SLIT3 gene expression by 1,25(OH)₂D₃ in an endometriosis stromal cell line

PDF

Download

Open document

Flip pages

Summarize

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

SLIT3 Gene Expression by 1,25(OH)
2
D
3
in an Endometriosis Stromal Cell Line

Benjamin Liu
USC ID: 6793786257
Advisor: Dr. Sue Ingles
Molecular Epidemiology
Master of Science
University of Southern California
August 2017

Table of Contents
Abstract. ......................................................................................................................................... 3
Introduction ................................................................................................................................... 4
Materials and Methods ............................................................................................................... 12
Cell Culture ................................................................................................................................... 12
Library Construction and Next-Generation Sequencing ............................................................... 12
Bioinformatic Analysis ................................................................................................................. 13
Tissue Samples.............................................................................................................................. 13
Quantitative RT-PCR .................................................................................................................... 14
Results .......................................................................................................................................... 15
Differential gene expression by 1,25(OH)2D3
in Endometriosis Stromal Cells .................................................................................................... 15
Gene Ontology and Pathway Analysis ......................................................................................... 20
Quantitative RT-PCR of SLIT3 expression in tissue samples ...................................................... 27
Discussion..................................................................................................................................... 29
Conclusion ................................................................................................................................... 32
Acknowledgements ..................................................................................................................... 33
References .................................................................................................................................... 34
Appendix ...................................................................................................................................... 40
Abstract
Endometriosis is a common and sometimes severe health disorder in women. This
disease is characterized by the invasion of endometrial cells into other areas of the body such as
the peritoneum, ovaries, bladder, and more. Symptoms of endometriosis include but are not
limited to severe dysmenorrhea, deep dyspareunia, chronic pelvic pain, ovulation pain, menstrual
pain, irregular flow, premenstrual spotting, infertility, chronic fatigue, or any combination of the
list. Vitamin D3 is a secosteroid hormone that plays an important role not only in bone
development and maintenance but also in the cell proliferation, differentiation, and apoptosis.
Calcitriol (1,25(OH)2D3) is the biologically active metabolite of vitamin D3. Several studies have
shown that a vitamin D system is active in certain tissues of the female reproductive system. In
this study, we examine the effects of 1,25(OH)2D3 on gene expression in endometrial tissue
obtained from patients with endometriosis. Stromal cell lines derived endometriosis patients
were treated with 1,25(OH)2D3. Using RNA-Seq, differential gene expression was compared
between treated and untreated groups. 1627 genes were at least two-fold differentially expressed.
SLIT3, part of the axonal guidance signaling pathway and one of the most down-regulated genes,
plays an important role in inhibiting cell migration and invasion and stimulating
neuroangiogenesis. Using quantitative RT-PCR, the more invasive endometriosis lesions were
determined to have higher SLIT3 expression compared to lesions in the uterine cavity. These
results provide evidence for SLIT3 as a gene target in endometriosis and vitamin D3 as a
treatment for this disease.
Keywords: Endometriosis, Vitamin D, 1,25(OH)2D3, Quantitative RT-PCR, SLIT3, Next-
Generation Sequencing, RNA-Seq, Differential Gene Expression
SLIT3 Gene Expression by 1,25(OH)2D3 4

Introduction

Endometriosis is a very common and possibly debilitating health disorder in women.
Those affected often suffer symptoms including but not limited to severe dysmenorrhea, deep
dyspareunia, chronic pelvic pain, ovulation pain, menstrual pain, irregular flow, premenstrual
spotting, infertility, and chronic fatigue. Based on the symptoms alone, diagnosis is difficult due
to differences in presentation and overlap with other conditions causing abdominal pain such as
irritable bowel syndrome and pelvic inflammatory disease (Kennedy et al., 2005). Intriguingly,
endometriosis also shares many notable characteristics with cancer such as uncontrolled cell
proliferation, invasion into other organs, peripheral metastasis, and inflammation (Ingles et al.,
2017). Endometriosis can be seen as a chronic disease with treatment options that are medical,
surgical, or a combination of both.
Endometriosis has a history going back at least to ancient Egyptian times. Back then, it
was believed that the wombs of women who did not fulfill their duties of motherhood would
wander and contribute to all sorts of diseases. Through the centuries, detailed descriptions and
symptoms of endometriosis were slowly put together by physicians and scientists such as the
Hippocratic Corpus, Claudius Galen, and William Harvey; the idea of a singular disease was not
yet accepted as much of the work done by each person was independent of the others.
With the advancement of surgical techniques in the 19
th
century, French pathologists were able
to macroscopically distinguish cases of endometriosis by correlating the clinical symptoms
discovered throughout the years with new postmortem findings. In 1860, Karl von Rokitansky
made the connection between previous macroscopic observations and the cellular basis of
endometriosis using microscopic histopathology techniques. Rokitansky gave detailed accounts
SLIT3 Gene Expression by 1,25(OH)2D3 5

