inducing the expression of several enzymes of the eicosanoid pathway and increased production of PGs, aligned with
similar effects on SSC differentiation markers, suggesting a possible role of PGs in the regulation of SSC differentiation
that is disrupted by GEN and MEHP.

74

2.5.1. SSCs express major eicosanoid pathway enzymes and produce PGs
To the best of our knowledge, this is the first study reporting that SSCs express major
eicosanoid pathway enzymes and produce PGs. Several studies have shown that PGs can play
a role in regulating fetal and neonatal testicular development in rats and humans [77, 195, 205].
In the present study, we found that SSCs, as modeled by the C18-4 cell line, exhibit RNA and
protein expression of Cox enzymes, as well as upstream regulator Pla2, responsible for
arachidonic acid formation from membrane phospholipids. We also identified the presence of
major PG synthases and measured secreted PG-D2, -E2, and -F2a and detected intracellular
pools of PGE2 and PGF2a. We chose the C18-4 cell line because it is a widely used in vitro model
of SSCs which has been characterized in multiple studies [206-208]. We further validated that the
C18-4 cell line express SSC markers and can be induced to differentiate by 1 μM RA, as exhibited
by decreases in classical SSC and undifferentiated spermatogonial markers (Thy1, Id4, and
Gfra1) and an increase in the differentiating spermatogonial marker Stra8. We also observed an
increasing trend towards in Kit expression, in line with cell differentiation. However, the transcript
levels of Sohlh1, another differentiating spermatogonia marker, was decreased by RA, suggesting
that the C18-4 cell line does not fully recapitulate the differentiation process of primary SSCs,
possibly due to cell line immortalization, which is a limitation of all cell line models [200]. Our
finding that Cox1 and Cox2 mRNA expression were downregulated by RA suggests that both
enzymes, and therefore PG production, are actively regulated during SSC differentiation. Besides
their typical role in inflammation and tumor formation, several studies have described the inhibition
of Coxs during cell differentiation, such as in epidermal cell and preadipocyte differentiation,
similarly as our findings in SSCs, as well as roles in cell fate [209-212]. Overall, our data showing
that the expression profile of eicosanoid pathway genes in the mouse C18-4 cell line are similar
to that of primary isolated PND8 rat spermatogonia and the finding that Coxs are altered during
RA-induced cell differentiation suggests that this model is adequate for use in our studies.

75

2.5.2. GEN and MEHP exert differential effects on the eicosanoid pathway in SSCs
Whereas Cox1 is constitutively expressed in all cells, the expression of Cox2 is cell-type
specific and its expression in SSCs has led us to question its role in these cells, and whether
EDCs can target this pathway as a mechanism behind some of their adverse effects on
spermatogenesis. Despite evidence of the involvement of eicosanoids in regulating male germ
cell development, and EDCs being known as disruptors of reproductive developmental processes,
no study has presently evaluated whether EDCs can modulate the eicosanoid pathway in relation
to the development of male germ cells. Our finding that Cox1 and 2 gene expression is altered by
RA, together with our previous findings that in utero exposure to the EDCs GEN and DEHP altered
the expression of spermatogonial genes and inflammatory markers, led us to postulate that the
synthesis of PGs by SSCs may play exclusive roles in regulating SSC own self-renewal and
differentiation, and that these processes may be disrupted by GEN and MEHP, the main
metabolite of DEHP. Indeed, we found that GEN and MEHP, alone and in mixtures, added at 10
μM and 100 μM for 24h, exhibited unique ways of altering the expression of eicosanoid pathway
enzymes and PG release in C18-4 cells, suggesting multiple modes of action of the individual
EDCs and their mixture in targeting this pathway and altering its downstream developmental
effects.
Our studies were first conducted in regular FBS conditions, and then replicated in conditions
in which PGs were removed from serum through a charcoal-based lipid stripping process [213].
Our data indicate that the effects of GEN and MEHP on upstream enzymes such as Cox enzymes
and Pla2 were mostly independent of the type of FBS used, but they also showed that switching
from Reg-FBS to CS-FBS impacted the basal and EDC-induced levels of Ptges and Ptgds
expression, while having no effect on Cbr1. While the effects on Ptges and Ptgds may be
attributed to the depletion of PGs in the treatment media, they may also be due to the reduction
by charcoal treatment of other hormones such as estrogens and growth factors inherent to FBS,
which may also interact downstream of the Cox pathway, as it has been established in the field

76

of female reproduction [214-216]. This suggests that PGs or other small molecules present in
Reg-FBS may play a role in the eicosanoid pathway homeostasis in SSCs, as well as altering the
effects of GEN and MEHP on individual PG synthases and PG production. Although using CS-
FBS is required in studies such as ours where the goal was to measure PGs, and in studies
examining the effects of steroid hormones, the reduction or removal of small molecules and
proteins by charcoal-stripping has been reported to alter specific genes or biological responses
in cell types such as endocrine cancer cells [217, 218]. Thus, one needs to pay close attention to
the constitution of the media used in vitro, especially when studying the effects of EDCs.
Regardless of the FBS type used, GEN alone or combined with MEHP at 100 μM upregulated the
major eicosanoid pathway genes Pla2 and Cox2 with similar amplitudes, indicating that these
effects were mainly driven by GEN. In contrast, Cox1 gene and protein expression were
downregulated in a consistent manner by these treatments as well as by 10 μM treatments.
Interestingly, at 10 μM, a concentration found in human blood, only the mixture significantly
decreased Cox1 protein levels, whereas at 100 μM, individual EDCs and mixtures reduced Cox1
protein. EDC-induced Cox2 overexpression suggests a compensatory feedback mechanism,
perhaps as a result of Cox1 downregulation. These data indicate that Cox1 is more sensitive to
dysregulation by GEN and MEHP than Cox2, and highlight the distinct impacts that these EDCs
have on Cox1 and Cox2, stressing the need of understanding the respective role of each enzyme,
to fully comprehend the consequences of exposing SSCs to these EDCs.
The comparative study made with the two types of FBS confirmed that a common effect
of 100 μM GEN was to upregulate the secretion of PGF2a, PGE2 and PGD2. Moreover, the
visualization of PGF2a and PGE2 in fixed cells indicated that PGF2a and PGE2 were also
increased inside the cells by the same treatment. While 100 μM GEN-MEHP significantly
increased PGF2a and PGD2 secretion, it only showed an increasing trend for PGE2. However,
the finding that PGE2 immunofluorescence signal was stronger in cells treated with 100 μM GEN-
MEHP suggests that a higher fraction of PGE2 might have accumulated inside the cells under

77

this condition. Moreover, the high retention of PGE2 inside cells treated with 100 μM GEN and
GEN-MEHP may explain the discrepancy between the large amplitude of Ptges increases and
the modest increases in PGE2 secretion in these conditions. Interestingly, PGE2 and PGF2a had
distinct subcellular localizations even in basal conditions, with PGE2 localized mainly in the
nucleus, whereas PGF2a was found in the nucleus, cytoplasm, and Golgi apparatus. These
findings suggest that secreted and intracellular fractions of PGs could have different roles in
SSCs, which will be interesting to examine in future studies.

2.5.3. GEN emerges as modulator of SSC differentiation concurrently to increasing PGs
Overall, changes in PG synthases and PGs induced by 100 μM GEN and GEN+MEHP
showed similar patterns as the changes observed in Foxo1 and Mcam genes, two markers of
committed spermatogonia progenitors, and in the spermatogonial differentiation marker Stra8.
These changes suggest that GEN and MEHP, particularly at the higher concentration, may induce
premature differentiation of SSCs. Although 100 μM is higher than levels of a single compound
measured in humans, reports of synergistic effects between various phthalates and between
estrogenic EDCs suggest that similar effects could be induced by mixtures of EDCs in which each
compound would be at low concentration [177, 219]. The correlation between GEN-driven
upregulation of the Cox2 pathway genes and increased spermatogonial differentiation markers
implies that the disruption to these processes by GEN is likely mediated by a greater eicosanoid
production via the induction of Cox2 activity. However, the impact of other molecular pathways
cannot be ruled out, as GEN has been shown to alter various signaling pathways involved in the
regulation of SSC development, which could interfere with the Cox pathway, as reported in other
cell or tissues types [196, 220].
The concurrence of GEN-driven upregulation of the expression of Cox2 pathway enzymes
and PGs and SSC progression toward more differentiated stages support the hypothesis that
these processes might be interrelated in SSCs, suggesting a novel function of the eicosanoid

78

pathway and prostaglandins in the regulation of SSC development. Moreover, GEN appears to
play a primary role in modulating Cox2 pathway expression in SSCs, which may be unrelated to
its anti-inflammatory role in other cell types. One study found GEN can decrease the expression
of Cox2 in both a human prostate cancer cell line and primary prostate epithelial cells, resulting
in decrease in PGE2 secretion, which is often elevated in cancers [196]. Similar results were
found in other cancer models as well [221]. However, other studies reported stimulatory effects
of phytoestrogens, including GEN, on PGE2 and PGF2a synthesis in non-cancer models such as
the bovine endometrium, similarly to the effects of GEN observed in our SSC model [222]. In
previous in vivo and in vitro studies, we found that GEN transiently disrupted signaling pathways
in neonatal testis and male germ cells [177, 183]. The results from the present study are also
consistent with our previous findings that maternal exposure to GEN and DEHP mixtures led to
increases in expression of inflammatory cell markers in testes of PND120 offspring [181, 187].
However, the dichotomy of GEN effects was highlighted by its ability to exert protective effects in
function of the developmental age examined, such as in our studies comparing the effects of in
utero exposure to GEN, DEHP and their mixtures on the gene expression and function of neonatal
and juvenile rat testes, demonstrating that GEN exerted anti-oxidant protective effects against
DEHP-induced oxidative stress in neonatal testis, but not in juvenile testis [182]. Moreover, high
expression of Cox2 was found to be linked to incidences of human testicular cancers [223],
suggesting that this pathway may play a crucial role in male reproduction and may be uniquely
susceptible to EDCs.

2.5.4. Distinctive effects of MEHP alone on PGD2 synthesis and SSC differentiation
Overall, treatments with MEHP-alone had little effects on changes in the expression of
genes of the eicosanoid pathway, but still upregulated PGD2 levels, induced the expression of
the differentiation marker Kit, and enhanced cell viability. In addition, 100 μM MEHP
downregulated Cox1 mRNA in regular FBS treatment conditions and decreased Ptgds mRNA to

79

the same level as that of 100 μM GEN in CS-FBS conditions. This suggest that these significant
effects of MEHP may be due to effects of this compound on cellular processes that indirectly
affect Cox pathway enzyme activities, particularly targeting PGD2 synthesis, as well as Kit
expression. Several studies support a role of MEHP in altering Cox2 expression and disrupting
PG secretion [197, 224]. In the present study on the C18-4 SSC cell line, 100 μM MEHP did not
alter Cox2 expression but it did significantly stimulate cell viability. These differences further
support the idea that the eicosanoid pathway and PGs exert specific and unique effects on
different phase of germ cell development, from fetal and early postnatal germ cells to
spermatogenic cells. Although these data were obtained with doses of MEHP higher than those
measured in human fluids or tissues and significantly higher than the levels detected in some
infants in the range of 10 μM, infants could be exposed to levels of phthalate mixtures other than
DEHP/MEHP reaching 100 μM or more, acting on the same receptors and signaling pathways,
due to simultaneous exposures to phthalates from multiple sources [219].

2.5.5. Synergistic effects of GEN and MEHP on PGD2 synthesis in SSCs
To date, the effects of EDC mixtures are largely underexplored, particularly in developing
infants who have risks of being exposed to these compounds via their environment and maternal
exposure. During neonatal life, the endocrine system regulates the development of a wide range
of biological systems, and the effects of EDC mixtures can involve complex interactions resulting
in synergistic and potentially more severe effects than individual compounds. We observed such
effect with 100 μM GEN+MEHP, which induced PGD2 production to a greater extent than GEN
or MEHP alone. EDC mixtures can induce synergistic or additive effects due to the ability to target
multiple molecular pathways, leading to greater reproductive toxicological effects from exposure
to mixtures. This was shown for phytoestrogens and anti-androgenic EDCs, including in our own
studies [174, 177, 187]. Similarly, lifelong combination treatment of low-dose GEN and vinclozolin,
an anti-androgenic fungicide, were reported to result in anomalies of the male reproductive tract

80

and fertility, including decreased sperm counts and reduced sperm motility [225]. Likewise, in the
present study, the induction of PGD2 secretion by GEN and MEHP used separately was relatively
low, but as a mixture, they were able to induce PGD2 secretion to a much greater extent. As we
did not observe higher expression levels of eicosanoid pathway enzymes with MEHP treatment
alone, and this synergistic effect was unique to PGD2, this suggests that MEHP potentiated the
upregulation of eicosanoid pathway enzyme activities by GEN by indirectly altering Cox activity
via another mode of action. High PGD2 levels have been shown to be negatively associated with
the differentiation of fetal male germ cells [195]. In the present study on spermatogonia, however,
higher PG levels were associated with greater differentiation of male germ cells. Considering that
the early phases of germ cell differentiation are known to be differentially regulated, and are
characteristically different, it is not surprising that they would be differently regulated by PGs [163,
199]. In fact, the trends observed from our published study on the inhibitory effect of analgesics
on both PG secretion and neonatal gonocyte differentiation is comparable with what is observed
here on spermatogonia, implying that the induction of PGs is positively correlated with both
neonatal gonocyte and SSC differentiation. Moreover, the finding that Cox1 and 2 were
differentially affected by RA and by GEN and MEHP implies that distinct roles of Cox1 and Cox2
in SSCs that warrant further investigation. Future studies will aim to investigate the direct effects
of PGs on SSC development.
In summary, we report for the first time that SSCs express major eicosanoid pathway
enzymes and produce PGs, and that the EDCs GEN and MEHP, alone and in combination,
uniquely disrupt the eicosanoid pathway and alter the expression of SSC differentiation markers.
This is substantial because it suggests that Cox2 may be important for SSC development, and
that PGs may be involved in regulating the differentiation processes of these cells. In this context,
the ability of two common EDCs GEN and MEHP to disrupt these processes requires greater
attention, especially concerning the exposure of male infants. Indeed, infancy is a critical period
for the establishment of the reservoir of germline stem cells required for male reproduction. Fetal

81

and infancy are well-established periods of sensitivity to EDCs, during which tightly regulated
cellular processes are required for proper organ development [226]. Given how environmental
levels of multiple EDCs can easily accumulate to reach the high levels, in part due to the
immaturity of metabolic tissues at these early ages, it is important to understand how EDCs can
impact molecular pathways that regulate neonatal germ cell development in baby boys. Improper
SSC regulation via disruption of self-renewal and differentiation processes can increase the risk
of infertility and testicular cancer, which could contribute to the rise of these diseases in the
western world.

2.6. SUMMARY
In the current chapter, we’ve reported that SSCs exhibit a functional eicosanoid system and we
were able to characterize this pathway in an in vitro model. Our results hinted at an ability of the
eicosanoid pathway to modulate effects of EDCs on spermatogonial differentiation, which would
suggest that this pathway could play a crucial role in regulating the development of germ cells in
support of our working hypothesis. In the next study, we will continue our exploration using
genomic approaches in the C18-4 cell line to further understand this pathway’s involvement with
SSC growth and development.

2.7. ACKNOWLEDGEMENTS
We acknowledge Renita Moradian and Haoyi Cui for their contributions as co-authors in the
current study. We thank Dr. Marie-Claude Hofmann for the generous gift of the C18-4 mouse
spermatogonial cell line. We are also grateful for the help from Dr. Gurpreet Manku who generated
the gene array data on primary rat spermatogonia during her PhD studies in Dr. Culty’s lab. These
gene data were used to examine the transcript levels of eicosanoid pathway-related genes, for
comparison with C18-4 cells.

82

CHAPTER 3: DIFFERENTIAL ROLES OF CYCLOOXYGENASE ENZYMES IN THE
REGULATION OF MURINE SPERMATOGONIAL STEM CELLS
1

OBJECTIVE (AIM 2): To elucidate a mechanism of action behind Cox inhibition on C18-4
spermatogonial stem cells using a combination of treatments with pharmacological Cox inhibitors
and gene silencing methods.

3.1 ABSTRACT
Acetaminophen (ace) and ibuprofen (ibu) are widely administered to babies due to their
presumed safety as over-the-counter drugs. Despite their overwhelmingly prevalent use, no
reports exist on the effects of cyclooxygenase (Cox) inhibitors on spermatogonial stem cells
(SSCs) required for sperm formation. Infancy represents a critical period for SSC formation and
disrupting SSCs or their precursors may be associated with infertility and testicular germ cell
tumor formation.
The current study examined the molecular and functional impact of Cox inhibition in SSCs
using pharmacological approaches and gene silencing in the C18-4 SSC model. Ace, but not ibu,
dose-dependently decreased retinoic acid-induced expression of the differentiation gene Stra8,
while the Cox2 selective inhibitor NS398 decreased the expression of Kit, another spermatogonia
differentiation marker. shRNA-based Cox1 silencing in C18-4 cells exhibited altered cellular
morphology and upregulation of Stra8 and Kit. Furthermore, RNA-Seq analysis of Cox1
knockdown cells indicated the activation of several signaling pathways including the TGFb, Wnt,
and Notch pathways, compared to control C18-4 cells. Notch pathway genes were uniquely
upregulated with 24-hour treatments of selective Cox1 and Cox2 pharmacological inhibitors
NS398, celecoxib, and FR122047 as well as with acetaminophen and ibuprofen. We report that
the Cox1 and 2 can regulate Notch3 expression, and the Notch pathway is a target of PGD2.
Overexpression of Notch3 was observed in human testicular cancer, suggesting that the

83

interaction between the eicosanoid and Notch signaling pathways may be critical for the proper
maintenance of spermatogenesis.

1
This chapter is derived from a manuscript in preparation entitled “Differential roles of cyclooxygenase enzymes in the
regulation of murine spermatogonial stem cells” by Tran-Guzman, et al.

84

3.2 INTRODUCTION
Acetaminophen (ace) and ibuprofen (ibu) are widely administered to babies due to their
presumed safety as over-the-counter drugs [227, 228]. Despite their prevalent use, studies have
suggested that treating infants with cyclooxygenase (Cox) inhibitors could jeopardize their long-
term reproductive functions [120, 229, 230]. However, no reports exist on the effects of Cox
inhibitors on spermatogonial stem cells (SSCs) required for sperm formation. Infancy represents
a critical period for SSC formation and disrupting SSCs or their precursors may be associated
with infertility and testicular germ cell tumor formation.
Testicular cancer and infertility are part of a series of disorders believed to have a common
developmental origin and known collectively as testicular dysgenesis syndrome (TDS). TDS can
develop during fetal life or within a few months after birth, during which a pool of pluripotent cells,
known as gonocytes, differentiate to form a reservoir of spermatogonial stem cells to ensure life-
long production of sperm. Molecular signals dictate the balance between differentiation and self-
renewal during the different stages of pre-meiotic spermatogenesis, ensuring that there is a
constant pool of spermatogonial stem cells that can be induced to differentiate to mature sperm
throughout a male’s life [17]. Steps of differentiation and meiosis in spermatocytes and spermatids
comprise the remainder of spermatogenesis. Improper germ cell development can result in
carcinoma in situ, the pathological precursor of testicular cancer which has origins in primordial
germ cells or gonocytes that have failed to differentiate into spermatogonium. In addition to
classical seminomas, another class of seminomas can arise from more differentiated postnatal
germ cells such as the SSCs, known as spermatocytic seminomas (SS). While SS can initially
form as a benign tumor, it can result in life-threatening disease if allowed to progress to more
serious sarcomas [231].
Eicosanoids are fatty-acid based signaling molecules that have effects ranging from the
regulation of physiological systems such as pain perception and cell growth, to specific roles in
females during pregnancy and childbirth [232, 233]. They are formed the through release of

85

esterified arachidonic acid (AA) from membrane phospholipids by an enzyme called
phospholipase A2 in response to inflammatory stimuli. Upon release, AA gets further oxidized by
various enzymes, epoxygenases, lipoxygenases, and cyclooxygenases. Two cyclooxygenases
exist in the cell, Cox1, which is constitutively expressed and Cox2, whose expression is inducible
and tissue specific. Binding of AA on the cyclooxygenase active site of the Cox enzymes is
responsible for its cyclization, forming prostaglandin (PG) H2. Due to the unstable nature of
PGH2, it is quickly converted to lipid-based mediators prostaglandins, prostacyclins, and
thromboxanes [234]. Several isoforms of prostaglandin synthases, which have varying expression
from tissue to tissue, isomerize PGH2 to various prostaglandins that exert both autocrine and
paracrine effects to nearby membrane receptors.
We previously reported that rat gonocytes and spermatogonia express Cox2 and that
gonocytes produce prostaglandins at significant levels [77]. We were also able to characterize in
a mouse spermatogonial stem cell model, the C18-4 cell line, expression of both Cox1 and Cox2
mRNA and protein, as well as other intermediates of the eicosanoid biosynthetic pathway
including phospholipase A2 (Pla2) and measurable levels of PGs -D2, E2, and F2a production
[235]. In this study, we are continuing to interrogate the role of the eicosanoid pathway on male
reproductive development using the C18-4 cell line as a relevant in vitro model, to elucidate the
mechanism of action of Cox inhibitors on spermatogonial development that could underlie
potentially toxic effects. Using several Cox1, Cox2, and non-selective pharmacological Cox
inhibitors, we have identified unique roles played by Cox1 and Cox2 on regulating SSC
differentiation and on downstream molecular signaling pathways. The upregulation of Notch
pathway genes and significant changes in cell morphology were observed in a Cox1 knockdown
model of the C18-4 cell line, suggesting that eicosanoids may regulate cellular maintenance and
differentiation. Interestingly, the NOTCH3 transcript was also upregulated in human seminomas
and embryonic carcinomas biopsies, which could hint at an important role that eicosanoids may

86

play during postnatal reproductive development, which when dysregulated, could contribute to
the cellular transformation of male germ cells.

3.3 MATERIALS AND METHODS
3.3.1 Cell Line and Treatments
The LTAg-immortalized mouse type A spermatogonia cell line was gifted by Marie-Claude
Hofmann (The University of Texas MD Anderson Cancer Center, Houston, Texas, USA). Cells
were cultured in Gibco™ DMEM containing 4.5 g/L d-Glucose, L-Glutamine and 110 mg/L of
sodium pyruvate (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% heat-
inactivated FBS (Sigma-Aldrich, St. Louis, MO, USA) and 1% Penicillin-Streptomycin (P/S
Solution 100X (Corning
™
) at 35°C, 7% CO2. The cells were allowed to grow in 25 cm
2
and 75cm
2
Corning
TM
cell culture-treated flasks and passed every other day.
Pharmacological Cox inhibitors NS398, celecoxib, and FR122047 were purchased from
Sigma Aldrich (St. Louis, MO, USA) and were dissolved in ultra-pure grade dimethyl sulfoxide
(DMSO from VWR International, Radnor, PA, USA) to obtain concentrations 5, 20, and 5 mg/ml
respectively. Treatment solutions were subsequently diluted in cell culture media to obtain 1, 10,
and 100 μM final concentrations. Acetaminophen and ibuprofen were purchased from Sigma
Aldrich (St. Louis, MO, USA). The drugs were dissolved in 100% ethanol to obtain concentrations
of 0.05 g/ml. Treatments were subsequently diluted in cell culture media to obtain 50 and 100 μM
final concentrations. All-trans Retinoic Acid purchased from Sigma Aldrich (St. Louis, MO, USA)
was dissolved in ultra-pure grade dimethyl sulfoxide (DMSO from VWR International, Radnor, PA,
USA) to obtain a final concentration of 1 μM. Supplemental PGD2 (Cat No. 538909) was
purchased from EMD Millipore Pore (Burlington, MA, USA). The compound was dissolved in
100% EtOH to obtain a stock concentration of 1 mg/ml. Treatments were subsequently diluted in

87

cell culture media to obtain 50 and 100 μM final concentrations. Stock solutions of the drugs were
stored at -20°C.

3.3.2. Cox1 shRNA mediated silencing
Cox1 shRNA plasmid for mouse (sc-35097-SH) and control shRNA plasmid (sc-108060) were
purchased from Santa Cruz Biotech (Dallas, TX, USA). Cells were transfected with 0.5 μg plasmid
and introduced into a 1.5 μL Lipofectamine 3000 mix (ThermoFisher Scientific, Waltham, MA)
diluted in Opti-MEM (ThermoFisher Scientific, Waltham, MA). Cells were treated with 4 μg/ml
puromycin for two weeks, and then successful silencing of Cox1 was validated with immunoblot
and qPCR assays before isolating cells into 1 cell/well concentrations into 96 well plates to select
for stably transfected clones. Cells were allowed to amplify for several weeks to obtain enough
cells for further downstream validation and analysis.

3.3.3. LC-MS of Prostaglandins
LC-MS of prostaglandins was conducted according to the previously reported protocol [77, 235].
In short, C18-4 cells were plated at 1,000,000 cells/well in 6-well Corning
TM
culture-treated
microplates overnight in culture medium consisting of DMEM medium supplemented with 10%
charcoal stripped-FBS (CS-FBS) and 1% Penicillin-Streptomycin (CS-culture medium). Cells
were treated with either medium containing DMSO/ethanol (used at the same final percentage as
in samples treated with Cox inhibitors) as vehicle control (VC), or with Cox inhibitors diluted in
CS-culture medium. Aliquots of CS-culture medium without cells were collected and stored at -
80°C to establish base line levels of PGs in LC-MS analysis. The conditioned media from vehicle
control cells and cells treated with Cox inhibitors were collected and stored at -80°C until LC-MS
analysis. Samples were analyzed using a Shimadzu Nexera ultra high-performance liquid
chromatograph (HPLC) coupled with a Sciex TripleTOF quadrupole time of flight mass
spectrometer (MS). PGs were detected using electrospray negative mode ionization followed by

88

MS/MS fragmentation. Sciex Analyst
®
v1.7 software was used for data acquisition. Sciex
MultiQuant
TM
v3.02 software was used to select peak area measurements from selected product
ions and to perform calibration curve regression analysis and sample quantifications. PG
concentrations were normalized to total protein/well and protein was measured by BCA
quantification as described below. Data represent the means ± SEM of two independent
experiments, each conducted in duplicates.

3.3.4. Prostaglandin D2 ELISA
For PGD2 ELISA quantification, C18-4 cells were plated at 400,000 cells/well in 12 well Corning
TM

culture-treated plates and treated with Cox inhibitors diluted in CS-culture medium or
DMSO/ethanol diluted in CS-culture medium (vehicle control). After treatments, cell culture
supernatants were collected and saved for prostaglandin analysis. Samples were stored at -80°C
until use. PGD2 levels were measured using the Prostaglandin D2 ELISA Kit (Cayman Chemical,
Ann Arbor, MI, USA) according to manufacturer’s instructions. ELISA standards detecting PGD2
at a range from 19.5 pg/ml to 2,500 pg/ml were prepared in CS-culture medium without cells. The
VICTOR
™
X5 Multilabel Plate Reader (PerkinElmer, Inc., Waltham, MA, USA) was used to detect
OD measurement of Ellman’s reagent at wavelength 405 nm. Data analysis was evaluated using
the Cayman PGD2 computer spreadsheet, and %B/B0 from standards S1-S8 versus PGD2
concentration was plotted in Prism version 7.0 (GraphPad Software, San Diego, CA) software
using a 4-parameter logistic fit. PGD2 concentrations were normalized to total protein/well
quantified with Bradford reagent according to the manufacturer’s protocol (VWR Life Science,
Solon, OH, USA). Cell layers were solubilized using 0.1 N NaOH and stored at -80°C for total
protein quantification. All experiments were performed using a minimum of three independent
experiments, with data normalized to VC. The mean ± SEM are shown.

89

3.3.5. EDU Proliferation Assay
Cells were plated in 96 well Corning
TM
culture-treated black bottom microplates at 8-10,000
cells/well density and allowed to grow overnight in serum-free DMEM containing 1% P/S. The
next day, cells were treated with CS-culture medium containing ethanol/DMSO (VC) or Cox
inhibitors diluted in CS-culture medium for 24 hours. For the last 6 hours of treatment, cells were
incubated with 10 μM EdU (5-ethynyl-2’-deoxyuridine) as recommended by the manufacturer
(Click-IT® EdU HCS Assay, Invitrogen, Waltham, MA, USA). Cells were washed with PBS and
fixed with 4% paraformaldehyde followed by a 0.1% Triton X-100 permeabilization surfactant. The
Click-IT reaction cocktail was then added to each well and allowed to incubate in the dark for 30
minutes. Cells were washed and 100 μL of 1:2000 HCS NuclearMask was added to stain DNA
for 30 minutes in the dark. The plates were imaged and quantified with the Cytation 5 Cell Imaging
Multi-Mode Reader (Biotek, Winooski, VT, USA).

3.3.6. RNA Extraction, cDNA Synthesis, and Quantitative Real Time PCR (qPCR)
C18-4 cells were plated at 100-200,000 cells/well and treated with CS-culture medium containing
ethanol/DMSO (VC) or with Cox inhibitors diluted in CS-culture medium for 24 hours. Cells were
collected after treatment and stored in RNA lysis buffer and extracted based on manufacturer
recommendations using the RNAqueous
TM
-Micro Total RNA Isolation Kit (Invitrogen, Carlsbad,
CA, USA). cDNA was synthesized from purified RNA using the PrimeScript
TM
RT Master Mix
(Takara Bio, Mountain View, CA, USA) also according to manufacturer’s instructions. qPCR was
performed using a LightCycler 480 with a SYBR Green PCR Master Mix kit (Roche Diagnostics).
Primers were designed using Primer-BLAST from the NCBI-NIH gene database and are listed in
Supplemental Table 1.
qPCR cycling conditions were as follows: initial step at 95°C followed by 40 cycles at 95°C
for 15 sec, 60°C for 1 minute. This was followed by both melting curves and cooling cycles. Direct
detection of PCR products was monitored by measuring the increase in fluorescence caused by

90

the binding of SYBR Green dye to double-stranded DNA, and the comparative threshold cycle
(Ct) method was used to analyze the data. Gapdh was used for data normalization for C18-4 cells.
Assays were performed in triplicates. All experiments were performed using a minimum of three
independent sample preparations and the mean ± SEM are shown.

3.3.7. Western Blotting
Cells were plated at 1,000,000 cells/well in 6-well Corning
TM
culture-treated microplates
overnight and treated as described above. Treated C18-4 cells were scraped from plates using
chilled PBS (Santa Cruz Biotechnology, TX, USA), centrifuged, and collected. Cell pellets were
stored in -80°C for western blot analysis.
Protein was extracted in 30 μL from cell pellets using RIPA lysis buffer (Santa Cruz
Biotechnology, TX, USA) containing Pierce
TM
protease and phosphatase inhibitors (Thermo
Scientific, Rockford, IL, USA). Contents were vortexed to lyse cells and protein extracts were
collected after centrifugation for 30 minutes at 4°C. Protein was quantified using the Pierce BCA
Protein Assay Kit following manufacturer’s instructions. Samples were calculated against the
standard curve diluted in RIPA lysis buffer based off the working range of 20-2,000 μg/mL BSA.
16 μg/sample of protein were mixed at 1:1 ratio with 2x Laemmli Sample Buffer + beta-
mercaptoethanol (Bio-Rad Laboratories, Hercules, CA, USA), and boiled for 10 minutes at 95°C
to allow for protein denaturation. Samples were frozen at -20°C until western blotting. Western
blotting was conducted according to protocol previously reported [235]. Antibodies used are as
follows: Cox1 (Cat No. 4841, Cell Signaling Technologies), Cox2 (Cat No. D5H5, Cell Signaling
Technologies), Notch3 (Cat No. ab23426, Abcam), Hes1 (Cat No. D6P2U, Cell Signaling
Technologies). Protein expression was normalized to alpha tubulin (Cat No. T9026, Sigma
Aldrich) or to total protein (TotalStain Q, Azure Biosystems, Dublin, CA, USA) according to the
manufacturer’s protocol.

91

3.3.8. Immunofluorescence
C18-4 cells were plated at 50,000 cells 96 well Corning
TM
culture-treated black bottom microplates
and were allowed to reach 80% confluency overnight. The next day, they were treated with CS-
culture medium containing ethanol/DMSO (VC) or with Cox inhibitors diluted in CS-culture media
for 24 hours. After treatment, the samples were fixed with cold 4% paraformaldehyde in PBS
(Santa Cruz Biotechnology, TX, USA) for 10 minutes at room temperature and washed with PBS.
The cells were then incubated with 1% Triton X-100 (Promega, Madison, WI, USA) in PBS for 10
minutes at room temperature, washed, and incubated in 5% donkey serum (Sigma-Aldrich, St.
Louis, MO, USA) in PBS for another 30 minutes. Cells were incubated with primary antibodies,
either Cox1 1:300 (Cat #4841 Cell Signaling Technology, Davers, MA, USA), Notch3 1:300 (Cat
No. ab23426, Abcam), or Kit 1:100 (SC168, Santa Cruz Biotech, TX, USA) diluted in 5% donkey
serum overnight at 4°C. On the next day, the wells were washed and corresponding Alexafluor
®

anti-mouse/ anti-rabbit secondary antibodies (Thermo Fisher, Eugene, OR, USA) were added at
1:400 in 5% donkey serum for 30 minutes at room temperature. The cells were washed one last
time, and 100 μL DAPI (D1306, Thermo Fisher, Eugene, OR, USA) at 300 nM was added to the
wells for 5 minutes. The Cytation 5 Cell Imaging Multi-Mode Reader (Biotek, Winooski, VT, USA)
was used to capture images and quantify protein expression levels after adjusting for background
signal.

3.3.9. Total RNA Sequencing
Total RNA of 3 samples representing different passages of Scrambled Controls and C18-4
Cox1-KD1

were sent to the USC Molecular Genomics Core lab for whole transcriptome RNA sequencing.
RNA was extracted from cells using the RNAqueous
TM
-Micro Total RNA Isolation Kit (Invitrogen,
Carlsbad, CA, USA) as previously described and quality control was performed using Agilent
BioAnalyzer 2100. Samples that passed quality checks (RIN > 8) were approved to move forward
for downstream analysis. Of the three scrambled controls, two passed quality control, resulting in

92

a sample size of 2 controls and 3 knockdown samples. cDNA libraries were prepared using
Takara SMARTer Stranded Total RNA-Seq Kit v2 – Pico Input Mammalian (Takara Bio Inc.,
Japan) following the manufacturer’s protocol. Prepared libraries were sequenced on the Illumina
Nextseq500 platform at a read length of 2x75 at 50 million paired reads.

3.3.10. Bioinformatics Analysis
Transcriptomic datasets of each sample were imported into Partek Flow
®
software (Partek Inc.
Chesterfield, MO, USA) as FASTQ files. Pre-alignment QA/QC was conducted, and bases were
trimmed based on a quality score (Phred) of 20. Bases were trimmed on both ends for a minimum
read length of 25 base pairs. The reminder reads were aligned to the mouse genome (mm10-
GENCODE Genes- release M24) using the STAR aligner tool. Post-alignment QA/QC was
conducted to confirm that over 70% of the reads were aligned to the reference genome. Gene
counts were conducted, and a noise reduction filter was applied to exclude features in which the
maximum value is less than equal to 10 (<10 counts/gene excluded). Upper quartile normalization
and gene specific analysis (GSA) were conducted, resulting in a list of 28,123 differentially
expressed genes between Scrambled controls and C18-4
Cox1-KD1
. Applying parameters of false
discovery rate greater than or equal to 0.1 and fold changes of -2 and 2 narrowed down the gene
list to 1,265 genes which were used for gene ontology and pathway analysis. The PANTHER
Classification System was used to evaluate gene ontologies and relationships [236, 237], and the
Qiagen Ingenuity Pathway Analysis Software (IPA; Qiagen, Hilden, Germany) was used to
evaluate functional pathways.

3.3.11. Human Testicular Tissue Collection and Gene Array
Non-tumoral human testicular “normal” and tumor biopsies were collected for gene array analysis
performed as previously reported [238]. In short, RNA was extracted with the Rneasy Protect Mini
Kit (Qiagen, Hilden, Germany) according to manufacturer’s instructions. Samples (“normal”

93

testicular tissue N=3, seminoma N=3, embryonic carcinoma N=2, unknown pathology N=1) were
analyzed using the HuGene-1_0-st-V1 array chip (Affymetrix, Santa Clara, CA, USA). Data were
normalized using a quantile normalization method correcting for background signal and
comparisons between samples were carried out using the Bayesian approach [239, 240]. All
samples were obtained under the supervision of the institutional ethics review boards and
informed consents were obtained from all sample providers.

3.4 RESULTS
3.4.1. Cox inhibitors decrease prostaglandin production in the C18-4 cell line
Confirmatory studies were conducted to validate that prostaglandin production was
inhibited with pharmacological Cox inhibitors as evidence that Cox enzymatic activity was blocked
as expected. Cells were plated overnight and treated 24 hours with various concentrations of
selective and non-selective Cox inhibitors in culture media supplemented with 10% CS-FBS,
which contains significantly lower amounts of PGs than conventional FBS used for cell culture.
NS398 and celecoxib are Cox2 selective inhibitors that target Cox2 at IC50= 3.8 µM and at IC50=
0.04 μM respectively [241, 242]. FR122047 is a selective Cox1 inhibitor (IC50= 28 nM), and
acetaminophen (ace) and ibuprofen (ibu) are non-selective Cox inhibitors (ace: IC50 = 26 μM for
Cox2, 114 μM for Cox1, ibu: IC50 = 12 μM for Cox1, IC50 = 80 μM for Cox2) [64, 243-245]. High
dose of NS398 had an ability to significantly decrease PGE2 levels when measured with LC-MS,
with the 100 μM dose inhibiting PGE2 levels by approximately 50% (Fig 12A). PGD2 levels were
decreased with high dose celecoxib, but only a trend towards a decrease was observed with
NS398 and FR122047 (Fig 12B-C). Validation studies using ELISA to measure PGD2 showed
significant decreases with both doses of NS398 and ibu, as well as a trend towards decrease with
50 μM of ace (Fig 12D-E).

94

Figure 12. Measurements of prostaglandin levels with Cox inhibitor treatment. Prostaglandin levels detected by LC/MS
with 24-hour treatment of (A, B) 10 and 100 uM NS398, (C) 1 and 10 uM celecoxib and FR12204. PGD2 levels
measured by ELISA with 24-hour treatment (D) NS398 and (E) acetaminophen and ibuprofen. LC/MS data are reported
as concentration (pg/mL); N=2 independent experiments conducted in duplicates. ELISA results are reported as fold
change/vehicle control (VC); N=3 independent experiments conducted in triplicates. Significant difference relative to
controls with one-way ANOVA test and multiple comparisons: * (p≤0.05), ** (p<0.01), *** (p<0.001).