of what he described as endometrial-like glands and stroma that he had unexpectedly found in
tissue samples. It was not until 1940’s that scientists began to understand endometriosis with the
depth that they do today. Most of this is due largely to advancements made in laboratory research
and clinical trials due to modern endocrinologists. In 1957, hormonal contraceptives were
introduced and prescribed as a form of birth control as well as a viable method of treating
endometriosis. With the understanding that pregnancy could calm the symptoms associated with
the disease, this treatment was one of the most pivotal breakthroughs in modern reproductive
medicine (Nezhat, Nezhat, & Nezhat, 2012).
Today, approximately 6-10% of the general female population suffer from this disease.
The incidence rate for endometriosis is 2.37-2.49 per 1000 per year (Bulletti, Coccia, Battistoni,
& Borini, 2010). Asymptomatic endometriosis accounts for about 20-25% of the cases. On the
other hand, women with symptoms of pain, infertility, or both account for 35-50% of cases.
Additionally, about 25-50% of women suffering from infertility have endometriosis and 30-50%
of women with endometriosis suffer from infertility (Bulletti et al., 2010).
Currently, there are several theories on the pathogenesis of endometriosis. Retrograde
menstruation is the most supported and accepted theory. It is thought that during menstruation
viable endometrial tissue is pushed through the fallopian tubes into the peritoneal cavity where
they can implant, grow, and invade other organs. This viable tissue is likely to contain
endometrial epithelial progenitor and mesenchymal stem cells. The probability of such
phenomenon increases with factors such as early age at menarche, long duration of menstrual
flows, and any biological alteration at the molecular level leading to the process of cell
implantation (Vercellini, Vigano, Somigliana, & Fedele, 2014). For endometriosis infiltrating the
SLIT3 Gene Expression by 1,25(OH)2D3 6

cul-de-sac and uterosacral ligaments as well as rare cases in the male urinary system, the
embryonic rest theory suggests that aberrant migration and differentiation of Müllerian cells
during fetal development result in dormant implants that are induced to form endometrial tissue
under the appropriate stimuli (Sasson & Taylor, 2008; Schrodt, Alcorn, & Ibanez, 1980). The
coelomic metaplasia theory proposes that the cells lining the ovaries and peritoneum undergo a
metaplastic change into endometrium due to an undetermined cause. Finally, a lymphovascular
metastasis theory suggested that endometrial cells could spread through lymphatic or vascular
means. Despite the numerous theories on the pathogenesis of endometriosis that have been
proposed and studied, no single one can explain all of the different variations of the disease.
Over the last few decades, evidence for genetic factors in endometriosis pathogenesis has
been accumulating. Simpson et al. first studied female siblings and mothers of patients with
endometriosis and found that these groups were more likely to be affected than the female
siblings and mothers of the respective patients’ husbands (Simpson, Elias, Malinak, & Buttram,
1980). More recent studies have estimated a relative risk to be 5.20 for sisters of patients and
1.56 for first cousins when compared to controls (Stefansson et al., 2002). Familial clustering of
endometriosis cases has shown the definite need to focus on the genetic aspects of endometriosis.
One genome-wide association study using 1,423 cases and 1,318 controls of Japanese
ancestry implicated the SNP rs10966235 in CDKN2B-AS1, a gene on chromosome 9p21.3. In
another GWA study based on samples of European ancestry from Australia and the UK,
rs12700667 on 7p15.2 and rs7521902 on 1p36.12 near WNT4 were found to be associated with
the disease. In a further meta-analysis of these two studies, five other SNPs were found to be
associated with the disease: rs13394619 at 2p25.1 in GREB1, rs10859871 at 12q22 near VEZT,
SLIT3 Gene Expression by 1,25(OH)2D3 7