PGD2 (ELISA)
VC 10 100
Fold Change/ VC
VC 100 50 100 50
**
*
*
*
PGD2 (ELISA)
D)
Dose (uM)
Acetaminophen Ibuprofen NS398
Dose (uM)
VC 10 100
PGE2 (LC/MS)
Concentra1on (pg/mL)
**
PGD2 (LC/MS)
VC 10 100
A)
NS398
Dose (uM)
NS398
Dose (uM)
B)
Dose (uM)
*
Concentra1on (pg/mL)
PGD2 (LC/MS)
VC 10 1 10 1
Celecoxib FR122047
C)
E)

95

3.4.2. Effects of pharmacological Cox inhibitors on cell proliferation and differentiation
The effects of 24-hour treatment of Cox inhibitors selectively targeted to Cox1, Cox2, or
both Cox enzymes, on proliferation and differentiation were evaluated. 10 μM NS398 slightly
induced cell proliferation, while 100 μM NS398 had no effect (Fig 13A-B). Up to 50 μM celecoxib
did not induce any effects, neither did 1-50 μM FR122047 (Fig 13C-D). 100 μM ace treatment
induced proliferation up to 20%, while 50 μM of ace and both doses of ibu did not significantly
alter cell proliferation (Fig 13E). Despite not altering the proliferation potential of cells, 50 μM
doses of celecoxib and FR122047 significantly reduced total cell counts to almost 50%,
suggesting apoptotic effects (Fig 13F-G). Therefore, our subsequent experiments were limited to
doses of 10-100 μM of NS398, 1- 10 μM of celecoxib and FR122047. We also chose to focus on
doses of 50 and 100 μM of ace/ibu as these corresponded to blood levels of ace and ibu derived
from a clinical trial conducted on feverish children treated with a single dose of ace and ibu [77,
245].
SSC, progenitor spermatogonia, and differentiated spermatogonial markers were used to
evaluate the ability of pharmacological Cox inhibitors to affect cell differentiation. Cells were
treated with medium or 1 μM retinoic acid (RA), used to induce SSC differentiation, alongside
medium, ace or ibu to evaluate the ability of the drugs to alter the differentiation status of SSCs.
24-hour treatment of 1 μM RA was able to reduce classical spermatogonial stem cell marker
Inhibitor Of DNA Binding 4 (Id4) expression (Fig 13H), and induce the expression of differentiated
spermatogonial marker Stimulated by retinoic acid 8 (Stra8) (Fig 13K), as expected. While there
was no effect of ace/ibu on the expression of Id4 or progenitor spermatogonial marker Forkhead
Box O1 (Foxo1) (Fig 13I), ace at 100 μM had an ability to significantly induce the expression of
progenitor marker Melanoma Cell Adhesion Molecule (Mcam) and inhibit the expression of Stra8
in the presence of RA (Fig 13J-K), which was not observed in non-differentiated conditions. We
hypothesized that the ability of ace to be a greater selective inhibitor of Cox2 may explain why
the effects were restricted to only ace, and this was confirmed when Kit expression was evaluated

96

with treatment of Cox2 selective inhibitor NS398. Upon 24-hour treatments, increasing
concentrations of NS398 dose dependently decreased Kit expression at a protein level (Fig 13L-
M), suggesting that Cox2 may uniquely exert pro-differentiation effects.
% of prolifera1ng cells (normalized)
% of prolifera1ng cells (normalized)
VC 10 100
NS398
Dose (uM)
VC 1
50
Celecoxib
Dose (uM)
10
VC 1 50
FR122047
Dose (uM)
10
VC 100 50 100 50
Dose (uM)
Acetaminophen Ibuprofen
***
***
A)
B)
250 uM
10 uM NS398
100 uM ace Nega1ve Control
Vehicle Control
C)
D) E)
GFP: EDU
Blue: DAPI

97

+ 1 uM Re7noic Acid No Re7noic Acid
Add nega(ve
control
Vehicle Control
10 uM NS398 100 uM NS398
NegaIve Control
GFP: Kit
Red: Alpha Tubulin
Blue: DAPI
M)
L)
Avg Cell Count (Normalized)
***
*
*
***
VC 10 50
Celecoxib
Dose (uM)
1 VC 10 50
FR122047
Dose (uM)
1
Avg Cell Count (Normalized)
average 2^dt
100 50 100 50
ACE IBU
100 50 100 50
ACE IBU
+ 1 uM Re7noic Acid No Re7noic Acid
100 50 100 50
ACE IBU
100 50 100 50
ACE IBU
Id4 Foxo1
+ 1 uM Re7noic Acid No Re7noic Acid
100 50 100 50
ACE IBU
100 50 100 50
ACE IBU
Mcam
average 2^dt
+ 1 uM Re7noic Acid No Re7noic Acid
100 50 100 50
ACE IBU
100 50 100 50
ACE IBU
Stra8
**
*
VC 10 100
NS398
Dose (uM)
**
Rela-ve Fluorescence (GFP)
F)
H)
G)
I)
J)
K)
50 uM
Kit

98

3.4.3. Morphological and transcriptomic characterization of C18-4
Cox1-KD1

To further characterize downstream effects of Cox inhibition on SSCs, we wanted to
interrogate the role of Cox1 by stably inhibiting its expression. Cox1 silencing was achieved using
shRNA-based transfection in which plasmids containing several silencer sequences targeted to
Cox1 were integrated into the C18-4 cellular genome. Control samples were treated with plasmids
containing scrambled sequences. Puromycin treatment at 4 μg/ml was used to select cells
positive for the transfection. We screened several candidates using different concentrations of
plasmid DNA and Lipofectamine 3000 and identified the clone knockdown “B” as the most
promising candidate for single cell isolation (Fig 14A). Knockdown B was plated into 96 well plates
in a density of a single cell per well and allowed to amplify. Isolated clones, referred now as C18-
4
Cox1-KD1
,

displayed significantly lower expression of Cox1 at both the protein and transcript level
of approximately 30% the expression of the Scrambled cells (Fig 14B-D).
We observed significant morphological differences in C18-4
Cox1-KD1
compared to the
scrambled controls (Fig 14E). The cells adopted a more fibroblastic or mesenchymal phenotype,
likely due to loss of integrity of adhesion between cells. Jam-1, junctional adhesion molecule 1,
an epithelial adhesion marker was significantly downregulated in C18-4
Cox1-KD1
compared to its
control (Fig 14F). Mmp2, matrix metalloproteinase 2, which encodes protein involved in
extracellular matrix remodeling, was also significantly downregulated. Silencing of Cox1 also
induced the expression of spermatogonial differentiation marker Stra8 and a trend towards an
increase in Kit, which would suggest that Cox1 is important for maintaining cells at an
Figure 13. Evaluation of pharmacological Cox inhibitors on C18-4 proliferation and differentiation. (A) Visualization of
EDU incorporation into C18-4 cells with 24-hour treatment of 10 uM NS398 and 100 uM acetaminophen. Green: EDU,
Blue: DAPI. Scale in µm. Quantification of % of proliferating cells normalized to vehicle control (VC) with treatment of
(B) 10 and 100 uM NS398, (C) 1-50 uM celecoxib, (D) 1-50 uM FR122047, (E) 50 and 100 uM acetaminophen and
ibuprofen. Quantification of average cell number is normalized to VC with treatment of up to 50 uM celecoxib (F) and
(G) FR122047. Effect of 24-hour treatment of acetaminophen and ibuprofen +/- 1 uM retinoic acid on mRNA expression
of (H) Id4, (I) Foxo1, (J) Mcam, (K) Stra8. (L, M) Visualization and quantification of Kit expression with 24-hour treatment
of 10 and 100 uM NS398. Green: Kit, Red: alpha tubulin, Blue: DAPI. Scale in µm. mRNA expression data were
normalized to Gapdh; N=3 independent experiments conducted in triplicates. Significant difference relative to controls
with one-way ANOVA test and multiple comparisons: * (p≤0.05), ** (p<0.01), *** (p<0.001).

99

undifferentiated state. Induction of cellular differentiation may also explain the observed changes
in the morphology as well.

Figure 14. Generation of based Cox1 knockdown cellular model C18-4
Cox1-KD1
.

(A) Expression of Cox1 on populations
of transfected cells with varying concentrations of shRNA and lipofectamine 3000 showing that knockdown “B” had
the most downregulated expression of Cox1. (B) Cox1 expression of isolated clonal population C18-4
Cox1-KD1
compared
to wildtype cells (WT), scrambled cells (SCR), and knockdown “B” (KD B). (C) Gene expression of Cox1 in C18-4
Cox1-
KD1
and Scrambled controls, (D) Immunofluorescence staining of Cox1 in C18-4
Cox1-KD1
and Scrambled controls. Green:
Cox1, red: alpha tubulin, blue: DAPI. Scale in µm. (E) Brightfield visualization illustrating morphological differences
between C18-4
Cox1-KD1
and Scrambled controls. (F) Changes in gene expression between C18-4
Cox1-KD1
and Scrambled
controls of Jam-1, Mmp2, Stra8, and Kit. Results are presented as fold change of controls. C18-4
Cox1-KD1
Cox1 protein
expression validated against total protein. mRNA expression data were normalized to Gapdh; N=3 independent
experiments conducted in triplicates. Significant difference relative to controls with t-test: * (p≤0.05), ** (p<0.01), ***
(p<0.001).

100

3.4.4. Ontology and pathway analysis of differentially expressed genes in C18-4
Cox1-KD1

To evaluate the signaling pathways altered by silencing Cox1, we next characterized C18-
4
Cox1-KD1
using total RNAseq (RNAseq). The integrity of RNA used for RNAseq of C18-4
Cox1-KD1

was assessed by agarose gel electrophoresis. Only samples with RIN above 8 were selected to
move forward with RNAseq, resulting in 2 scrambled controls (SCR-1, SCR-3) and 3 knockdown
samples (KD1, 2, 3). FASTQ files were imported into Partek Flow and the two control samples
averaged approximately 12,422,840 total reads whereas the knockdown samples averaged
approximately 13,353,310 total reads. Number of reads removed in the controls were ~1410
(0.011%) and in the knockdown ~2070 (0.015%). Cleaned reads were aligned to the mouse
genome assembly, and counts were normalized. This resulted in 436 down- and 829 up-regulated
genes in C18-4
Cox1-KD1
illustrated by the volcano plot (Fig 15A). Overall, we identified 1,265
differentially expressed genes (DEG) using a false discovery rate of <0.1 and fold change
parameters of -2 to 2. Hierarchical clustering analysis of the three individual C18-4
Cox1-KD1
samples
and two scrambled samples show high similarity in the DEGs between the controls and
knockdown samples (Fig 15B).

101

102

Figure 15. RNAseq characterization of C18-4
Cox1-KD1
. (A) Volcano plot and (B) hierarchical clustering of 1,265
differentially expressed genes (DEGs) between C18-4
Cox1-KD1
and Scrambled controls evaluated by total RNA sequencing
and analyzed using Partek Flow (FDR <0.1, Fold change -2 to 2), (C-D) Gene ontology analysis of DEGs analyzed using
PANTHER GO, (E, F) Major pathways (inhibition of MMP, Regulation of EMT, and Prostanoid Biosynthesis) and classical
molecular signaling pathways (TGFb, Notch, and Wnt) were predicted to be altered by Cox1 silencing analyzed using
Ingenuity Pathway Analysis. (G) qPCR validation of TGFb and Notch pathway genes: Tgfbr3, Tgfb2, Tgfb3, Smad3,
Smad9, Notch1, Notch3. mRNA expression data were normalized to Gapdh; N=3 independent experiments conducted
in triplicates. Significant difference relative to controls with t-test: * (p≤0.05), ** (p<0.01), *** (p<0.001).

103

The PANTHER classification system was employed for functional and gene ontology
annotation of differentially expressed genes based on the gene list of 1,265 genes compiled by
Partek Flow. Top biological processes identified by PANTHER included cellular processes (557
genes), biological regulation (376 genes), metabolic processes (347 genes), response to stimulus
(221 genes), and signaling (176 genes) (Fig 15C). As for the molecular function analysis, the
greatest percentage of genes were annotated with binding (350 genes), followed by catalytic
activity (242 genes), and molecular function regulation (163 genes) (Fig 15D). 995 genes were
annotated as pathway related genes, with 26 genes associated with the Wnt signaling pathway,
12 genes associated with the TGFb signaling pathway, and 4 genes associated with the Notch
signaling pathway.
To further interrogate the signaling pathways that were altered in our data set, we next
imported the same gene list of 1,265 genes into Ingenuity Pathway Analysis (IPA) to identify
significantly altered signaling pathways. Amongst other top canonical pathways altered by Cox1
knockdown included regulation of epithelial-mesenchymal transition pathway (192 genes),
inhibition of matrix metalloproteases (39 genes), and inhibition of prostanoid synthesis (10 genes)
(Fig 15E). Wnt, TGFb, and Notch signaling pathways that were identified as significantly
differentially altered by PANTHER classification were also identified as top pathways altered via
IPA. Interestingly, all three pathways were predicted to be involved in the regulation of epithelial
to mesenchymal transition (EMT), one of the top canonical pathways activated in our data set
(Fig 15F).
Top DEGs associated with EMT include several Wnt family members Wnt11, Wnt5a,
Wnt7a, and Wnt9a, TGFb genes Tgfb2, Tgfb3, and Smad3, and Notch receptors Notch1 and
Notch3 (Supp. Table 2). Next, qPCR was performed to verify RNAseq expression values. We
chose to verify the genes that were most relevant to our model and that were amongst the most
significantly altered. Amongst the genes that were identified with the major signaling pathways
that were predicted to be altered, TGFb and Notch signaling, Tgfbr3 was significantly increased

104

by approximately 2-fold as determined by RNAseq, but only trends towards increases in Tgfb2
and Tgfb3 mRNA expression were observed (Fig 15G). Smad3 and Smad9, genes that are also
associated with TGFb signaling, were significantly increased when measured with qPCR
consistent with RNAseq expression values. Notch1 and Notch3 transcripts were significantly
upregulated as well. Overall, qPCR data were consistent with what was being measured using
RNAseq.

3.4.5. Activation of Notch3 signaling pathway in C18-4
Cox1-KD1

Immunofluorescence analysis showed increased expression of Notch3 in C18-4
Cox1-KD1
compared to the scrambled controls (Fig 16A), confirming the induction of this signaling pathway
when Cox1 expression is reduced. To further investigate the effect of Cox1 knockdown on Notch
pathway expression, we evaluated protein levels of Notch3 and its downstream target,
transcription factor Hes1 using western blotting (Fig 16B). Compared to the scrambled controls,
Notch3 receptor expression was upregulated in C18-4
Cox1-KD1
by approximately 50% (Fig 16C).
Hes1 was significantly increased in C18-4
Cox1-KD1
compared to the scrambled controls as well. The
alteration of Cox2 expression was consistent with what was observed using RNAseq and qPCR
(Supp Table 3).

105

Figure 16. Notch3 pathway is upregulated in C18-4
Cox1-KD1
. (A) Immunofluorescence image of Notch3
expression in C18-4
Cox1-KD1
and Scrambled controls. Green: Notch3, red: alpha tubulin, blue: DAPI. Scale in
µm. (B) Western blots of Notch3, Hes1, Cox1 and Cox2 expression in C18-4
Cox1-KD1
and Scrambled controls
are cropped at indicated bands and protein expressions are normalized to alpha tubulin. (C) Quantification
of Notch3, Hes1, Cox1, and Cox2 protein expression normalized against total protein. Results are presented
as fold change of controls. N=3 independent experiments. Significant difference relative to controls with t-
test: * (p≤0.05), ** (p<0.01), *** (p<0.001).

106

3.4.6. Differential effects of pharmacological Cox inhibitors on Notch pathway expression
To validate the effects of blocking eicosanoid synthesis on Notch signaling, wildtype C18-
4 cells were treated for 24 hours with pharmacological Cox inhibitors that target Cox enzymes
selectively and non-selectively. Both drugs targeting Cox2 significantly induced the expression of
Notch3- NS398 at 50 and 100 μM, and Celecoxib at 1 and 10 μM (Fig 17A-B). NS398 reduced
Notch1 expression at 50 μM but celecoxib had no effect on the expression of this gene. Celecoxib
induced the expression of Hes1 at both 1 and 10 μM, whereas only trends towards an increase
was observed with NS398 treatment, which only reached significance at the 50 μM (p=0.054).
Interestingly, 10 μM NS398 induced Hey1 expression while the drug concentration had to reach
100 μM before a decrease in the expression of this gene was observed. On the other hand,
celecoxib decreased Hey1 expression at 1 μM, but the effect was not significant at 10 μM.
FR122047 was used in the same conditions as the Cox2 selective inhibitors to evaluate
any differential effects between Cox1 and Cox2 selective inhibition. Similar to what was observed
with the Cox2 selective inhibitors, FR122047 treatment over 24 h induced the gene expression of
Notch3 and Hes1 at both doses of the drugs, 1 and 10 μM (Fig 17C). There was no effect of
FR122047 on Notch1 expression, and similarly to what was observed with 100 μM NS398, high
dose FR122047 induced a significant decrease in Hey1 expression.
Lastly, ace and ibu were chosen as representative non-selective Cox inhibitors at doses
of 50 and 100 μM, which were converted from human-relevant doses [245]. Both 100 μM ace and
ibu significantly upregulated Notch3 expression, while only 100 μM ibu induced Hes1 expression
(Fig 17D). An increase was observed with 100 μM ace on Hes1 expression, though not significant.
Interestingly, while not observed with the Cox selective inhibitors, 100 μM ace induced the
expression of Notch1, which indicates that ace is capable of targeting both types of Notch
receptors, perhaps due to independent effects outside Cox inhibition. 100 μM ibu was also able
to induce Hey1 expression, which is another effect not observed with the selective Cox inhibitors
at high doses and may be attributed uniquely to this drug.

107

Evaluation of Notch3 protein expression using immunofluorescence staining illustrated
similar effects, with NS398 and FR122047 inducing the expression of Notch3 visualized as bright
green fluorescence positioned within the nucleus (Fig 17E). Quantification of positive cells
showed significant increases with the high doses of 100 uM and 10 uM NS398 and FR122047,
respectively (Fig 17F-G). Western blot validation showed similar results, however induction of
Notch3 expression was not observed with either dose of celecoxib (Fig 17H).

108

Notch3
Notch1
Hes1
Hey1
B)
Notch3 Notch1
Hes1 Hey1
VC 10 50 100
NS98
Dose (uM)
Rela-ve g ene expr es sion
(fold change of contr o l )
VC 10 50 100
NS398
Dose (uM)
Rela-ve g ene expr es sion
(fold change of contr o l )
Rela-ve g ene expr es sion
(fold change of contr o l )
Rela-ve g ene expr es sion
(fold change of contr o l )
VC 10 50 100
NS398
Dose (uM)
VC 10 50 100
NS398
Dose (uM)
*
***
*
*
**
***
*
*
*
**
VC 1 10
Celecoxib
Dose (uM)
Rela-ve g ene expr es sion
(fold change of contr o l )
Rela-ve g ene expr es sion
(fold change of contr o l )
Rela-ve g ene expr es sion
(fold change of contr o l )
Rela-ve g ene expr es sion
(fold change of contr o l )
VC 1 10
Celecoxib
Dose (uM)
VC 1 10
Celecoxib
Dose (uM)
VC 1 10
Celecoxib
Dose (uM)
A) C)
Rela-ve g ene expr es sion
(fold change of contr o l )
Rela-ve g ene expr es sion
(fold change of contr o l )
Rela-ve g ene expr es sion
(fold change of contr o l )
Rela-ve g ene expr es sion
(fold change of contr o l )
Notch3 Notch1
Hes1 Hey1
VC 1 10
FR122047
Dose (uM)
VC 1 10
FR122047
Dose (uM)
VC 1 10
FR122047
Dose (uM)
VC 1 10
FR122047
Dose (uM)
*
*
*
*
***
D)
VC 50 100 50 100
Dose (uM)
Acetaminophen Ibuprofen
Notch3 Notch1
Hes1 Hey1
VC 50 100 50 100
Dose (uM)
Acetaminophen Ibuprofen
VC 50 100 50 100
Dose (uM)
Acetaminophen Ibuprofen
VC 50 100 50 100
Dose (uM)
Acetaminophen Ibuprofen
Rela-ve g ene expr es sion
(fold change of contr o l )
Rela-ve g ene expr es sion
(fold change of contr o l )
Rela-ve g ene expr es sion
(fold change of contr o l )
Rela-ve g ene expr es sion
(fold change of contr o l )
**
*
*
* *
E)
VC 10 100
NS398
Dose (uM)
% Posi7ve Cells
(Fold change/ VC)
Notch3
*
VC 1 10
FR122047
Dose (uM)
% Posi7ve Cells
(Fold change/ VC)
Notch3
*
F) G)
Vehicle Control 10 uM FR122047
100 uM NS398 Vehicle Control
50 uM GFP: Notch3
Red: Alpha T ub ulin

109

3.4.7. PGD2 negatively regulates Notch signaling in C18-4 spermatogonial cells
We next wanted to explore whether prostaglandins are involved in mediating the
interaction between the eicosanoid and Notch signaling pathways, and whether they are
responsible in the activation of the Notch pathway observed with silencing Cox1. Wildtype C18-4
cells were treated for 24 hours with 1, 10, and 50 μM of supplemental PGD2. PGD2 treatment
resulted in the suppression of Notch3 mRNA expression (Fig 18A), and a similar effect was
observed with Hes1 expression as suppression was also observed at the 10 and 50 μM doses
(Fig 18C). 10 μM PGD2 downregulated Notch1 and Hey1 expression (Fig 18B, D), but this effect
was lost at the highest dose at 50 μM, which may be due to compensatory effects associated with
the higher dose of PGD2.
Figure 17. Effect of 24-hr treatment of Cox inhibitors on expression of Notch pathway genes. Effects of treatments
on Notch3, Notch1, Hes1, and Hey1 with (A) 10-100 μM NS398, (B) 1-10 μM celecoxib, (C) 1-10 μM FR122047,
and (D) 50 and100 μM acetaminophen and ibuprofen. Results are presented as fold change of vehicle control (VC).
mRNA expression data is normalized to Gapdh. (E) Immunofluorescence image of Notch3 expression after 24-hour
treatment with 100 uM NS398 and 10 uM FR122047 against their respective vehicle controls. Green: Notch3, red:
alpha tubulin, blue: DAPI. Scale in µm. Quantification of % positive cells is normalized to vehicle controls with (F) 10
and 100 uM NS398 and (G) 1 and 10 uM FR122047 24-hour treatment. N=3 independent experiments conducted
in triplicates. Significant difference relative to VC with one-way ANOVA test and multiple comparisons: * (p≤0.05), **
(p<0.01), *** (p<0.001). (H) Representative western blots are cropped to show Notch3 expression with 24-treatment
of NS398, celecoxib, and FR122047. Expression is normalized against total protein.

110

Figure 18. Supplemental PGD2 negatively regulates Notch3 and Notch pathway components. mRNA expression of
(A) Notch3, (B) Notch1, (C) Hes1, and (D) Hey1 with 24-hour treatment of 1, 10, and 50 uM PDG2. Results are
presented as fold change of vehicle control (VC). mRNA expression data is normalized to Gapdh. Significant difference
relative to VC with one-way ANOVA test and multiple comparisons: * (p≤0.05), ** (p<0.01), *** (p<0.001).

3.4.8. Notch signaling is disrupted in testicular cancer
Gene microarray was performed on human non-tumoral “normal” and testicular germ cell
tumor (TGCT) biopsies and the TCam2 (TGCT) cell line samples. In total, 9 pathological samples
(containing Tcam2 human TGCT cell line samples (N=2), seminomas N=3, and non-seminomas,
embryonic carcinomas (EC) N=2, and 1 of unknown pathology) and 3 normal specimens were
collected from patients. Gene counts between normal and tumor paired samples were normalized
using quantile normalization corrected for background signal and comparisons were carried out
using the Bayesian statistical approach as described in Manku et al., 2016 [238]. Gene microarray
analysis showed an increase in the expression of NOTCH3 corresponding with a decrease in
APH1B, which encodes gamma secretase subunit APH-1B, in tumor samples compared to
“normal” human samples (Fig 19A-B). As the gamma secretase complex is responsible for
releasing the Notch intracellular domain from its respective receptor, a decrease in expression
Rela7ve gene e xpression
(fold change of control)
VC 10 50
PGD2
Dose (uM)
1
Notch3
VC 10 50
PGD2
Dose (uM)
1
Rela7ve gene e xpression
(fold change of control)
VC 10 50
PGD2
Dose (uM)
1
Rela7ve gene e xpression
(fold change of control)
VC 10 50
PGD2
Dose (uM)
1
Notch1
Hes1 Hey1
***
***
**
***
**
**
***
A) B)
C)
D)

111

suggests disruptions of downstream Notch signaling targets [246]. These data indicate that the
constitutive upregulation of Notch3 expression may be a novel hallmark of testicular cancer.
Altogether, the data suggest that the Notch3 signaling pathway may be involved in regulating the
proper development of spermatogonia, and its dysregulated activation may contribute to the
malignant transformation of healthy spermatogonia or their precursor cells to tumor cells.

Figure 19. Expression of NOTCH3 and APH1B in human testicular cancer. Expression of (A) NOTCH3 and (B) APH1B
transcripts measured by gene microarray and normalized between normal specimens (N=3) and pathological samples
(N=9) using the Bayesian statistical approach: * (Bayes p<0.05).

3.5. DISCUSSION
The objective of this study was to explore the role of the eicosanoid pathway in
spermatogonial development. We chose to target the Cox enzymes due to their properties as rate
limiting enzymes of the eicosanoid pathway and as targets of over-the-counter analgesics
acetaminophen and ibuprofen. As Cox1 is the more prevalent of the two enzymes, and being one
that is constitutively expressed in all cell types [49], we focused our efforts towards interrogating
the role of Cox1. We believe that a greater understanding of its involvement in regulating
spermatogonial development will also aid in elucidating potentially toxic properties of analgesic
drugs that target the eicosanoid pathway. Observations from Cox1 knockdown studies were
confirmed with pharmacological Cox inhibitors that target Cox1 and Cox2 both selectively and
Rela%ve Expression Units
Normal Cancer
NOTCH3 APH1B
Normal Cancer
Rela%ve Expression Units
A)
B)
*
*

112

non-selectively. A summary of our results is shown in Figure 20. Our data indicate that blocking
Cox enzymes either with shRNA-based silencing or pharmaceutical Cox inhibitors can alter the
cellular properties of the spermatogonial stem cell, as modeled by the C18-4 cell line. Cox2 was
found to be positively associated with differentiation, while Cox1 was negatively associated with
differentiation. Previous studies conducted by our lab found that ace and ibu also provoked
functional changes in neonatal gonocytes, and we reported that treatments of these drugs were
able to induce gonocyte proliferation and partially inhibit differentiation after 24 hrs. In the current
study, we report that targeting Cox1 may have an opposite effect on SSCs than in gonocytes by
inducing the differentiation of the germ cells during this stage of development, while trends with
Cox2 inhibitors are consistent. But it is also important to note that gonocytes and SSCs represent
two unique stages of development with the establishment of germ cells on the basement
membrane being the major differentiator between these two cell types. These cells also exhibit
unique roles, with the gonocytes’ main purpose being to establish the foundational SSC pool and
SSCs being to maintain lifelong production of sperm. Their unique roles are reflected in the
distinctive molecular profile of these two populations of cells and their different fate trajectories
[17, 247-249]. Therefore, it is possible that drugs targeting the same enzymes within the gonocyte
and spermatogonial populations may have different effects on regulating differentiation. However,
we cannot disregard that our findings are based on a cell-line model, which, despite being able to
exhibit differentiating properties that are characteristic of SSCs, may not be able to recapitulate
all the properties of a non-immortalized mammalian germ cell population.

113

Figure 20. Schematic of crosstalk between eicosanoid and Notch pathways and predicted downstream effects.

From the total RNAseq analysis, we identified several signaling pathways that were
altered by Cox1 silencing. Of those, the Notch signaling pathway was consistently upregulated in
both knockdown and drug studies, suggesting that this pathway is a major downstream target of
eicosanoids. This is confirmed with treatment of PGD2 on wildtype C18-4 cells, illustrating that
PGD2 can negatively regulate the expression of Notch pathway genes. By suppressing PGD2
levels, Cox inhibitors may be releasing a brake on Notch pathway activity, which subsequently
results in constitutive activation of the pathway. The Notch signaling pathway is a well conserved
pathway that has been widely characterized for its involvement of cell fate determination [250].
Upon ligand binding, the Notch receptor undergoes proteolytic cleavage leading to the
cytoplasmic release of its intracellular domain (NICD) and subsequent translocation into the
nucleus where it interacts with transcriptional regulators and activate target genes such as
Hairy/Enhancer of Split (Hes) and Hes-related (Hey) genes [251]. Cox1 silencing led to increased
expression of both Notch3 and Hes1, while pharmacological Cox inhibitors induced differential
effects on Notch receptors and their downstream targets. Interestingly, NOTCH3 was also
upregulated in biopsies of testicular germ cell tumors, in addition to disruption of APH1B, which
is involved in NICD release. These results suggest that this pathway may play an important role
in spermatogonia maintenance and may contribute to the pathogenesis of testicular cancer.
Notch3 signaling involvement in testicular cancer may be associated with the
morphological changes observed in the C18-4
Cox1-KD1
cells. Silencing of Cox1 in the C18-4 cell

114

line prompted morphological changes that seemed to reflect an epithelial to mesenchymal
transition (EMT) both in the alteration of C18-4
Cox1-KD1
cell physical appearance, decreases in the
expression of cellular adhesion and matrix metalloproteinase genes, and in the molecular
pathways that were predicted to be altered from IPA. EMT is characterized by a transition from
an epithelial phenotype to an elongated fibroblastic or mesenchymal phenotype leading to
increased motility and invasion, which is also characteristic of tumors. During embryonic
development, EMT is related to cellular plasticity, and this process plays critical roles in stem cell
behavior and induction of pluripotency [252]. While more studies will have to be conducted to
evaluate downstream Notch signaling effects before definitively determining a role of this
pathway’s involvement with EMT observed in the C18-4
Cox1-KD1
cells, we can presently turn to the
literature to speculate a possible relationship.
The Notch pathway has been found to be implicated in many human cancers and several
Notch pathway receptors and have been shown to be prognostic markers in different types of
cancers such as breast, gastric, and prostate [253, 254]. Whereas many studies found positive
correlations between Notch signaling activation and cancer progression, some studies evaluating
the involvement of Notch signaling on EMT showed that this pathway is negatively correlated with
the EMT transition of breast cancer cells [255, 256]. The overexpression of the Notch intracellular
domain (N3ICD) decreased mesenchymal marker vimentin and increased levels of tight junctional
proteins E-cadherin and E-catenin. Perhaps in spermatogonia, this signaling pathway can also
play a significant role in the EMT of the SSC, and may be involved in the malignant transformation
of spermatogonia, which may aid in the pathogenesis of spermatocytic seminomas [231].
Other studies support the involvement of the Notch signaling pathway in spermatogenesis
either in affecting the development of germ cells themselves or being involved in the regulation of
germ cell development by supporting Sertoli cells. Notch3 and Hes1 were found to be expressed
in pup and prepuberal spermatogonia, consistent with our findings, and only weakly in adult
spermatogonia [257]. As for the role of Notch signaling in Sertoli cells, the cells responsible for

115

supporting the proper maturation of spermatogonia, the Notch-Hes1 signaling axis was
dispensable for spermatogenesis as Notch knockout mice retained normal spermatogenesis and
fertility [258]. However, more recently, studies have shown that Notch signaling activation in
Sertoli cells is important in regulating SSC development [259, 260]. Undifferentiated
spermatogonia can modulate the activation of the Notch1-Hey1 axis to downregulate CYP26b1
and remove its blockade on retinoic acid production to drive SSC differentiation [207]. Our study
indicates that the Notch signaling pathway is important to the development of SSCs, supporting
a similar role to that of Sertoli cells.
Alternatively, one cannot rule out the role of TGFb signaling in our model as several TGFb
signaling family members were upregulated in the C18-4
Cox1-KD1
cells as well. The TGFb signaling
pathway plays a critical role in establishing the germline during mouse embryogenesis, in addition
to steroidogenesis, cord formation, and gonocyte behavior [261, 262]. Introduction of bone
morphogenetic protein 4 (BMP4) was found to induce differentiation of male chicken embryonic
stem cells into PGCs and SSCs, whereas inhibition of TGFb signaling interfered with the ability
of these cells to differentiate to SSCs [263]. Furthermore, TGFb was able to induce Cox2
expression during EMT in breast cancer progression, and Cox2 had the ability to enhance
oncogenic TGFb signaling by inhibiting Smad 2/3 activity via PGE2 [264]. Both the Notch and
TGFb signaling pathways are known to converge on the regulation of differentiation events during
development. For example, both Notch and TGFb are known to inhibit myogenesis [265, 266].
Furthermore, not only do the two pathways interact, but significant crosstalk exists between them.
Stimulation of neural stem cells and myoblasts with TGFb induced a rapid increase in Hes1
expression, and reciprocally, Notch signaling was able to influence the expression of TGFb target
genes [267]. Similar findings were observed in studies evaluating the ability of both pathways to
block myogenic differentiation [268] and activate the Hes-related gene, Herp2, in endothelial cells
[269]. While we chose to focus our efforts on characterizing Notch signaling with respect to
pharmacological Cox inhibitors, we note that TGFb signaling may play an important role as well,

116

and its ability to crosstalk with Notch signaling pathways can also contribute to downstream
functional effects.
In future studies, we aim to explore the mechanistic interactions between eicosanoid and
Notch signaling pathways that are supported by the literature. In a gastric cancer cell line,
characteristics of cancer progression that were enhanced by Notch1 pathway activation were
suppressed when the cells were treated with NS398 or by Cox2 knockdown [270]. It was also
found that PGE2 receptor acts downstream of the Notch signaling pathway to inhibit differentiation
of human skeletal muscle progenitors [271]. Interestingly, while these studies suggest that the
eicosanoid pathway acts downstream of the Notch signaling pathway, we show that eicosanoids
can also regulate the expression of Notch signaling pathway genes. Indeed, prostaglandins may
play crucial roles as biochemical mediators that regulate molecular signaling pathways that are
important for cell fate determination of SSCs.
The roles of non-selective Cox inhibitors on male germ cell development are only recently
beginning to be deciphered. While several groups have reported that Cox inhibitors can decrease
the expression of fetal germ cell markers [103, 118, 127, 272], no reports currently exist on the
effects of Cox inhibitors on SSCs, which are critical for sperm formation and fertility beyond
puberty. Infancy represents a critical period for SSC formation and disrupting SSCs or their
precursors may be associated with infertility and testicular cancer [273-275]. The development of
germ cells is highly dynamic during this period of development, and disruptions to the balance
between self-renewal, proliferation, and apoptosis have been proposed to initiate testicular germ
cell tumor formation [276]. Therefore, a greater understanding of how Cox inhibitors, such as
acetaminophen and ibuprofen, might impact the development of neonatal germ cells could
provide insights into the origins of testicular cancer and male infertility that have been afflicting
young men at an alarming rate.
In conclusion, we report that Cox inhibition can alter SSC differentiation and lead to the
activation of the Notch signaling pathway primarily via the Notch3-Hes1 axis. We speculate that

117

the Notch3 pathway underlies a mechanism behind the ability of pharmacological Cox inhibitors
to disrupt normal fate determination by inducing the morphological transformation of SSCs. Future
studies will explore interactions between eicosanoid and Notch signaling pathways to further
understand the role of Notch3 in regulating SSC development.

3.6 SUMMARY
In this chapter, we showed that the Cox enzymes can alter SSC differentiation in different ways,
with Cox2 positively affecting differentiation and Cox1 acting in the opposite direction. We have
unveiled a pathway that Cox inhibitors can activate in SSCs and speculate that the Notch pathway
may regulate fate determination. In the next study, we take this investigation further by evaluating
whether there are any long-term effects with postnatal treatment with pharmacological Cox
inhibitors, focusing primarily on the in vivo effects of commonly used analgesic drugs
acetaminophen and ibuprofen on male reproductive development.

3.7. ACKNOWLEDGEMENTS
We acknowledge co-author Amina Khan for her contribution to the current study. We thank Dr.
Marie-Claude Hofmann for the generous gift of the C18-4 mouse spermatogonia cell line, as well
as the help from Dr. Gupreet Manku for the gene array data from the human testicular cancer
samples. We would also like thank the bioinformatics team at USC Norris library in aiding with our
total RNAseq analysis, the USC Norris Molecular Genomics Core for sequencing our samples for
the Cox1 knockdown study, and Ari Gritsas and team at the McGill University Health Centre for
conducting the prostaglandin LC/MS studies.

118

CHAPTER 4: CHARACTERIZING THE EFFECTS OF EARLY POSTNATAL ACE AND IBU
ADMINISTRATION ON IMMEDIATE AND LONG-TERM MALE REPRODUCTIVE
DEVELOPMENT

OBJECTIVE (AIM 3): To examine the effects of human relevant doses of acetaminophen and
ibuprofen in relation to germ cell development in vivo and assess their potential link to male
reproductive pathologies.

4.1 ABSTRACT
Several studies, including ours, in the reproductive toxicology field have suggested that
the eicosanoid system can play a role in regulating male reproductive development, and that
pharmacological inhibitors of this pathway, such as acetaminophen (ace) and ibuprofen (ibu), can
impact the proper development of the testes. The postnatal timeframe (PND1-7) of rodent
development is a critical period in which neonatal testicular gonocyte population proliferate and
migrate toward the basement membrane of the seminiferous tubules where they subsequently
transition to spermatogonial stem cells (SSCs) as well as the first-wave spermatogonia over the
course of several days. In this study, we aim to assess whether the administration of human-
relevant doses of ace and ibu to rats for the first seven days after birth, targeting these crucial
developmental events, would lead to any immediate or long-term adverse effects in the function
and morphology of the various cell types and structures in rat testes.
To explore the in vivo effects of ibu and ace on male reproduction, we treated Sprague
Dawley rats from postnatal day (PND)1-7 with ace at 0.7 and 1.4 mg/day (low/high dose), ibu at
0.4 and 0.7 mg/day (low/high dose) or vehicle. Single cell RNA sequencing (scRNAseq) analysis
on PND8 testes identified distinct cell populations, with spermatogonia expressing characteristic
markers Id4, Dazl, Sohlh1, and DDX4. 7-day treatment of high dose ace and ibu induced
differential expression of 137 and 431 genes, respectively. These genes are involved in signaling
pathways related to cell migration and survival, amino acid degradation, and DNA repair. 77

119

upregulated and 11 downregulated genes were shared between the two drugs, such as genes
involved in GPCR activity, Jun/Fos signaling, and fatty acid metabolism. Long-term effects were
observed in PND90 rats, including vacuolation in seminiferous tubules, and increases in
testosterone and luteinizing hormone levels by the high dose treatments.

120

4.2 INTRODUCTION
The recent decades have seen increases in male reproductive disorders and diseases,
with total fertility rates trending downward in western countries, correlating with increases in
ailments such as poor semen quality, cryptorchidisms, and hypospadias [1]. Cyclooxygenase
(Cox) inhibitors or non-steroidal anti-inflammatory drugs (NSAIDs) have been gaining greater
attention recently for their widespread use during pregnancy and contribution to significant rates
of male reproductive disorders [7, 193, 227, 229]. Two common over-the-counter Cox inhibitors,
acetaminophen (ace) and ibuprofen (ibu), can non-selectively target the two isoforms of Cox
enzymes, Cox1 and Cox2, and block prostaglandin (PG) production. Ibu competitively inhibits
Cox substrate, arachidonic acid, for binding at the active site of both enzymes, and has greater
selectivity for Cox1 [62, 63]. Ace, alternatively, can take advantage of a side pocket in Cox2 that
is inaccessible in Cox1 and therefore can selectivity target Cox2 to a greater extent [64]. Little is
currently known about the consequences of neonatal administration of these two drugs on later-
life reproductive disorders. However, the potential toxic effects of NSAID and analgesic drug are
particularly important to explore in neonates, who are often administered these drugs for the
treatment of various conditions.
We have previously reported the presence of a functional Cox pathway and detectable
PG production both in gonocyte precursors and spermatogonia in mouse and rat models, and our
studies support a role of these enzymes to regulate the differentiation and maintenance of these
germ cell types [77, 235]. In this study, we are aiming to evaluate similar treatment conditions in
a rodent model to assess the immediate and long-term effects of neonatal administration of Cox
inhibitors at significant stages of germ cell development and spermatogenesis. These include the
development of neonatal gonocytes, the establishment of the spermatogonial stem cell pool, and
spermatogenesis and sperm production in adults. To mimic the real-life scenario of a hospitalized
neonate who is administered ace and ibu on a consistent basis, we dosed rat pups from postnatal
day (PND) 1 to 7 to target early significant male reproductive events leading from gonocyte to

121

SSC/spermatogonia. The therapeutic doses we have chosen to use are reflective of human
relevant doses derived from clinical trials conducted on children and from recommended doses
of product labels [245]. These consist of a high dose and low dose, to explore a range of
observable outcomes that can be induced at a toxicological dose and one that is more relevant
to a level exposed by most children in this setting. To capture significant reproductive events
leading to sperm formation, we have characterized the effects of these drugs on testicular tissue
during the establishment of neonatal gonocyte proliferation (PND4), establishment of
spermatogonia (PND8), and in adulthood (PND90).
To further support whole tissue characterization of the testes, we have taken a
transcriptomic approach that has allowed for mechanistic assessments of effects at a single cell
level. Conventional gene-based methods will often give average readout of genes due to the
heterogeneity of tissues like the testes, and this issue can be resolved using high-throughput
techniques such as single cell RNA sequencing (scRNAseq), which has the advantage of being
able to define cell-type specific changes in gene expression. We have leveraged this technique
to help us explore differentially expressed genes that were altered with 7-day high dose
administration of ace and ibu on PND8 rats and to characterize signaling pathways that have also
been predicted to be altered with these treatments using bioinformatic-based computational tools.
Our analysis has been restricted to the different cell populations of the testes (spermatogonia,
Sertoli, and Leydig cells) which are characterized by well-established markers to compare the
effects of ace and ibu on the individual cell types.
Lastly, we evaluated testicular function in the adults to determine whether long-term
reproductive toxicological effects were observed with the neonatal administration of these drugs
by evaluating serum levels of testosterone and pituitary luteinizing hormone, as well as
characterizing the effects of the treatments on testicular morphology and signaling. Overall, we
report morphological alterations in testicular structure, signaling, and changes in long-term
production of hormones, which would suggest unintended long-term consequences of neonatal

122

administration of these drugs in addition to immediate effects on changes in molecular signaling
with 7-day postnatal administration of ace and ibu. In light of these findings, we believe that
greater consideration must be taken to monitor the administration of pharmacological Cox
inhibitors at early ages, as we have observed significant toxicological effects on the male
reproductive system with these seemingly harmless drugs, which could have long-term
consequence for infants in NICUs and other prolonged hospitalized settings.

4.3 MATERIALS AND METHODS.
4.3.1 Animals and Treatments
Neonatal male Sprague Dawley rat pups and their foster mothers were purchased from Charles
River Labs (Wilmington, MA, USA) and were treated daily starting at postnatal day (PND) 1 with
acetaminophen (0.7, 1.4 mg/day), ibuprofen (0.4, 0.7 mg/day), or vehicle control (containing
EtOH) up until sacrifice at PND4 or until completion of treatment on day 7, at which point rats
were sacrificed on PND8. One cohort of animals of each treatment group were allowed to age
until 90-100 days (PND90) to evaluate the effect of seven-day treatment on long-term male
reproductive effects. Pharmacological Cox inhibitors acetaminophen and ibuprofen were
purchased from Sigma Aldrich (St. Louis, MO, USA). Drugs were dissolved in ethanol to obtain
stock solutions of 0.17 g/mL. Stock solutions of ace and ibu were subsequently diluted in puppy
formula (Esbilac Powder Milk Replacer, PetAg, Inc., Illinois, USA) to obtain the abovementioned
final concentrations. Vehicle controls contained 4.2% ethanol. Drugs were administered twice
daily to pups using plastic droppers.

4.3.2 Single Cell RNA Sequencing
Testicular cells from PND8 rats that were treated for 7 days with high dose 1.4 mg/day
acetaminophen (high ace), 0.7 mg/day ibuprofen (high ibu), and vehicle control (VC) were isolated
into single cell preparations according to modified protocol previously described [77]. In brief, cells

123

were isolated by sequential enzymatic digestion from decapsulated testes in RMPI 1640 medium
(Invitrogen, Burlington, ON, CA) with 5% heat-inactivated FBS (Sigma-Aldrich, St. Louis, MO,
USA), 2% Penicillin-Streptomycin Solution 100X (Corning
™
), and 1% amphotericin B (CellGro).
Single cell RNA sequencing was prepared using 10x Genomics 3’ v3.1 platform (Cat. #1000092,
10x Genomics, Pleasanton, CA) according to the manufacturer’s protocol. Samples were checked
for cell viability, and only those containing >70% live cells were allowed to move forward with
sequencing. 1 animal per treatment was sequenced, for a total of 3 samples. Samples were
parsed into individual cells and individual cDNA libraries were constructed using the 10x
Chromium Controller. Libraries were sequenced using the Illumina Noveseq 6000 platform
(Illumina, San Diego, CA) at a read length of 28x90 and read depth of 80,000 reads/cell for 6000-
8000 cells. Sequencing data were packaged and analyzed using the 10x Genomics cell ranger
analysis tool.

4.3.3 Bioinformatics Analysis
scRNAseq bioinformatics analysis was conducted using the Partek Flow
®
software.
FASTQ files were imported, and tags were trimmed to allow for the removal of approximately 10%
of reads per sample. The STAR 2.7.3a aligner was used to align the reads to the rat whole
genome (Rattus norvegicus (rat)- rn6), and the deduplicate UMI task was run using annotation
model rn6- Ensembl Transcripts release 100. Alignment tasks resulted in samples of: VC- 8,022
cells (93.24% reads, 16,320 median reads per cell), high ace- 13,259 cells (91.96% reads, 16,465
median reads per cell), and high ibu- 16,614 cells (95.18% reads, 14,753 median reads per cell).
Barcodes are next filtered based off knee plots generated from the deduplicated reads task
according to default parameters recommended by the software. Next, the reads for each barcode
were quantified to the rnf_ensembl_release100_v2 gene/feature annotation model with minimum
reads of 30, 50% minimum read overlap, and forward-reverse strand specificity. Reads were
further filtered to be within 500-4,000 expressed genes, 10% maximum mitochondrial reads,

124

approximately 30% ribosomal reads, and between 1000-15,000 total counts. After filtering out
approximately 30% of cells that did not meet these requirements, 2,500 cells (VC), 4,133 (high
ace), and 7,232 (high ibu) were used in the final analysis.
Next noise reduction was used to filter genes that were not expressed by any cell in the
data set but were included in the matrix file. The data were then normalized according to
recommended parameters which accounted for differences in unique molecular identifier (UMI)
counts per cell and subsequently, the data were log transformed. To reduce the dimensionality of
the data prior to cluster, the results of PCA and Scree plots were analyzed. A cut off was set to
20 PCs, which is the point at which additional PCs offered little additional information. Graph
based clustering based off our parameters resulted in 9-11 clusters in each sample, which was
visualized with the t-SNE plots. Clusters were reidentified based off characteristic markers of
different cell types in the testes. Downstream analysis to evaluate differentially expressed genes
were conducted using parameters FDR <0.01, and fold change between -2 to 2. Qiagen Ingenuity
Pathway Analysis Software (IPA; Qiagen, Hilden, Germany) was also used to evaluate functional
pathways using lists of DEGs generated from Partek Flow.

4.3.4 Tissue Histology
Tissue processing, embedding, and sectioning
Isolated testes from PND4, PND8, and PND90 rats were rinsed with sterile PBS, fixed
with 4% PFA in PBS (Santa Cruz Biotech, Dallas, TX) overnight, and kept in 70% EtOH at 4°C
until processing. Thereafter, they were automatically processed by the Spin Tissue Processor
Microm STP-120 (Thermo Fisher Scientific, Waltham, MA) with a program to process tissues
using the following reagents (90min/step): dehydrant (70%, 80%, 95%, 100%) (Richard-Allan
Scientific), 2 cycles of xylene (Cancer Diagnostics, Durham, NC) and 2 rounds of paraffin (Type9,
Thermo Fisher Scientific, Waltham, MA). Tissues were embedded using the Tissue Embedding
Center EC-350 (Thermo Fisher Scientific, Waltham, MA) at 60°C in tissue block molds filled with

125

hot paraffin. Tissue blocks were cooled down at room temperature for 30 minutes and stored at
4°C for sectioning.
Tissues were trimmed and rehydrated on ice water for 30 minutes and were sectioned at
5 µm using the Rotary Microtome HM-310 (Thermo Fisher Scientific, Waltham, MA). Next,
sections were transferred to the water bath (14792V, VWR Life Science, Solon, OH) at 37°C, and
picked up by VistaVision™ HistonBond
®
slides (16004406, VWR Life Science, Solon, OH).
Sections were left to dry overnight.

H&E Staining
Slides were stained using the automated Shandon Varistain
®
Gemini ES (Thermo Fisher
Scientific, Waltham, MA) with the following reagents: Clear-Rite 3, Clarifier, Flex 95, Flex 100,
Eosin-Y, Richard-Allan Scientific Hematoxylin 721, Shandon Bluing Reagent. All reagents were
purchased from Thermo Fisher Scientific. Slides were immediately mounted with Richard-Allen
Mounting Medium (Thermo Fisher Scientific, Waltham, MA) and dried overnight in the fume hood.

Immunofluorescent Staining
Immunostaining of testicular sections was performed as previously described [77]. In short, slides
were first dewaxed and rehydrated using Citrosol and Trilogy solutions (Cell Marque, Rocklin,
CA). Next, DAKO Target Antigen Retrieval Solution (Agilent Technologies Inc., Santa Clara, CA)
was used for epitope retrieval prior to one hour blocking with PBS containing 10% BSA and 10%
donkey serum. Once overnight primary antibody (Cox1 4841 Cell Signaling Technology 1:100,
Cox2 ab15191 Abcam 1:100, Notch3 ab23426 Abcam, 1:100) incubation was complete, slides
were incubated for one hour with secondary antibody (donkey anti-rabbit IgG Alexa Fluor 488, A-
21206, Thermo Fisher Scientific). The slides were co-stained with DAPI (Vector Labs,
Burlingame, CA) for 5 minutes, and mounted with one drop of Fluoromount-G (Electron

126

Microscopy Sciences, Hatfield, PA). Slides were imaged using the Revolve-G130 fluorescent
microscope after adjusting for background signal.

4.3.5 Testosterone and luteinizing hormone ELISAs
PND90 rats were bled, and blood samples were spun down at 1300 RCF for 10 minutes for serum
collection. Hormone levels were measured with the following ELISA kits: Testosterone (582701,
Cayman Chemical, Ann Arbor, MI) and luteinizing hormone (CSB-E12654r, Cusabio Technology,
Houston, TX) according to manufacturer’s instructions. Samples were diluted at 1:2 concentration
in PBS for LH measurements. OD measurements were detected by the VICTOR
TM
X5 Multilabel
Plate Reader (PerkinElmer, Inc., Waltham, MA, USA) per manufacturer’s recommendations. Data
were analyzed using the %B/B0 values and concentrations of standards vs. hormone levels were
plotted in Prism version 7.0 (GraphPad Software, San Diego, CA) software using a 4-parameter
logistic fit.

4.3.6 Statistics
Statistical analysis was conducted by one-way ANOVA with post-hoc multiple comparison
analysis using Prism Graphpad Software (San Diego, CA). Statistical significance was considered
at values of * p ≤ 0.05, ** p ≤ 0.01 and *** p ≤ 0.001 indicating difference of treatments compared
to vehicle control samples. Results are shown as the mean ± SEM. Statistical analysis of
bioinformatics data was conducted as described in the abovementioned section “Bioinformatics
Analysis”.

4.4 RESULTS
4.4.1 No effects of treatment on animal weights or AGD
To assess immediate and long-term reproductive developmental effects of ace and ibu on
postnatal male reproductive function, PND1 rats were treated for seven days with two human

127

infant relevant doses of ace or ibu in addition to a solvent-matched vehicle control. The current
infant recommended dose of acetaminophen (Tylenol
®
) according to the recommended intake
[277-279] is 40 mg per 10-pound infant, corresponding to a max daily dose of ~35.2 mg/kg/day.
Multiplying by 4 to give the rat pup equivalent dose resulted in 140.8 mg/kg per day in rats or a
1.4 mg/day dose for a 10 g pup, which is identified as our “high dose”. The “low dose” corresponds
to half the high dose, at 0.7 mg/day. As the recommended dose of ibuprofen (Motrin
®
) is lower
than ace at 0.5 mg per 15-pound infant with a max daily intake at 0.28/mg/day and based off of
similar in vivo studies evaluating ibu at half the doses of ace [119, 277, 278], we have determined
a high dose of ibu at 0.7 mg/day and a low dose at 0.4 mg/day. Pups were sacrificed at PND4, at
a period of proliferation in gonocytes and migration to the basement membrane, at PND8 when
the spermatogonial stem cell population is established, and at PND90 at adulthood. Organs were
collected and weights were assessed to evaluate any potential toxicities associated with the
treatments prior to downstream analyses. Our animal study schematic is outlined in Figure 21.

Animal weights were monitored and recorded daily for each group of treatment conditions:
vehicle control (N=30 rats), 0.7 mg/day ace “low ace” (N=15 rats), 0.4 mg/day ibu “low ibu” (N=15
Figure 21. Schematic of animal study design. Sprague Dawley rat pups were dosed from postnatal (PND) 1 to 7 with
vehicle (VC), acetaminophen at 1.4 mg/day (high ace) and 0.7 mg/day (low ace), and ibuprofen at 0.7 mg/day (high
ibu) and 0.4 mg/day (low ibu).

128

rats), 1.4 mg/day ace “high ace” (N=25 rats), and 0.7 mg/day ibu “high ibu” (N=24 rats) (Figure
22A). No changes were observed in animal weights between days 1 to 5 of treatments, but
significant decreases in body weights were observed on day 6 and day 7 of low ace and low ibu
treatments (Figure 22B, C). However, these decreases in body weight were recovered by PND90
as there were no significant alterations by adulthood (Figure 22D). Testes, kidneys, livers, and
brains were collected in PND4 and PND8 pups and were weighed to further assess organ-specific
toxicities (Figure 22E-H). No decreases in organ weights were observed with any treatments,
further confirming that the doses of treatments administered did not have a generalized adverse
effect on organ weights.
Anogenital Distance (AGD) measurement is an indication of reproductive health as shorter
AGDs in men can reflect decreased fetal androgen levels and has been found to be associated
with other conditions of testicular dysgenesis syndrome [1, 47]. AGDs were measured in PND4,
PND8, and PND90 rats to assess any ostensible changes in reproductive health (Figure 23A-B).
While some slight increases were observed with the high dose treatments in pups, by adulthood
no significant changes were observed in AGD measurements. Overall, we report no adverse
effects suggesting obvious organ-specific toxicity to the animals at the doses that were chosen
for this study.

129

A
B C D
Figure 22. Effects of treatments on animal weights. A) Graphical representation of pup body weights (g) during drug
treatment from PND1 to PND7, B) at PND6, C) PND7, and D) PND90. Organ weights at PND4 and PND8 of E) testis,
F) kidney, G) liver, and H) brain. Results are presented as average of at least 3 individual animals per treatment group.
Significant difference relative to vehicle control with One way ANOVA test and multiple comparisons: * (p≤0.05), **
(p<0.01), *** (p<0.001).

130

Figure 23. Effects of treatments on anogenital distances. Measurements of AGD (mm/cm) in rats treated with low ace,
low ibu, high ace, and high ibu relative to VC at age A) PND4, B) PND8, and C) PND90. Results are presented as
average of at least 3 individual animals per treatment group. Significant difference relative to vehicle control with One
way ANOVA test and multiple comparisons: * (p≤0.05), ** (p<0.01), *** (p<0.001).

4.4.2 Ace and ibu induces pronounced vacuolation in PND4 and PND8 pup testes
Histological sections from treated PND4 and PND8 pups stained with hematoxylin and
eosin (H&E) were used to assess changes in testicular morphology. H&E stains of tissues indicate
the presence of vacuoles within the tubules that could suggest alterations in normal germ cell
maturation and development (Figure 24A).
Upon quantifications of the images consisting of at least 5 images per animal, and at least
5 animals per treatment group, significant inductions of pronounced vacuolation were observed.
All treatment groups had significantly higher percentages of pronounced vacuolation when
compared to the vehicle control at PND4, or after 3 days of treatment (Figure 24B). High ace and
high ibu groups exhibited recovery by PND8, after 7 days of treatment, but rats treated with the
low doses continued to exhibit higher levels of vacuolation (Figure 24C). In fact, the number of
tubules containing vacuoles were much greater at PND8, with at least 2-fold greater percentages
of pathology in the low ace and low ibu groups when compared to the VC, than at PND4 at which
the vacuolation did not surpass 20% of all tubules quantified. As adverse effects in testicular
morphology were apparent in the short-term administration of these drugs, we next wanted to
further assess the molecular changes underlying these changes, with a focus on the high ace and
high ibu conditions at PND8.
A
B C

131

Figure 24. Effect of treatments on testicular morphology in pups. A) H&E staining of PND4 low ace treated testis
compared to vehicle control illustrating the presence of morphological abnormalities as vacuolization indicated at the
red arrows. Quantification of morphologically abnormal tubules relative to total tubules (% degenerated tubules) in pups
at B) PND4 and C) PND8. Results are presented as average of at least 3 individual animals per treatment group.
Significant difference relative to vehicle control with One way ANOVA test and multiple comparisons: * (p≤0.05), **
(p<0.01), *** (p<0.001).

4.4.3 Characterization of molecular effects of ace and ibu on testicular cell populations of
PND8 pups using scRNAseq
To assess underlying molecular changes that could explain immediate morphological
alterations observed in the pups upon treatment, we resorted to using single cell RNA sequencing
(scRNAseq) to characterize differential genetic alterations between the treatment populations
relative to the VC. Testes of VC, high ace, and high ibu treated PND8 pups were dissociated into
single cell suspensions and sent for genetic sequencing. Quality control checks were conducted

132

to confirm >70% live cells prior to sequencing and to ensure that >50% of reads were mapped
onto exonic regions after sequencing. The number of cells used for the analysis were
approximately 6,300 (VC), 11,000 (high ace), and 14,600 (high ibu). The Partek Flow software
was employed for bioinformatics analysis. Reads were trimmed to identify the unique molecular
identifiers of each paired read and then were aligned to the rat genome. Reads were then further
filtered to remove PCR artifacts and barcodes that corresponded to dead or dying cells. Cells with
greater than 10% mitochondrial counts were excluded, as well as cells beyond the range of
approximately 5-25% ribosomal counts. Of the original cell population containing the three
samples combined, which constituted of approximately 37,000 cells, 13,864 cells were deemed
satisfactory for downstream analysis.
Clustering analysis grouped cells into 9-12 clusters, which were then identified based off
the expression profiles of a specific cluster to a signature gene list of markers representing Sertoli
cells, spermatogonial stem cells (SSCs)/ spermatogonial cells, Leydig cells, macrophages, and
peritubular myoid cells (Figure 25A-B). Cell populations that did not express uniquely the genes
associated with the various testicular cell populations were categorized as “N/A” and were not
selected for analysis. The Partek Flow
®
software identified unique populations of cells that
corresponded to the expected transcriptomic profiles of the different testicular cell populations
and in the proportion expected of a PND8 rat testis. Controls exhibited approximately 40% Sertoli
cells, 16% Leydig cells, 7.5% spermatogonial cells, 18% peritubular myoid cells, and 1.2%
macrophages (Table 3). Treated samples deviated slightly from these values in proportion but
remained relatively consistent in their cell numbers of the different populations, likely due to the
significant deviations in total cell numbers of the different samples (VC: ~2,500, ace: ~4,100, ibu:
~7,200). Despite this, we are confident that the results are likely reflective of the cell populations
in the testes. Furthermore, this experiment has been repeated and so the results presented will
be further validated in subsequent studies to ensure reproducibility of the findings.

133

Table 3. Table showing the proportions of cells identified in each population of the testes using scRNAseq.

Sample
name
Leydig
Cells
Macroph
ages
Myoid
Cells
Sertoli
Cells
Spermat
ogonia
N/A Total
VC 401 31 449 1005 188 425 2499
High ace 490 0 730 1941 131 841 4133
High ibu 977 229 1037 4095 157 737 7232

Sample
name
Leydig
Cells
Macroph
ages
Myoid
Cells
Sertoli
Cells
Spermat
ogonia
N/A Total
VC 16% 1% 18% 40% 8% 17% 100%
High ace 12% 0% 18% 47% 3% 20% 100%
High ibu 14% 3% 14% 57% 2% 10% 100%

All
samples
1868 260 2216 7041 476 2003 13864

134

Figure 25. scRNAseq characterization of treated PND8 testicular cell suspensions. A) Dot plot illustrating expression
of various signature genetic markers characteristic of the various cell types in the testis (spermatogonial, Sertoli cells,
Leydig cells, peritubular myoid cells, and macrophages). The blue color represents the relative expression level, with
the darker color representing higher expression of the gene marker in the specific cluster listed on the right of the plot.
The circle represents the number of cells in the cluster expressing a specific gene. B) tSNE plots of the samples (VC,
high ace, high ibu) with clusters recategorized by the expression level of signature genes characteristic of the different
cell types in the testes.

4.4.4 Characterization of molecular changes in spermatogonia
A significant number of genes were altered in the spermatogonial cell population with the
treatments relative to the vehicle control. Using cut off parameters of false discovery rate (FDR)
<0.01 and fold change -2 and 2, 97 genes were upregulated, and 40 genes were downregulated
in the high ace dose group. Alternatively, in the high ibu dose group, 416 genes were upregulated
compared to 15 downregulated genes (Figure 26A).
Some of the top differentially expressed genes (DEGs) that were altered with either
treatment include several genes involved with GTPase activity such as rnd2 and rasl10a, which
were significantly upregulated with both high ace and high ibu (Table 4). Other genes that were
upregulated by high ace include fos, which is involved in Jun-Fos signaling, Mt3 involved in heavy
metal binding, and pmp22, which encodes a peripheral myelin protein. Genes that were
downregulated by high ace include one that encodes a protease inhibitor, wfdc1, an LDL receptor
related protein, lrpap1, a Cox assembly mitochondrial protein, Cmc1, a gene involved in nucleic
acid binding, r3hdm4, and a transporter of pyruvate into the mitochondria, mpc1. On the other
hand, genes that were upregulated by high ibu include kcne1, which regulates activity of
potassium channels, and cldn11, which encodes a member of the claudin family of tight junctional
proteins and is a major claudin in Sertoli cells. Genes that were highly downregulated include
clic1, which encodes a chloride channel, mdb3, which is involved in methyl-CpG binding, gsn,
which is involved in the regulation of actin filaments, and rnf10 and ube2v2, which both encode
proteins involved in ubiquitination.

135

Figure 26. scRNAseq characterization of the spermatogonial cell cluster in treated PND8 pups. A) Volcano plot
illustrating up and downregulated differentially expressed genes in the spermatogonial cell cluster with high ace and
high ibu treatment relative to the vehicle control. List of pathways predicted to be altered using the Ingenuity Pathway
Analysis software by B) high ace and C) high ibu treatment. D) Venn diagram of common genes altered by both high
ace and high ibu treatment.

136

Table 4. List of top differentially expressed genes upregulated and downregulated by high ace and high ibu treatment in the spermatogonial cell cluster and their
respective p-values, false discovery rate, fold change, and functions.
Gene ID
P-value (VC vs. High
ace)
FDR (VC vs. High
ace)
Fold change (VC vs.
High ace)
Function
Fos 9.75E-47 2.82E-44 255.32
Involved with TF complex AP-1 (Fos-Jun
pathway)
Rnd2 2.27E-51 7.63E-49 153.32 Encodes Rho GTPase
Mt3 1.27E-27 1.83E-25 150.25 Metallothionein (binds heavy metal ions)
Rasl10a 1.50E-09 5.11E-08 119.33 Involved with GTP binding and GTPase activity
Pmp22 3.78E-13 1.98E-11 101.97 Encodes peripheral myelin protein 22

Wfdc1 1.63E-07 4.07E-06 -75.07 Functions as protease inhibitor
Lrpap1 6.21E-15 3.89E-13 -74.58 Interacts with LDL receptor related protein
Cmc1 1.47E-18 1.25E-16 -73.43 COX assembly mitochondrial protein homolog
R3hdm4 3.45E-06 6.50E-05 -72.93 Involved in nucleic acid binding
Mpc1 1.67E-05 2.62E-04 -72.54 Responsible for transporting pyruvate into MT

Gene ID
P-value (VC vs. High
ibu)
FDR (VC vs. High
ibu)
Fold change (VC vs.
High ibu)
Function
Rps15al
4
0.00E+00 0.00E+00 209.09 Encodes structural component of ribosomes
Rasl10a 1.18E-33 6.85E-32 174.64 Involved with GTP binding and GTPase activity
Kcne1 6.69E-147 2.56E-144 174.33 Regulates activity of potassium channels
Rnd2 3.49E-74 5.80E-72 172.79 Encodes Rho GTPase
Cldn11 1.15E-43 9.39E-42 166.74 Encodes member of Claudin family

Clic1 6.24E-61 7.87E-59 -85.25 Encodes chloride channel
Rnf10 4.45E-91 8.89E-89 -75.15 Encodes ring finger protein (ubiquitination)
Mbd3 1.56E-75 2.65E-73 -71.53 Nuclear protein involved methyl-CpG binding
Ube2v2 2.22E-04 1.03E-03 -70.76 Encodes unbiquitin conjugating enzyme
Gsn 3.02E-04 1.37E-03 -70.76
Involved in assembly/disassembly of actin
filaments

137

Importing lists of differentially expressed genes into Ingenuity Pathway Analysis (IPA),
show activation of several signaling pathways (Figure 26B-C). Treatment of high ace triggered
the activation of the CXCR4 signaling pathway, NRF2-mediated oxidative stress response, ILK
signaling, and cholecystokinin/Gastrin-mediated signaling. The TGFb signaling pathway was
predicted to be the most activated upon high ace treatment in spermatogonial cells. As for
treatment with high ibu, EIF2 signaling, the nucleotide excision repair pathway, and integrin repair
pathway were predicted to be activated, with the super pathway of methionine degradation as the
pathway that was predicted to be the most highly activated.
As both high ace and high ibu treatments triggered activation of GTPase related genes in
spermatogonial cells, we were curious to see what other genes were commonly altered between
the two treatments in similar directions. 77 genes were upregulated by both high ace and high ibu
treatments whereas 11 genes were commonly downregulated by both treatments (Figure 26D).
Of the 77 genes that were upregulated in addition to rnd2 and rasl10a include tspo (mitochondrial
translocator protein), mt3 (metallothionein 3), fabp3 (fatty acid-binding protein 3), and fos (Table
5). Downregulated genes include two genes involved in ubiquitination, uqcc2 (ubiquinol-
cytochrome c reductase assembly factor 2) and ube2v2 (ubiquitin conjugating enzyme E2 V2)
amongst others such as scpep1 (serine carboxypeptidase 1), rpl17 (ribosomal protein L17), and
gsn (gelsolin). Both treatments had the same extent of downregulation, which is likely an artifact
of the statistical analysis rather than a true effect. Analytical replication of this experiment is
ongoing and will confirm these findings.

138

Table 5. List of shared differentially expressed genes altered in the same direction with high ace and high ibu
treatment and their respective fold changes and functions.
Gene
name
Fold change (VC
vs. High Ace)
Fold change (VC
vs. High Ibu)
Function
Rnd2 153.32 172.79 Encodes Rho GTPase
Tspo 73.54 72.12 Mitochondrial translocator protein
Mt3 150.25 141.93
Metallothionein (binds heavy metal
ions)
Rasl10
a
119.33 174.64
Involved with GTP binding and
GTPase activity
Rps15
al4
98.67 209.09
Encodes structural component of
ribosomes
Fabp3 2.35 2.71 Encodes fatty acid-binding protein 3
Fos 255.32 88.66
Involved with TF complex AP-1 (Fos-
Jun pathway)
Rpl17 -2.39 -2.22 Encodes ribosomal protein L17
Uqcc2 -69.92 -69.92
Encodes ubiquinol-cytochrome c
reductase assembly factor 2
Scpep1 -69.60 -69.60 Encodes serine carboxypeptidase 1
Ube2v
2
-70.76 -70.76
Encodes unbiquitin conjugating
enzyme
Gsn -70.76 -70.76
Involved in assembly/disassembly of
actin filaments

4.4.5 Characterization of molecular changes in Sertoli cells
As for the Sertoli cell population, comprising of the largest proportion of cells in the testes
at PND8, high ace and high ibu differentially altered 110 and 152 genes, respectively (Figure 27).
The gene that was the topmost upregulated by high ace treatment is Fosb, which encodes Fos
protein of the Jun-Fos signaling cascade implicated in cell proliferation, differentiation,
transformation, suggesting that in Sertoli cells high ace treatment may alter the normal properties
of this cell type (Table 6). Apoe, Apolipoprotein E, involved in the synthesis of lipoproteins and
Foxs1, a member of the forkhead family of transcription factors, comprised of several other genes
that were upregulated with high ace treatment. Amongst the top downregulated genes include
Khdrbs3, involved in RNA binding, Fxyd5, involved in actin binding and regulation of sodium
channels, and Mfsd10, which encodes a member of the major facilitator superfamily of transporter
proteins.

139

Figure 27. Volcano plot illustrating up and downregulated differentially expressed genes in the Sertoli cell cluster with
high ace and high ibu treatment relative to the vehicle control.

Interestingly, both high ace and high ibu treatment had the ability to upregulate Amh
expression to approximately the same extent, ~80-fold. Amh is a glycoprotein produced by
immature Sertoli cells that is responsible for male sex differentiation [280] and regulates the fate
of Sertoli cells by promoting proliferation at low levels and apoptosis at high levels as well as
regulating the Sertoli cell development [281]. Significant increases in the expression of Amh due
to ace and ibu exposure would suggest possible Sertoli cell dysfunction, which could have long-
term effects in germ cells as the primary role of Sertoli cells is to support spermatogenesis.
Other genes impacted by high ibu treatment include Akr1c19, which encodes a member
of the aldo-keto reductase family, Lgmn, which is involved in cell proliferation, and Ripor2,
involved in G protein signaling. Top downregulated genes include Rcan1, involved in calcineurin-
dependent signaling, Rnf10, involved in ubiquitin-protein transferase activity, and Mt2A, which
plays a role in the control of metals in the cell.

Table 6. List of top differentially expressed genes upregulated and downregulated by high ace and high ibu treatment in the Sertoli cell cluster and their respective
p-values, false discovery rate, fold change, and functions.
Gene ID
P-value (VC
vs. High ace)
FDR (VC vs.
High ace)
Fold change (VC vs.
High ace)
Function
Fosb 2.17E-41 7.02E-39 227.12
Encodes Fos proteins implicated as regulators of cell proliferation,
differentiation, and transformation
Apoe 1.23E-47 5.76E-45 81.71 Encodes apolipoprotein E involved in the synthesis of lipoproteins
Foxs1 9.30E-25 1.49E-22 81.22 Encodes a members of the forkhead family of transcription factors
Pmp22 9.26E-73 7.25E-70 79.90 Encodes peripheral myelin protein 22
Amh 2.99E-15 2.63E-13 79.90 Provides instructions for male sex differentiation
Khdrbs3 3.81E-94 4.77E-91 -74.52
Involved in RNA binding, may negatively regulate cell growth and
inhibit cell proliferation
Fxyd5 7.75E-10 3.76E-08 -74.61 Involved in actin binding and sodium channel regulator activity
Mfsd10 1.85E-45 7.41E-43 -75.22
Encodes a member of the major facilitator superfamily of
transporter proteins
Atp6v1d 4.06E-30 7.79E-28 -76.84 Mediates acidification of eukaryotic intracellular organelles
Mrpl43 2.95E-28 5.49E-26 -76.96 Involved in protein synthesis within the mitochondrion

Gene ID
P-value
(VC vs.
High ibu)
FDR (VC vs.
High ibu)
Fold change (VC vs.
High ibu)
Function
Akr1c19 8.95E-202 1.29E-198 165.07 Encodes aldo-keto reductase family 1
Lgmn 4.94E-62 1.16E-59 88.92
Plays a role in the regulation of cell proliferation via its role in
EGFR degradation
Pmp22 1.55E-89 6.08E-87 83.40 Encodes peripheral myelin protein 22
Ripor2 3.80E-93 1.55E-90 83.32
This gene encodes an atypical inhibitor of the small G protein
RhoA
Amh 1.15E-18 8.09E-17 82.81 Provides instructions for male sex differentiation
Tmem184c 2.56E-20 1.93E-18 -78.51 Possible tumor suppressor which may play a role in cell growth
Rcan1 6.03E-23 5.20E-21 -79.25
Interacts with calcineurin A and inhibits calcineurin-dependent
signaling pathway
Rnf10 8.40E-79 2.59E-76 -79.79 Involved in ubiquitin-protein transferase activity
Ppp1r1b 1.48E-25 1.45E-23 -81.59 Inhibitor of protein-phosphatase 1
Mt2A 5.71E-33 6.88E-31 -83.36 Plays role in the homeostatic control of metal in the cell

141

4.4.6 Characterization of molecular changes in Leydig cells
Despite the Leydig cell population constituting only a small proportion of the testes (a large
proportion of Leydig cells have also been removed during processing, and thus constituting an
even smaller percent of total testicular cells in our study), we were able to identify 31 DEGs altered
by high ace treatment and 37 by high ibu treatment (Figure 28). Of those altered by high ace
treatment, Ube2s encodes a member of the ubiquitin-conjugating enzyme family, Psmg3 encodes
a chaperone protein, and Tmsb10 plays an important role inf the organization of the cytoskeleton
(Table 7). Amongst some of the genes downregulated by high ace treatment include Tiprl, which
is a negative regulator of PP2A, Mfsd10, which encodes a member of transporter proteins, and
Krt42, involved in cytoskeletal maintenance.

Figure 28. Volcano plot illustrating up and downregulated differentially expressed genes in the Leydig cell cluster with
high ace and high ibu treatment relative to the vehicle control.

Tmsb10 was the topmost upregulated gene with high ibu treatment and was amongst the
top genes altered by high ace treatment in Leydig cells, suggesting that targeting eicosanoid
synthesis in Leydig cells can affect the cytoskeletal organization of this cell type. Among other top

142

upregulated genes with high ibu treatment include Sparcl1, involved in calcium ion binding, clic2,
which encodes a chloride intracellular channel protein, and Hspe1, which encodes a heat shock
chaperone protein. Amongst some of the downregulated genes include Mrpl43, which is involved
in mitochondrial protein synthesis, Mbd3, which encodes a nuclear protein involved in methyl-
CpG binding, and Rnf10, involved in ubiquitin-protein transferase activity.

Table 7. List of top differentially expressed genes upregulated and downregulated by high ace and high ibu treatment in the Leydig cell cluster and their respective
p-values, false discovery rate, fold change, and functions.
Gene ID
P-value (VC vs.
High ace)
FDR (VC vs.
High ace)
Fold change (VC
vs. High ace) Function
Ube2s 4.54E-19 4.26E-16 103.41
Member of the ubiquitin-conjugating enzyme family
Psmg3 2.34E-14 1.63E-11 92.94
Chaperone protein which promotes assembly of the 20S proteasome
Tmsb10 4.74E-10 1.78E-07 82.45
Plays an important role in the organization of the cytoskeleton
Egr1 5.96E-06 8.18E-04 81.14
Functions as a transcriptional regulator required for differentitation
and mitogenesis
Nt5c3b 1.34E-07 3.06E-05 74.75
Required for nucleotide binding and 5'-nucleotidase activity
Tiprl 3.79E-05 3.67E-03 -73.57
Inhibitory regulator of protein phosphatase-2A (PP2A)
Mfsd10 1.23E-12 6.78E-10 -74.93
Encodes a member of the major facilitator superfamily of transporter
proteins
Mbd3 8.07E-13 4.89E-10 -79.01
Nuclear protein characterized by the presence of a methyl-CpG
binding domain (MBD)
Krt42 3.32E-06 4.95E-04 -79.99
Keratin, invovled in cytoskeletal maintenance
Lgals5 1.23E-15 9.63E-13 -141.29
Galectin-5, involved in carbohydrate binding

Gene ID
P-value (VC vs.
High ibu)
FDR (VC vs.
High ibu)
Fold change (VC
vs. High ibu) Function
Tmsb10 8.20E-19 2.27E-16 92.30
Plays an important role in the organization of the cytoskeleton
Sparcl1 1.27E-10 1.82E-08 89.91
Involved in calcium ion binding
Clic2 7.96E-06 5.32E-04 84.80
Encodes a chloride intracellular channel protein
Sult5a1 2.85E-05 1.68E-03 79.16
Involved in sulfotransferase activity
Hspe1 6.00E-11 9.02E-09 77.31
Encodes a major heat shock protein which functions as a chaperonin
Mrpl43 1.44E-47 2.08E-44 -73.52
Involved in protein synthesis within the mitochondrion
Leprotl1 1.62E-05 1.02E-03 -74.67
Encodes Leptin Receptor Overlapping Transcript Like 1
Mfsd10 3.03E-24 1.27E-21 -74.93
Encodes a member of the major facilitator superfamily of transporter
proteins
Mbd3 5.69E-18 1.49E-15 -79.01
Nuclear protein characterized by the presence of a methyl-CpG
binding domain (MBD)
Rnf10 1.81E-27 8.73E-25 -79.71
Involved in ubiquitin-protein transferase activity

144

4.4.7 Histological characterization of adverse morphological alterations in treated adult
rats
Histological staining was conducted on testicular sections to observe changes in testicular
morphology in the adult animals, which were allowed to age for 3-4 months after 7-day postnatal
exposure of high and low doses of ace and ibu. H&E images of PND90 tissues show greater
vacuolation in tubules with low ibu treatment group compared to controls, which upon
quantification was statistically significant (Figure 29A-B). A trend towards an increase in the
percent of pronounced vacuolation was observed with the low ace treatment group as well. No
observable differences were found with either of the high dose treatment groups.
Immunofluorescent stains of several proteins of interest in the eicosanoid and Notch
signaling pathways were conducted to assess changes in adult testes. There were no strong
differences in the expression of Cox1 between the VC or any of the treatment groups, although
in some tubules, there was an apparent brighter signal in areas of the germinal epithelium, that
will require quantification comparing the same spermatogenic stages (Figure 29C). However, the
expression of Cox2 was clearly highly upregulated in the high ace group and differences were
also observed with the high ibu group (Figure 29D). This suggests that the targeted effects of the
drugs on the inhibition of the eicosanoid pathway at early ages may have induced increases in
Cox2 expression throughout life, and up to adulthood, perhaps as a compensatory mechanism.
Notch3 is another protein of interest to us since it was found to be upregulated in our in
vitro studies with several pharmacological Cox inhibitors (Tran-Guzman et al., in preparation).
Interestingly, Notch3 expression was highly induced by high ibu treatment and slightly increased
with high ace treatment as well (Figure 29E). Though the role of the Notch signaling pathway
remains to be elucidated in the testes, we have also observed overexpression of Notch3 in
testicular cancer patients, which would suggest an involvement of this pathway to the malignant
transformation of germ cells, the cell type of which Notch3 seems to be localized in the testes.
Ongoing studies are being conducted to assess the stages of spermatogenesis that are correlated

145

with Notch3 overexpression, and to further understand the role of this signaling pathway in relation
to treatment with pharmacological Cox inhibitors.

146

147

148

4.4.8 High dose ace and ibu induces testosterone and LH levels in adult rats
To evaluate testicular function, the circulating levels of testosterone and luteinizing
hormone (LH), the pituitary regulating testosterone production, were measured to assess whether
steroidogenesis has been affected by the treatments. All treatments induced increases in
testosterone levels, but only changes with high ace treatment were significant, exhibiting almost
2.5-fold greater levels of serum testosterone compared to controls (Figure 30A). LH levels trended
in the same direction, with high ibu and high ace treatments inducing levels of LH at significant
levels, at approximately 8-fold higher inductions (Figure 30B). Despite an increase in the levels
of testosterone, the levels were not high enough to inhibit LH production, which is characteristic
of constitutively high testosterone levels [282]. Taking together the results from the adult rats, we
can conclude that postnatal administration of ace and ibu at a critical window of spermatogonial
cell establishment can result in alterations in testicular morphology and function long after drug
exposure. Future experiments will aim to assess whether these long-term alterations in testicular
function will affect the reproductive potential of the exposed animals.
Figure 29. Morphological and histological evaluation of treated adult testes. A) H&E staining of PND90 low ibu treated
testis compared to vehicle control illustrating the presence of morphological abnormalities as vacuolization indicated at
the red arrow. B) Quantification of morphologically abnormal tubules relative to total tubules (% degenerated tubules) in
adult rats. Results are presented as average of at least 3 individual animals per treatment group. Significant difference
relative to vehicle control with One way ANOVA test and multiple comparisons: * (p≤0.05), ** (p<0.01), *** (p<0.001). C)
Visualization of Cox1 expression in VC, high ace, and high ibu treated adult rat testes. FITC: Cox1, Blue: DAPI. D)
Visualization of Cox2 expression in VC, low ace, high ace, low ibu, and high ibu treated adult rat testes. FITC: Cox2,
Blue: DAPI. E) Visualization of Notch3 expression in VC, low ace, high ace, low ibu, and high ibu treated adult rat testes.
FITC: Notch3, Blue: DAPI. Scale in µm.