rs4141819 at 2p14, rs7739264 at 6p22.3, and rs1537377 at 9p21.3 (Nyholt et al., 2012). A third
GWAS using Italian cases and controls confirmed the two main results of the previous European
based study. Additionally, CDKN2B-AS1 was again found to have a significant association with
the disease but now at rs1333049. Furthermore, an association for fibronectin 1 (FN1) rs1250248
was found only in severe cases of endometriosis. Interestingly, an epistatic interaction between
the FN1 and WNT4 SNPs was found particularly in cases of endometriosis at the ovarian site
(Pagliardini et al., 2013). WNT4 encodes a member of the wingless-type secreted protein family.
It is necessary for the development of the female reproductive tract and its suppression results in
male sexual development. CDKN2B-AS1 encodes a long noncoding RNA that is transcribed in
the antisense direction of the CDKN2B and CDKN2A gene cluster which itself encodes three
tumor suppressor proteins: p15, p16-INK4a, and p14ARF. GREB1 is an early response gene in
the estrogen receptor pathway that has also been found to have elevated expression in peritoneal
endometriotic lesions (Vercillini et al., 2014; Pellegrini et al., 2012). The complexity and
variability of endometriosis from case to case are evidence that the disease involves more than
just these 8 genes.
In normal menstrual cycles, shedding of the uterine lining is followed by endometrial
proliferation under estrogenic stimulation. Estradiol (E2) is one important promoter of this
proliferation and the ovaries serve as its main source for healthy endometrium. The endometrium
itself is capable of producing E2 as well as estrone, estrone sulfate, and estrone-3-sulfate (Dassen
et al., 2007). Estrogenic effects are mediated through ERα and ERβ, two distinct estrogen
receptors. E2 and other estrogens bind to these nuclear receptors resulting in individual cell
growth, cell proliferation, and endometrial thickening. Progesterone secretion and receptor
SLIT3 Gene Expression by 1,25(OH)2D3 8

binding brings this growth and proliferation in the endometrium to a halt after ovulation during
the luteal phase of the menstrual cycle (Deligdisch, 2000).
With estrogen-dependent proliferative cycling of disease-free tissue, it is no surprise that
the diseased-state is estrogen-dependent as well. In healthy uterine tissue, ERα is the dominant
receptor, induces proliferation, and activates progesterone receptor expression; ERβ is thought to
oppose the actions of ERα when induced. Disease-free uterine and endometriotic tissue can be
distinctly characterized by the mRNA expression ratio of the two receptors (Bukulmez, Hardy,
Carr, Word, & Mendelson, 2008). Some studies have shown that ERα mRNA and protein levels
are drastically lowered while ERβ levels are extremely high in endometriotic stromal cells,
endometriomas, and peritoneal lesions (Bukulmez et al. 2008; Bulun et al., 2010). ERβ induction
by E2 has even been shown to suppress ERα expression (Bulun et al., 2010). These observations
could logically explain the decreased expression of progesterone receptors seen in endometriosis
cases (Bulun et al., 2006). Despite this logic, other studies have shown opposite trends for
estrogen receptor responses in stromal and epithelial lesions: lesions may contain a much higher
ratio of ERα to ERβ compared to healthy and eutopic tissue (Matsuzaki et al., 2000).
In addition of the role that the ERα/ERβ ratio plays in endometriosis, progesterone
sensitivity appears to be equally important to the nature of the disease. There are two isoforms of
progesterone receptor: the 94-kDa PR-A and the 114-kDa PR-B. The functions of both are
specific to cell type and promoter region. PR-A and PR-B are transcribed from two distinct
promoters when induced by estrogen resulting in PR-B having an additional 164 amino acids.
For the most part, PR-A acts as a transcriptional repressor for progesterone-induced genes while
PR-B acts as an activator of the same genes. PR-A’s ability to act as a repressor comes from an
SLIT3 Gene Expression by 1,25(OH)2D3 9