149

Figure 30. Effect of treatment on serum testosterone and luteinizing hormone levels in PND90 rats. A) Serum
testosterone concentration (pg/ml) and B) serum Luteinizing hormone concentrations (mIU/ml) of adult rats treated with
low ace, low ibu, high ace and high ibu relative to VC. Results are presented as average of at least 3 individual animals
per treatment group. Significant difference relative to vehicle control with One way ANOVA test and multiple
comparisons: * (p≤0.05), ** (p<0.01), *** (p<0.001).

4.5 DISCUSSION
To recapitulate the conditions of newborns being exposed to NSAIDs and analgesics
during infancy, we exposed Sprague Dawley rats for the first seven days of life to two commonly
used over-the-counter pharmacological COX inhibitors, ace and ibu. The doses used in this study
reflected human exposure levels of doses that can realistically be reached by infants taking such
drugs on a regular basis or those being hospitalized and are administered these drugs as part of
standard care. Body and organ weights were monitored over the course of the study to ensure
that the doses did not reach levels of toxicity that would compromise the physiological systems of
the animals and confound our findings. We note that the kidney and brain weights were
significantly higher with high ibu treatment, but the differences were slight, at <1g, and thus
unlikely to reflect adverse toxicity. AGD is an androgen sensitive parameter that was measured
to assess apparent reproductive toxicity, as shorter distances between the anus to the base of
the scrotum has been found to be associated with reduced sperm counts and morphology in
rodents and human [283]. AGDs were not altered by the treatments either, despite some
A B

150

increases with high ace and ibu treatments in the earlier ages, further confirming that no obvious
reproductive toxicities were observed with the doses used in this study.
Despite the lack of obvious toxicity observed in weights or via AGD measurements,
morphological changes can be observed in the testicular morphology of the animals upon closer
inspections. The presence of the morphological alterations was greater in the treatment groups
at earlier ages, likely due to the greater sensitivity of the testes to drug insult at PND4. By PND8,
only the low ace and ibu groups were significantly impacted by the treatments, but the presence
of vacuolation constitutes a greater proportion of the total tubules at this age. By PND8, the pups
have been exposed to the full seven days of treatment and the vacuolation observed likely reflect
the toxicological impact of the exposure on critical windows of reproductive development.
Vacuolation may result from accumulation of fluids, lipids, or phospholipids or can be a result of
loss of germ cells. Such toxicity of the testis may be localized to germ cells as a result of Sertoli
cell injury, cytotoxicity, hypoxia or inflammation [284]. Ongoing studies are being conducted in
attempts to identify whether these alterations are localized to specific cell types. Interestingly, the
low doses had a greater impact than the high doses, and effects such as these have been
reported in other toxicity studies [27, 174, 177], likely due to the body’s ability to compensate for
the greater impact of more toxic doses.
To assess the impact of seven days of ace and ibu administration on testicular function,
high ace and high ibu treated PND8 rats were chosen at random to evaluate the effects of the
treatments on isolated cell populations of the testes. The spermatogonia represents a unique
population of cells characterized by markers such as Dazl, Sohlh1, and DDX4/vasa. Amongst
some of the top genes that were differentially altered by the treatments compared to the vehicle
controls include several that are involved in Jun-Fos signaling and G-protein coupled receptor
(GPCR) activity, which are pathways that are well-established regulators of cell growth. Studies
have found that the constitutive activation of these pathways can lead to oncogenic transformation
of cells and subsequent tumor formation as a result of uncontrolled growth [285, 286]. The

151

Ingenuity Pathway Analysis software was used to further assess the signaling pathways that were
predicted to be altered with high ace or high ibu treatment. Several pathways of interest include
CXCR4 and TGFb signaling, predicted to be activated by high ace treatment, and the methionine
degradation pathway, predicted to be activated by high ibu treatment. The TGFb signaling
pathway is of great interest to us as it is a pathway both found to be activated from our in vitro
studies and this in vivo study by pharmacological Cox inhibitors. As discussed in detail in the prior
chapter, TGFb is critical for embryogenesis and SSC differentiation, and has also been found to
be implicated in oncogenic signaling through interaction with PGE2 [261-264].
As both ace and ibu target the eicosanoid pathway, we were interested to see whether
there would be any overlapping effects of both treatments. 77 genes were upregulated between
both treatments and 11 genes were downregulated. Amongst some of the genes that were
upregulated, several were involved in GPCR and Jun-Fos signaling and methionine degradation
important for maintaining cellular homeostasis. Those that were downregulated include several
genes involved in ubiquitination pathways important for DNA repair, apoptosis, and maintaining
the protein levels in the cell [287, 288].
Sertoli and Leydig cell-specific effects are of interest to us as well, and our data suggests
that the germ cells are not the only population that can be targeted by NSAIDs and analgesic
drugs in the testes. One finding of much significance is the upregulation of Amh independently by
both high ace and high ibu treatment. Several studies conducted in human testes explants also
found that ibu had the ability to alter AMH hormone levels with fetal exposure [82] and adults
[128]. In these studies, AMH levels were inhibited, which is in contrast to what was observed in
our studies. However, these studies have been conducted on whole testes tissues that were
exposed to the drugs during adulthood and fetal development, suggesting that the administration
and dosing window may be critical to cellular responses. Furthermore, current reports have only
evaluated ibu effects on AMH levels, which may not necessarily mimic the changes in gene
expression. Nonetheless, our finding still remains significant in that it would suggest potential

152

changes in testicular development and function, which is regulated by AMH production by Sertoli
cells [289].
To characterize long-term effects of the postnatal treatments on male reproductive
development, pups were allowed to grow into adulthood and were sacrificed between 90-100
days for morphological and functional assessments of reproductive health. Morphological
characterization of the testes shows the presence of vacuolation, which upon quantification was
significantly upregulated with low ibu treatment. As these rats have only been exposed to high ibu
at the first seven days of life, morphological alterations such as these suggest that targeting critical
stages of reproductive development during infancy can lead to long-term adverse effects on the
testes. Further assessments into the expression of major eicosanoid rate-limiting enzymes Cox1
and Cox2 showed that while there were no obvious differences on the treatments on Cox1
expression, these drugs induced the expression of Cox2 quite dramatically. Despite the ability of
these drugs to inhibit Cox2 activity as part of their mechanism of action, the overexpression of
this enzyme in the adults would suggest a compensatory effect, which can be observed long after
the rats of been exposed to the drug. Cox2 overexpression has been implicated in testicular
cancer and is positively associated with other male reproductive toxicities [68, 69], so it would be
of interest to further assess how the overexpression of eicosanoid signaling would affect testicular
function. One possible consequence of Cox2 increased expression in testis could be an
exacerbated sensitivity of the germ cells to subsequent chemicals, drugs, or microbial exposure
of the animals, in view of Cox2 involvement in inflammatory processes.
Another protein of interest to us is Notch3, which has been shown to be upregulated in
our in vitro Cox inhibitor studies. PND90 rats exposed for their first week of life to high ibu were
found to highly overexpress Notch3, consistent with our in vitro findings. Although this effect does
not seem to be as pronounced with high ace treatment, a similar trend can be observed. As
Notch3 was also found to be overexpressed in testicular cancer tissues, the data presented here
adds to the weight of evidence supporting our hypothesis that Notch3 signaling can be activated

153

upon Cox inhibition and is likely involved in the cellular transformation that could contribute to
testicular cancers. This study is ongoing, as our next objective is to confirm the localization of
Notch3 in the testes, which we hypothesize is likely confined to the germ cell population, and to
assess the stage of spermatogenesis that is correlated with the highest expression of this
receptor. One study has followed the expression of Notch3 through testicular development and
concluded that it is localized in postnatal and prepubertal stages of spermatogenesis, so we will
focus our initial efforts at these earlier stages as well [257].
Lastly, hormonal assessments were conducted to evaluate overall testicular health on a
functional level. As normal steroidogenesis is important for healthy testicular function, alterations
in testosterone and LH levels in the blood can indicate abnormalities in testicular function. LH is
particularly important during late fetal development as it is the main stimulus for testosterone
production, responsible for testicular descent and penile growth [290]. We report that exposure
during the first week after birth to high ace significantly increased testosterone production in the
adult, with trends towards increases with other treatments as well, and both high ace and ibu had
an ability to induce LH levels. While adverse male reproductive conditions are often attributed to
low testosterone and LH levels, high testosterone and LH levels can also indicate dysregulation
of hypothalamic-pituitary-testicular axis, as indicated by symptoms of an infertile patient with
elevated serum testosterone [282]. Taken together, adverse effects on testicular morphology and
function can be observed in adult rats exposed to human-relevant doses of ace and ibu during
the first postnatal week, at a period corresponding to infancy in human based on the timing of
SSC formation. However, whether these findings can be related to infertility remains to be
determined.
To our knowledge, this is the first report that exposure in the first postnatal week to
commonly used OTC drugs ace and ibu can have immediate adverse effects on testicular
morphology and molecular signaling pathways on the different cell types of the testes. These
adverse events remain in effect long-term as damaged seminiferous tubules, overexpression of

154

eicosanoid and Notch pathway genes, and induction of testosterone and LH hormone levels that
were significantly altered in adults. While this study is a work in progress, as we understand the
limitation of having only N=1 sample for single cell sequencing and our immunohistochemical
images, the multiple approaches taken to assess overall reproductive function supports our
hypothesis that targeting eicosanoid biosynthesis during critical stages of spermatogenesis can
affect overall reproductive health.
Future studies will continue to decipher the roles that eicosanoid and Notch signaling
pathways can play in spermatogonial development and evaluate how targeting these pathways
can adversely affect male reproductive function and potential malignant transformation. We will
also further explore the functional endpoints of sperm quality and number to assess whether these
consequences can result in reduced capacity of the adults to produce offspring. Overall, the work
described here supports the growing body of evidence in the reproductive field reporting that
pharmacological Cox inhibitors can adversely affect male reproductive health when administered
in early postnatal ages. Therefore, caution must be taken to closely monitor the doses of these
drugs when given to infants, which could potentially be contributing to the trends of poor male
reproductive health that have been increasing in modern times.

4.6 SUMMARY
In this chapter, we conducted an in vivo study to assess whether consumption of pharmacological
COX inhibitors ace and ibu could affect long-term reproductive health consistent with in vitro
findings. Despite no changes in AGD at the different stages of development, changes in testicular
morphology were observed after 4 and 7-days of treatment as well as alterations in gene
expressions of various cell types in the testes at PND8. Adverse effects in testicular morphology
were observed in the low ibu treatment group at adulthood as well as inductions in testosterone
and LH levels with high ace and high ibu treatment groups. Overall, these findings support our

155

hypothesis that pharmacological COX inhibitors can adversely affect the proper development of
testis and spermatogonial stem cell.

4.7 ACKNOWLEDGEMENTS
The authors would like to thank Amina Khan for her involvement in the histological work on this
project and Chantal Sottas for her involvement in the planning and aiding with the execution of
the animal study. We would also like thank the bioinformatics team at USC Norris library in aiding
with our scRNAseq analysis and the USC Norris Molecular Genomics Core for sequencing our
samples. Lastly, we would like to thank the Translational Research lab at the USC School of
Pharmacy for sectioning the tissue samples and conducting the H&E stains for us.

156

CHAPTER 5: SUMMARY, CHALLENGES, AND FUTURE PERSPECTIVES

5.1 SUMMARY
The objective of this project was to investigate whether the eicosanoid biosynthetic
pathway could regulate spermatogonial stem cell development. We addressed this objective
using three specific aims to characterize and understand the role of this pathway behind
spermatogonial development using both in vitro and in vivo models. Firstly, we characterized the
eicosanoid biosynthetic pathway in a well-established cellular model of the SSC. Next, we
evaluated cellular effects of inhibiting eicosanoid signaling using shRNA and pharmacological
COX inhibitors. Lastly, we conducted an animal study and assessed whether any immediate or
long-term adverse effects on male reproductive development can be observed. These three
specific aims have allowed us to obtain a comprehensive picture of the role of eicosanoid
biosynthesis on male reproductive development in rodent models. Despite our studies being
carried in the rodent, the doses we have chosen for these studies are human-relevant doses. The
relevancy of our findings suggests that careful consideration must be taken when administering
pharmacological Cox inhibitors to infants.

5.1.1 Summary of Chapter 2
In our first project described in Chapter 2, we sought to characterize the eicosanoid
pathway in the C18-4 cell line and determine whether this pathway can be modulated by the EDCs
genistein and MEHP. Prior to beginning our studies, we validated the differentiation potential of
C18-4 cells by stimulating the cells with RA. We also found that RA could alter Cox1 and Cox2
mRNA expression, giving us the first hint that these enzymes are actively regulated during SSC
differentiation. We further characterized the mRNA and protein expression levels of major
eicosanoid pathway components in the C18-4 cell line. We found that the expression levels were

157

similar to that of isolated primary spermatogonia from rats and verified that the C18-4 cells were
able to produce detectable levels of PGD2, PGE2, and PGF2a.
Next, we aimed to determine whether synthesis of PGs by SSCs can regulate its own self-
renewal and differentiation, and whether these processes can be disrupted by two common
endocrine disruptors, the soy phytoestrogen GEN and the phthalate plasticizer metabolite MEHP,
to which babies are exposed. We found that GEN and MEHP, alone and in mixtures, could
uniquely alter expression of eicosanoid pathway enzymes and PG release in C18-4 cells after 24
hours of treatment. Our studies were first conducted in media containing regular FBS conditions
and then replicated in CS- FBS conditions to study eicosanoid signaling without the interference
of confounding effects from PGs inherent to regular FBS. Interestingly, the switch in regular FBS
to CS-FBS impacted the expression of several prostaglandin synthases. However, regardless of
the FBS type used, we observed GEN-driven inductions of Pla2 and Cox2, and inhibition of Cox1
mRNA. 100 μM GEN was able to upregulate secretion of PGF2a, PGE2, and PGD2. The modest
increase in PGE2 may be attributed by high retention of PGE2 inside cells treated with 100 μM
GEN and GEN+MEHP.
Changes in PG synthases and PGs induced by 100 μM GEN and GEN+MEHP showed
similar patterns as changes in markers of committed spermatogonia progenitors, Foxo1 and
Mcam, and spermatogonial differentiation marker Stra8, suggesting that GEN and MEHP may
induce premature differentiation of SSCs. The correlation between GEN-driven upregulation of
eicosanoid pathway genes and differentiation markers could suggest that disruption of SSC
differentiation by GEN is likely mediated by greater eicosanoid production via induction of Cox2
activity. This supports the hypothesis that these processes may be interrelated in SSCs,
suggesting a novel function of the eicosanoid pathway and PGs in the regulation of SSC
development.
In contrast to GEN-driven effects on eicosanoid and SSC differentiation genes from GEN-
MEHP treatments, MEHP-alone treatments had little effects on changes in the expression of

158

eicosanoid pathway components, but still upregulated PGD2 levels, induced expression of
differentiation marker Kit, and enhanced cell viability. We also observed synergistic effects with
100 μM GEN+MEHP, which induced PGD2 production to a greater extent than GEN or MEHP
alone. Overall, this suggests that EDCs and EDC mixtures can affect eicosanoid biosynthesis,
which may also involve complex interactions on multiple signaling pathways that have yet to be
fully understood.

5.1.2 Summary of Chapter 3

Following Chapter 2, where we reported for the first time that SSCs express major
eicosanoid pathway components and produce PGs, we concluded that the eicosanoid pathway
may be important for SSC development and PGs could regulate the differentiation processes of
these cells. Thus, we take this study one step further in Chapter 3 by using gene silencing and
pharmacological approaches to further understand the mechanisms behind the role of eicosanoid
and SSC development. We were interested in deciphering the role of the COX enzymes in the
C18-4 cell line using pharmacological Cox inhibitors that are selective for Cox1, Cox2, or could
target both enzymes non-selectively. Confirmatory studies were first conducted to validate that
PG production was reduced upon treatment with pharmacological COX inhibition. Next, we
evaluated the drugs’ abilities to alter differentiation and proliferation. Both 10 μM NS398 and 100
μM ace had the ability to slightly induce proliferation, whereas 50 μM celecoxib and FR122047
were inhibitory to growth. While we speculated that Cox2 activity was correlated to differentiation
in Chapter 2, here we definitively showed that inhibition of Cox2 with 100 μM ace and 100 μM
NS398 was associated with blocking differentiation, providing definitive evidence that the
eicosanoid pathway is important for regulating SSC development.
To take a deeper dive into the mechanisms behind the COX enzymes’ role in SSC
development, we designed a Cox1 knockdown cell line (C18-4
Cox1-KD1
). Cox1 knockdown of C18-
4 cells resulted in significant decreases in cellular adhesion and induction of differentiation,

159

suggesting that Cox1 is important for maintaining cells at an undifferentiated state, which opposes
the role of Cox2. Subjecting the cells to total RNAseq showed activation of several pathways
including the epithelial-mesenchymal transition pathway, inhibition of matrix metalloproteases,
and inhibition of prostanoid synthesis. Amongst the signaling pathways associated with EMT,
Wnt, TGFb, and Notch pathways were the top ones activated, which we were able to further
validate using qPCR.
Next, we decided to study the Notch3 signaling pathway in greater detail as both Notch3
and its downstream target, Hes1, were upregulated in C18-4
Cox1-KD1
cells. Notch3 and Hes1 mRNA
expression was consistently upregulated with pharmacological Cox inhibitors NS398, celecoxib,
FR122047, ace and ibu. Depending on the pharmacological inhibitor used, Notch1 and Hey1
expression were differentially altered. On the protein level, only 100 μM NS398 and 10 μM
FR122047 were able to significantly increase Notch3 expression, which was not observed with
celecoxib.
Interactions between eicosanoid and Notch pathways may be mediated by PGs, and so
we next evaluated the effect of PGD2 treatment on mRNA expression of components of the Notch
signaling pathway. Increasing concentrations of PGD2 dose dependently inhibited Notch3 and
Hes1, whereas only the 10 μM dose was inhibitory to Notch1 and Hey1 expression. Interestingly,
50 μM PGD2 induced Hey1 mRNA levels. Thus, it is likely that PGD2 is a negative regulator of
Notch3 expression, and that pharmacological Cox inhibitors are able release the PGD2 blockade
on Notch3 and enhance its activation.
Constitutive activation of the Notch signaling pathway has been implicated in cancers, and
we report that NOTCH3 is upregulated in human testicular cancer samples. Gamma secretase
gene APH1B was also found to be decreased, suggesting that dysregulation of the Notch
signaling pathway may be a hallmark of testicular cancer. With support from our in vitro studies
of Cox1 knockdown cells resulting in activation of EMT (decreases in expression of cellular
adhesion and matrix metalloproteinase genes and predicted activation of EMT using IPA), we

160

speculate that that blocking eicosanoid signaling could disrupt the proper SSC development such
that it may be inducing the malignant transformation of the germ cells. Taken together, the results
from Chapter 3 take our investigations from our first study one step further, confirming that
eicosanoid signaling is important for spermatogonial development. The Cox enzymes play a role
in the differentiation of SSCs, and PG regulation of Notch pathway may be important for the proper
development and fate determination of germ cells, which when disrupted, could result in germ cell
malignancy.

5.1.3 Summary of Chapter 4

In light of our in vitro findings from Chapter 2 and 3, Chapter 4 focuses on the animal study
designed to answer the question of whether postnatal administration of human-relevant doses of
commonly used pharmacological COX inhibitors ace and ibu can alter testicular development.
Animals were treated with high doses of ace (1.4 mg/day) and ibu (0.7 mg/day) and low doses of
ace (0.7 mg/day) and ibu (0.4 mg/day) for the first 7 days of life and testes were evaluated at
PND4, PND8, and PND90 at adulthood. We monitored animal and organ weights throughout
study duration to ensure that the doses of used for treatment did not reach levels that would result
in systemic toxicity. We did not observe any effects of the treatments on AGD either.
Assessment of testicular morphology showed the presence of pronounced vacuolation
with all treatment groups at PND4 and with low ace and low ibu at PND8, suggesting that the
treatments had adverse effects after 4 days and up to 7 days in the testes. High ace and high ibu
treated PND8 testicular single cell suspensions were randomly chosen for further analysis using
scRNAseq. The presence of major testicular cell types Sertoli cells, SSCs/spermatogonia, Leydig
cells, macrophages, and peritubular myoid cells were recovered in single cell suspensions upon
evaluation of cellular expression profiles of several cell type specific markers.
In the spermatogonial cell cluster, genes that were upregulated with both the treatments
compared to VC included genes involved with GTPase activity. Genes that were differentially

161

altered with high ace treatment include Jun-Fos signaling genes, those involved in heavy metal
binding, protease inhibition, and nuclear acid binding. Alternatively, genes that were differentially
altered with high ibu treatment include those that regulate activity of potassium channels, tight
junctions, chloride channels, and ubiquitination. Pathways that are predicted to be activated by
high ace treatment include the CXCR4 signaling pathway, NRF2-mediated oxidative stress
response, and TGFb signaling. EIF2 signaling, the nucleotide excision repair pathway, and
integrin repair pathway are amongst those that were predicted to be activated with high ibu
treatment. Commonly altered genes between both treatments include 77 upregulated genes and
11 downregulated genes. Genes that were commonly upregulated included tspo (mitochondrial
translocator protein), mt3 (metallothionein 3), and fabp3 (fatty acid-binding protein 3). Amongst
those that were commonly downregulated between both treatments include two genes involved
in ubiquitination, uqcc2 (ubiquinol-cytochrome c reductase assembly factor 2) and ube2v2
(ubiquitin conjugating enzyme E2 V2) amongst others such as scpep1 (serine carboxypeptidase
1), rpl17 (ribosomal protein L17), and gsn (gelsolin).
Characterization of differentially expressed genes altered in the Sertoli cell cluster with
high ace treatment compared to VC include several genes involved in Jun-Fos signaling,
synthesis of lipoproteins, RNA binding, and sodium channel regulation. Genes altered by high ibu
treatment include several genes involved in cell proliferation, G protein signaling, calcineurin-
dependent signaling, and ubiquitin-protein transferase activity. Both treatments had ability to
upregulate Amh expression, which could indicate Sertoli cell dysfunction as AMH levels are
important for regulating Sertoli cell development. As Sertoli cells support spermatogenesis,
changes in Amh expression could have indirect adverse consequences for the germ cells as well.
As for the Leydig cell population, high ace treatment alters expression of genes involved in
ubiquitin conjugation, protein chaperone, and cytoskeletal organization. High ibu treatment also
alters cytoskeletal organization as well as calcium ion binding, regulation of chloride channels,
mitochondrial protein synthesis, and methyl-CpG binding.

162

Overall, the treatments were able to independently alter the expression of genes relative
to VC in spermatogonia, Sertoli and Leydig cell clusters that could have ramifications to the
functional properties of the various cell types at PND8. So, we next wanted to assess whether the
changes observed in short-term had consequences for long-term male reproductive health. Rats
were allowed to grow to adulthood after 7-day treatment, where they were sacrificed at
approximately 90-100 days to evaluate testicular morphology. Significant vacuolation was
observed with low ibu treatment compared to VC, and a trend towards an increase was observed
with low ace treatment as well, though not statically significant. IF staining were conducted on
proteins of interest representing eicosanoid and Notch pathways, Cox1, Cox2, and Notch3. No
clear changes were observed in Cox1 expression, but Cox2 expression was upregulated in the
high ace and high ibu groups compared to VC. Notch3 was also highly upregulated with the high
ibu group compared to the VC, in support of our in vitro findings.
As for functional assessments of testicular health, we measured hormone levels of
testosterone and LH. High ace treated animals had significantly greater levels of testosterone
compared to VC, and both high dose treatments were able to induce LH levels in adults as well.
High testosterone and LH levels can signify dysregulation of the hypothalamic-pituitary-testicular
axis, and it would be interesting to see whether such abnormalities would have adverse effects
on fertility potential of the adults after first postnatal week exposure.
Taken together, both our in vitro and in vivo data support an important role of the
eicosanoid pathway in regulating spermatogonial stem cell development. Administration of
pharmacological COX inhibitors directly to C18-4 cells affected their ability to differentiate properly
and altered signaling pathways that are important to regulating cell fate. When rats were treated
with these drugs during critical stages of spermatogonial stem cell formation, we observed
immediate adverse effects in signaling pathways regulating growth and proliferation of germ cells
in addition to the functionality of other testicular cell types as well. As for long-term consequences,
treated adult testes were characterized by adverse phenotypes, alterations in Cox2 and Notch3

163

expression, and upregulation of testosterone and LH levels. The weight of evidence presented
here thus far suggests that the eicosanoid pathway is important for the normal development of
the spermatogonial stem cell and pharmacological COX inhibitors targeted to this cell type or the
testis could result in abnormal germ cell development, affecting not only spermatogenesis, but
the health of the overall testes as well.

5.2 CHALLENGES
5.2.1 Translatability between rodent models to human

As the majority of the work presented here has been conducted in rodent models, a
challenge of this project is the translatability of the findings to human models of disease. Currently,
the rodent remains the ideal model to study reproductive biology due to limited access to human
testicular samples for our study. Furthermore, as described in Chapter 1.3, the human and rodent
spermatogenic cycles are similar, with both cycles comprising of phases of quiescence,
proliferation, migration, and differentiation. Given the limited accessibility to available human-
derived samples, we chose to study a well-established cell line and attempted to validate our
findings in an in vivo model. However, the rodent and human are biologically unique species, with
slight differences in their reproductive systems, and so we acknowledge that our findings may not
directly translate to the human. Future studies can aim to address this issue by designing
epidemiological studies to follow the reproductive health of infants with high exposure to ace and
NSAIDs over time or take advantage of advances in human spermatogonial stem cell culture to
replicate studies in a human cell-derived model [291].

5.2.2 Cox2 knockdown C18-4 model for comparison
Another challenge of our study is the lack of a Cox2 knockdown study for comparison. In
both Chapters 2 and 3, our results suggest that Cox2 is positively correlated with differentiation
based on treatment of the C18-4 cell line with EDCs and pharmacological COX inhibitors.

164

Generating a Cox2 knockdown model similarly to what has been done for Cox1 would have aided
to definitively confirm this effect and provide greater insight into the differences between the roles
of Cox2 and Cox1 in the spermatogonial stem cell. We utilized two pharmacological Cox inhibitors
that are selective to Cox2, NS398 and celecoxib, to resolve the issue of a lack of a Cox2 specific
knockdown cell line.

5.2.3 Defining a role of Notch3 to the malignant transformation of SSCs
While we were able to illustrate that the Notch pathway is downstream of eicosanoid
signaling in SSCs, we could only speculate on the role of Notch3 in relation to the fate
determination of germ cells as we have not designed specific experiments to explore its effects.
In Chapter 3.5, we reported on several studies that showed the Notch pathway’s involvement with
EMT and cellular transformation of various cancers. As for future experiments that could be
designed to address this limitation, treating the C18-4 cell line with gamma secretase inhibitors
and evaluating changes in markers of germ cell tumors or knocking down Notch3 in the cells and
evaluating its proliferation may aid in this exploration.

5.3 FUTURE PERSPECTIVES
5.3.1 Contribution to the field of reproductive biology and toxicology
Despite the growing body of knowledge investigating the adverse effects of
pharmacological Cox inhibitors on male reproductive development, the mechanisms behind their
roles are just beginning to be revealed. As reviewed in great detail in Chapter 1, studies evaluating
the effects of NSAIDs and analgesic drugs have been primarily limited to investigating fetal and
early neonatal time points prior to the development of SSCs in the testes. Our study contributed
to the field in that it was the first to report the presence of a functional eicosanoid pathway in
SSCs, and we have since shown that this pathway can play important roles in regulating cellular
differentiation, maturation, and overall testicular function. When eicosanoid signaling was

165

disrupted with genetic modifications or via pharmacological Cox inhibitors, we observed changes
in differentiation, alterations in molecular pathways regulating growth, and toxicity in testicular
morphology. Whether these adverse effects could affect reproductive potential remains to be
elucidated. Overall, our work provides a greater understanding of the impact that NSAID and
analgesic drugs can have on SSC development, which is significant to advancing the fields of
reproductive biology and toxicology. Given how commonly infants and children are being exposed
to antipyretic/analgesic drugs, there is a great need for studies such as this to combat the
reproductive pathologies that are on the rise in the developed world.

5.3.2 Role of Notch signaling in spermatogonial development
As discussed in Chapter 3.5 and briefly in Chapter 5.2.3, the role of Notch3 and the Notch
signaling pathway remains to be elucidated in SSC development. While we were not able to study
the function of Notch3 and its downstream effectors in greater detail, we did observe its
upregulation in human testicular cancer tissues and interestingly, in testes of rats treated with
high ibu from our in vivo study. This pathway is implicated in cancer and the constitutive activation
of the Notch pathway has been found to be prognostic in several cancer types [253, 254]. While
it is currently not well established in the development of SSCs and germ cells, it has been found
to be important in Sertoli cells for the regulation of RA production to support spermatogenesis
[259, 260]. Future work that can be conducted to further explore the role of this pathway could be
to silence Notch3 in our in vitro model to investigate its effects on properties such as growth,
differentiation, and changes in the expression of spermatocytic seminoma markers, which could
signify malignant transformation of these cells.
We have preliminary data to also suggest crosstalk between Notch3 and TGFb signaling
pathways, which is another pathway involved in regulating cell growth and proliferation [292].
Interestingly, TGFb signaling was also found to be activated by pharmacological Cox inhibitors in
C18-4 cells as well as with high ace treatment in rat spermatogonial cells in our in vivo study, and

166

therefore can add another layer of complexity in our understanding of the interplay between the
role of eicosanoids and COX enzymes in spermatogonial growth and development. Treatment of
C18-4 cells with DAPT, a y-secretase inhibitor which blocks the signaling downstream of the
Notch receptors, was able to dose-dependently decrease the expression of Tgfb3 with trends
towards decreases in Tgfb2 and Smad3 in our preliminary investigations into the role of Notch3
(data not shown). Therefore, future studies could aim to decipher the effects of
COX/PGD2/Notch/TGFb signaling axis on cellular growth and function of SSCs.

5.3.3 Interaction between Notch and eicosanoid signaling
In Chapter 3.4.7 we report that supplemental PGD2 could negatively regulate Notch3 and
Hes1 expression, and to an extent Notch1 and Hey1 after 24 hours of treatment in C18-4 cells.
We were only able to conduct qPCR analysis and have yet to validate these results in immunoblot
assays. Given the extent of the decreases in the mRNA expression of Notch3, with approximately
50% downregulation with 50 uM PGD2, we can expect to see similar results in protein studies.
Future studies could continue to follow this study by exploring direct interactions between
prostaglandin types and Notch receptors. Though we only evaluated the effects of PGD2, we
acknowledge that PGE2 and PGF2a are also produced by SSCs and may have similar or
contrasting actions on Notch signaling.
Furthermore, studies could be done to evaluate in greater detail how PGD2 is inhibiting
expression levels of the Notch3 receptor. PGD2 is acting as a brake on Notch3, which gets
released when eicosanoid biosynthesis is blocked by pharmacological Cox inhibitors. So PGD2
may directly interact with Notch3, but it is more likely that PGD2 is acting on regulators of the
signaling pathway such as gamma secretase, which is responsible for the activation of the
pathway. Perhaps PGD2 negatively regulates the activity of gamma secretase during basal
conditions, and Cox inhibition can deplete PGD2 levels and release a block on gamma secretase,
which would lead to the activation of the Notch signaling cascade. Experiments that can be

167

designed to evaluate these potential interactions include evaluating changes in gamma secretase
activity with PGD2 supplementation and subsequent changes in the expression of different
components of the Notch pathway. Exploring the direct interactions between PGD2 and gamma
secretase or other Notch pathway components using immunoprecipitation or similar assays could
also be used to definitively show the direct interactions between the two signaling pathways in
support of our hypothesis.

5.3.4 Further exploration of pathways revealed by scRNAseq
scRNAseq analysis of our in vivo study has revealed numerous numbers of differentially
expressed genes and altered pathways in PND8 rat pups treated with high ace and high ibu for
the first seven days of life. We recognize that the replicates of this study are currently being
processed, which may reveal new sets of DEGs and altered signaling pathways. Nonetheless,
the current data set can still guide future explorations for this project. Of several signaling
pathways to investigate in greater detail is the TGFb signaling pathway, which is activated by high
ace treatment in spermatogonial cells is of great interested to us as discussed in the prior section.
Another intriguing direction of exploration is the change in Amh expression by both high
ace and high ibu treatment in Sertoli cells, particularly because AMH is important for regulating
the fate of Sertoli cells, and as a result, is also important for the development of germ cells. The
ability of both drugs to induce Amh expression would suggest that this likely a consequence of
blocking eicosanoid synthesis. Ongoing studies in our lab is focusing on the effects of ace and
ibu on Sertoli cell function and can further confirm this effect in the in vitro TM4 immature Sertoli
cell model and in primary isolated Sertoli cells from rats.
While also still preliminary, IF staining in PND90 testicular tissues illustrates upregulation
of Notch3 with several of our drug treatments, particularly with high ibu, which is also consistent
with our in vitro findings discussed in the previous section. Whether the overexpression of Notch3
can affect testicular function and reproductive potential in the adult rats remain to be explored.

168

Sperm quality and count could also be evaluated to assess whether spermatogenesis has been
adversely affected. Future studies can be designed to pair the treated male rats with females to
evaluate the reproductive health of the litters and in future generations.

5.3 CONCLUSION
Taken together, our review of the current state of understanding behind the eicosanoid
biosynthetic pathway and effects of pharmacological Cox inhibitors on male reproductive
development revealed that studies conducted in spermatogonial stem cells were lacking.
Therefore, we were able to address a gap in knowledge by studying the effects of NSAID and
analgesic drugs on the SSCs. We showed in several independent studies that SSCs express a
functional eicosanoid pathway and that COX inhibitors can alter SSC differentiation and pathways
regulating cell growth. Then we validated these findings using an in vivo model to illustrate that
while these drugs have immediate effects on altering cellular function, in the long-term, they also
could induce potentially toxic effects on testicular morphology and function. More importantly, this
work, though not completely conclusive, has opened many avenues of exploration for future and
necessary studies into the effects of blocking eicosanoid synthesis on spermatogenic
development. We have only touched the surface of the effects of these drugs on overall testicular
function, steroidogenesis, and other important cell types in the testes. So, it remains to be
determined whether these findings can translate to human subjects, and whether regulatory
actions will need to be taken in order to curtail the administration of NSAIDs and analgesic drugs
to infants to combat rising instances of male reproductive pathologies and infertility.

169

REFERENCES

[1] N.E. Skakkebaek, E. Rajpert-De Meyts, G.M. Buck Louis, J. Toppari, A.-M. Andersson, M.L.
Eisenberg, T.K. Jensen, N. Jørgensen, S.H. Swan, K.J. Sapra, Male reproductive disorders and
fertility trends: influences of environment and genetic susceptibility, Physiological reviews 96(1)
(2016) 55-97.
[2] A.M. Andersson, K. Bay, H. Frederiksen, N. Skakkebaek, Endocrine disrupters: we need
research, biomonitoring and action, Wiley Online Library, 2016.
[3] K. Boisen, M. Kaleva, K. Main, H. Virtanen, A. Haavisto, I. Schmidt, M. Chellakooty, I.
Damgaard, C. Mau, M. Reunanen, Difference in prevalence of congenital cryptorchidism in
infants between two Nordic countries, The Lancet 363(9417) (2004) 1264-1269.
[4] A. Mehta, A.K. Nangia, J.M. Dupree, J.F. Smith, Limitations and barriers in access to care for
male factor infertility, Fertility and sterility 105(5) (2016) 1128-1137.
[5] H. Levine, N. Jørgensen, A. Martino-Andrade, J. Mendiola, D. Weksler-Derri, I. Mindlis, R.
Pinotti, S.H. Swan, Temporal trends in sperm count: a systematic review and meta-regression
analysis, Human reproduction update 23(6) (2017) 646-659.
[6] G. Bandoli, K. Palmsten, C. Chambers, Acetaminophen use in pregnancy: Examining
prevalence, timing, and indication of use in a prospective birth cohort, Paediatric and perinatal
epidemiology 34(3) (2020) 237-246.
[7] D.M. Kristensen, U. Hass, L. Lesné, G. Lottrup, P.R. Jacobsen, C. Desdoits-Lethimonier, J.
Boberg, J.H. Petersen, J. Toppari, T.K. Jensen, Intrauterine exposure to mild analgesics is a
risk factor for development of male reproductive disorders in human and rat, Human
reproduction 26(1) (2011) 235-244.
[8] G.S. Berkowitz, R.H. Lapinski, Risk factors for cryptorchidism: a nested case-control study,
Paediatric and perinatal epidemiology 10(1) (1996) 39-51.
[9] M.S. Jensen, C. Rebordosa, A.M. Thulstrup, G. Toft, H.T. Sørensen, J.P. Bonde, T.B.
Henriksen, J. Olsen, Maternal use of acetaminophen, ibuprofen, and acetylsalicylic acid during
pregnancy and risk of cryptorchidism, Epidemiology (2010) 779-785.
[10] FDA, FDA recommends avoiding use of NSAIDs in pregnancy at 20 weeks or later because
they can result in low amniotic fluid, in: U.F.a.D. Administration (Ed.) 2020.
[11] S. Siu, J. Yeung, T. Lau, A study on placental transfer of diclofenac in first trimester of
human pregnancy, Human Reproduction 15(11) (2000) 2423-2425.
[12] G.G. Briggs, R.K. Freeman, S.J. Yaffe, Drugs in pregnancy and lactation: a reference guide
to fetal and neonatal risk, Lippincott Williams & Wilkins2012.
[13] M. Culty, Gonocytes, the forgotten cells of the germ cell lineage, Birth Defects Res C
Embryo Today 87(1) (2009) 1-26.
[14] C.M. Spiller, J. Bowles, Sex determination in mammalian germ cells, Asian J Androl 17(3)
(2015) 427-32.
[15] H. Li, G. MacLean, D. Cameron, M. Clagett-Dame, M. Petkovich, Cyp26b1 expression in
murine Sertoli cells is required to maintain male germ cells in an undifferentiated state during
embryogenesis, PLoS One 4(10) (2009) e7501.

170

[16] T.M. Plant, A.J. Zeleznik, Knobil and Neill's physiology of reproduction, Academic
Press2014.
[17] G. Manku, M. Culty, Mammalian gonocyte and spermatogonia differentiation: recent
advances and remaining challenges, Reproduction 149(3) (2015) R139-R157.
[18] T.L. Gaskell, A. Esnal, L.L.L. Robinson, R.A. Anderson, P.T.K. Saunders,
Immunohistochemical profiling of germ cells within the human fetal testis: identification of three
subpopulations, Biology of reproduction 71(6) (2004) 2012-2021.
[19] R.P. Piprek, Molecular mechanisms of cell differentiation in gonad development,
Springer2016.
[20] M.D. Griswold, Spermatogenesis: The Commitment to Meiosis, Physiol Rev 96(1) (2016) 1-
17.
[21] H. Chen, D. Mruk, X. Xiao, C.Y. Cheng, Human spermatogenesis and its regulation, Male
Hypogonadism, Springer2017, pp. 49-72.
[22] X. Gu, S.Y. Li, T. DeFalco, Immune and vascular contributions to organogenesis of the
testis and ovary, FEBS J (2021).
[23] A. Meinhardt, M. Wang, C. Schulz, S. Bhushan, Microenvironmental signals govern the
cellular identity of testicular macrophages, J Leukoc Biol 104(4) (2018) 757-766.
[24] J.C. Hutson, Changes in the concentration and size of testicular macrophages during
development, Biology of reproduction 43(5) (1990) 885-890.
[25] J.C. Hutson, Development of cytoplasmic digitations between Leydig cells and testicular
macrophages of the rat, Cell and tissue research 267(2) (1992) 385-389.
[26] S. Jones, A. Boisvert, S. Francois, L. Zhang, M. Culty, In utero exposure to di-(2-ethylhexyl)
phthalate induces testicular effects in neonatal rats that are antagonized by genistein
cotreatment, Biol Reprod 93(4) (2015) 92.
[27] C. Walker, S. Ghazisaeidi, B. Collet, A. Boisvert, M. Culty, In utero exposure to low doses
of genistein and di-(2-ethylhexyl) phthalate (DEHP) alters innate immune cells in neonatal and
adult rat testes, Andrology 8(4) (2020) 943-964.
[28] C. Walker, S. Garza, V. Papadopoulos, M. Culty, Impact of endocrine-disrupting chemicals
on steroidogenesis and consequences on testicular function, Molecular and Cellular
Endocrinology 527 (2021) 111215.
[29] P.M. Foster, Disruption of reproductive development in male rat offspring following in utero
exposure to phthalate esters, International journal of andrology 29(1) (2006) 140-147.
[30] E. Rajpert-de Meyts, C.E. Hoei-Hansen, From gonocytes to testicular cancer: the role of
impaired gonadal development, Annals of the New York Academy of Sciences 1120(1) (2007)
168-180.
[31] R.M. Sharpe, Pathways of endocrine disruption during male sexual differentiation and
masculinisation, Best practice & research Clinical endocrinology & metabolism 20(1) (2006) 91-
110.
[32] M. Kanatsu-Shinohara, K. Inoue, J. Lee, M. Yoshimoto, N. Ogonuki, H. Miki, S. Baba, T.
Kato, Y. Kazuki, S. Toyokuni, Generation of pluripotent stem cells from neonatal mouse testis,
Cell 119(7) (2004) 1001-1012.

171

[33] K. Guan, K. Nayernia, L.S. Maier, S. Wagner, R. Dressel, J.H. Lee, J. Nolte, F. Wolf, M. Li,
W. Engel, Pluripotency of spermatogonial stem cells from adult mouse testis, Nature 440(7088)
(2006) 1199-1203.
[34] F. Izadyar, F. Pau, J. Marh, N. Slepko, T. Wang, R. Gonzalez, T. Ramos, K. Howerton, C.
Sayre, F. Silva, Generation of multipotent cell lines from a distinct population of male germ line
stem cells, Reproduction 135(6) (2008) 771.
[35] R. Habert, H. Lejeune, J.M. Saez, Origin, differentiation and regulation of fetal and adult
Leydig cells, Mol Cell Endocrinol 179(1-2) (2001) 47-74.
[36] P. Sararols, I. Stevant, Y. Neirijnck, D. Rebourcet, A. Darbey, M.K. Curley, F. Kuhne, E.
Dermitzakis, L.B. Smith, S. Nef, Specific Transcriptomic Signatures and Dual Regulation of
Steroidogenesis Between Fetal and Adult Mouse Leydig Cells, Front Cell Dev Biol 9 (2021)
695546.
[37] Y. Shima, K. Miyabayashi, S. Haraguchi, T. Arakawa, H. Otake, T. Baba, S. Matsuzaki, Y.
Shishido, H. Akiyama, T. Tachibana, K. Tsutsui, K. Morohashi, Contribution of Leydig and
Sertoli cells to testosterone production in mouse fetal testes, Mol Endocrinol 27(1) (2013) 63-73.
[38] P.J. O’shaughnessy, P. Baker, U. Sohnius, A.M. Haavisto, H.M. Charlton, I. Huhtaniemi,
Fetal development of Leydig cell activity in the mouse is independent of pituitary gonadotroph
function, Endocrinology 139(3) (1998) 1141-1146.
[39] F.-p. Zhang, T. Pakarainen, F. Zhu, M. Poutanen, I. Huhtaniemi, Molecular characterization
of postnatal development of testicular steroidogenesis in luteinizing hormone receptor knockout
mice, Endocrinology 145(3) (2004) 1453-1463.
[40] P.J. O’Shaughnessy, P.J. Baker, M. Heikkila, S. Vainio, A.P. McMahon, Localization of 17β-
hydroxysteroid dehydrogenase/17-ketosteroid reductase isoform expression in the developing
mouse testis—androstenedione is the major androgen secreted by fetal/neonatal Leydig cells,
Endocrinology 141(7) (2000) 2631-2637.
[41] Y. Shima, K. Miyabayashi, S. Haraguchi, T. Arakawa, H. Otake, T. Baba, S. Matsuzaki, Y.
Shishido, H. Akiyama, T. Tachibana, Contribution of Leydig and Sertoli cells to testosterone
production in mouse fetal testes, Molecular endocrinology 27(1) (2013) 63-73.
[42] L. Li, B.R. Zirkin, V. Papadopoulos, Leydig cell androgen synthesis, Encyclopedia of
reproduction 1 (2018) 215-221.
[43] T.G. Travison, A.B. Araujo, A.B. O’Donnell, V. Kupelian, J.B. McKinlay, A population-level
decline in serum testosterone levels in American men, The Journal of Clinical Endocrinology &
Metabolism 92(1) (2007) 196-202.
[44] WHO, WHO laboratory manual for the examination and processing of human semen, World
Health Organization, Geneva, 2010.
[45] D.S. Guzick, J.W. Overstreet, P. Factor-Litvak, C.K. Brazil, S.T. Nakajima, C. Coutifaris,
S.A. Carson, P. Cisneros, M.P. Steinkampf, J.A. Hill, Sperm morphology, motility, and
concentration in fertile and infertile men, New England Journal of Medicine 345(19) (2001)
1388-1393.
[46] S. Sathyanarayana, L. Beard, C. Zhou, R. Grady, Measurement and correlates of ano-
genital distance in healthy, newborn infants, International journal of andrology 33(2) (2010) 317-
323.

172

[47] A. Thankamony, N. Lek, D. Carroll, M. Williams, D.B. Dunger, C.L. Acerini, K.K. Ong, I.A.
Hughes, Anogenital distance and penile length in infants with hypospadias or cryptorchidism:
comparison with normative data, Environmental health perspectives 122(2) (2014) 207-211.
[48] C.G. Bornehag, F. Carlstedt, B.A. Jonsson, C.H. Lindh, T.K. Jensen, A. Bodin, C. Jonsson,
S. Janson, S.H. Swan, Prenatal phthalate exposures and anogenital distance in Swedish boys,
Environ Health Perspect 123(1) (2015) 101-7.
[49] W.L. Smith, D.L. DeWitt, R.M. Garavito, Cyclooxygenases: structural, cellular, and
molecular biology, Annual review of biochemistry 69(1) (2000) 145-182.
[50] J. Vane, Y. Bakhle, R. Botting, CYCLOOXYGENASES 1 AND 2, Annual review of
pharmacology and toxicology 38(1) (1998) 97-120.
[51] G. Cirino, Multiple controls in inflammation: extracellular and intracellular phospholipase A2,
inducible and constitutive cyclooxygenase, and inducible nitric oxide synthase, Biochemical
pharmacology 55(2) (1998) 105-111.
[52] W.L. Smith, R.C. Murphy, The eicosanoids: cyclooxygenase, lipoxygenase and
epoxygenase pathways, Biochemistry of lipids, lipoproteins and membranes, Elsevier2016, pp.
259-296.
[53] M. Murakami, I. Kudo, Recent advances in molecular biology and physiology of the
prostaglandin E2-biosynthetic pathway, Progress in lipid research 43(1) (2004) 3-35.
[54] R. Langenbach, S.G. Morham, H.F. Tiano, C.D. Loftin, B.I. Ghanayem, P.C. Chulada, J.F.
Mahler, C.A. Lee, E.H. Goulding, K.D. Kluckman, Prostaglandin synthase 1 gene disruption in
mice reduces arachidonic acid-induced inflammation and indomethacin-induced gastric
ulceration, Cell 83(3) (1995) 483-492.
[55] M.-J. Seo, D.-K. Oh, Prostaglandin synthases: Molecular characterization and involvement
in prostaglandin biosynthesis, Progress in lipid research 66 (2017) 50-68.
[56] J.A. Mitchell, T.D. Warner, Cyclo-oxygenase-2: pharmacology, physiology, biochemistry
and relevance to NSAID therapy, British journal of pharmacology 128(6) (1999) 1121-1132.
[57] K.L. Pierce, J.W. Regan, Prostanoid receptor heterogeneity through alternative mRNA
splicing, Life sciences 62(17-18) (1998) 1479-1483.
[58] J. Chakraborty, S. Das, J. Wang, S. Dey, Developmental expression of the cyclo-
oxygenase-1 and cyclo-oxygenase-2 genes in the peri-implantation mouse uterus and their
differential regulation by the blastocyst and ovarian steroids, Journal of molecular endocrinology
16 (1996) 107-122.
[59] W.F. O'Brien, The role of prostaglandins in labor and delivery, Clinics in perinatology 22(4)
(1995) 973-984.
[60] R. Sawdy, D. Slater, N. Fisk, D.K. Edmonds, P. Bennett, Use of a cyclo-oxygenase type-2-
selective non-steroidal anti-inflammatory agent to prevent preterm delivery, The Lancet
350(9073) (1997) 265-266.
[61] O. Laneuville, D.K. Breuer, D.L. Dewitt, T. Hla, C.D. Funk, W.L. Smith, Differential inhibition
of human prostaglandin endoperoxide H synthases-1 and-2 by nonsteroidal anti-inflammatory
drugs, Journal of Pharmacology and Experimental Therapeutics 271(2) (1994) 927-934.
[62] D.L. DeWitt, Cox-2-selective inhibitors: the new super aspirins, Molecular Pharmacology
55(4) (1999) 625-631.

173

[63] L.J. Marnett, S.W. Rowlinson, D.C. Goodwin, A.S. Kalgutkar, C.A. Lanzo, Arachidonic Acid
Oxygenation by COX-1 and COX-2: MECHANISMS OF CATALYSIS AND INHIBITION* 210,
Journal of Biological Chemistry 274(33) (1999) 22903-22906.
[64] B. Hinz, O. Cheremina, K. Brune, Acetaminophen (paracetamol) is a selective
cyclooxygenase-2 inhibitor in man, The FASEB journal 22(2) (2008) 383-390.
[65] G.G. Graham, M.J. Davies, R.O. Day, A. Mohamudally, K.F. Scott, The modern
pharmacology of paracetamol: therapeutic actions, mechanism of action, metabolism, toxicity
and recent pharmacological findings, Inflammopharmacology 21(3) (2013) 201-232.
[66] M. Ouellet, M.D. Percival, Mechanism of acetaminophen inhibition of cyclooxygenase
isoforms, Archives of Biochemistry and Biophysics 387(2) (2001) 273-280.
[67] H. Lim, B.C. Paria, S.K. Das, J.E. Dinchuk, R. Langenbach, J.M. Trzaskos, S.K. Dey,
Multiple female reproductive failures in cyclooxygenase 2–deficient mice, Cell 91(2) (1997) 197-
208.
[68] T. Hase, R. Yoshimura, M. Matsuyama, Y. Kawahito, S. Wada, K. Tsuchida, H. Sano, T.
Nakatani, Cyclooxygenase-1 and-2 in human testicular tumours, European Journal of Cancer
39(14) (2003) 2043-2049.
[69] M.B. Frungieri, S. Weidinger, V. Meineke, F.M. Köhn, A. Mayerhofer, Proliferative action of
mast-cell tryptase is mediated by PAR2, COX2, prostaglandins, and PPARγ: possible relevance
to human fibrotic disorders, Proceedings of the National Academy of Sciences 99(23) (2002)
15072-15077.
[70] M.E. Matzkin, A. Mayerhofer, S.P. Rossi, B. Gonzalez, C.R. Gonzalez, S.I. Gonzalez-
Calvar, C. Terradas, R. Ponzio, E. Puigdomenech, O. Levalle, Cyclooxygenase-2 in testes of
infertile men: evidence for the induction of prostaglandin synthesis by interleukin-1β, Fertility
and sterility 94(5) (2010) 1933-1936.
[71] W.R. Winnall, U. Ali, M.K. O'Bryan, J.J. Hirst, P.A. Whiley, J.A. Muir, M.P. Hedger,
Constitutive expression of prostaglandin-endoperoxide synthase 2 by somatic and
spermatogenic cells is responsible for prostaglandin E2 production in the adult rat testis, Biology
of reproduction 76(5) (2007) 759-768.
[72] M.M. Frungieri MB, Gonzalez-Calvar SI, Mayerhofer A, Calandra RS. , Stimulatory effect of
interleukin 1β (IL1β) on testicular cyclooxygenase 2 (COX2) expression: possible relevance to
male fertility. , (Abstracts of the 89th Annual Meeting of the Endocrine Society) (2007) 730.
[73] M.E. Matzkin, S.I. Gonzalez Calvar, A. Mayerhofer, R.S. Calandra, M.B. Frungieri,
Testosterone induction of prostaglandin-endoperoxide synthase 2 expression and prostaglandin
F (2alpha) production in hamster Leydig cells, (2009).
[74] M.B. Frungieri, R.S. Calandra, A. Mayerhofer, M.E. Matzkin, Cyclooxygenase and
prostaglandins in somatic cell populations of the testis, Reproduction 149(4) (2015) R169-80.
[75] M.E. Matzkin, E.H. Pellizzari, S.P. Rossi, R.S. Calandra, S.B. Cigorraga, M.B. Frungieri,
Exploring the cyclooxygenase 2 (COX2)/15d-Delta(12,14)PGJ(2) system in hamster Sertoli
cells: regulation by FSH/testosterone and relevance to glucose uptake, Gen Comp Endocrinol
179(2) (2012) 254-64.
[76] V. Rey-Ares, S.P. Rossi, K.-G. Dietrich, F.-M. Köhn, J.U. Schwarzer, H. Welter, M.B.
Frungieri, A. Mayerhofer, Prostaglandin E2 (PGE2) is a testicular peritubular cell-derived factor
involved in human testicular homeostasis, Molecular and cellular endocrinology 473 (2018) 217-
224.

174

[77] G. Manku, P. Papadopoulos, A. Boisvert, M. Culty, Cyclooxygenase 2 (COX2) expression
and prostaglandin synthesis in neonatal rat testicular germ cells: Effects of acetaminophen and
ibuprofen, Andrology 8(3) (2020) 691-705.
[78] A. Tran-Guzman, R. Moradian, H. Cui, M. Culty, In vitro impact of genistein and mono(2-
ethylhexyl) phthalate (MEHP) on the eicosanoid pathway in spermatogonial stem cells, Reprod
Toxicol 107 (2022) 150-165.
[79] X. Wang, M.T. Dyson, Y. Jo, D.M. Stocco, Inhibition of cyclooxygenase-2 activity enhances
steroidogenesis and steroidogenic acute regulatory gene expression in MA-10 mouse Leydig
cells, Endocrinology 144(8) (2003) 3368-3375.
[80] H. Kubota, S. Sasaki, Y. Kubota, Y. Umemoto, Y. Yanai, K. Tozawa, Y. Hayashi, K. Kohri,
Cyclooxygenase-2 Protects Germ Cells Against Spermatogenesis Disturbance in Experimental
Cryptorchidism Model Mice, Journal of andrology 32(1) (2011) 77-85.
[81] R.M. Sharpe, Androgens and the masculinization programming window: human–rodent
differences, Biochemical Society Transactions 48(4) (2020) 1725-1735.
[82] M.B. Maamar, L. Lesné, K. Hennig, C. Desdoits-Lethimonier, K.R. Kilcoyne, I. Coiffec, A.D.
Rolland, C. Chevrier, D.M. Kristensen, V. Lavoué, Ibuprofen results in alterations of human fetal
testis development, Scientific reports 7(1) (2017) 1-15.
[83] S.K. Amateau, M.M. McCarthy, Induction of PGE 2 by estradiol mediates developmental
masculinization of sex behavior, Nature neuroscience 7(6) (2004) 643-650.
[84] J.M. Schwarz, M.M. McCarthy, Cellular mechanisms of estradiol-mediated masculinization
of the brain, The Journal of steroid biochemistry and molecular biology 109(3-5) (2008) 300-
306.
[85] N. Ohashi, T. Kohno, Analgesic effect of acetaminophen: a review of known and novel
mechanisms of action, Frontiers in Pharmacology 11 (2020) 1916.
[86] N. Ohashi, D. Uta, M. Sasaki, M. Ohashi, Y. Kamiya, T. Kohno, Acetaminophen metabolite
N-acylphenolamine induces analgesia via transient receptor potential vanilloid 1 receptors
expressed on the primary afferent terminals of C-fibers in the spinal dorsal horn, Anesthesiology
127(2) (2017) 355-371.
[87] R. Cenedella, Prostaglandins and male reproductive physiology, Advances in sex hormone
research 1 (1975) 325-358.
[88] L.C.H. Ellis, J.L., Prostaglandins. In The Testis, Eds A. D. Johnson, W. R. Gomes and N. L.
Van Demark. (1977) 289-313.
[89] M. Prasad, M. Rajalakshmi, G. Gupta, N. Dinakar, R. Arora, T. Karkun, Epididymal
environment and maturation of spermatozoa, Male Fertility and Sterility (1974) 459-478.
[90] A. Hoffmann, D. Bächner, N. Betat, J. Lauber, G. Gross, Developmental expression of
murine β-trace in embryos and adult animals suggests a function in maturation and
maintenance of blood-tissue barriers, Developmental dynamics 207(3) (1996) 332-343.
[91] R.l. Gerena, N. Eguchi, Y. Urade, G.J. Killian, Stage and region-specific localization of
lipocalin-type prostaglandin D synthase in the adult murine testis and epididymis, Journal of
andrology 21(6) (2000) 848-854.
[92] S. Kugathas, K. Audouze, S. Ermler, F. Orton, E. Rosivatz, M. Scholze, A. Kortenkamp,
Effects of common pesticides on prostaglandin D2 (PGD2) inhibition in SC5 mouse Sertoli cells,
evidence of binding at the COX-2 active site, and implications for endocrine disruption,
Environmental health perspectives 124(4) (2016) 452-459.

175

[93] P. Baker, P. O Shaughnessy, Expression of prostaglandin D synthetase during
development in the mouse testis, REPRODUCTION-CAMBRIDGE- 122(4) (2001) 553-559.
[94] I.R. Adams, A. McLaren, Sexually dimorphic development of mouse primordial germ cells:
switching from oogenesis to spermatogenesis, (2002).
[95] D. Gerashchenko, C.T. Beuckmann, Y. Kanaoka, N. Eguchi, W.C. Gordon, Y. Urade, N.G.
Bazan, O. Hayaishi, Dominant expression of rat prostanoid DP receptor mRNA in
leptomeninges, inner segments of photoreceptor cells, iris epithelium, and ciliary processes,
Journal of neurochemistry 71(3) (1998) 937-945.
[96] C. Sorrentino, B. Silvestrini, L. Braghiroli, S.S. Chung, S. Giacomelli, M.-G. Leone, Y.-b.
Xie, Y.-p. Sui, M.-y. Mo, C.Y. Cheng, Rat prostaglandin D2 synthetase: its tissue distribution,
changes during maturation, and regulation in the testis and epididymis, Biology of reproduction
59(4) (1998) 843-853.
[97] E.T. Samy, J.C. Li, J. Grima, W.M. Lee, B. Silvestrini, C.Y. Cheng, Sertoli cell prostaglandin
D2 synthetase is a multifunctional molecule: its expression and regulation, Endocrinology 141(2)
(2000) 710-721.
[98] S. Malki, S. Nef, C. Notarnicola, L. Thevenet, S. Gasca, C. Méjean, P. Berta, F. Poulat, B.
Boizet-Bonhoure, Prostaglandin D2 induces nuclear import of the sex-determining factor SOX9
via its cAMP-PKA phosphorylation, The EMBO journal 24(10) (2005) 1798-1809.
[99] T. Ishikawa, P.L. Morris, A multistep kinase-based sertoli cell autocrine-amplifying loop
regulates prostaglandins, their receptors, and cytokines, Endocrinology 147(4) (2006) 1706-
1716.
[100] D.R. Cooper, M.P. Carpenter, Sertoli-cell prostaglandin synthesis. Effects of (follitropin)
differentiation and dietary vitamin E, Biochemical Journal 241(3) (1987) 847-855.
[101] M.B. Frungieri, S.I. Gonzalez-Calvar, F. Parborell, M. Albrecht, A. Mayerhofer, R.S.
Calandra, Cyclooxygenase-2 and prostaglandin F2α in Syrian hamster Leydig cells: inhibitory
role on luteinizing hormone/human chorionic gonadotropin-stimulated testosterone production,
Endocrinology 147(9) (2006) 4476-4485.
[102] C.-C. Wu, R.-Y. Shyu, C.-H. Wang, T.-C. Tsai, L.-K. Wang, M.-L. Chen, S.-Y. Jiang, F.-M.
Tsai, Involvement of the prostaglandin D2 signal pathway in retinoid-inducible gene 1 (RIG1)-
mediated suppression of cell invasion in testis cancer cells, Biochimica et Biophysica Acta
(BBA)-Molecular Cell Research 1823(12) (2012) 2227-2236.
[103] B. Moniot, S. Ujjan, J. Champagne, H. Hirai, K. Aritake, K. Nagata, E. Dubois, S. Nidelet,
M. Nakamura, Y. Urade, Prostaglandin D2 acts through the Dp2 receptor to influence male
germ cell differentiation in the foetal mouse testis, Development 141(18) (2014) 3561-3571.
[104] L. Walch, E. Clavarino, P.L. Morris, Prostaglandin (PG) FP and EP1 receptors mediate
PGF2α and PGE2 regulation of interleukin-1β expression in Leydig cell progenitors,
Endocrinology 144(4) (2003) 1284-1291.
[105] C. Schell, M.B. Frungieri, M. Albrecht, S.I. Gonzalez-Calvar, F.M. Köhn, R.S. Calandra, A.
Mayerhofer, A prostaglandin D2 system in the human testis, Fertility and sterility 88(1) (2007)
233-236.
[106] M.B. Frungieri, S.I. Gonzalez Calvar, M.F.A. Parborell, M. Albrecht, A. Mayerhofer, R.S.
Calandra, Cyclooxygenase-2 (COX-2) and prostaglandin F2a (PGF2a) in Syrian Hamster
Leydig cells: inhibitory role on LH/hCG-stimulated testosterone production, (2006).

176

[107] D. Wilhelm, R. Hiramatsu, H. Mizusaki, L. Widjaja, A.N. Combes, Y. Kanai, P. Koopman,
SOX9 regulates prostaglandin D synthase gene transcription in vivo to ensure testis
development, Journal of Biological Chemistry 282(14) (2007) 10553-10560.
[108] B. Moniot, F. Declosmenil, F. Barrionuevo, G. Scherer, K. Aritake, S. Malki, L. Marzi, A.
Cohen-Solal, I. Georg, J.r. Klattig, The PGD2 pathway, independently of FGF9, amplifies SOX9
activity in Sertoli cells during male sexual differentiation, (2009).
[109] P. Philibert, B. Boizet-Bonhoure, A. Bashamboo, F. Paris, K. Aritake, Y. Urade, J. Leger,
C. Sultan, F. Poulat, Unilateral cryptorchidism in mice mutant for Ptgds, Human mutation 34(2)
(2013) 278-282.
[110] Y. Tokugawa, I. Kunishige, Y. Kubota, K. Shimoya, T. Nobunaga, T. Kimura, F. Saji, Y.
Murata, N. Eguchi, H. Oda, Lipocalin-type prostaglandin D synthase in human male
reproductive organs and seminal plasma, Biology of reproduction 58(2) (1998) 600-607.
[111] C. Schell, M. Albrecht, S. Spillner, C. Mayer, L. Kunz, F. Kohn, U. Schwarzer, A.
Mayerhofer, 15-Deoxy-Δ12-14-prostaglandin-J2 induces hypertrophy and loss of contractility in
human testicular peritubular cells: implications for human male fertility, Endocrinology 151(3)
(2010) 1257-1268.
[112] C. Kampfer, S. Spillner, K. Spinnler, J. Schwarzer, C. Terradas, R. Ponzio, E.
Puigdomenech, O. Levalle, F. Köhn, M. Matzkin, Evidence for an adaptation in ROS scavenging
systems in human testicular peritubular cells from infertility patients, International journal of
andrology 35(6) (2012) 793-801.
[113] C. Gupta, A.S. Goldman, The arachidonic acid cascade is involved in the masculinizing
action of testosterone on embryonic external genitalia in mice, Proceedings of the National
Academy of Sciences 83(12) (1986) 4346-4349.
[114] M. Axelstad, S. Christiansen, J. Boberg, M. Scholze, P.R. Jacobsen, L.K. Isling, A.
Kortenkamp, U. Hass, Mixtures of endocrine-disrupting contaminants induce adverse
developmental effects in preweaning rats, Reproduction 147(4) (2014) 489-501.
[115] A. Dean, W. Mungall, C. McKinnell, R.M. Sharpe, Prostaglandins, masculinization and its
disorders: effects of fetal exposure of the rat to the cyclooxygenase inhibitor-indomethacin, Plos
one 8(5) (2013) e62556.
[116] A.M. Stutz, RD Ruiz, M. Fiol De Cuneo, JL Lacuara, G, L. Munoz, Functional activity of
mouse sperm was not affected by low doses of aspirin-like drugs, Archives of andrology 44(2)
(2000) 117-128.
[117] M.G. Barbosa, B.C. Jorge, J. Stein, D.A. Santos Ferreira, A.C.d.S. Barreto, A.C.C. Reis,
S.D.S. Moreira, L.C.D.L. Inocencio, L.F.B. Macorini, A.C. Arena, Pre-pubertal exposure to
ibuprofen impairs sperm parameters in male adult rats and compromises the next generation,
Journal of Toxicology and Environmental Health, Part A 83(15-16) (2020) 559-572.
[118] A. Dean, S. Van Den Driesche, Y. Wang, C. McKinnell, S. Macpherson, S.L. Eddie, H.
Kinnell, P. Hurtado-Gonzalez, T.J. Chambers, K. Stevenson, Analgesic exposure in pregnant
rats affects fetal germ cell development with inter-generational reproductive consequences,
Scientific reports 6(1) (2016) 1-12.
[119] M. Rossitto, C. Marchive, A. Pruvost, E. Sellem, A. Ghettas, S. Badiou, T. Sutra, F. Poulat,
P. Philibert, B. Boizet-Bonhoure, Intergenerational effects on mouse sperm quality after in utero
exposure to acetaminophen and ibuprofen, The FASEB Journal 33(1) (2019) 339-357.
[120] P. Hurtado-Gonzalez, R.A. Anderson, J. Macdonald, S. van den Driesche, K. Kilcoyne, A.
Jørgensen, C. McKinnell, S. Macpherson, R.M. Sharpe, R.T. Mitchell, Effects of exposure to

177

acetaminophen and ibuprofen on fetal germ cell development in both sexes in rodent and
human using multiple experimental systems, Environmental health perspectives 126(4) (2018)
047006.
[121] D.C.K. Ribeiro, M.T. Passoni, H. Meldola, T.Z. Curi, G.N. da Silva, S.E.L. Tolouei, G.S.
Hey, N. Grechi, A.C. Dos Santos, R.I.C. Souza, Prenatal diclofenac exposure delays pubertal
development and induces behavioral changes in rats, Reproductive Toxicology 96 (2020) 380-
389.
[122] A. Vyas, A. Purohit, H. Ram, Assessment of dose-dependent reproductive toxicity of
diclofenac sodium in male rats, Drug and chemical toxicology 42(5) (2019) 478-486.
[123] S. van Den Driesche, J. Macdonald, R.A. Anderson, Z.C. Johnston, T. Chetty, L.B. Smith,
C. McKinnell, A. Dean, N.Z. Homer, A. Jorgensen, Prolonged exposure to acetaminophen
reduces testosterone production by the human fetal testis in a xenograft model, Science
translational medicine 7(288) (2015) 288ra80-288ra80.
[124] P. da Silva Balin, B.C. Jorge, A.R.R. Leite, C.S. Borges, E. Oba, E.J.R. Silva, A.L. de
Barros, J.d.A.C. Horta-Júnior, A.C. Arena, Maternal exposure to ibuprofen can affect the
programming of the hypothalamus of the male offspring, Regulatory Toxicology and
Pharmacology 111 (2020) 104576.
[125] D.M. Kristensen, L. Lesné, V. Le Fol, C. Desdoits-Lethimonier, N. Dejucq-Rainsford, H.
Leffers, B. Jégou, Paracetamol (acetaminophen), aspirin (acetylsalicylic acid) and indomethacin
are anti-androgenic in the rat foetal testis, International journal of andrology 35(3) (2012) 377-
384.
[126] J.B. Holm, C. Chalmey, H. Modick, L.S. Jensen, G. Dierkes, T. Weiss, B.A.H. Jensen,
M.M. Nørregård, K. Borkowski, B. Styrishave, Aniline is rapidly converted into paracetamol
impairing male reproductive development, Toxicological Sciences 148(1) (2015) 288-298.
[127] O. Albert, C. Desdoits-Lethimonier, L. Lesné, A. Legrand, F. Guillé, K. Bensalah, N.
Dejucq-Rainsford, B. Jégou, Paracetamol, aspirin and indomethacin display endocrine
disrupting properties in the adult human testis in vitro, Human reproduction 28(7) (2013) 1890-
1898.
[128] D.M. Kristensen, C. Desdoits-Lethimonier, A.L. Mackey, M.D. Dalgaard, F. De Masi, C.H.
Munkbøl, B. Styrishave, J.-P. Antignac, B. Le Bizec, C. Platel, Ibuprofen alters human testicular
physiology to produce a state of compensated hypogonadism, Proceedings of the National
Academy of Sciences 115(4) (2018) E715-E724.
[129] P. Gaudriault, S. Mazaud-Guittot, V. Lavoué, I. Coiffec, L. Lesné, N. Dejucq-Rainsford, M.
Scholze, A. Kortenkamp, B. Jégou, Endocrine disruption in human fetal testis explants by
individual and combined exposures to selected pharmaceuticals, pesticides, and environmental
pollutants, Environmental health perspectives 125(8) (2017) 087004.
[130] S. Mazaud-Guittot, C.N. Nicolaz, C. Desdoits-Lethimonier, I. Coiffec, M.B. Maamar, P.
Balaguer, D.M. Kristensen, C. Chevrier, V. Lavoué, P. Poulain, Paracetamol, aspirin, and
indomethacin induce endocrine disturbances in the human fetal testis capable of interfering with
testicular descent, The Journal of Clinical Endocrinology & Metabolism 98(11) (2013) E1757-
E1767.
[131] M.T. Passoni, M.N. Kristensen, R.N. Morais, C. Woitkowiak, A.C. Boareto, B.A. da Silva
Amaral, N. Grechi, P.R. Dalsenter, C.H. Munkboel, B. Styrishave, Assessment of the analgesic
dipyrone as a possible (anti) androgenic endocrine disruptor, Toxicology letters 285 (2018) 139-
147.

178

[132] D.M. Kristensen, S. Mazaud-Guittot, P. Gaudriault, L. Lesné, T. Serrano, K.M. Main, B.
Jégou, Analgesic use—prevalence, biomonitoring and endocrine and reproductive effects,
Nature Reviews Endocrinology 12(7) (2016) 381-393.
[133] A.Z. Bauer, S.H. Swan, D. Kriebel, Z. Liew, H.S. Taylor, C.-G. Bornehag, A.M. Andrade, J.
Olsen, R.H. Jensen, R.T. Mitchell, Paracetamol use during pregnancy—a call for precautionary
action, Nature Reviews Endocrinology 17(12) (2021) 757-766.
[134] H. Arslan, A. Aktaş, E. Elibol, O. Esener, A. Türkmen, K. Yurt, M. Onger, B. Altunkaynak,
S. Kaplan, Effects of prenatal diclofenac sodium exposure on newborn testis: a
histomorphometric study, Biotechnic & Histochemistry 91(4) (2016) 277-282.
[135] T. DeFalco, I. Bhattacharya, A.V. Williams, D.M. Sams, B. Capel, Yolk-sac–derived
macrophages regulate fetal testis vascularization and morphogenesis, Proceedings of the
National Academy of Sciences 111(23) (2014) E2384-E2393.
[136] C.A. Snijder, A. Kortenkamp, E.A. Steegers, V.W. Jaddoe, A. Hofman, U. Hass, A.
Burdorf, Intrauterine exposure to mild analgesics during pregnancy and the occurrence of
cryptorchidism and hypospadia in the offspring: the Generation R Study, Human Reproduction
27(4) (2012) 1191-1201.
[137] C. Philippat, L. Giorgis-Allemand, C. Chevrier, S. Cordier, B. Jégou, M.-A. Charles, R.
Slama, Analgesics during pregnancy and undescended testis, Epidemiology 22(5) (2011) 747-
749.
[138] J.D. Interrante, E.C. Ailes, J.N. Lind, M. Anderka, M.L. Feldkamp, M.M. Werler, L.G.
Taylor, J. Trinidad, S.M. Gilboa, C.S. Broussard, Risk comparison for prenatal use of analgesics
and selected birth defects, National Birth Defects Prevention Study 1997–2011, Annals of
epidemiology 27(10) (2017) 645-653. e2.
[139] C. Rebordosa, M. Kogevinas, E. Horváth-Puhó, B. Nørgård, M. Morales, A.E. Czeizel, H.
Vilstrup, H.T. Sørensen, J. Olsen, Acetaminophen use during pregnancy: effects on risk for
congenital abnormalities, American journal of obstetrics and gynecology 198(2) (2008) 178. e1-
178. e7.
[140] J.N. Lind, S.C. Tinker, C.S. Broussard, J. Reefhuis, S.L. Carmichael, M.A. Honein, R.S.
Olney, S.E. Parker, M.M. Werler, N.B.D.P. Study, Maternal medication and herbal use and risk
for hypospadias: data from the National Birth Defects Prevention Study, 1997–2007,
Pharmacoepidemiology and drug safety 22(7) (2013) 783-793.
[141] D.V. Lind, K.M. Main, H.B. Kyhl, D.M. Kristensen, J. Toppari, H.R. Andersen, M.S.
Andersen, N.E. Skakkebæk, T.K. Jensen, Maternal use of mild analgesics during pregnancy
associated with reduced anogenital distance in sons: a cohort study of 1027 mother–child pairs,
Human reproduction 32(1) (2017) 223-231.
[142] B. Fisher, A. Thankamony, I.A. Hughes, K. Ong, D.B. Dunger, C.L. Acerini, Prenatal
paracetamol exposure is associated with shorter anogenital distance in male infants, Human
Reproduction 31(11) (2016) 2642-2650.
[143] A. Ernst, N. Brix, L.L. Lauridsen, J. Olsen, E.T. Parner, Z. Liew, L.H. Olsen, C.H. Ramlau-
Hansen, Acetaminophen (paracetamol) exposure during pregnancy and pubertal development
in boys and girls from a Nationwide Puberty Cohort, American journal of epidemiology 188(1)
(2019) 34-46.
[144] M.M. Smarr, K.L. Grantz, R. Sundaram, J.M. Maisog, M. Honda, K. Kannan, G.M. Buck
Louis, Urinary paracetamol and time-to-pregnancy, Human Reproduction 31(9) (2016) 2119-
2127.

179

[145] J.A. Halpern, R.J. Fantus, C. Chang, M.K. Keeter, B. Helfand, N.E. Bennett, R.E.
Brannigan, Effects of nonsteroidal anti-inflammatory drug (NSAID) use upon male gonadal
function: A national, population-based study, Andrologia 52(4) (2020) e13542.
[146] B. Høyer, C. Ramlau-Hansen, J. Bonde, S. Larsen, G. Toft, Use of non-prescription
analgesics and male reproductive function, Reproductive Toxicology 74 (2017) 70-76.
[147] R. Gerena, D. Irikura, Y. Urade, N. Eguchi, D. Chapman, G. Killian, Identification of a
fertility-associated protein in bull seminal plasma as lipocalin-type prostaglandin D synthase,
Biology of reproduction 58(3) (1998) 826-833.
[148] K. Fujimori, T. Inui, N. Uodome, K. Kadoyama, K. Aritake, Y. Urade, Zebrafish and chicken
lipocalin-type prostaglandin D synthase homologues: conservation of mammalian gene
structure and binding ability for lipophilic molecules, and difference in expression profile and
enzyme activity, Gene 375 (2006) 14-25.
[149] A. Pradhan, P.-E. Olsson, Juvenile ovary to testis transition in zebrafish involves inhibition
of ptges, Biology of reproduction 91(2) (2014) 33, 1-15.
[150] D. Wilhelm, F. Martinson, S. Bradford, M.J. Wilson, A.N. Combes, A. Beverdam, J.
Bowles, H. Mizusaki, P. Koopman, Sertoli cell differentiation is induced both cell-autonomously
and through prostaglandin signaling during mammalian sex determination, Developmental
biology 287(1) (2005) 111-124.
[151] A. Jørgensen, J.E. Nielsen, B.F. Nielsen, J.E. Morthorst, P. Bjerregaard, H. Leffers,
Expression of prostaglandin synthases (pgds and pges) during zebrafish gonadal differentiation,
Comparative Biochemistry and Physiology Part A: Molecular & Integrative Physiology 157(1)
(2010) 102-108.
[152] C. Bereketoglu, A. Pradhan, P.-E. Olsson, Nonsteroidal anti-inflammatory drugs (NSAIDs)
cause male-biased sex differentiation in zebrafish, Aquatic Toxicology 223 (2020) 105476.
[153] A. Gravel, M.M. Vijayan, Salicylate disrupts interrenal steroidogenesis and brain
glucocorticoid receptor expression in rainbow trout, Toxicological Sciences 93(1) (2006) 41-49.
[154] M.G. Wade, G. Van Der Kraak, M. Gerrits, J. Ballantyne, Release and steroidogenic
actions of polyunsaturated fatty acids in the goldfish testis, Biology of reproduction 51(1) (1994)
131-139.
[155] B. Moniot, B. Boizet-Bonhoure, F. Poulat, Male specific expression of lipocalin-type
prostaglandin D synthase (cPTGDS) during chicken gonadal differentiation: relationship with
cSOX9, Sexual Development 2(2) (2008) 96-103.
[156] A.-B. Muñiz-González, Ibuprofen as an emerging pollutant on non-target aquatic
invertebrates: Effects on Chironomus riparius, Environmental Toxicology and Pharmacology 81
(2021) 103537.
[157] Z. Liew, A. Ernst, Intrauterine exposure to acetaminophen and adverse developmental
outcomes: Epidemiological Findings and methodological issues, Current Environmental Health
Reports (2021) 1-11.
[158] S. Lymperi, A. Giwercman, Endocrine disruptors and testicular function, Metabolism 86
(2018) 79-90.
[159] R.M. Sharpe, N.E. Skakkebaek, Are oestrogens involved in falling sperm counts and
disorders of the male reproductive tract?, Lancet 341(8857) (1993) 1392-5.