inhibitory domain that is also found in PR-B; the difference in this domain’s activity is likely due
to the isoforms interacting with different proteins. Investigators have shown that PR-A can act as
a transrepressor of PR-B activity as well as estrogen, androgen, and mineralocorticoid
transcriptional activity (Giangrande & McDonnell, 1999). Progesterone can be viewed as a
growth limiting hormone in the endometrium as it can inhibit and even reverse the estrogen-
induced endometrial growth, hyperplasia, or adenocarcinoma (Bulun et al., 2006). Additionally,
in healthy tissue, progesterone binding in endometrial stromal cells induces paracrine signaling.
In turn, this stimulates epithelial cells to convert E2 into the weakly estrogenic estrone (E1) by
expressing 17B-hydroxysteroid dehydrogenase (17BHSD) type 2 (Zeitoun et al., 1998). All of
this activity relies on PR-A and PR-B. Because of progesterone’s repressive effects, some cases
of endometriosis may be treated with progestin while other cases fail to show regression
indicating an issue with progesterone receptors (Attia et al., 2000). Without proper mediation by
the progesterone receptors, endometrial tissue may undergo unchecked proliferation like what is
seen in endometriosis.
Like estrogen and progesterone, vitamin D is a steroid hormone that plays an important
role in the function of healthy endometria. In humans, vitamin D2 and D3 are the most
important. Vitamin D2 is human-made for the most part and used to fortify foods. Vitamin D3,
also known as cholecalciferol, is a secosteroid generated from a process beginning with 7-
dehydrocholesterol when sunlight is absorbed in the skin of animals. Vitamin D3 itself does not
have much biological activity as the classical model explains that it must first be converted to 25-
hydroxyvitamin D3 (25(OH)D3) by 25-hydroxylase (25-OHase) in the liver. Then, 25(OH)D3 is
converted to the biologically active calcitriol (1,25-dihydroxyvitamin D3 or 1,25(OH)2D3) in the
SLIT3 Gene Expression by 1,25(OH)2D3 10

kidney by 1alpha-hydroxylase (1α-OHase). Once synthesized, calcitriol binds to vitamin D
binding protein (DBP) and is transported to its target organs. The effects of calcitriol are
mediated by binding to vitamin D receptors (VDR) located in the nuclei of target cells all across
the human body. Once activated, VDR heterodimerizes with the retinoid-X receptor (RXR) and
acts as a transcription factor for genes containing vitamin D response elements (VDRE) in their
promoter regions. For this reason, vitamin D is seen as a prohormone.
The biological effects of vitamin D that involve bone development and maintenance are
better known. However, the non-calciotropic effects in tissue are currently more intriguing and
less clear. Calcitriol has been implicated in the regulation of cell proliferation, differentiation,
and apoptosis (Ross & U.S. Institute of Medicine, 2011). These properties allow 1,25(OH)2D3 to
exert anti-tumor effects in certain tissue through different pathways (Ma, Johnson & Trump,
2016). Researchers have shown calcitriol to be immunomodulatory, promoting cultured human
macrophages to synthesize certain antimicrobial peptides. Furthermore, the inactive 25(OH)D3
has been shown to induce the same response in macrophages showing that extrarenal synthesis of
the 1,25(OH)2D3 is possible (Hewison & Adams, 2008). In the reproductive system, decidualized
stromal endometrial cells from early pregnancies were shown to be able to convert 25(OH)D3 to
1,25(OH)2D3 (Kachkache, Rebut-Bonneton, Demignon, Cynober, & Garabedian, 1993). In the
human placenta, 1,25(OH)2D3 stimulates estradiol and progesterone secretion in a dose-
dependent manner (Barrera et al., 2007). In the mouse model, VDR null mice of both sexes
showed gonadal insufficiencies that were only restored by estradiol supplementation indicating
that vitamin D is an important factor in estrogen synthesis within both female and male gonads
SLIT3 Gene Expression by 1,25(OH)2D3 11