180

[160] H.W. Bennetts, E.J. Uuderwood, F.L. Shier, A specific breeding problem of sheep on
subterranean clover pastures in Western Australia, Australian veterinary journal 22(1) (1946) 2-
12.
[161] E. Diamanti-Kandarakis, J.P. Bourguignon, L.C. Giudice, R. Hauser, G.S. Prins, A.M.
Soto, R.T. Zoeller, A.C. Gore, Endocrine-disrupting chemicals: an Endocrine Society scientific
statement, Endocr Rev 30(4) (2009) 293-342.
[162] M.A. Kamrin, Phthalate risks, phthalate regulation, and public health: a review, J Toxicol
Environ Health B Crit Rev 12(2) (2009) 157-74.
[163] M. Culty, Gonocytes, from the fifties to the present: is there a reason to change the
name?, Biology of reproduction 89(2) (2013) 46-1.
[164] E. Rajpert-de Meyts, C.E. Hoei-Hansen, From gonocytes to testicular cancer: the role of
impaired gonadal development, Ann N Y Acad Sci 1120(1) (2007) 168-80.
[165] R. Waheeb, M.C. Hofmann, Human spermatogonial stem cells: a possible origin for
spermatocytic seminoma, Int J Androl 34(4 Pt 2) (2011) e296-305; discussion e305.
[166] N.E. Skakkebaek, E. Rajpert-De Meyts, G.M. Buck Louis, J. Toppari, A.M. Andersson,
M.L. Eisenberg, T.K. Jensen, N. Jorgensen, S.H. Swan, K.J. Sapra, S. Ziebe, L. Priskorn, A.
Juul, Male Reproductive Disorders and Fertility Trends: Influences of Environment and Genetic
Susceptibility, Physiol Rev 96(1) (2016) 55-97.
[167] M. Hendryx, J. Luo, Children’s environmental chemical exposures in the USA, NHANES
2003–2012, Environmental Science and Pollution Research 25(6) (2018) 5336-5343.
[168] K.K. Rozman, J. Bhatia, A.M. Calafat, C. Chambers, M. Culty, R.A. Etzel, J.A. Flaws, D.K.
Hansen, P.B. Hoyer, E.H. Jeffery, J.S. Kesner, S. Marty, J.A. Thomas, D. Umbach, NTP-
CERHR expert panel report on the reproductive and developmental toxicity of genistein, Birth
Defects Res B Dev Reprod Toxicol 77(6) (2006) 485-638.
[169] R. Green, R. Hauser, A.M. Calafat, J. Weuve, T. Schettler, S. Ringer, K. Huttner, H. Hu,
Use of di (2-ethylhexyl) phthalate–containing medical products and urinary levels of mono (2-
ethylhexyl) phthalate in neonatal intensive care unit infants, Environmental health perspectives
113(9) (2005) 1222-1225.
[170] G.P. Daston, J.C. Cook, R.J. Kavlock, Uncertainties for endocrine disrupters: our view on
progress, Toxicological Sciences 74(2) (2003) 245-252.
[171] A. Kortenkamp, Low dose mixture effects of endocrine disrupters: implications for risk
assessment and epidemiology, Int J Androl 31(2) (2008) 233-40.
[172] R.T. Zoeller, L. Doan, B. Demeneix, A.C. Gore, A. Nadal, S. Tan, Update on Activities in
Endocrine Disruptor Research and Policy, Endocrinology 160(7) (2019) 1681-1683.
[173] I. Silins, J. Hogberg, Combined toxic exposures and human health: biomarkers of
exposure and effect, Int J Environ Res Public Health 8(3) (2011) 629-47.
[174] D.J. Spurgeon, O.A. Jones, J.L. Dorne, C. Svendsen, S. Swain, S.R. Sturzenbaum,
Systems toxicology approaches for understanding the joint effects of environmental chemical
mixtures, Sci Total Environ 408(18) (2010) 3725-34.
[175] S.F. Arnold, D.M. Klotz, B.M. Collins, P.M. Vonier, L.J. Guillette, Jr., J.A. McLachlan,
Synergistic activation of estrogen receptor with combinations of environmental chemicals,
Science 272(5267) (1996) 1489-92.

181

[176] K. Ramamoorthy, F. Wang, I.C. Chen, J.D. Norris, D.P. McDonnell, L.S. Leonard, K.W.
Gaido, W.P. Bocchinfuso, K.S. Korach, S. Safe, Estrogenic activity of a dieldrin/toxaphene
mixture in the mouse uterus, MCF-7 human breast cancer cells, and yeast-based estrogen
receptor assays: no apparent synergism, Endocrinology 138(4) (1997) 1520-1527.
[177] R. Thuillier, M. Mazer, G. Manku, A. Boisvert, Y. Wang, M. Culty, Interdependence of
PDGF and estrogen signaling pathways in inducing neonatal rat testicular gonocytes
proliferation, Biol Reprod 82 (2010) 825-836.
[178] G.A. LeBlanc, L.J. Bain, V.S. Wilson, Pesticides: multiple mechanisms of
demasculinization, Mol Cell Endocrinol 126(1) (1997) 1-5.
[179] E. Ribeiro, C. Ladeira, S. Viegas, EDCs Mixtures: A Stealthy Hazard for Human Health?,
Toxics 5(1) (2017) 5.
[180] M. Culty, R. Thuillier, W. Li, Y. Wang, D.B. Martinez-Arguelles, C.G. Benjamin, K.M.
Triantafilou, B.R. Zirkin, V. Papadopoulos, In utero exposure to di-(2-ethylhexyl) phthalate
exerts both short-term and long-lasting suppressive effects on testosterone production in the rat,
Biology of reproduction 78(6) (2008) 1018-1028.
[181] S. Jones, A. Boisvert, T.B. Duong, S. Francois, P. Thrane, M. Culty, Disruption of rat testis
development following combined in utero exposure to the phytoestrogen genistein and
antiandrogenic plasticizer di-(2-ethylhexyl) phthalate, Biol Reprod 91(3) (2014) 64.
[182] S. Jones, A. Boisvert, S. Francois, L. Zhang, M. Culty, In utero exposure to di-(2-
ethylhexyl) phthalate induces testicular effects in neonatal rats that are antagonized by genistein
cotreatment, Biology of reproduction 93(4) (2015) 92-1.
[183] Y. Wang, R. Thuillier, M. Culty, Prenatal estrogen exposure differentially affects estrogen
receptor-associated proteins in rat testis gonocytes, Biol Reprod 71(5) (2004) 1652-64.
[184] D. Roberts, D.N.R. Veeramachaneni, W.D. Schlaff, C.A. Awoniyi, Effects of chronic dietary
exposure to genistein, a phytoestrogen, during various stages of development on reproductive
hormones and spermatogenesis in rats, Endocrine 13(3) (2000) 281-286.
[185] S. Christiansen, J. Boberg, M. Axelstad, M. Dalgaard, A.M. Vinggaard, S.B. Metzdorff, U.
Hass, Low-dose perinatal exposure to di(2-ethylhexyl) phthalate induces anti-androgenic effects
in male rats, Reprod Toxicol 30(2) (2010) 313-21.
[186] L.D. Zhang, H.C. Li, T. Chong, M. Gao, J. Yin, D.L. Fu, Q. Deng, Z.M. Wang, Prepubertal
exposure to genistein alleviates di-(2-ethylhexyl) phthalate induced testicular oxidative stress in
adult rats, Biomed Res Int 2014 (2014) 598630.
[187] C. Walker, S. Ghazisaeidi, B. Collet, A. Boisvert, M. Culty, In utero exposure to low doses
of genistein and di-(2-ethylhexyl) phthalate (DEHP) alters innate immune cells in neonatal and
adult rat testes, Andrology 8(4) (2020) 943-964.
[188] D.D. Apa, S. Cayan, A. Polat, E. Akbay, Mast cells and fibrosis on testicular biopsies in
male infertility, Arch Androl 48(5) (2002) 337-44.
[189] M.B. Frungieri, R.S. Calandra, L. Lustig, V. Meineke, F.M. Kohn, H.J. Vogt, A. Mayerhofer,
Number, distribution pattern, and identification of macrophages in the testes of infertile men,
Fertil Steril 78(2) (2002) 298-306.
[190] M.E. Matzkin, E. Riviere, S.P. Rossi, R. Ponzio, E. Puigdomenech, O. Levalle, C.
Terradas, R.S. Calandra, A. Mayerhofer, M.B. Frungieri, beta-adrenergic receptors in the up-
regulation of COX2 expression and prostaglandin production in testicular macrophages:
Possible relevance to male idiopathic infertility, Mol Cell Endocrinol 498 (2019) 110545.

182

[191] C. Schell, M.B. Frungieri, M. Albrecht, S.I. Gonzalez-Calvar, F.M. Kohn, R.S. Calandra, A.
Mayerhofer, A prostaglandin D2 system in the human testis, Fertil Steril 88(1) (2007) 233-6.
[192] S.B. Miller, Prostaglandins in health and disease: an overview, Elsevier, 2006, pp. 37-49.
[193] D.M. Kristensen, C. Desdoits-Lethimonier, A.L. Mackey, M.D. Dalgaard, F. De Masi, C.H.
Munkbol, B. Styrishave, J.P. Antignac, B. Le Bizec, C. Platel, A. Hay-Schmidt, T.K. Jensen, L.
Lesne, S. Mazaud-Guittot, K. Kristiansen, S. Brunak, M. Kjaer, A. Juul, B. Jegou, Ibuprofen
alters human testicular physiology to produce a state of compensated hypogonadism, Proc Natl
Acad Sci U S A 115(4) (2018) E715-E724.
[194] M. Ben Maamar, L. Lesne, K. Hennig, C. Desdoits-Lethimonier, K.R. Kilcoyne, I. Coiffec,
A.D. Rolland, C. Chevrier, D.M. Kristensen, V. Lavoue, J.P. Antignac, B. Le Bizec, N. Dejucq-
Rainsford, R.T. Mitchell, S. Mazaud-Guittot, B. Jegou, Ibuprofen results in alterations of human
fetal testis development, Sci Rep 7 (2017) 44184.
[195] B. Moniot, S. Ujjan, J. Champagne, H. Hirai, K. Aritake, K. Nagata, E. Dubois, S. Nidelet,
M. Nakamura, Y. Urade, Prostaglandin D2 acts through the Dp2 receptor to influence male
germ cell differentiation in the foetal mouse testis, Development (2014) dev-103408.
[196] S. Swami, A.V. Krishnan, J. Moreno, R.S. Bhattacharyya, C. Gardner, J.D. Brooks, D.M.
Peehl, D. Feldman, Inhibition of prostaglandin synthesis and actions by genistein in human
prostate cancer cells and by soy isoflavones in prostate cancer patients, Int J Cancer 124(9)
(2009) 2050-9.
[197] T.M. Onorato, P.W. Brown, P.L. Morris, Mono-(2-ethylhexyl) phthalate increases
spermatocyte mitochondrial peroxiredoxin 3 and cyclooxygenase 2, J Androl 29(3) (2008) 293-
303.
[198] D.M. Kristensen, M.L. Skalkam, K. Audouze, L. Lesne, C. Desdoits-Lethimonier, H.
Frederiksen, S. Brunak, N.E. Skakkebaek, B. Jegou, J.B. Hansen, S. Junker, H. Leffers, Many
putative endocrine disruptors inhibit prostaglandin synthesis, Environ Health Perspect 119(4)
(2011) 534-41.
[199] G. Manku, M. Culty, Mammalian gonocyte and spermatogonia differentiation: recent
advances and remaining challenges, Reproduction (Cambridge, England) 149(3) (2015) R139-
57.
[200] J.M. Oatley, R.L. Brinster, Regulation of spermatogonial stem cell self-renewal in
mammals, Annu Rev Cell Dev Biol 24 (2008) 263-86.
[201] G. Manku, M. Mazer, M. Culty, Neonatal testicular gonocytes isolation and processing for
immunocytochemical analysis, Methods Mol Biol 825 (2012) 17-29.
[202] G. Manku, S.S. Wing, M. Culty, Expression of the ubiquitin proteasome system in neonatal
rat gonocytes and spermatogonia: role in gonocyte differentiation, Biol Reprod 87(2) (2012) 44.
[203] G. Manku, M. Culty, Dynamic changes in the expression of apoptosis-related genes in
differentiating gonocytes and in seminomas, Asian J Androl 17(3) (2015) 403-14.
[204] J.P. Novak, M.C. Miller, 3rd, D.A. Bell, Variation in fiberoptic bead-based oligonucleotide
microarrays: dispersion characteristics among hybridization and biological replicate samples,
Biol Direct 1 (2006) 18.
[205] D.M. Kristensen, L. Lesne, V. Le Fol, C. Desdoits-Lethimonier, N. Dejucq-Rainsford, H.
Leffers, B. Jegou, Paracetamol (acetaminophen), aspirin (acetylsalicylic acid) and indomethacin
are anti-androgenic in the rat foetal testis, Int J Androl 35(3) (2012) 377-84.

183

[206] M.-C. Hofmann, Gdnf signaling pathways within the mammalian spermatogonial stem cell
niche, Molecular and cellular endocrinology 288(1-2) (2008) 95-103.
[207] P.A. Parekh, T.X. Garcia, R. Waheeb, V. Jain, P. Gandhi, M.L. Meistrich, G. Shetty, M.-C.
Hofmann, Undifferentiated spermatogonia regulate Cyp26b1 expression through NOTCH
signaling and drive germ cell differentiation, The FASEB Journal 33(7) (2019) 8423-8435.
[208] R. Tian, C. Yao, C. Yang, Z. Zhu, C. Li, E. Zhi, J. Wang, P. Li, H. Chen, Q. Yuan, Z. He, Z.
Li, Fibroblast growth factor-5 promotes spermatogonial stem cell proliferation via ERK and AKT
activation, Stem Cell Res Ther 10(1) (2019) 40.
[209] A.K. Yadav, B.C. Jang, Inhibition of Lipid Accumulation and Cyclooxygenase-2 Expression
in Differentiating 3T3-L1 Preadipocytes by Pazopanib, a Multikinase Inhibitor, Int J Mol Sci 22(9)
(2021).
[210] Y. Deng, L. Li, J.H. Zhu, P.P. Li, Y.X. Deng, H.H. Luo, Y.Y. Yang, B.C. He, Y. Su, COX-2
promotes the osteogenic potential of BMP9 through TGF-beta1/p38 signaling in mesenchymal
stem cells, Aging (Albany NY) 13(8) (2021) 11336-11351.
[211] H.F. Tiano, C.D. Loftin, J. Akunda, C.A. Lee, J. Spalding, A. Sessoms, D.B. Dunson, E.G.
Rogan, S.G. Morham, R.C. Smart, R. Langenbach, Deficiency of either cyclooxygenase (COX)-
1 or COX-2 alters epidermal differentiation and reduces mouse skin tumorigenesis, Cancer Res
62(12) (2002) 3395-401.
[212] H. Moon, A.C. White, A.D. Borowsky, New insights into the functions of Cox-2 in skin and
esophageal malignancies, Exp Mol Med 52(4) (2020) 538-547.
[213] Z. Cao, C. West, C.S. Norton-Wenzel, R. Rej, F.B. Davis, P.J. Davis, R. Rej, Effects of
resin or charcoal treatment on fetal bovine serum and bovine calf serum, Endocr Res 34(4)
(2009) 101-8.
[214] A. Hassouna, E. Obaia, S. Marzouk, M. Rateb, M. Haidara, The role of sex hormones in
induced-systemic inflammation in female albino rats, Acta Physiol Hung 101(1) (2014) 112-27.
[215] Q. Li, S. Zhang, W. Mao, C. Fu, Y. Shen, Y. Wang, B. Liu, J. Cao, 17beta-estradiol
regulates prostaglandin E2 and F2alpha synthesis and function in endometrial explants of cattle,
Anim Reprod Sci 216 (2020) 106466.
[216] A.K. Goff, Steroid hormone modulation of prostaglandin secretion in the ruminant
endometrium during the estrous cycle, Biol Reprod 71(1) (2004) 11-6.
[217] C. Tu, M.V. Fiandalo, E. Pop, J.J. Stocking, G. Azabdaftari, J. Li, H. Wei, D. Ma, J. Qu,
J.L. Mohler, L. Tang, Y. Wu, Proteomic Analysis of Charcoal-Stripped Fetal Bovine Serum
Reveals Changes in the Insulin-like Growth Factor Signaling Pathway, J Proteome Res 17(9)
(2018) 2963-2977.
[218] M.J. Sikora, M.D. Johnson, A.V. Lee, S. Oesterreich, Endocrine Response Phenotypes
Are Altered by Charcoal-Stripped Serum Variability, Endocrinology 157(10) (2016) 3760-3766.
[219] K.L. Howdeshell, A.K. Hotchkiss, L.E. Gray, Jr., Cumulative effects of antiandrogenic
chemical mixtures and their relevance to human health risk assessment, Int J Hyg Environ
Health 220(2 Pt A) (2017) 179-188.
[220] T. Akiyama, J. Ishida, S. Nakagawa, H. Ogawara, S. Watanabe, N. Itoh, M. Shibuya, Y.
Fukami, Genistein, a specific inhibitor of tyrosine-specific protein kinases, J Biol Chem 262(12)
(1987) 5592-5.
[221] F. Ye, J. Wu, T. Dunn, J. Yi, X. Tong, D. Zhang, Inhibition of cyclooxygenase-2 activity in
head and neck cancer cells by genistein, Cancer Lett 211(1) (2004) 39-46.

184

[222] I. Woclawek-Potocka, M.M. Bah, A. Korzekwa, M.K. Piskula, W. Wiczkowski, A. Depta,
D.J. Skarzynski, Soybean-derived phytoestrogens regulate prostaglandin secretion in
endometrium during cattle estrous cycle and early pregnancy, Exp Biol Med (Maywood) 230(3)
(2005) 189-99.
[223] T. Hase, R. Yoshimura, M. Matsuyama, Y. Kawahito, S. Wada, K. Tsuchida, H. Sano, T.
Nakatani, Cyclooxygenase-1 and -2 in human testicular tumours, Eur J Cancer 39(14) (2003)
2043-9.
[224] L.M. Tetz, D.M. Aronoff, R. Loch-Caruso, Mono-ethylhexyl phthalate stimulates
prostaglandin secretion in human placental macrophages and THP-1 cells, Reprod Biol
Endocrinol 13(1) (2015) 56.
[225] M.L. Vilela, E. Willingham, J. Buckley, B.C. Liu, K. Agras, Y. Shiroyanagi, L.S. Baskin,
Endocrine disruptors and hypospadias: role of genistein and the fungicide vinclozolin, Urology
70(3) (2007) 618-21.
[226] M. Mallozzi, G. Bordi, C. Garo, D. Caserta, The effect of maternal exposure to endocrine
disrupting chemicals on fetal and neonatal development: A review on the major concerns, Birth
Defects Res C Embryo Today 108(3) (2016) 224-242.
[227] D.M. Kristensen, S. Mazaud-Guittot, P. Gaudriault, L. Lesné, T. Serrano, K.M. Main, B.
Jégou, Analgesic use — prevalence, biomonitoring and endocrine and reproductive effects,
Nature Reviews Endocrinology 12(7) (2016) 381-393.
[228] R.K. Hernandez, M.M. Werler, P. Romitti, L. Sun, M. Anderka, Nonsteroidal
antiinflammatory drug use among women and the risk of birth defects, American Journal of
Obstetrics and Gynecology 206(3) (2012) 228.e1-228.e8.
[229] D.M. Kristensen, L. Lesné, V. Le Fol, C. Desdoits-Lethimonier, N. Dejucq-Rainsford, H.
Leffers, B. Jégou, Paracetamol (acetaminophen), aspirin (acetylsalicylic acid) and indomethacin
are anti-androgenic in the rat foetal testis, International Journal of Andrology 35(3) (2012) 377-
384.
[230] A. Dean, W. Mungall, C. McKinnell, R.M. Sharpe, Prostaglandins, Masculinization and Its
Disorders: Effects of Fetal Exposure of the Rat to the Cyclooxygenase Inhibitor- Indomethacin,
PLoS ONE 8(5) (2013) e62556.
[231] R. Waheeb, M.C. Hofmann, Human spermatogonial stem cells: a possible origin for
spermatocytic seminoma, International Journal of Andrology 34(4pt2) (2011) e296-e305.
[232] M. Bloor, M. Paech, Nonsteroidal anti-inflammatory drugs during pregnancy and the
initiation of lactation, Anesthesia & Analgesia 116(5) (2013) 1063-1075.
[233] J. Reese, X. Zhao, W.-G. Ma, N. Brown, T.J. Maziasz, S.K. Dey, Comparative Analysis of
Pharmacologic and/or Genetic Disruption of Cyclooxygenase-1 and Cyclooxygenase-2 Function
in Female Reproduction in Mice*, Endocrinology 142(7) (2001) 3198-3206.
[234] D.L. Simmons, R.M. Botting, T. Hla, Cyclooxygenase Isozymes: The Biology of
Prostaglandin Synthesis and Inhibition, Pharmacological Reviews 56(3) (2004) 387-437.
[235] A. Tran-Guzman, R. Moradian, H. Cui, M. Culty, In vitro impact of genistein and mono(2-
ethylhexyl) phthalate (MEHP) on the eicosanoid pathway in spermatogonial stem cells,
Reproductive Toxicology 107 (2022) 150-165.
[236] H. Mi, A. Muruganujan, P.D. Thomas, PANTHER in 2013: modeling the evolution of gene
function, and other gene attributes, in the context of phylogenetic trees, Nucleic acids research
41(D1) (2012) D377-D386.

185

[237] P.D. Thomas, M.J. Campbell, A. Kejariwal, H. Mi, B. Karlak, R. Daverman, K. Diemer, A.
Muruganujan, A. Narechania, PANTHER: a library of protein families and subfamilies indexed
by function, Genome research 13(9) (2003) 2129-2141.
[238] G. Manku, A. Hueso, F. Brimo, P. Chan, P. Gonzalez-Peramato, N. Jabado, T. Gayden,
M. Bourgey, Y. Riazalhosseini, M. Culty, Changes in the expression profiles of claudins during
gonocyte differentiation and in seminomas, Andrology 4(1) (2016) 95-110.
[239] J.P. Novak, M.C. Miller, D.A. Bell, Variation in fiberoptic bead-based oligonucleotide
microarrays: dispersion characteristics among hybridization and biological replicate samples,
Biology direct 1(1) (2006) 1-17.
[240] G. Manku, S.S. Wing, M. Culty, Expression of the ubiquitin proteasome system in neonatal
rat gonocytes and spermatogonia: role in gonocyte differentiation, Biology of reproduction 87(2)
(2012) 44-1.
[241] N. Futaki, S. Takahashi, M. Yokoyama, I. Arai, S. Higuchi, S. Otomo, NS-398, a new anti-
inflammatory agent, selectively inhibits prostaglandin G/H synthase/cyclooxygenase (COX-2)
activity in vitro, Prostaglandins 47(1) (1994) 55-59.
[242] T.D. Penning, J.J. Talley, S.R. Bertenshaw, J.S. Carter, P.W. Collins, S. Docter, M.J.
Graneto, L.F. Lee, J.W. Malecha, J.M. Miyashiro, Synthesis and biological evaluation of the 1,
5-diarylpyrazole class of cyclooxygenase-2 inhibitors: identification of 4-[5-(4-methylphenyl)-3-
(trifluoromethyl)-1 H-pyrazol-1-yl] benzenesulfonamide (SC-58635, celecoxib), Journal of
medicinal chemistry 40(9) (1997) 1347-1365.
[243] T. Ochi, Y. Motoyama, T. Goto, The analgesic effect profile of FR122047, a selective
cyclooxygenase-1 inhibitor, in chemical nociceptive models, European journal of pharmacology
391(1-2) (2000) 49-54.
[244] M. Kato, S. Nishida, H. Kitasato, N. Sakata, S. Kawai, Cyclooxygenase-1 and
cyclooxygenase-2 selectivity of non-steroidal anti-inflammatory drugs: investigation using
human peripheral monocytes, Journal of Pharmacy and Pharmacology 53(12) (2001) 1679-
1685.
[245] R.D. Brown, J.T. Wilson, G.L. Kearns, V.F. Eichler, V.A. Johnson, K.M. Bertrand, Single-
dose pharmacokinetics of ibuprofen and acetaminophen in febrile children, The Journal of
Clinical Pharmacology 32(3) (1992) 231-241.
[246] T. Dejaegere, L. Serneels, M.K. Schäfer, J. Van Biervliet, K. Horré, C. Depboylu, D.
Alvarez-Fischer, A. Herreman, M. Willem, C. Haass, Deficiency of Aph1B/C-γ-secretase
disturbs Nrg1 cleavage and sensorimotor gating that can be reversed with antipsychotic
treatment, Proceedings of the National Academy of Sciences 105(28) (2008) 9775-9780.
[247] M. Culty, Gonocytes, the forgotten cells of the germ cell lineage, Birth Defects Research
Part C: Embryo Today: Reviews 87(1) (2009) 1-26.
[248] J.M. Oatley, R.L. Brinster, Regulation of spermatogonial stem cell self-renewal in
mammals, Annual review of cell and developmental biology 24 (2008) 263-286.
[249] N.C. Law, M.J. Oatley, J.M. Oatley, Developmental kinetics and transcriptome dynamics
of stem cell specification in the spermatogenic lineage, Nature communications 10(1) (2019) 1-
14.
[250] J.A. Harper, J.S. Yuan, J.B. Tan, I. Visan, C.J. Guidos, Notch signaling in development
and disease, Clinical genetics 64(6) (2003) 461-472.

186

[251] M. Klüppel, J.L. Wrana, Turning it up a Notch: cross-talk between TGFβ and Notch
signaling, Bioessays 27(2) (2005) 115-118.
[252] D.H. Kim, T. Xing, Z. Yang, R. Dudek, Q. Lu, Y.-H. Chen, Epithelial mesenchymal
transition in embryonic development, tissue repair and cancer: a comprehensive overview,
Journal of clinical medicine 7(1) (2018) 1.
[253] Z. Aburjania, S. Jang, J. Whitt, R. Jaskula-Stzul, H. Chen, J.B. Rose, The role of notch3 in
cancer, The oncologist 23(8) (2018) 900.
[254] M.V. Giuli, E. Giuliani, I. Screpanti, D. Bellavia, S. Checquolo, Notch signaling activation
as a hallmark for triple-negative breast cancer subtype, Journal of oncology 2019 (2019).
[255] J. Tan, X. Zhang, W. Xiao, X. Liu, C. Li, Y. Guo, W. Xiong, Y. Li, N3ICD with the
transmembrane domain can effectively inhibit EMT by correcting the position of tight/adherens
junctions, Cell adhesion & migration 13(1) (2019) 203-218.
[256] X. Zhang, X. Liu, J. Luo, W. Xiao, X. Ye, M. Chen, Y. Li, G.J. Zhang, Notch3 inhibits
epithelial–mesenchymal transition by activating Kibra-mediated Hippo/YAP signaling in breast
cancer epithelial cells, Oncogenesis 5(11) (2016) e269-e269.
[257] R. Okada, M. Fujimagari, E. Koya, Y. Hirose, T. Sato, Y. Nishina, Expression profile of
NOTCH3 in mouse Spermatogonia, Cells Tissues Organs 204(5-6) (2017) 283-292.
[258] Z. Huang, B. Rivas, A.I. Agoulnik, NOTCH1 gain of function in germ cells causes failure of
spermatogenesis in male mice, PLoS One 8(7) (2013) e71213.
[259] T.X. Garcia, T. DeFalco, B. Capel, M.-C. Hofmann, Constitutive activation of NOTCH1
signaling in Sertoli cells causes gonocyte exit from quiescence, Developmental biology 377(1)
(2013) 188-201.
[260] T.X. Garcia, M.-C. Hofmann, NOTCH signaling in Sertoli cells regulates gonocyte fate,
Cell cycle 12(16) (2013) 2538-2545.
[261] K.L. Loveland, V. Dias, S. Meachem, E. Rajpert-De Meyts, The transforming growth
factor-β superfamily in early spermatogenesis: potential relevance to testicular dysgenesis,
International journal of andrology 30(4) (2007) 377-384.
[262] J.C. Young, S. Wakitani, K.L. Loveland, TGF-β superfamily signaling in testis formation
and early male germline development, Elsevier, pp. 94-103.
[263] Q. Zuo, K. Jin, Y. Zhang, J. Song, B. Li, Dynamic expression and regulatory mechanism of
TGF-β signaling in chicken embryonic stem cells differentiating into spermatogonial stem cells,
Bioscience reports 37(4) (2017) BSR20170179.
[264] J.R. Neil, K.M. Johnson, R.A. Nemenoff, W.P. Schiemann, COX-2 inactivates Smad
signaling and enhances EMT stimulated by TGF-β through a PGE 2-dependent mechanisms,
Carcinogenesis 29(11) (2008) 2227-2235.
[265] K. Kuroda, S. Tani, K. Tamura, S. Minoguchi, H. Kurooka, T. Honjo, Delta-induced Notch
signaling mediated by RBP-J inhibits MyoD expression and myogenesis, Journal of Biological
Chemistry 274(11) (1999) 7238-7244.
[266] J. Massagué, S. Cheifetz, T. Endo, B. Nadal-Ginard, Type beta transforming growth factor
is an inhibitor of myogenic differentiation, Proceedings of the National Academy of Sciences
83(21) (1986) 8206-8210.

187

[267] A. Blokzijl, C. Dahlqvist, E. Reissmann, A. Falk, A. Moliner, U. Lendahl, C.F. Ibáñez,
Cross-talk between the Notch and TGF-β signaling pathways mediated by interaction of the
Notch intracellular domain with Smad3, The Journal of cell biology 163(4) (2003) 723-728.
[268] C. Dahlqvist, A. Blokzijl, G. Chapman, A. Falk, K. Dannaeus, C.F. Ibâñez, U. Lendahl,
Functional Notch signaling is required for BMP4-induced inhibition of myogenic differentiation,
(2003).
[269] F. Itoh, S. Itoh, M.J. Goumans, G. Valdimarsdottir, T. Iso, G.P. Dotto, Y. Hamamori, L.
Kedes, M. Kato, P.t. Dijke, Synergy and antagonism between Notch and BMP receptor signaling
pathways in endothelial cells, The EMBO journal 23(3) (2004) 541-551.
[270] T.-S. Yeh, C.-W. Wu, K.-W. Hsu, W.-J. Liao, M.-C. Yang, A.F.-Y. Li, A.-M. Wang, M.-L.
Kuo, C.-W. Chi, The activated Notch1 signal pathway is associated with gastric cancer
progression through cyclooxygenase-2, Cancer research 69(12) (2009) 5039-5048.
[271] F. Sakai-Takemura, K.i. Nogami, A. Elhussieny, K. Kawabata, Y. Maruyama, N.
Hashimoto, S.i. Takeda, Y. Miyagoe-Suzuki, Prostaglandin EP2 receptor downstream of Notch
signaling inhibits differentiation of human skeletal muscle progenitors in differentiation
conditions, Communications biology 3(1) (2020) 1-13.
[272] M. Ben Maamar, L. Lesné, K. Hennig, C. Desdoits-Lethimonier, K.R. Kilcoyne, I. Coiffec,
A.D. Rolland, C. Chevrier, D.M. Kristensen, V. Lavoué, J.-P. Antignac, B. Le Bizec, N. Dejucq-
Rainsford, R.T. Mitchell, S. Mazaud-Guittot, B. Jégou, Ibuprofen results in alterations of human
fetal testis development, Scientific Reports 7(1) (2017) 44184.
[273] S.R. Singh, O. Burnicka-Turek, C. Chauhan, S.X. Hou, Spermatogonial stem cells,
infertility and testicular cancer, Journal of Cellular and Molecular Medicine 15(3) (2011) 468-
483.
[274] D.M. Kristensen, S.B. Sonne, A.M. Ottesen, R.M. Perrett, J.E. Nielsen, K. Almstrup, N.E.
Skakkebaek, H. Leffers, E.R.-D. Meyts, Origin of pluripotent germ cell tumours: The role of
microenvironment during embryonic development, Molecular and Cellular Endocrinology 288(1-
2) (2008) 111-118.
[275] K. Almstrup, S.B. Sonne, C.E. Hoei-Hansen, A.M. Ottesen, J.E. Nielsen, N.E.
Skakkebaek, H. Leffers, E.R.-D. Meyts, From embryonic stem cells to testicular germ cell
cancer - should we be concerned?, International Journal of Andrology 29(1) (2006) 211-218.
[276] B. Hayes-Lattin, C.R. Nichols, Testicular Cancer: A Prototypic Tumor of Young Adults,
Seminars in Oncology 36(5) (2009) 432-438.
[277] T. Wong, A.S. Stang, H. Ganshorn, L. Hartling, I.K. Maconochie, A.M. Thomsen, D.W.
Johnson, Combined and alternating paracetamol and ibuprofen therapy for febrile children,
Evidence-Based Child Health: A Cochrane Review Journal 9(3) (2014) 675-729.
[278] J.T. Wilson, R.D. Brown, G.L. Kearns, V.F. Eichler, V.A. Johnson, K.M. Bertrand, B.A.
Lowe, Single-dose, placebo-controlled comparative study of ibuprofen and acetaminophen
antipyresis in children, The Journal of pediatrics 119(5) (1991) 803-811.
[279] M. Rossitto, M. Ollivier, S. Déjardin, A. Pruvost, C. Brun, C. Marchive, A.L. Nguyen, A.
Ghettas, C. Keime, B. de Massy, In utero exposure to acetaminophen and ibuprofen leads to
intergenerational accelerated reproductive aging in female mice, Communications biology 2(1)
(2019) 1-13.
[280] N. Josso, Anti-Müllerian hormone and Sertoli cell function, Hormone Research in
Paediatrics 38(Suppl. 2) (1992) 72-76.

188

[281] Z. ur Rehman, T. Worku, J.S. Davis, H.S. Talpur, D. Bhattarai, I. Kadariya, G. Hua, J. Cao,
R. Dad, T. Hussain, Role and mechanism of AMH in the regulation of Sertoli cells in mice, The
Journal of Steroid Biochemistry and Molecular Biology 174 (2017) 133-140.
[282] H.L. Ashby, R.M. Gama, H. Sur, J. Inglis, C. Ford, R. Gama, Hypergonadotrophic
hypogonadism due to testicular adrenal rest tumours presenting with hypogonadotrophic
hypergonadism, Annals of clinical biochemistry 49(5) (2012) 497-499.
[283] L. Priskorn, A. Bang, L. Nordkap, M. Krause, J. Mendiola, T. Jensen, A. Juul, N.
Skakkebaek, S. Swan, N. Jørgensen, Anogenital distance is associated with semen quality but
not reproductive hormones in 1106 young men from the general population, Human
Reproduction 34(1) (2019) 12-24.
[284] D. Creasy, A. Bube, E.d. Rijk, H. Kandori, M. Kuwahara, R. Masson, T. Nolte, R. Reams,
K. Regan, S. Rehm, Proliferative and nonproliferative lesions of the rat and mouse male
reproductive system, Toxicologic Pathology 40(6_suppl) (2012) 40S-121S.
[285] J.S. Gutkind, Cell growth control by G protein-coupled receptors: from signal transduction
to signal integration, Oncogene 17(11) (1998) 1331-1342.
[286] H. van Dam, M. Castellazzi, Distinct roles of Jun: Fos and Jun: ATF dimers in
oncogenesis, Oncogene 20(19) (2001) 2453-2464.
[287] A. Ciechanover, The ubiquitin–proteasome pathway: on protein death and cell life, The
EMBO journal 17(24) (1998) 7151-7160.
[288] J. Myung, K.B. Kim, C.M. Crews, The ubiquitin-proteasome pathway and proteasome
inhibitors, Medicinal research reviews 21(4) (2001) 245-273.
[289] M. Szarras-Czapnik, M. Gajewska, J. Ksiazyk, R. Janas, M. Ginalska-Malinowska, Anti-
Müllerian hormone (AMH) measurements in the assessment of testicular function in prepubertal
boys and in sexual differentiation disorders, Endokrynologia, diabetologia i choroby przemiany
materii wieku rozwojowego: organ Polskiego Towarzystwa Endokrynologow Dzieciecych 12(3)
(2006) 195-199.
[290] R.P. Grinspon, R.A. Rey, Anti-müllerian hormone and sertoli cell function in paediatric
male hypogonadism, Hormone research in paediatrics 73(2) (2010) 81-92.
[291] M.G. Sahare, H. Imai, Recent advances of in vitro culture systems for spermatogonial
stem cells in mammals, Reproductive medicine and biology 17(2) (2018) 134-142.
[292] H.L. Moses, C. Arteaga, M. Alexandrow, L. Dagnino, M. Kawabata, D. Pierce Jr, R. Serra,
TGF beta regulation of cell proliferation, Princess Takamatsu Symposia, 1994, pp. 250-263.

189

APPENDICES
Appendix 1: Supplemental Table 1. List of primers used for qPCR.

Gene Forward Primer Reverse Primer
Gapdh AAGGTCATCCCAGAGCTGAA CTGCTTCACCACCTTCTTGA
Cox1 CCTCTTTCCAGGAGCTCACA TCGATGTCACCGTACAGCTC
Foxo1 CTTCAAGGATAAGGGCGACA GACAGATTGTGGCGAATTGA
Gapdh AAGGTCATCCCAGAGCTGAA CTGCTTCACCACCTTCTTGA
Id4 CAGGGTGACAGCATTCTCTG CCGGTGGCTTGTTTCTCTTA
Kit AGCAAATGTCACAACAACCT CCTCGTATTCAACAACCAAA
Mcam CAAACTGGTGTGCGTCTTCTT CTTTTCCTCTCCTGGCACAC
Stra8 GCCTCAAAGTGGCAGGTACTG CTTATCCAGGCTTTCTTCCTGTTC
Jam1 CAAGGCAAGGGTTCGGTGTA TAGGGAGCTGTGATCTGGCT
Mmp2 GCCCCCATGAAGCCTTGTTT TAGCGGTCTCGGGACAGAAT
Tgbfr3 CCTCCGCAGTACAGACCAAG AACCCTCCGAAACCAGGAAG
Tgbf2 TCCCCTCCGAAAATGCCATC TGCTATCGATGTAGCGCTGG
Tgbf3 ATGACCCACGTCCCCTATCA CAGACGGCCAGTTCATTGTG
Smad3 CAGCCATGTCGTCCATCCTG CCATCCAGTGACCTGGGGAT
Smad9 CCTGAGCTCTGCCTCCTATG ACACACTTGCTAGGCTGACC
Notch1 TGCCATATACAGGAGCCACG ATTGGTGTTCTGGCAGGAGG
Notch3 AGGTGGTCACAGACTTGAATGA GTGGGGTGAAGCCATCAGG
Hes1 CGGAATCCCCTGTCTACCTC CTTGGAATGCCGGGAGCTAT
Hey1 CCACTGCAGTTAACTCCTCCT CGCGTCAAAATAACCTTTCCCT

190

Appendix 2: Supplemental Table 2. List of differentially expressed genes involved in signaling
pathways predicted to be altered by Cox1 silencing.

Gene Name
Gene
Symbol FDR
Fold
Change
peroxisome proliferator activated receptor gamma Pparg 2.16E-05 32.7
inhibitor of DNA binding 2 Id2 1.80E-03 6.185
phosphoinositide-3-kinase regulatory subunit 5 Pik3r5 3.13E-04 5.868
platelet derived growth factor receptor beta Pdgfrb 1.17E-03 4.673
twist family bHLH transcription factor 2 Twist2 7.60E-04 4.627
Wnt family member 7A Wnt7a 3.41E-02 3.653
phosphoinositide-3-kinase regulatory subunit 1 Pik3r1 1.33E-03 3.488
notch receptor 3 Notch3 1.26E-03 2.785
transforming growth factor beta 2 Tgfb2 2.78E-03 2.761
notch receptor 1 Notch1 1.84E-03 2.653
jagged canonical Notch ligand 1 Jag1 7.95E-04 2.632
fibroblast growth factor 2 Fgf2 9.14E-02 2.565
fibroblast growth factor 5 Fgf5 2.04E-02 2.536
SMAD family member 3 Smad3 7.14E-04 2.308
transforming growth factor beta 3 Tgfb3 4.72E-03 2.268
phosphoinositide-3-kinase regulatory subunit 3 Pik3r3 2.15E-02 2.059
AKT serinethreonine kinase 3 Akt3 1.28E-03 -2.007
epithelial splicing regulatory protein 2 Esrp2 2.56E-02 -2.171
Wnt family member 5A Wnt5a 6.41E-03 -2.462
par-6 family cell polarity regulator alpha Pard6a 1.39E-02 -2.471
Wnt family member 9A Wnt9a 5.83E-04 -2.8
Wnt family member 11 Wnt11 5.22E-04 -2.994
matrix metallopeptidase 2 Mmp2 1.47E-04 -3.751
fibroblast growth factor 22 Fgf22 8.97E-02 -4.516
fibroblast growth factor 21 Fgf21 6.90E-04 -11.362

191

Appendix 3: Supplemental Table 3. Complete list of 1,265 differentially expressed genes
altered by Cox1 silencing.

Gene P-value FDR Fold change
Gpnmb 1.08E-08 2.16E-05 115.140
Gm3776 1.26E-06 2.65E-04 108.375
Gsta1 7.77E-06 8.18E-04 100.600
Aebp1 9.89E-09 2.16E-05 90.035
Syt13 8.83E-06 8.87E-04 87.451
Megf10 3.14E-06 4.83E-04 70.012
Ltbp1 3.46E-06 5.06E-04 52.109
Fosl1 2.68E-08 3.00E-05 49.578
Mmp3 1.81E-05 1.43E-03 49.345
Gm10660 2.78E-05 1.89E-03 46.806
Mmp10 3.31E-06 4.95E-04 46.175
Nrp2 1.50E-07 8.16E-05 42.102
Spp1 2.24E-12 6.30E-08 36.828
Pparg 1.13E-08 2.16E-05 32.678
Ccl2 8.65E-07 2.10E-04 31.036
Pcdh17 3.14E-10 4.42E-06 30.628
Bex2 5.55E-08 4.22E-05 29.552
Lancl3 2.92E-06 4.65E-04 27.352
Ccl7 3.11E-04 8.62E-03 24.208
Plin4 1.75E-08 2.35E-05 22.056
Ccn5 1.50E-05 1.27E-03 21.595
Prkg2 9.82E-09 2.16E-05 21.323
Serpina3h 3.62E-04 9.52E-03 21.147
Sox9 2.27E-04 6.96E-03 20.742
Aqp1 7.36E-05 3.43E-03 19.909
Bglap3 2.20E-08 2.58E-05 19.859
Hoxb13 2.29E-06 3.83E-04 19.569
Ly6e 5.69E-05 2.97E-03 18.860
Gm12840 2.94E-04 8.33E-03 18.329
Ppargc1a 6.88E-09 2.16E-05 17.902
Ebf1 4.91E-05 2.74E-03 17.565
Ly6a 1.34E-04 5.00E-03 17.028
Sdc3 5.23E-05 2.83E-03 16.905
Dnah6 1.50E-04 5.35E-03 16.065
Angpt4 8.46E-06 8.62E-04 16.058
Gm29216 2.43E-03 3.23E-02 15.888
Tph2 6.24E-03 6.07E-02 14.730
Gm48681 3.67E-03 4.24E-02 14.540
Gm15398 1.42E-05 1.23E-03 14.333
Cxcl12 2.82E-06 4.55E-04 14.087
Chst1 1.24E-03 2.11E-02 13.649
Plpp3 3.89E-07 1.44E-04 13.368
Itgb7 1.86E-05 1.45E-03 13.265
Gm21941 4.47E-03 4.86E-02 13.068
Fabp4 9.49E-04 1.79E-02 12.979
Pde2a 1.27E-03 2.15E-02 12.137
Col12a1 1.95E-06 3.57E-04 12.003
Zfp979 5.74E-07 1.65E-04 11.957
Stra6 1.56E-04 5.45E-03 11.465
Gda 1.54E-06 3.13E-04 11.441
Eda2r 8.56E-05 3.83E-03 11.305
Xdh 3.95E-08 3.54E-05 11.116
Apol9b 5.61E-06 6.69E-04 10.927
Gm6093 1.17E-03 2.04E-02 10.780
Ifit1 7.36E-06 7.90E-04 10.720
Ott 3.90E-03 4.42E-02 10.671
Mir100hg 3.10E-07 1.28E-04 10.578
Runx3 1.27E-08 2.16E-05 10.577
Gm11843 1.26E-02 9.60E-02 10.348
Tafa5 3.52E-03 4.12E-02 10.145
Rsad2 2.93E-05 1.97E-03 10.003

192

Ang2 9.78E-04 1.82E-02 9.919
4930486L24Rik 2.23E-06 3.76E-04 9.862
Apol9a 4.29E-06 5.66E-04 9.598
Rtp4 4.43E-04 1.08E-02 9.567
Angptl2 1.05E-07 6.91E-05 9.513
Cd24a 1.56E-09 1.10E-05 9.386
Vcan 6.13E-05 3.08E-03 9.353
Arhgap22 1.06E-07 6.91E-05 9.309
Tmem171 2.11E-03 2.92E-02 9.186
Ifi211 2.03E-07 1.00E-04 9.093
Nrn1 3.79E-03 4.33E-02 8.895
Adssl1 2.30E-04 7.04E-03 8.810
Slfn10-ps 1.09E-03 1.95E-02 8.807
Cyp1b1 1.18E-06 2.53E-04 8.738
Col8a1 1.40E-07 8.16E-05 8.672
Fer1l6 2.94E-06 4.65E-04 8.522
Adam12 1.41E-03 2.30E-02 8.489
S1pr1 1.11E-04 4.43E-03 8.403
Plekhg4 1.34E-05 1.18E-03 8.402
Pcolce2 3.63E-06 5.19E-04 8.265
Angpt1 1.18E-02 9.21E-02 8.227
Plxna2 2.17E-06 3.73E-04 8.217
Fahd2a 6.89E-04 1.44E-02 8.157
Hmga2 3.89E-09 2.16E-05 8.004
Ddx60 4.37E-04 1.06E-02 7.942
Tcim 1.44E-04 5.24E-03 7.852
Gm8178 9.83E-04 1.82E-02 7.735
Gm14137 1.01E-05 9.60E-04 7.712
Slpi 3.36E-05 2.15E-03 7.645
Scara3 3.60E-03 4.19E-02 7.622
Gm46620 6.32E-08 4.56E-05 7.608
Gpd1 5.94E-06 6.90E-04 7.503
Npepl1 5.47E-04 1.24E-02 7.429
Lgals3bp 2.85E-07 1.24E-04 7.419
Ccn3 1.92E-08 2.40E-05 7.401
Slc38a4 1.75E-04 5.86E-03 7.378
Ifi204 7.08E-09 2.16E-05 7.370
Zfp454 5.66E-03 5.68E-02 7.082
Pla1a 7.06E-03 6.57E-02 7.012
Slfn2 3.90E-05 2.36E-03 6.983
Cd300lb 2.64E-03 3.41E-02 6.967
Nr1h4 2.55E-03 3.32E-02 6.916
Gm49337 6.60E-04 1.40E-02 6.897
Oas1a 8.77E-03 7.55E-02 6.859
Oasl2 2.57E-05 1.80E-03 6.766
Gm36940 9.02E-05 3.96E-03 6.617
Serpina3i 1.01E-02 8.30E-02 6.568
Gpc6 1.71E-07 8.76E-05 6.551
Hlx 1.57E-04 5.47E-03 6.502
Ifi44 3.93E-04 9.99E-03 6.490
Ly6c1 1.10E-02 8.78E-02 6.442
Coro6 1.08E-04 4.39E-03 6.394
Prkar2b 7.88E-09 2.16E-05 6.335
S100a4 5.09E-05 2.78E-03 6.319
Usp18 5.54E-04 1.25E-02 6.242
Htra3 2.16E-03 2.97E-02 6.230
Id2 2.58E-05 1.80E-03 6.185
Prrx2 1.18E-04 4.59E-03 6.175
Nav3 6.48E-03 6.19E-02 6.152
Aqp5 1.19E-03 2.07E-02 6.087
Serpina3g 5.79E-03 5.77E-02 6.087
Cdh6 4.89E-03 5.16E-02 6.056
Parp14 3.68E-04 9.59E-03 5.981
Podnl1 3.65E-04 9.57E-03 5.946
Gbp3 5.37E-04 1.23E-02 5.899
Thbs1 5.91E-07 1.68E-04 5.873

193

Pik3r5 1.57E-06 3.13E-04 5.868
Zfp469 6.20E-08 4.56E-05 5.834
Gm10324 3.03E-03 3.72E-02 5.796
Cp 1.90E-07 9.55E-05 5.752
Reep1 1.28E-03 2.17E-02 5.725
Gm45407 1.12E-03 1.98E-02 5.723
Psmb8 5.12E-03 5.32E-02 5.666
Cbr3 1.74E-05 1.40E-03 5.663
Itgb4 4.25E-04 1.05E-02 5.613
Gm42778 2.37E-03 3.18E-02 5.596
Runx1 1.18E-08 2.16E-05 5.580
Slc16a2 6.30E-07 1.72E-04 5.568
Fam189a2 3.47E-06 5.06E-04 5.528
Robo1 8.48E-03 7.38E-02 5.505
Npr3 4.79E-08 3.96E-05 5.505
Gper1 1.19E-04 4.61E-03 5.487
Runx2os2 9.49E-03 7.94E-02 5.483
Has2 4.86E-03 5.14E-02 5.474
Gm37018 1.07E-03 1.92E-02 5.422
Ndnf 2.17E-05 1.61E-03 5.414
Adamtsl3 3.28E-05 2.12E-03 5.414
Eif4e3 6.74E-03 6.37E-02 5.411
Sh3kbp1 5.04E-07 1.54E-04 5.393
Ptp4a3 2.15E-04 6.69E-03 5.370
Gm37783 3.93E-03 4.44E-02 5.327
Tdrd9 1.34E-04 5.00E-03 5.307
Cd68 1.90E-04 6.15E-03 5.296
Gm14636 1.96E-04 6.28E-03 5.287
Pla2g5 1.48E-07 8.16E-05 5.277
AY036118 8.42E-03 7.34E-02 5.275
Gm37524 3.89E-03 4.41E-02 5.258
Atxn1 5.74E-07 1.65E-04 5.125
Duox1 1.22E-02 9.40E-02 5.121
Igf1 7.08E-04 1.47E-02 5.081
Dclk1 1.96E-03 2.80E-02 5.075
Ccdc80 7.04E-07 1.81E-04 5.072
Tnc 4.38E-06 5.76E-04 5.060
Gm42427 2.47E-06 4.10E-04 5.047
Dock3 5.15E-07 1.54E-04 5.039
Synpo 1.63E-05 1.34E-03 5.007
4933413C19Rik 2.27E-03 3.07E-02 4.881
9230105E05Rik 1.44E-03 2.32E-02 4.867
Cplx2 7.96E-06 8.32E-04 4.844
Mal2 1.09E-05 1.02E-03 4.823
Runx2 1.21E-07 7.43E-05 4.822
Ccnd1 9.90E-07 2.30E-04 4.790
Itgb2 6.91E-03 6.48E-02 4.780
Crlf1 1.11E-02 8.81E-02 4.771
A4galt 6.27E-03 6.07E-02 4.761
Cd6 7.90E-03 7.03E-02 4.753
Tle2 1.39E-05 1.21E-03 4.717
Pdgfrb 1.33E-05 1.17E-03 4.673
Cd44 4.61E-07 1.47E-04 4.652
Lrp8 4.45E-08 3.79E-05 4.643
Abi3bp 5.93E-04 1.30E-02 4.638
Twist2 6.91E-06 7.60E-04 4.627
E230020D15Rik 7.61E-05 3.52E-03 4.627
Hmcn2 1.60E-04 5.54E-03 4.609
Elk3 1.61E-07 8.55E-05 4.600
4930517G19Rik 9.12E-04 1.74E-02 4.596
Osmr 1.49E-08 2.18E-05 4.580
Dmpk 2.14E-05 1.60E-03 4.574
Gm38381 1.09E-03 1.95E-02 4.565
Gm48887 8.92E-03 7.62E-02 4.534
Neat1 9.02E-03 7.68E-02 4.525
Neurl3 8.98E-03 7.66E-02 4.512

194

Kirrel3 2.98E-04 8.40E-03 4.495
Lifr 2.26E-05 1.67E-03 4.490
Gdnf 2.66E-03 3.42E-02 4.484
Mndal 2.18E-04 6.76E-03 4.468
Vmn1r43 1.82E-05 1.43E-03 4.441
Eva1c 1.60E-04 5.54E-03 4.440
H2-T24 3.57E-04 9.41E-03 4.368
Pgm5 5.85E-04 1.29E-02 4.343
Myom3 1.06E-02 8.55E-02 4.337
Gm44260 3.78E-04 9.74E-03 4.281
Nrep 4.50E-06 5.83E-04 4.271
Gm11639 1.16E-03 2.03E-02 4.249
Oasl1 2.84E-06 4.57E-04 4.239
Gm15907 5.54E-03 5.59E-02 4.226
Gm48996 2.10E-04 6.57E-03 4.223
Kbtbd11 5.09E-03 5.30E-02 4.184
Dach1 2.37E-03 3.17E-02 4.173
Adgrl3 1.09E-02 8.70E-02 4.153
1700006J14Rik 4.26E-03 4.70E-02 4.139
Ncald 3.76E-08 3.54E-05 4.132
G430095P16Rik 1.62E-03 2.51E-02 4.117
Gm47664 9.29E-03 7.81E-02 4.093
Cpne2 5.23E-08 4.09E-05 4.083
Cmklr1 1.38E-03 2.26E-02 4.073
Gm48225 4.15E-03 4.62E-02 4.070
Oxr1 2.78E-08 3.00E-05 4.069
Gm47920 9.86E-04 1.82E-02 4.067
Gm43258 5.77E-03 5.76E-02 4.049
Mx2 3.16E-03 3.83E-02 4.031
Fxyd5 8.94E-06 8.88E-04 4.004
Trib1 7.23E-06 7.85E-04 3.992
Gm44981 4.17E-04 1.03E-02 3.991
Myo10 4.70E-07 1.48E-04 3.989
Slc25a48 1.73E-03 2.61E-02 3.964
Gm38046 3.43E-03 4.04E-02 3.946
Gm43795 1.87E-03 2.72E-02 3.945
Gm49353 6.78E-04 1.43E-02 3.945
C130075A20Rik 3.95E-04 1.00E-02 3.943
Zan 2.63E-05 1.82E-03 3.937
Por 9.07E-07 2.18E-04 3.933
2810457G06Rik 7.96E-03 7.06E-02 3.928
St3gal1 2.93E-06 4.65E-04 3.921
Rilp 1.47E-03 2.36E-02 3.909
Crabp2 2.49E-03 3.28E-02 3.905
Sh3pxd2b 1.55E-08 2.18E-05 3.895
Ldlr 2.53E-05 1.78E-03 3.877
Vim 4.96E-06 6.23E-04 3.873
Adamts2 1.10E-02 8.79E-02 3.872
Gm42898 2.55E-03 3.32E-02 3.840
Marcks 3.39E-08 3.53E-05 3.828
Gm45074 1.04E-02 8.46E-02 3.827
Tfrc 5.18E-08 4.09E-05 3.800
Nkain1 1.53E-04 5.38E-03 3.798
Gm45073 9.84E-04 1.82E-02 3.793
Olfr55 1.12E-02 8.86E-02 3.790
Sytl2 3.20E-06 4.89E-04 3.777
Ccdc170 1.01E-05 9.60E-04 3.773
Gm43009 2.52E-03 3.30E-02 3.771
Nsg1 1.15E-04 4.55E-03 3.765
Ankrd29 7.28E-03 6.68E-02 3.730
Gm43682 7.44E-04 1.52E-02 3.716
Gm48314 2.98E-03 3.68E-02 3.715
Rasgef1a 1.53E-03 2.41E-02 3.711
9030624G23Rik 4.05E-05 2.40E-03 3.708
Pknox2 1.10E-02 8.80E-02 3.689
Sgk1 1.76E-06 3.30E-04 3.687

195

Gbp7 9.99E-04 1.84E-02 3.681
Nrg1 7.59E-07 1.92E-04 3.679
Gm37105 2.36E-03 3.16E-02 3.672
Gm37589 7.99E-04 1.58E-02 3.659
Wnt7a 2.65E-03 3.41E-02 3.653
Bod1-ps 5.89E-03 5.84E-02 3.636
Gm43260 4.82E-03 5.11E-02 3.629
Ssc5d 3.35E-06 4.99E-04 3.627
Rorb 2.83E-03 3.56E-02 3.624
Bach2 1.24E-05 1.12E-03 3.610
Ifi203 3.72E-05 2.29E-03 3.609
Id3 6.31E-05 3.15E-03 3.608
Gm43259 5.31E-04 1.21E-02 3.606
Filip1l 2.55E-06 4.22E-04 3.595
Gm10097 5.03E-03 5.26E-02 3.580
Cep295nl 1.29E-03 2.18E-02 3.579
C730045M19Rik 1.02E-02 8.35E-02 3.562
Mtus2 6.77E-05 3.25E-03 3.555
Cela1 4.80E-04 1.14E-02 3.555
Gm49890 1.02E-04 4.24E-03 3.550
Sned1 1.59E-05 1.33E-03 3.536
Gm4961 8.83E-03 7.57E-02 3.532
Tchh 4.40E-07 1.47E-04 3.525
Adam11 1.68E-03 2.57E-02 3.522
Chst11 4.38E-05 2.56E-03 3.516
Gm48602 8.03E-03 7.10E-02 3.500
Gm42603 3.92E-03 4.43E-02 3.498
Pik3r1 1.60E-05 1.33E-03 3.488
Ifit3b 1.14E-03 2.01E-02 3.488
Gnb4 8.06E-06 8.33E-04 3.482
Cnr1 1.09E-02 8.70E-02 3.478
Gm49705 2.54E-03 3.31E-02 3.476
Tnfaip3 1.09E-04 4.42E-03 3.466
Irs1 3.63E-05 2.26E-03 3.458
Gm5424 1.04E-06 2.35E-04 3.457
Elmo1 1.08E-06 2.36E-04 3.454
Hivep3 3.18E-07 1.30E-04 3.446
Uba7 2.37E-04 7.18E-03 3.440
Gm37747 7.20E-03 6.62E-02 3.430
C530043K16Rik 6.51E-04 1.39E-02 3.424
Mecom 2.69E-07 1.20E-04 3.422
Angpt2 5.52E-04 1.25E-02 3.375
Snhg3 1.25E-05 1.12E-03 3.372
Gm49783 1.43E-04 5.24E-03 3.360
Txlnb 4.49E-04 1.08E-02 3.359
Nuak1 6.71E-05 3.23E-03 3.352
Gm43010 2.28E-03 3.07E-02 3.348
Syne3 2.25E-07 1.04E-04 3.345
Lrrc32 5.83E-03 5.79E-02 3.342
Dact1 1.69E-03 2.57E-02 3.340
Tnfrsf19 7.67E-03 6.89E-02 3.336
Gm7265 1.49E-03 2.37E-02 3.332
Ass1 4.03E-08 3.54E-05 3.323
Syt12 5.20E-07 1.54E-04 3.322
Gm43011 6.19E-03 6.04E-02 3.316
1110002E22Rik 7.06E-03 6.56E-02 3.311
Catsperg2 1.09E-02 8.73E-02 3.310
Ttc39a 6.34E-05 3.15E-03 3.308
Pde9a 6.47E-03 6.19E-02 3.297
Phlda1 2.22E-05 1.65E-03 3.277
9530086O07Rik 7.04E-03 6.55E-02 3.263
Nlrc5 1.12E-04 4.49E-03 3.262
4932442E05Rik 1.96E-03 2.80E-02 3.261
Gm28875 9.94E-03 8.21E-02 3.253
Gm49204 5.24E-03 5.41E-02 3.244
Prickle1 1.09E-04 4.42E-03 3.243

196

Baiap2 3.27E-06 4.95E-04 3.241
Camk1d 3.05E-07 1.28E-04 3.234
Nav2 8.71E-04 1.69E-02 3.231
Dusp4 8.02E-06 8.33E-04 3.227
Gm48035 4.60E-05 2.65E-03 3.222
Pax2 1.50E-03 2.39E-02 3.217
Iffo2 4.58E-07 1.47E-04 3.206
Gm48804 5.84E-04 1.29E-02 3.197
Gm37257 1.17E-02 9.15E-02 3.187
Synm 9.72E-07 2.28E-04 3.184
Smad9 1.71E-04 5.77E-03 3.181
Irf7 7.92E-04 1.58E-02 3.178
Gm10389 5.65E-04 1.26E-02 3.173
Gm21781 1.57E-04 5.46E-03 3.171
Bid 7.10E-07 1.81E-04 3.168
Enc1 1.56E-06 3.13E-04 3.162
Ptprk 4.02E-07 1.47E-04 3.162
Adamtsl4 6.47E-05 3.18E-03 3.161
Gm10388 1.57E-03 2.46E-02 3.160
Pbx1 5.08E-04 1.19E-02 3.139
Gm43256 1.74E-03 2.61E-02 3.135
Ak1 3.04E-07 1.28E-04 3.104
Gm48870 1.20E-03 2.09E-02 3.102
Gm43359 6.65E-03 6.29E-02 3.079
Gm49979 3.99E-04 1.01E-02 3.074
Lncppara 7.77E-04 1.56E-02 3.071
Fam81a 3.19E-03 3.85E-02 3.054
S100a16 4.51E-03 4.88E-02 3.032
Gm38357 4.95E-04 1.16E-02 3.031
Myof 2.08E-05 1.58E-03 3.029
Frmd6 3.26E-07 1.31E-04 3.021
Zfp978 5.05E-07 1.54E-04 3.020
Gm2026 1.36E-03 2.24E-02 3.017
Gm14418 5.97E-03 5.91E-02 3.014
Psmb10 7.95E-05 3.63E-03 3.011
Insig1 1.81E-05 1.43E-03 3.010
Slc37a2 8.37E-05 3.77E-03 3.001
Lncpint 3.76E-04 9.74E-03 3.000
Unc13d 6.39E-04 1.38E-02 2.995
Gm43517 1.69E-03 2.57E-02 2.987
Phospho1 1.75E-04 5.86E-03 2.980
Gm43312 2.80E-03 3.53E-02 2.978
9930017N22Rik 1.62E-03 2.51E-02 2.978
Slc41a2 5.02E-06 6.24E-04 2.978
Cdkn2b 6.55E-06 7.31E-04 2.969
Deptor 1.61E-06 3.17E-04 2.969
Axin2 1.51E-03 2.40E-02 2.962
Gm17334 2.86E-03 3.58E-02 2.960
1700109K24Rik 9.67E-03 8.05E-02 2.960
Bahcc1 1.79E-03 2.65E-02 2.959
3110039M20Rik 3.82E-05 2.32E-03 2.958
1700109H08Rik 5.10E-04 1.19E-02 2.955
Arsb 1.48E-07 8.16E-05 2.952
Six3 1.20E-02 9.29E-02 2.950
Anxa2 2.47E-07 1.12E-04 2.948
Glis3 2.22E-06 3.76E-04 2.944
Cxcl1 7.30E-03 6.68E-02 2.936
Gm43792 7.67E-03 6.88E-02 2.935
Sp8 3.83E-04 9.84E-03 2.933
Apcdd1 3.64E-04 9.55E-03 2.922
Sntb2 3.43E-07 1.32E-04 2.909
Gm43984 2.26E-03 3.06E-02 2.906
Gm43519 1.16E-02 9.07E-02 2.895
Prodh 1.99E-04 6.38E-03 2.895
Gm6611 2.63E-03 3.40E-02 2.891
Tgtp1 8.88E-03 7.59E-02 2.890

197

Gm43953 4.57E-03 4.92E-02 2.889
2510016D11Rik 6.24E-03 6.07E-02 2.882
Zfp804a 1.25E-02 9.51E-02 2.882
Dusp27 1.97E-03 2.80E-02 2.881
Gm9996 1.84E-03 2.70E-02 2.877
Matn2 7.76E-04 1.56E-02 2.866
Sfrp4 3.34E-03 3.97E-02 2.861
Gxylt2 1.09E-03 1.95E-02 2.854
Rps13-ps2 9.10E-03 7.72E-02 2.853
Tnfrsf1b 3.78E-06 5.29E-04 2.850
Atp8b5 8.84E-04 1.71E-02 2.850
Srcin1 4.32E-03 4.73E-02 2.849
Ccdc141 8.01E-06 8.33E-04 2.842
Ccdc85b 5.02E-06 6.24E-04 2.840
Gbe1 9.55E-05 4.09E-03 2.836
Faxc 1.36E-03 2.24E-02 2.833
Gm42731 8.50E-03 7.39E-02 2.820
Tbc1d8 5.57E-05 2.94E-03 2.808
A330074K22Rik 1.53E-03 2.41E-02 2.807
Pvt1 3.13E-04 8.67E-03 2.806
Gm13092 7.62E-04 1.55E-02 2.803
Ddit4l 1.69E-03 2.57E-02 2.797
4833412C15Rik 1.10E-03 1.96E-02 2.795
Gm37091 5.28E-04 1.21E-02 2.795
Ifitm3 2.42E-05 1.74E-03 2.794
Sdc1 1.18E-07 7.40E-05 2.789
Notch3 1.47E-05 1.26E-03 2.785
Ntn4 4.65E-06 5.98E-04 2.784
Slc4a4 1.12E-05 1.04E-03 2.784
Ankrd1 4.50E-05 2.61E-03 2.783
Plxdc2 6.47E-05 3.18E-03 2.782
Gm47572 6.87E-03 6.44E-02 2.781
Fblim1 2.86E-07 1.24E-04 2.781
Gm43658 6.46E-03 6.19E-02 2.777
Gm13988 2.23E-03 3.03E-02 2.770
Gm37855 3.31E-04 8.99E-03 2.770
Gm12905 3.69E-04 9.59E-03 2.770
Camk2b 3.94E-05 2.37E-03 2.765
Fam110b 4.28E-04 1.05E-02 2.764
C430019N01Rik 1.05E-02 8.47E-02 2.761
Tgfb2 5.11E-05 2.78E-03 2.761
Gm9752 1.13E-02 8.95E-02 2.759
Gm49308 8.77E-03 7.55E-02 2.755
Ifi202b 2.23E-06 3.76E-04 2.750
Parp9 1.72E-04 5.78E-03 2.749
Mylk 1.25E-04 4.78E-03 2.738
Grhl2 7.29E-03 6.68E-02 2.738
5330406M23Rik 7.53E-03 6.80E-02 2.729
Gm43637 3.45E-03 4.06E-02 2.728
Tnfsf15 1.67E-03 2.56E-02 2.726
Gm43484 1.99E-03 2.83E-02 2.726
Dtx3l 3.30E-05 2.12E-03 2.726
Apln 6.32E-03 6.10E-02 2.724
Mt1 3.70E-04 9.61E-03 2.722
Gm49322 1.21E-03 2.09E-02 2.714
Gm49510 1.05E-02 8.49E-02 2.711
9430034N14Rik 1.14E-02 8.99E-02 2.708
Gm19972 3.70E-03 4.26E-02 2.707
Tnfrsf23 3.23E-04 8.83E-03 2.706
Nrp1 1.51E-05 1.27E-03 2.705
Gm43421 6.78E-03 6.38E-02 2.702
Cda 2.78E-03 3.52E-02 2.696
Gm42702 5.79E-03 5.77E-02 2.688
Fat4 2.11E-06 3.68E-04 2.686
Cadm1 4.85E-06 6.12E-04 2.686
Gm38312 6.45E-04 1.38E-02 2.686

198

Them6 1.03E-04 4.27E-03 2.683
Gm42546 1.47E-03 2.36E-02 2.680
Ank 2.13E-06 3.68E-04 2.679
Gm43336 7.65E-03 6.88E-02 2.674
Anxa8 2.26E-04 6.96E-03 2.669
Gm13502 3.71E-05 2.29E-03 2.665
Gm48958 4.47E-03 4.86E-02 2.664
Ugdh 2.09E-07 1.01E-04 2.661
Slc6a6 2.43E-05 1.75E-03 2.661
B930095G15Rik 4.62E-04 1.10E-02 2.657
Il18bp 2.28E-03 3.08E-02 2.657
Lynx1 3.53E-04 9.37E-03 2.656
Osgin1 6.40E-05 3.17E-03 2.655
Notch1 2.67E-05 1.84E-03 2.653
Gm49759 2.09E-03 2.91E-02 2.652
Krt80 4.00E-06 5.53E-04 2.649
Parm1 4.39E-05 2.56E-03 2.649
Myo1e 3.35E-07 1.31E-04 2.649
2700078F05Rik 2.61E-03 3.37E-02 2.639
Il1r1 1.25E-04 4.77E-03 2.637
Jag1 7.44E-06 7.95E-04 2.632
Ier5l 7.22E-04 1.49E-02 2.627
Rragd 1.70E-05 1.38E-03 2.624
Gm37978 4.62E-03 4.95E-02 2.621
Gm20186 3.77E-04 9.74E-03 2.619
Ext1 8.12E-07 2.02E-04 2.617
9330162G02Rik 4.88E-04 1.15E-02 2.617
Nhsl2 3.09E-05 2.04E-03 2.614
Anxa5 6.84E-07 1.78E-04 2.612
Gm43813 1.13E-02 8.94E-02 2.611
Asap1 5.75E-06 6.79E-04 2.611
BC037039 3.65E-04 9.57E-03 2.606
Pde4a 3.66E-04 9.58E-03 2.604
Acsl5 5.12E-07 1.54E-04 2.603
Pxylp1 1.28E-05 1.14E-03 2.599
Herc3 1.46E-04 5.26E-03 2.599
Map3k5 2.53E-03 3.31E-02 2.590
Lgals3 1.47E-05 1.25E-03 2.586
Pak1 3.27E-04 8.93E-03 2.584
Mthfd2l 6.34E-05 3.15E-03 2.582
Itgb8 4.95E-05 2.76E-03 2.581
Foxq1 7.52E-04 1.54E-02 2.580
Il13ra1 4.07E-06 5.55E-04 2.579
Gm44822 5.91E-03 5.86E-02 2.578
Dpysl3 5.88E-05 3.00E-03 2.577
Bhlhe41 4.69E-05 2.68E-03 2.576
Tnfrsf22 5.28E-04 1.21E-02 2.576
Timp3 1.44E-03 2.33E-02 2.574
Steap3 2.13E-05 1.60E-03 2.573
Gm9732 9.55E-04 1.79E-02 2.570
Slc12a7 7.75E-06 8.18E-04 2.569
Gm37186 7.50E-03 6.78E-02 2.568
Gadd45g 5.68E-05 2.97E-03 2.566
Fgf2 1.17E-02 9.14E-02 2.565
Ldha 6.38E-07 1.72E-04 2.561
G930009F23Rik 7.27E-04 1.50E-02 2.558
Heg1 6.74E-07 1.77E-04 2.553
Ablim1 8.37E-07 2.05E-04 2.551
Gm15998 3.63E-03 4.21E-02 2.549
Atg9b 5.87E-03 5.83E-02 2.549
Arhgef10l 5.61E-06 6.69E-04 2.548
Idi1 2.60E-06 4.28E-04 2.547
Gm19557 1.84E-03 2.70E-02 2.545
Tmem86a 5.79E-03 5.77E-02 2.544
Fgf5 1.17E-03 2.04E-02 2.536
Sod3 9.35E-05 4.05E-03 2.534

199

Ecm1 9.61E-06 9.31E-04 2.533
Elovl6 2.64E-05 1.82E-03 2.530
Ttyh2 9.31E-07 2.22E-04 2.516
Gm10277 4.56E-03 4.92E-02 2.516
Gm15337 5.42E-03 5.53E-02 2.514
Gm14288 2.45E-05 1.75E-03 2.513
Gm49484 4.17E-03 4.62E-02 2.512
Dnah7b 6.21E-05 3.11E-03 2.512
Rmnd1 2.63E-05 1.82E-03 2.511
Hivep2 5.86E-06 6.85E-04 2.500
Inf2 1.46E-04 5.26E-03 2.500
Gm43571 1.26E-02 9.56E-02 2.500
Slc24a3 1.19E-04 4.60E-03 2.493
Ereg 3.23E-03 3.87E-02 2.493
Rab32 9.82E-06 9.46E-04 2.492
Apol7a 5.30E-04 1.21E-02 2.491
Anxa1 9.64E-06 9.31E-04 2.489
Arid5b 1.45E-04 5.24E-03 2.489
Gm48478 9.84E-03 8.16E-02 2.488
Ctnnal1 2.38E-05 1.73E-03 2.488
Gng11 7.11E-03 6.58E-02 2.482
Retreg1 9.72E-07 2.28E-04 2.476
Carmil3 5.40E-03 5.51E-02 2.472
Dap 3.68E-07 1.38E-04 2.470
Ugcg 1.16E-05 1.06E-03 2.470
Serpine1 7.24E-05 3.39E-03 2.467
Gm16153 5.08E-03 5.29E-02 2.466
Lgr4 2.64E-06 4.32E-04 2.462
Gm48027 1.31E-02 9.84E-02 2.457
Mt2 2.54E-03 3.31E-02 2.453
Erap1 4.80E-05 2.70E-03 2.452
1700034J05Rik 5.53E-03 5.59E-02 2.451
Cx3cl1 2.62E-04 7.73E-03 2.450
Lurap1l 6.70E-05 3.23E-03 2.442
Glipr1 1.12E-05 1.04E-03 2.441
Aspg 1.84E-03 2.70E-02 2.440
Adamts1 3.38E-04 9.13E-03 2.437
Gsap 1.66E-04 5.63E-03 2.436
Eps8 6.42E-07 1.72E-04 2.436
Prcd 1.07E-02 8.59E-02 2.431
Fam185a 1.90E-03 2.76E-02 2.429
Eml1 2.37E-05 1.73E-03 2.426
Gm42819 1.40E-03 2.28E-02 2.424
B130046B21Rik 3.78E-03 4.33E-02 2.418
Fhl3 5.61E-05 2.96E-03 2.417
Mmp28 7.71E-04 1.55E-02 2.416
Rnf150 7.74E-04 1.56E-02 2.416
Gm37606 7.69E-03 6.90E-02 2.414
C130089K02Rik 5.25E-03 5.41E-02 2.413
Hmgcs1 5.23E-06 6.43E-04 2.412
Gm12693 4.21E-03 4.66E-02 2.412
Gas6 5.14E-05 2.80E-03 2.409
Cpt1a 6.73E-07 1.77E-04 2.408
C030015A19Rik 2.43E-04 7.31E-03 2.407
Pdzrn3 1.30E-04 4.90E-03 2.406
Nptxr 3.04E-06 4.72E-04 2.405
Gm45532 1.15E-02 9.03E-02 2.405
Gbp2b 4.26E-03 4.70E-02 2.404
Ddit4 1.54E-04 5.41E-03 2.401
Gm19409 8.82E-03 7.57E-02 2.401
Gm44187 1.14E-02 8.99E-02 2.400
Cubn 1.47E-04 5.27E-03 2.392
Zfp36l2 8.26E-07 2.04E-04 2.391
Masp1 6.01E-06 6.92E-04 2.389
Dipk1b 1.75E-03 2.62E-02 2.388
Zmiz1os1 1.30E-02 9.74E-02 2.385

200

Pqlc2 1.39E-03 2.27E-02 2.384
Cpt1b 4.80E-03 5.09E-02 2.378
Car5b 9.28E-05 4.04E-03 2.377
Abr 7.55E-06 8.04E-04 2.372
Pfkp 1.39E-05 1.21E-03 2.372
Thsd7a 1.23E-05 1.11E-03 2.369
Lgals1 1.10E-05 1.03E-03 2.367
Btc 4.45E-03 4.85E-02 2.365
Tgfbr3 7.63E-05 3.52E-03 2.365
Ccdc68 1.91E-03 2.76E-02 2.365
Gm11960 1.06E-02 8.55E-02 2.364
Gm44699 2.68E-03 3.43E-02 2.364
Mmp15 6.44E-04 1.38E-02 2.362
Hipk2 1.07E-06 2.36E-04 2.360
Dusp10 5.09E-05 2.78E-03 2.358
Gjb4 1.41E-03 2.30E-02 2.358
Parp3 6.86E-06 7.59E-04 2.358
Peak1 5.09E-05 2.78E-03 2.355
AI506816 9.93E-04 1.83E-02 2.353
Gm37621 1.28E-02 9.68E-02 2.350
Gm47087 2.68E-03 3.43E-02 2.348
Zfp972 1.04E-02 8.46E-02 2.342
Fbxo4 4.99E-05 2.77E-03 2.333
Ccnd2 1.15E-04 4.55E-03 2.333
Col4a1 1.52E-06 3.13E-04 2.329
Gm10524 2.14E-04 6.69E-03 2.329
Col16a1 2.10E-04 6.57E-03 2.326
Insyn2a 2.18E-03 2.99E-02 2.319
Btbd2 3.77E-06 5.29E-04 2.315
Nmnat2 1.20E-03 2.09E-02 2.310
Ptchd4 1.37E-05 1.20E-03 2.309
Gm42748 1.33E-02 9.90E-02 2.309
Smad3 6.27E-06 7.14E-04 2.308
9-Sep 3.94E-06 5.49E-04 2.301
Mknk2 4.81E-06 6.10E-04 2.298
3110056K07Rik 1.30E-03 2.18E-02 2.296
Isg15 6.49E-03 6.20E-02 2.291
Abhd14b 5.81E-03 5.78E-02 2.288
Tiam2 2.38E-05 1.73E-03 2.286
Slc27a1 7.61E-06 8.07E-04 2.283
Gm4767 1.23E-02 9.46E-02 2.280
Chst10 3.86E-03 4.38E-02 2.278
Pdgfra 2.48E-04 7.42E-03 2.277
Serpinb8 2.00E-03 2.83E-02 2.274
Ppm1h 4.45E-04 1.08E-02 2.274
9330151L19Rik 1.07E-02 8.58E-02 2.273
Zmiz1 1.62E-03 2.52E-02 2.273
Btg1 3.74E-05 2.30E-03 2.272
D10Wsu102e 2.00E-06 3.63E-04 2.271
Gchfr 7.10E-03 6.58E-02 2.268
Mllt3 4.22E-06 5.63E-04 2.268
Slc30a4 3.55E-06 5.11E-04 2.268
Tgfb3 1.23E-04 4.72E-03 2.268
Snx29 4.39E-04 1.07E-02 2.267
Esr1 7.06E-03 6.56E-02 2.264
A730011C13Rik 6.62E-03 6.28E-02 2.260
Prtg 3.21E-04 8.81E-03 2.258
Klf8 7.16E-03 6.61E-02 2.256
Prkca 1.83E-04 6.02E-03 2.253
Gm42798 1.74E-03 2.61E-02 2.251
Dhrs9 1.16E-04 4.58E-03 2.251
Gm45769 8.19E-03 7.21E-02 2.247
Trim65 2.72E-03 3.46E-02 2.242
Gm17491 4.08E-03 4.56E-02 2.242
Gfod1 2.89E-03 3.61E-02 2.235
Nfkbia 1.45E-03 2.34E-02 2.229

201

Plk2 1.69E-05 1.38E-03 2.229
Stat5a 3.00E-06 4.69E-04 2.228
Mettl25 1.18E-03 2.06E-02 2.228
Pde8a 9.19E-05 4.01E-03 2.228
Aox1 4.97E-03 5.21E-02 2.228
Gm49599 1.98E-03 2.82E-02 2.222
4632427E13Rik 6.63E-03 6.28E-02 2.222
Rab27b 1.09E-04 4.42E-03 2.221
Esyt2 2.16E-05 1.61E-03 2.221
Kcnmb4 6.58E-03 6.26E-02 2.221
Il33 1.85E-03 2.71E-02 2.220
Gm37760 3.47E-03 4.07E-02 2.219
Trp53inp1 2.52E-04 7.52E-03 2.219
Ttc28 2.97E-05 1.99E-03 2.218
Gm42653 2.79E-03 3.53E-02 2.218
Polm 1.92E-04 6.19E-03 2.215
Gm37940 7.13E-03 6.59E-02 2.215
Ahnak2 7.85E-04 1.57E-02 2.214
11-Sep 4.65E-05 2.66E-03 2.213
Syt1 5.04E-03 5.27E-02 2.213
Zfp950 1.91E-03 2.76E-02 2.212
Gm49405 9.56E-04 1.79E-02 2.212
Sh3bp2 9.11E-05 3.99E-03 2.211
Cox16 6.83E-05 3.27E-03 2.211
Timp2 6.37E-04 1.38E-02 2.207
Eppk1 8.08E-04 1.60E-02 2.207
Gm4316 3.86E-03 4.38E-02 2.207
Mir99ahg 8.84E-03 7.58E-02 2.205
Nln 4.77E-06 6.08E-04 2.203
Gm45110 1.15E-02 9.02E-02 2.203
Tnxb 6.93E-04 1.45E-02 2.203
Gm14443 5.24E-03 5.41E-02 2.202
Gm21992 3.06E-04 8.57E-03 2.202
Ctla2b 1.31E-02 9.82E-02 2.202
Rasa4 6.67E-03 6.31E-02 2.202
Parp12 2.18E-04 6.75E-03 2.202
Gm42611 1.29E-02 9.73E-02 2.201
Gm29361 7.76E-03 6.94E-02 2.199
Gm38082 2.69E-03 3.44E-02 2.197
Id1 3.22E-03 3.87E-02 2.197
Trim16 9.94E-06 9.51E-04 2.195
Morrbid 3.08E-03 3.76E-02 2.191
Tbccd1 2.31E-05 1.70E-03 2.188
Ppp1r12b 2.44E-05 1.75E-03 2.187
Gm13966 1.03E-02 8.41E-02 2.187
Pip5k1b 2.52E-04 7.50E-03 2.186
Slc25a24 1.31E-06 2.73E-04 2.185
Npnt 5.09E-03 5.30E-02 2.185
Gm45779 7.25E-03 6.66E-02 2.184
Tmco4 1.43E-03 2.32E-02 2.182
Mertk 1.41E-04 5.17E-03 2.182
Kitl 3.34E-03 3.97E-02 2.179
9330159F19Rik 1.10E-03 1.96E-02 2.178
Akr1c14 1.09E-02 8.71E-02 2.174
Rasl11a 2.75E-03 3.49E-02 2.174
Pdgfc 4.32E-04 1.06E-02 2.167
Fam124a 2.15E-03 2.96E-02 2.166
Rc3h1 7.47E-05 3.47E-03 2.166
Gm37254 7.52E-03 6.79E-02 2.163
Dhrs3 7.08E-03 6.57E-02 2.158
Ptgfrn 5.36E-06 6.49E-04 2.158
Gm44168 4.15E-03 4.62E-02 2.155
Rere 6.75E-04 1.42E-02 2.153
Airn 4.38E-03 4.79E-02 2.153
Rin1 6.53E-06 7.31E-04 2.151
Tle3 1.07E-05 1.01E-03 2.148