(Kinuta et al., 2000). It is clear from such studies that certain tissues such as the uterus have their
own system to regulate vitamin D metabolism.
Many have observed discrepancies in vitamin D and related proteins in cases and controls
for diseases such as endometriosis. Serum 25(OH)D3 and 1,25(OH)2D3 levels have been
investigated to explore such differences. 1-alphahydroxylase has also been shown to be higher in
diseased compared to healthy tissue (Agic et al., 2007). Some studies have shown that patients
have lower circulating levels of 25(OH)D3 while at least one other study has shown the opposite
(Miyashita, 2016; Somigliana et al., 2007; Harris, Chavarro, Malspeis, Willet, & Missmer 2012).
While one study shows no difference in 1,25(OH)2D3 serum levels, another found that patients
have elevated levels (Hartwell, Rodbro, Jensen, Thomsen, & Christiansen, 1990). For patients
with endometriosis, the vitamin D system appears to be dysregulated, however, the direction of
dysregulation remains uncertain. The conflicting results on endometriosis and the vitamin D
system may speak on the many different causes of endometriosis and give insight into how each
case may receive personalized treatment.
The complex relationship between vitamin D and endometriosis requires further
investigation. In this study, we treated endometriosis stromal cell lines with 1,25(OH)2D3 and
used next-generation sequencing to look at global gene expression differences between treated
and untreated lines. Our objective was to identify genes and pathways in these cell lines whose
expression changed significantly due to vitamin D treatment. Using quantitative reverse
transcriptase polymerase chain reaction, we examined the expression of two genes that are
strongly regulated by vitamin D according to our initial results, SLIT3 and CYP24A1, in tissue
from endometriosis lesions and healthy, eutopic endometrium.
SLIT3 Gene Expression by 1,25(OH)2D3 12

Materials and Methods
Cell Culture
The endometriosis stromal cell line ESC22B used in this study were provided by Dr.
Doerthe Brueggmann and derived from peritoneal endometriosis lesions. The cell culture media
consisted of Dulbeco’s Modified Eagle’s Medium/F12 (DMEM/F12) (BioWhittaker
®
, Lonza,
Walkersville, MD, USA) , 10% fetal bovine serum (FBS) (Gibco
®
, Invitrogen Life Technology,
Calsbad, CA, USA), 5% L-glutamine (Cellgro
®
, Mediatech Corning, Manassas, VA, USA), and
antibiotics (Penicillin and Streptomycin) (Cellgro
®
, Mediatech Corning, Manassas, VA, USA).
Cells were cultured for 25 days in a humidifier at 5% CO2, 95% air, and 37oC. Cells were
passaged twice and divided into 6 tissue culture dishes, 3 treatment and 3 control. Then, the cells
were allowed to proliferate to approximately 60-80% confluency 24 hours prior to treatment.
Next, the cells were placed into the culture media described above minus FBS in order to
eliminate the possibility of 1,25(OH)2D3 being present in FBS. The 1,25(OH)2D3 (0.1
uM)(Sigma) was dissolved in DMSO and added to each treatment dish to simulate a supra-
physiologic concentration. Cells were then moved back into the incubator to culture for 4 more
hours before undergoing RNA extraction using RNeasy
®
Mini Kits (Qiagen, Valencia, CA,
USA).
Library Construction and Next-Generation Sequencing
cDNA library construction was carried out at the USC Epigenome Center. RNA libraries
were generated using Illumina TruSeq
®
RNA-Sample Preparation Kits (Illumina, La Jolla, CA,
USA) following the Illumina protocol with 1 ug of total RNA as starting material. RNA quality
SLIT3 Gene Expression by 1,25(OH)2D3 13

was determined using an Agilent 2100 Bioanalyzer (Agilent, Santa Clara, CA, USA).
Sequencing was done at the USC Genome and Cytometry Core. The Illumina HiSeq
®
2500
platform (Illumina, La Jolla, CA, USA) was used to generate 50 base pair paired-end reads (50
PE). Libraries were applied to an Illumina V3 flow cell at a concentration of 16 pM. An average
of 41 million reads per sample (range 38-44) was obtained.
Bioinformatic Analysis
Genome alignment, quality control, and differential gene expression was analyzed with
Partek
®
Flow and Partek
®
Genomic Suite (Partek, ST. Louis, MO, USA). Raw reads (both ends)
were trimmed based on a minimum quality score of 20 and minimum read length of 25.
Trimmed reads were aligned to the human genome hg38 using Star2.4.1d with the help of the
annotation model Gencode v24 (Dobin et al., 2013). Aligned reads were quantified to the model
using the Partek E/M method. Counts were normalized to the upper quartile with offset 1 and
analyzed for differential expression with the Gene-specific analysis (GSA) method. Genes were
considered to be differentially expressed if the false discovery rate (FDR) was less than 0.05 and
there as at least a 2-fold difference in expression. Further gene ontology and pathway analysis
was done by Ingenuity
®
Pathway Analysis (IPA) (Ingenuity, Redwood City, CA, USA).
Tissue Samples
A total of 43 tissue samples (38 endometriosis and 5 controls) were collected from
women of reproductive age. Of the cases, 6 were from endometriomas, 6 from bladder lesions, 5
from other extra-uterine sites (2 from rectal and 1 each from cervical, peritoneum, and uterine-
sacral) and 21 from within the uterine lining. All samples were treated with 1,25(OH)2D3 (0.1
uM)(Sigma). Pathology reports confirmed endometriosis diagnoses in cases and excluded
SLIT3 Gene Expression by 1,25(OH)2D3 14

endometrial pathology in controls. All subjects were between 23 and 49 years of age, reported
regular menstrual cycles, were not pregnant, and had not been on hormonal treatment. Each
patient’s menstrual cycle phase was recorded per the patient’s verbal history as follicular (cycle
day 5-14) versus luteal (cycle day 15-30). Endometrial pathology was excluded in healthy
controls. The study was approved by the University of Southern California Institutional Review
board (#HS 032005).
Quantitative RT-PCR
Total RNA was extracted using Qiagen Mini Kits (Qiagen, Valencia CA). cDNA was
generated using the Qiagen RT2 First Strand kit (Qiagen, Valencia CA) and then stored at -40
o

C. Forward and reverse primers for the genes of interest (CYP24A1 and SLIT3) and the
housekeeping gene (GADPH) were designed by Qiagen. GADPH expression was found to be
relatively unchanged between case and control cell lines allowing for its use in the comparative
Ct method. SYBR Green with ROX passive reference, HotStart DNA Taq polymerase, and
nucleotides were all also provided by Qiagen in an optimized real-time PCR buffer. Quantitative
PCR was performed on an Applied Biosystems 7900HT Fast Real-Time PCR System at the USC
Epigenome Center. All samples were run in triplicate for the gene of interest and housekeeping
gene.
Triplicate Ct readings were averaged and ΔCt values were calculated for each sample as
the difference between CYP24A1 and GADPH Ct values. Mean ΔCt values were compared
between endometrial and control tissues using Student’s t test. One-way analysis of variance
was used to compare tissue sites (uterine cavity, endometrioma, bladder, miscellaneous ectopic
sites, control). Post-hoc comparisons were carried out using the Bonferroni method. Relative
SLIT3 Gene Expression by 1,25(OH)2D3 15

CYP24 and SLIT3 expression (endometrial vs. control) was calculated as difference in ΔCt
values.
Results
Differential gene expression by 1,25(OH)
2
D
3
in endometriosis stromal cells
In order to maximize the quality of our sequencing results, we merged data from a
previous run (UPC batch) of the same samples with this current run (HSC batch) for differential
expression analysis. By doing so, we treat both batches as technical replicates and substantially
increase the sequencing depth for each sample. Post-alignment QA/QC data were generated by
Partek
®
Flow to assess base call and mapping quality, which were both very good (Table 1). The
combined total reads from both batches ranged from 79-107 million for each sample. Alignment
was greater than 93% for all samples while only control 2 from the HSC batch was below 94%.
In the table, Phred quality scores are used. In short, parameters related to the shape and
resolution at each base are