202

Pstpip1 8.05E-03 7.11E-02 2.147
Tlr4 1.18E-04 4.59E-03 2.145
Efnb2 1.92E-03 2.76E-02 2.144
Cdkn2a 2.80E-06 4.55E-04 2.143
Hbegf 7.12E-05 3.36E-03 2.141
Frk 9.10E-05 3.99E-03 2.141
Acap3 2.77E-05 1.89E-03 2.138
Slc5a3 7.36E-06 7.90E-04 2.137
Emx2 1.32E-02 9.88E-02 2.137
Mbnl1 1.50E-05 1.27E-03 2.134
Ahr 9.71E-04 1.81E-02 2.133
Pakap 4.11E-04 1.02E-02 2.132
Lasp1 7.96E-06 8.32E-04 2.129
mt-Nd6 3.88E-03 4.40E-02 2.129
Mgst1 1.85E-03 2.71E-02 2.128
Gm11839 1.23E-02 9.47E-02 2.128
4732440D04Rik 5.73E-04 1.27E-02 2.127
Gm42863 9.12E-03 7.73E-02 2.127
Gm37422 3.42E-03 4.03E-02 2.125
Ece1 2.41E-05 1.74E-03 2.125
Tomm5 9.10E-06 8.98E-04 2.124
P3h2 6.22E-04 1.36E-02 2.124
Ifit3 6.60E-03 6.26E-02 2.124
Vav3 6.74E-05 3.24E-03 2.124
Cdk18 4.00E-05 2.39E-03 2.124
A430110C17Rik 8.86E-03 7.59E-02 2.121
Col7a1 6.08E-03 5.97E-02 2.121
Lratd2 1.18E-03 2.07E-02 2.117
Gm50147 2.68E-03 3.43E-02 2.116
A630089N07Rik 6.29E-05 3.14E-03 2.114
Cdc42ep2 2.53E-04 7.54E-03 2.114
B4galt5 1.01E-03 1.84E-02 2.110
Gm17137 1.25E-02 9.53E-02 2.109
Rgs3 1.59E-03 2.48E-02 2.108
9530056E24Rik 9.17E-03 7.75E-02 2.107
Ceacam1 8.43E-03 7.35E-02 2.106
Rbm47 3.22E-05 2.10E-03 2.104
Mxra8 1.81E-05 1.43E-03 2.104
Rbm4b 1.57E-03 2.46E-02 2.102
Itpr3 1.85E-04 6.08E-03 2.101
Tdrd7 2.37E-05 1.73E-03 2.096
Fasn 5.79E-05 2.98E-03 2.096
Tm4sf1 7.26E-06 7.85E-04 2.096
Xk 1.27E-02 9.62E-02 2.094
Plxnd1 1.68E-05 1.38E-03 2.094
B230307C23Rik 3.83E-04 9.84E-03 2.094
9330159M07Rik 1.27E-02 9.61E-02 2.094
Ctsb 6.55E-06 7.31E-04 2.091
Magi2 3.89E-05 2.36E-03 2.089
Ankrd13b 3.71E-05 2.29E-03 2.084
Irf9 1.37E-04 5.09E-03 2.079
Gm50012 5.81E-03 5.78E-02 2.079
Gdf11 6.47E-05 3.18E-03 2.078
C030034I22Rik 7.04E-03 6.55E-02 2.078
St5 7.78E-05 3.56E-03 2.077
Epha4 3.17E-03 3.83E-02 2.077
Trim2 1.26E-04 4.79E-03 2.077
Nr1d1 1.66E-04 5.64E-03 2.076
Asic1 3.86E-03 4.39E-02 2.076
Rapgef3 2.68E-04 7.84E-03 2.075
Cdk14 1.35E-05 1.18E-03 2.071
1500004A13Rik 1.70E-03 2.58E-02 2.071
Ywhaq-ps3 7.48E-03 6.76E-02 2.070
Mcc 1.71E-04 5.77E-03 2.070
Pik3r3 1.27E-03 2.15E-02 2.059
9630013D21Rik 7.58E-03 6.83E-02 2.059

203

Rai14 2.00E-05 1.53E-03 2.059
Col27a1 1.31E-04 4.94E-03 2.056
4930478M13Rik 8.85E-03 7.59E-02 2.056
Lonrf3 6.54E-05 3.19E-03 2.056
Lima1 1.91E-05 1.48E-03 2.055
Gm29609 4.94E-03 5.19E-02 2.052
Cd9 8.31E-06 8.49E-04 2.050
Afap1l2 1.15E-04 4.55E-03 2.048
Hes1 1.76E-03 2.63E-02 2.048
Gm5512 9.23E-03 7.78E-02 2.047
Syt11 6.72E-05 3.24E-03 2.045
Azin1 4.06E-06 5.55E-04 2.045
4933407K13Rik 7.07E-03 6.57E-02 2.045
Kcnab2 5.63E-05 2.96E-03 2.043
Eya4 1.48E-03 2.37E-02 2.040
F2r 2.28E-04 6.97E-03 2.038
Foxf1 5.11E-04 1.19E-02 2.038
Gm7435 4.40E-03 4.80E-02 2.038
Gm43817 1.29E-02 9.72E-02 2.037
Rasa3 3.27E-05 2.12E-03 2.037
Ddx58 1.55E-03 2.43E-02 2.036
Nqo1 5.10E-03 5.31E-02 2.036
Eya1 3.31E-04 8.99E-03 2.033
Epdr1 8.65E-05 3.86E-03 2.032
Fam43a 1.38E-04 5.11E-03 2.030
Fbln2 6.50E-04 1.39E-02 2.030
Fmnl2 1.47E-05 1.25E-03 2.029
Fam102b 2.95E-05 1.98E-03 2.025
Eno1 2.83E-05 1.91E-03 2.025
Stxbp5 1.52E-04 5.37E-03 2.024
Mmp19 7.16E-04 1.48E-02 2.024
Ttc41 7.35E-03 6.70E-02 2.024
Gm7074 4.70E-03 5.02E-02 2.019
Gm28635 3.78E-03 4.32E-02 2.019
Fbxo36 2.72E-03 3.46E-02 2.018
4931428F04Rik 8.11E-04 1.60E-02 2.018
Gm44901 7.80E-03 6.97E-02 2.016
Myc 1.51E-05 1.27E-03 2.015
Cpped1 1.73E-05 1.39E-03 2.004
Htr1b 4.52E-03 4.89E-02 2.003
Arhgef9 8.02E-03 7.10E-02 2.003
S100a1 1.14E-03 2.01E-02 2.001
Zfp334 5.05E-05 2.78E-03 -2.002
Gypc 8.77E-05 3.90E-03 -2.005
Cnpy2 5.79E-06 6.81E-04 -2.005
Akt3 1.53E-05 1.28E-03 -2.007
Osbpl2 3.98E-05 2.38E-03 -2.007
Usp2 4.40E-03 4.80E-02 -2.007
Klf5 1.13E-05 1.04E-03 -2.008
Herpud1 3.48E-04 9.30E-03 -2.009
Mcam 5.95E-05 3.03E-03 -2.010
Gpm6a 1.14E-02 8.99E-02 -2.011
Klhdc3 1.10E-05 1.03E-03 -2.012
Metrn 5.26E-04 1.21E-02 -2.013
Gm24407 2.28E-03 3.08E-02 -2.014
Ypel5 2.09E-05 1.59E-03 -2.017
Mettl5 2.18E-04 6.75E-03 -2.017
Cenpq 1.37E-04 5.11E-03 -2.021
Kif21b 4.43E-04 1.07E-02 -2.021
Slc1a5 4.22E-06 5.63E-04 -2.022
Gm26707 3.64E-03 4.22E-02 -2.024
Slc25a40 5.87E-06 6.85E-04 -2.024
Sptbn2 1.71E-03 2.58E-02 -2.025
Leprotl1 8.66E-06 8.76E-04 -2.032
Glis2 1.79E-03 2.65E-02 -2.035
Dzip1 4.47E-04 1.08E-02 -2.038

204

Pcdh10 2.81E-03 3.54E-02 -2.039
Mr1 1.69E-03 2.57E-02 -2.042
Foxp2 7.14E-04 1.48E-02 -2.043
Zrsr1 1.92E-05 1.49E-03 -2.043
Dag1 2.16E-03 2.97E-02 -2.044
Gm42826 5.16E-03 5.35E-02 -2.046
Txndc16 1.02E-04 4.24E-03 -2.046
Frs3 2.32E-03 3.12E-02 -2.047
Spin4 3.06E-03 3.75E-02 -2.048
Poli 5.21E-04 1.20E-02 -2.048
Selenom 5.08E-06 6.29E-04 -2.048
Ptprq 9.12E-03 7.73E-02 -2.048
Ada 1.19E-04 4.60E-03 -2.055
Tspan15 6.83E-05 3.27E-03 -2.055
Obsl1 3.62E-06 5.19E-04 -2.056
Dram2 3.93E-04 9.99E-03 -2.061
Klc4 5.55E-05 2.94E-03 -2.062
Gm8010 1.01E-02 8.28E-02 -2.066
Mbd6 3.22E-03 3.87E-02 -2.066
Zcwpw1 2.87E-05 1.94E-03 -2.068
Polr3k 4.40E-04 1.07E-02 -2.068
Kcnip3 1.16E-02 9.07E-02 -2.068
1700105P06Rik 1.15E-02 9.03E-02 -2.086
Mybl2 1.89E-04 6.14E-03 -2.087
Atad2b 3.43E-06 5.05E-04 -2.095
Arhgef26 9.35E-03 7.85E-02 -2.101
Dus2 6.36E-04 1.38E-02 -2.104
Otub2 7.85E-04 1.57E-02 -2.106
Prr15l 7.97E-03 7.07E-02 -2.107
Myrf 7.18E-03 6.61E-02 -2.110
Ttc9 5.99E-03 5.92E-02 -2.114
Dgat2 3.17E-04 8.75E-03 -2.119
Sall3 6.19E-03 6.04E-02 -2.121
Sema3b 3.69E-04 9.59E-03 -2.121
Hmgxb4 5.31E-06 6.49E-04 -2.121
Pcdh19 9.04E-03 7.69E-02 -2.121
Gm6863 3.54E-06 5.11E-04 -2.122
Nudt18 1.10E-04 4.43E-03 -2.126
Phyhd1 6.87E-03 6.44E-02 -2.128
Cacna1g 9.90E-04 1.83E-02 -2.132
Dnajc3 1.06E-03 1.91E-02 -2.132
Zfp354c 3.04E-03 3.74E-02 -2.139
Zfp449 8.22E-04 1.62E-02 -2.140
Gnao1 1.24E-03 2.11E-02 -2.142
Fbn2 2.86E-04 8.22E-03 -2.142
Naca 2.40E-05 1.74E-03 -2.148
Irx5 1.20E-02 9.28E-02 -2.150
Tmcc2 1.18E-04 4.59E-03 -2.151
Tmem121 2.39E-03 3.20E-02 -2.154
Lmx1b 3.61E-05 2.26E-03 -2.155
Rbpms 9.45E-05 4.07E-03 -2.157
Slc2a8 2.51E-03 3.29E-02 -2.160
Med16 5.66E-06 6.71E-04 -2.163
Igf2 2.93E-04 8.32E-03 -2.165
Usp29 6.26E-03 6.07E-02 -2.166
Pde6d 4.74E-05 2.69E-03 -2.166
Ankrd44 2.52E-05 1.78E-03 -2.167
Fem1b 6.03E-06 6.92E-04 -2.169
Esrp2 1.67E-03 2.56E-02 -2.171
B4galt4 2.49E-05 1.77E-03 -2.172
Pdia4 1.34E-03 2.22E-02 -2.183
Mlana 3.57E-05 2.26E-03 -2.184
Nemp1 8.18E-06 8.43E-04 -2.184
Gm25939 1.88E-04 6.13E-03 -2.184
Trpt1 2.39E-03 3.20E-02 -2.189
Esyt3 6.37E-04 1.38E-02 -2.196

205

Plxnb1 1.24E-02 9.49E-02 -2.198
8430429K09Rik 1.19E-03 2.07E-02 -2.199
Smyd1 2.38E-03 3.18E-02 -2.206
Nfia 1.09E-03 1.95E-02 -2.210
Gm47571 3.04E-03 3.73E-02 -2.212
Tdrkh 1.08E-05 1.02E-03 -2.212
Tysnd1 1.00E-03 1.84E-02 -2.213
Arhgap44 1.69E-03 2.57E-02 -2.213
Eno3 4.86E-04 1.15E-02 -2.215
Neu3 4.03E-03 4.53E-02 -2.216
R3hdm2 8.22E-06 8.44E-04 -2.226
Gm26330 2.64E-03 3.41E-02 -2.227
Prdm6 4.77E-04 1.13E-02 -2.229
Ccn1 1.05E-04 4.30E-03 -2.234
Etl4 1.12E-03 1.98E-02 -2.234
Pcsk5 1.20E-03 2.08E-02 -2.238
Thnsl2 1.91E-03 2.76E-02 -2.245
Plppr3 1.07E-02 8.63E-02 -2.259
Gbx2 2.93E-03 3.64E-02 -2.261
Rnf126 2.04E-06 3.65E-04 -2.263
Chn2 6.43E-05 3.18E-03 -2.267
Gdf15 4.23E-04 1.05E-02 -2.267
Sh3tc2 1.62E-05 1.34E-03 -2.268
Gls2 1.79E-03 2.65E-02 -2.270
D330041H03Rik 4.27E-05 2.51E-03 -2.273
Dnajb9 2.06E-06 3.67E-04 -2.275
Bnc1 1.20E-02 9.29E-02 -2.277
Mmd 1.82E-04 6.01E-03 -2.281
Trnp1 7.35E-04 1.51E-02 -2.281
Gm23444 3.17E-04 8.74E-03 -2.286
Pex6 5.39E-05 2.89E-03 -2.290
Slit2 6.02E-05 3.06E-03 -2.295
Tead4 7.02E-05 3.33E-03 -2.298
Plekha4 5.67E-03 5.68E-02 -2.300
Peg3 1.21E-02 9.33E-02 -2.301
Gm7623 1.41E-03 2.30E-02 -2.303
Csdc2 8.03E-03 7.10E-02 -2.304
Styk1 6.20E-04 1.35E-02 -2.307
R3hdm4 1.32E-06 2.73E-04 -2.307
Prr5 5.03E-04 1.17E-02 -2.311
Suox 3.77E-04 9.74E-03 -2.312
Snord17 6.06E-04 1.33E-02 -2.313
Slc45a3 3.20E-03 3.86E-02 -2.321
Baz2a 7.02E-05 3.33E-03 -2.321
Paqr3 2.96E-04 8.36E-03 -2.325
Mipol1 1.02E-04 4.24E-03 -2.328
Ptges3 7.25E-06 7.85E-04 -2.330
Stk39 4.45E-05 2.59E-03 -2.331
Palm 1.87E-06 3.49E-04 -2.333
Tagln 1.70E-03 2.58E-02 -2.333
Nup210 7.69E-05 3.54E-03 -2.339
4921524J17Rik 2.96E-06 4.65E-04 -2.343
Ptgis 6.51E-05 3.19E-03 -2.344
Alox5ap 8.81E-03 7.56E-02 -2.345
Pde3b 1.42E-04 5.21E-03 -2.347
Tspan2 9.51E-05 4.08E-03 -2.351
Gca 4.99E-05 2.77E-03 -2.352
Steap1 2.77E-05 1.89E-03 -2.356
Pycr1 1.11E-03 1.98E-02 -2.356
Arhgap28 1.81E-05 1.43E-03 -2.356
Slc46a3 2.92E-03 3.63E-02 -2.368
Itih2 3.21E-03 3.87E-02 -2.369
Sulf1 9.96E-04 1.83E-02 -2.374
Scarna6 1.42E-04 5.20E-03 -2.376
n-R5s106 1.18E-02 9.21E-02 -2.380
Ptgs2 5.35E-03 5.49E-02 -2.383

206

Msi1 5.00E-05 2.77E-03 -2.385
Zbtb18 4.11E-06 5.56E-04 -2.386
Syde2 1.19E-05 1.08E-03 -2.387
Col4a4 2.06E-03 2.89E-02 -2.393
Upk3b 2.11E-03 2.92E-02 -2.393
Pcdh11x 7.93E-04 1.58E-02 -2.403
Gm24497 2.11E-03 2.92E-02 -2.407
Tex15 1.46E-05 1.25E-03 -2.420
Des 1.18E-04 4.59E-03 -2.427
Gpc4 1.04E-03 1.88E-02 -2.428
Tmem259 7.64E-07 1.92E-04 -2.428
Cpq 3.30E-06 4.95E-04 -2.429
Gm24265 1.36E-03 2.24E-02 -2.431
Satb1 4.09E-06 5.56E-04 -2.435
Atp5b 4.07E-05 2.41E-03 -2.436
Gm23201 1.38E-03 2.26E-02 -2.440
Amer1 1.42E-03 2.32E-02 -2.455
Ero1lb 3.75E-04 9.72E-03 -2.455
Wnt5a 2.01E-04 6.41E-03 -2.462
Sft2d2 4.42E-07 1.47E-04 -2.471
Pard6a 6.50E-04 1.39E-02 -2.471
Polr2e 1.88E-06 3.49E-04 -2.486
Gm45716 2.06E-04 6.49E-03 -2.492
Pip4k2c 4.36E-07 1.47E-04 -2.494
Cnn1 1.19E-02 9.24E-02 -2.495
Pmaip1 1.70E-05 1.38E-03 -2.501
Snord22 6.45E-05 3.18E-03 -2.502
Cracr2b 3.81E-03 4.35E-02 -2.505
Arhgap29 8.92E-06 8.88E-04 -2.515
Tuba4a 2.36E-04 7.16E-03 -2.515
Rnf144b 3.45E-05 2.19E-03 -2.525
Pdlim4 5.24E-05 2.83E-03 -2.532
Efnb1 1.78E-04 5.94E-03 -2.534
Rny3 3.66E-05 2.27E-03 -2.537
Dctn2 6.36E-07 1.72E-04 -2.538
Gm50321 5.11E-03 5.32E-02 -2.538
Fstl3 1.32E-02 9.85E-02 -2.545
Ints10 4.51E-05 2.61E-03 -2.551
Ptgs1 8.39E-03 7.33E-02 -2.556
Gm26232 4.23E-03 4.67E-02 -2.557
Insl6 1.02E-04 4.24E-03 -2.562
Rpia 2.13E-06 3.68E-04 -2.567
Polrmt 1.65E-06 3.18E-04 -2.583
Jazf1 5.57E-04 1.25E-02 -2.585
Fkbp1b 1.07E-02 8.61E-02 -2.588
Myrip 3.22E-04 8.81E-03 -2.600
Bsg 1.70E-06 3.25E-04 -2.604
2810408A11Rik 1.59E-03 2.48E-02 -2.607
Dpysl5 9.80E-05 4.16E-03 -2.607
Chac1 3.13E-06 4.83E-04 -2.608
Irx2 4.03E-04 1.01E-02 -2.610
Ccdc158 2.50E-03 3.29E-02 -2.610
Eps8l2 7.69E-03 6.90E-02 -2.614
B4galnt2 4.14E-06 5.58E-04 -2.640
Eci2 1.04E-04 4.27E-03 -2.644
Cpox 1.30E-05 1.16E-03 -2.647
Ptges3-ps 1.24E-06 2.62E-04 -2.652
Krt7 7.90E-03 7.03E-02 -2.654
Clgn 1.33E-02 9.93E-02 -2.657
Man2a1 4.22E-05 2.49E-03 -2.659
Gdf6 1.30E-03 2.19E-02 -2.661
Tll1 9.40E-05 4.05E-03 -2.662
Zbtb7c 6.51E-03 6.21E-02 -2.670
Lpcat2 1.30E-05 1.16E-03 -2.677
Prxl2c 1.19E-06 2.54E-04 -2.679
Gadd45a 2.12E-07 1.01E-04 -2.696

207

Tcf24 1.19E-05 1.08E-03 -2.697
Gm24950 7.84E-04 1.57E-02 -2.704
Mgarp 5.62E-03 5.65E-02 -2.711
Gm14230 1.17E-03 2.05E-02 -2.712
Dgka 8.49E-05 3.81E-03 -2.722
Gm23849 1.33E-03 2.21E-02 -2.729
Asrgl1 9.28E-06 9.12E-04 -2.729
Kdm5b 4.75E-07 1.48E-04 -2.732
Reep2 1.00E-03 1.84E-02 -2.745
Tmem88 1.15E-05 1.06E-03 -2.758
Ugt8a 1.18E-04 4.59E-03 -2.759
F5 1.99E-03 2.82E-02 -2.765
Slc39a8 7.89E-04 1.58E-02 -2.767
Nectin2 1.18E-06 2.53E-04 -2.769
Hr 1.70E-05 1.38E-03 -2.771
Pla2g12a 9.51E-06 9.29E-04 -2.783
Ddx25 7.02E-04 1.46E-02 -2.786
Rab6b 4.07E-04 1.02E-02 -2.786
Mtus1 3.76E-05 2.30E-03 -2.786
Nudt11 3.60E-03 4.19E-02 -2.793
Cyb5r1 1.67E-07 8.68E-05 -2.796
Wnt9a 4.48E-06 5.83E-04 -2.800
Scn1b 1.05E-02 8.52E-02 -2.803
Ptk7 4.75E-06 6.07E-04 -2.803
B4galnt1 2.59E-05 1.80E-03 -2.810
Nupr1 2.09E-06 3.68E-04 -2.811
Zxda 1.64E-04 5.60E-03 -2.821
Gm25313 6.44E-04 1.38E-02 -2.824
2410006H16Rik 1.25E-03 2.13E-02 -2.824
Tmem108 1.29E-03 2.18E-02 -2.841
Nrg2 3.98E-03 4.48E-02 -2.852
Hcn2 1.08E-06 2.36E-04 -2.853
Gpr85 4.08E-04 1.02E-02 -2.854
Gch1 7.75E-04 1.56E-02 -2.862
Aplp1 1.64E-06 3.18E-04 -2.866
Gm23804 2.11E-04 6.61E-03 -2.869
Snora24 4.53E-04 1.09E-02 -2.877
Extl1 1.11E-04 4.43E-03 -2.881
Acot2 3.35E-07 1.31E-04 -2.889
Stard5 2.21E-07 1.03E-04 -2.904
Sort1 6.91E-06 7.60E-04 -2.905
Spats2l 2.70E-03 3.45E-02 -2.914
Atp7b 7.16E-04 1.48E-02 -2.916
Ak5 1.02E-04 4.24E-03 -2.920
Ly9 7.69E-03 6.90E-02 -2.924
Kank4 1.34E-02 9.96E-02 -2.925
Zfp125 3.45E-05 2.19E-03 -2.930
Gli1 8.83E-04 1.71E-02 -2.938
Rdm1 8.83E-05 3.92E-03 -2.947
Rprd1a 5.33E-06 6.49E-04 -2.962
AU018091 4.62E-03 4.96E-02 -2.975
Hoxd9 2.25E-04 6.94E-03 -2.975
Bmpr1b 1.82E-03 2.68E-02 -2.980
Slain1 6.53E-05 3.19E-03 -2.993
Wnt11 3.68E-06 5.22E-04 -2.994
Gm49394 1.99E-05 1.52E-03 -3.011
S1pr5 3.92E-03 4.43E-02 -3.013
Gm23971 3.75E-04 9.71E-03 -3.023
Chrd 3.89E-04 9.93E-03 -3.029
Reep6 2.13E-05 1.60E-03 -3.031
Piwil2 4.65E-03 4.98E-02 -3.034
Shisa8 1.05E-03 1.90E-02 -3.036
Otog 3.12E-03 3.80E-02 -3.037
Gm22614 1.61E-03 2.50E-02 -3.056
C130021I20Rik 7.23E-05 3.39E-03 -3.066
Flrt1 1.02E-03 1.86E-02 -3.074

208

Wdr18 1.72E-06 3.26E-04 -3.079
Gpc3 2.66E-04 7.82E-03 -3.094
Reck 5.50E-05 2.93E-03 -3.098
Kdm7a 9.25E-08 6.35E-05 -3.101
Kcnd2 4.83E-03 5.11E-02 -3.105
B3gnt5 7.02E-06 7.69E-04 -3.115
Pou4f1 5.28E-05 2.84E-03 -3.154
Avil 5.14E-04 1.19E-02 -3.184
Galnt18 1.35E-03 2.23E-02 -3.204
Pkhd1l1 7.56E-03 6.82E-02 -3.210
Rgs7 5.67E-03 5.68E-02 -3.211
Lrrc75a 1.65E-06 3.18E-04 -3.226
Ubxn2a 1.93E-06 3.54E-04 -3.229
Limch1 1.10E-07 7.05E-05 -3.232
Cyp4b1 9.31E-03 7.82E-02 -3.248
Map3k9 8.15E-05 3.70E-03 -3.250
Dthd1 3.90E-04 9.93E-03 -3.278
Gm7489 3.75E-03 4.30E-02 -3.286
Cpne1 5.18E-06 6.39E-04 -3.286
Tgm2 1.19E-02 9.27E-02 -3.298
Gm15559 7.93E-04 1.58E-02 -3.299
Irf5 2.46E-05 1.75E-03 -3.300
Rhbdl3 8.25E-04 1.63E-02 -3.309
Arl4d 1.84E-05 1.44E-03 -3.367
Fbln5 4.15E-07 1.47E-04 -3.371
Prim1 4.51E-07 1.47E-04 -3.382
Trib3 2.03E-06 3.65E-04 -3.383
Bst1 7.07E-03 6.57E-02 -3.393
Ciart 6.61E-06 7.35E-04 -3.404
Arxes2 1.07E-03 1.92E-02 -3.421
Ppp1r15a 6.30E-07 1.72E-04 -3.426
Cdh3 1.32E-03 2.21E-02 -3.435
Gas1 2.63E-03 3.40E-02 -3.440
n-R5s193 2.16E-03 2.97E-02 -3.495
Adrb1 5.42E-04 1.23E-02 -3.495
Caskin1 5.22E-05 2.83E-03 -3.507
Pltp 6.39E-06 7.22E-04 -3.531
Erbb3 8.21E-08 5.78E-05 -3.555
Ocln 9.37E-05 4.05E-03 -3.569
Prss12 5.76E-05 2.98E-03 -3.574
4932441J04Rik 1.29E-02 9.69E-02 -3.580
Cdsn 4.01E-06 5.53E-04 -3.604
Cth 8.95E-05 3.96E-03 -3.610
Zfp185 2.79E-05 1.89E-03 -3.627
Calml4 2.24E-04 6.92E-03 -3.642
Hoxd4 9.07E-03 7.71E-02 -3.643
Wnk2 3.28E-04 8.94E-03 -3.643
Arid3a 1.03E-06 2.34E-04 -3.665
Ccdc36 3.20E-05 2.09E-03 -3.665
Gm15127 6.30E-03 6.09E-02 -3.730
Fbxw18 2.22E-03 3.02E-02 -3.744
Mmp2 4.44E-07 1.47E-04 -3.751
Gtpbp2 1.47E-08 2.18E-05 -3.791
Zfp521 9.88E-06 9.48E-04 -3.851
Gm3716 1.05E-02 8.54E-02 -3.871
Fbxo2 5.70E-05 2.97E-03 -3.903
Shmt2 1.01E-06 2.32E-04 -3.904
Arhgef25 6.07E-07 1.71E-04 -3.941
Tfcp2l1 5.72E-05 2.97E-03 -3.956
Dtx3 4.46E-06 5.83E-04 -3.996
Mycl 1.65E-04 5.62E-03 -4.029
Hoxd8 3.63E-07 1.38E-04 -4.072
Chchd10 1.57E-06 3.13E-04 -4.085
Lsr 3.18E-05 2.09E-03 -4.096
Pex5l 3.29E-05 2.12E-03 -4.107
Mars 1.51E-07 8.16E-05 -4.136

209

Folr1 3.25E-06 4.94E-04 -4.155
Col3a1 2.27E-04 6.96E-03 -4.159
Gm23472 2.57E-04 7.62E-03 -4.227
Ltbp4 4.28E-06 5.66E-04 -4.237
Gm15287 2.09E-06 3.68E-04 -4.245
Derl3 1.97E-03 2.80E-02 -4.254
Gm25890 3.43E-06 5.05E-04 -4.258
1810010H24Rik 1.07E-03 1.92E-02 -4.260
Col8a2 2.68E-04 7.84E-03 -4.264
Rftn1 6.24E-06 7.13E-04 -4.301
Adam22 1.32E-04 4.95E-03 -4.354
Ddit3 4.01E-08 3.54E-05 -4.421
Lgals7 1.16E-02 9.09E-02 -4.457
Tnfrsf11a 2.65E-04 7.80E-03 -4.461
Fgf22 1.14E-02 8.97E-02 -4.516
Gnb3 4.81E-03 5.10E-02 -4.571
Igsf9 2.15E-05 1.61E-03 -4.594
Hoxd3os1 4.87E-04 1.15E-02 -4.599
Prr32 8.04E-05 3.67E-03 -4.635
Plac9b 1.01E-03 1.84E-02 -4.639
Pnma5 1.08E-06 2.36E-04 -4.647
Coch 1.32E-02 9.89E-02 -4.672
Nudt10 1.84E-04 6.05E-03 -4.705
Tnfaip6 3.98E-03 4.48E-02 -4.715
Il11 7.99E-03 7.08E-02 -4.823
Acox2 3.02E-05 2.01E-03 -4.854
Aldh1a2 2.94E-03 3.65E-02 -4.910
Lcn2 3.56E-03 4.15E-02 -4.915
Gm24830 6.92E-05 3.30E-03 -4.954
Zbtb32 1.02E-02 8.34E-02 -5.000
Hunk 2.33E-05 1.71E-03 -5.045
Prph 1.25E-04 4.76E-03 -5.052
Luzp4 1.51E-03 2.40E-02 -5.088
Angptl6 2.56E-05 1.80E-03 -5.169
Mfsd2b 2.45E-03 3.24E-02 -5.183
Gm41318 4.30E-03 4.73E-02 -5.223
P2rx3 5.66E-03 5.68E-02 -5.241
Arxes1 1.15E-03 2.02E-02 -5.337
Tex19.1 4.54E-07 1.47E-04 -5.360
Wfdc1 3.52E-04 9.36E-03 -5.559
n-R5s56 8.39E-03 7.33E-02 -5.566
EU599041 3.41E-03 4.02E-02 -5.569
Nkx6-2 1.30E-02 9.78E-02 -5.641
Rspo1 1.55E-03 2.43E-02 -5.704
Gm15091 7.53E-03 6.80E-02 -5.714
Neto1 5.63E-07 1.65E-04 -5.764
Sptb 1.40E-07 8.16E-05 -5.954
Slc27a6 1.97E-08 2.40E-05 -6.113
Ptgds 4.20E-04 1.04E-02 -6.209
Dach2 4.34E-07 1.47E-04 -6.269
Zfp141 6.16E-09 2.16E-05 -6.371
Arhgap9 1.58E-06 3.13E-04 -6.602
Vaultrc5 9.41E-04 1.78E-02 -6.605
Bcam 9.38E-09 2.16E-05 -6.893
Gm10439 1.27E-05 1.14E-03 -7.055
Ppef1 3.73E-05 2.29E-03 -7.112
Lrrn4 9.15E-04 1.74E-02 -7.289
Gsta3 2.39E-04 7.20E-03 -7.512
Rbms3 8.91E-03 7.62E-02 -7.555
Arhgdib 1.31E-08 2.16E-05 -7.627
Lrp2 4.65E-06 5.98E-04 -7.718
Gm2381 6.60E-03 6.26E-02 -7.764
Lbp 1.03E-06 2.34E-04 -8.162
Ppp1r16b 7.80E-10 7.31E-06 -8.165
Pcdh7 4.03E-08 3.54E-05 -8.258
Map7d2 1.71E-03 2.58E-02 -8.675

210

Igfbp2 4.59E-07 1.47E-04 -8.725
Gm24044 6.19E-03 6.04E-02 -8.848
n-R5s120 1.07E-03 1.92E-02 -9.030
Apela 8.19E-05 3.71E-03 -9.914
Bdh1 3.45E-04 9.26E-03 -11.086
Ppfibp2 4.74E-03 5.05E-02 -11.146
Fgf21 5.96E-06 6.90E-04 -11.362
Nkain4 6.58E-05 3.20E-03 -11.516
Gm13544 6.27E-03 6.07E-02 -12.244
Otogl 4.35E-03 4.77E-02 -14.222
Car6 1.81E-03 2.66E-02 -14.745
Slc47a1 5.55E-06 6.69E-04 -21.556
2400006E01Rik 2.35E-04 7.15E-03 -34.779
Klhl13 8.99E-05 3.96E-03 -34.932
Pgc 2.21E-03 3.01E-02 -46.721
Ccnb3 1.75E-06 3.30E-04 -67.238

211

Appendix 4: Supplemental Figure 1. Mycoplasma results for C18-4 cell line passage 29 (p29).

2% Agarose in TAE
Mycoplasma PCR Detection Kit abm #G238
DNA
Marker
1
18
Legend
1. MA10 Start
2. MA10 p8 Culty T1
3. MA10 p8 Culty T1
4. MA10 p3 Papa. T1
5. MA10 p3 Papa. T1
6. MA10 p10 Culty T2
7. MA10 p5 Papa. T2
8. MA10 Culty T3
9. MA10 Culty T3
10. MA10 Papa. T3
11. MA10 Papa. T3
12. MA10 Culty T4
13. MA10 Culty T4
14. MA10 Papa. T4
15. MA10 Papa. T4
16. C18-4 p29
17. Negative Control
18. Positive Control
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

Abstract (if available)

Abstract The objective of this thesis is to investigate the role of the eicosanoid biosynthetic pathway in regulating neonatal and juvenile stages of spermatogenesis, and the potential impact of compounds targeting this pathway on male reproductive health. These include prevalent environmental toxicants such as plasticizer metabolite Mono(2-ethylhexyl) Phthalate (MEHP) and soy-derived isoflavone genistein as well as commonly used drugs acetaminophen (ace) and ibuprofen (ibu). We hypothesize that the eicosanoid pathway plays a role in spermatogonial stem cell (SSC) fate and that chemicals disrupting eicosanoid synthesis can disturb germ cell growth, of which Cox enzymes and prostaglandin effectors can play important roles in regulating. Maintaining a balance between self-renewal and differentiation in the SSC pool is critical for sustaining male fertility. Starting at puberty, some SSCs differentiate to form spermatozoa, whereas others retain their stem cell identity to maintain an adequate SSC pool. Disruption to proper germ cell development has been shown to be a source of infertility and germ cell derived seminomas, emphasizing the relevance of this study for male reproductive health.
Three specific aims have been developed to address the objective of this project and our hypothesis. Firstly, we have characterized the eicosanoid pathway in the C18-4 cell line, an SSC model, and established that SSCs express eicosanoid pathway components and can produce measurable amounts of PGs. We also discovered that MEHP and GEN can modulate eicosanoid pathway expression and alter PG levels. Next, using a combination of treatments with pharmacological Cox inhibitors and gene silencing methods, we unveiled a mechanism of action behind Cox inhibition on SSCs, which involves Notch3 and the Notch signaling pathway. Lastly, 7-day postnatal treatment with human relevant doses of ace and ibu had potential to alter testis morphology and the transcriptome and functional pathways of several testicular cell types, including spermatogonia, with long-term consequences on testicular morphology and hormonal regulation in adult rats. Taken together these results support our hypothesis, and thus greater caution should be taken when exposing drugs that target eicosanoid biosynthesis to infants.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Effect of acetaminophen and ibuprofen on spermatogenesis and cell signaling mechanisms

PDF

Disruption of immature Sertoli cell functions by endocrine disruptors and analgesic drugs

PDF

Effects of endocrine disrupting chemicals and pharmaceuticals on rodent immature Sertoli cell functions

PDF

Identification of functional pathways altered in testis by in utero exposure to low doses of genistein and DEHP

PDF

Mitochondrial dynamics regulate Leydig cell health and integrity

PDF

Role of steroidogenic acute regulatory protein (STAR) and sterol carrier protein-x (SCPx) in the transport of cholesterol and other lipids

PDF

Pten regulates beta-cell regeneration intrinsically and independently of development

PDF

Reevaluation of the pregnenolone biosynthesis pathway in the human brain

PDF

Short term high fat diet (HFD) stimulates β cell proliferation through mTOR while the prolonged treatment induces β cell senescence via p27

PDF

The role of translocator protein (TSPO) in mediating the effects of mitotane in human adrenocortical carcinoma cells

PDF

Transcriptional regulation of lipid metabolism and exploring the specific roles of AKT isoforms

PDF

The role of CD99 in T cells

PDF

Fetal exposure to genistein and DEHP mixtures alters the expression of genes involved in critical testicular functions in adult rats

PDF

Effect of a mixture of two environmental chemicals on the interactions between macrophages and spermatogonia

PDF

PTEN loss antagonizes aging through promoting regeneration and prevents oxidative stress induced cell death

PDF

Steatosis induces Wnt/β-catenin pathway to stimulate proliferation of hepatic tumor initiating cells and promote liver cancer development

PDF

Akt1 deletion decrease proliferation in aged pancreatic beta-cells by arresting cell cycle in S phase

PDF

Investigation of butyrate’s effects on colonic stem cell development

PDF

Regulation of mitochondrial bioenergetics via PTEN (phosphatase and tensin homolog deleted on chromosome 10)/estrogen-related receptor alpha (ERRα) signaling

PDF

Towards DNA-directed assembly of pMHC multimers for detection of low-affinity T cells

Asset Metadata

Creator Tran-Guzman, Amy (author)

Core Title Eicosanoid regulation of spermatogonial stem cell development

School School of Pharmacy

Degree Doctor of Philosophy

Degree Program Molecular Pharmacology and Toxicology

Degree Conferral Date 2022-08

Publication Date 07/27/2024

Defense Date 04/07/2022

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag acetaminophen,COX1,COX2,eicosanoid pathway,endocrine disrupting compounds,genistein,ibuprofen,MEHP,NSAIDs,OAI-PMH Harvest,prostaglandins,spermatogonial stem cell

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Culty, Martine (committee chair), Gopalakrishna, Rayudu (committee member), Papadopoulos, Vassilios (committee member)

Creator Email amyt@usc.edu,amyt917@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC111375427

Unique identifier UC111375427

Legacy Identifier etd-TranGuzman-11022

Document Type Dissertation

Format application/pdf (imt)

Rights Tran-Guzman, Amy

Type texts

Source 20220728-usctheses-batch-962 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu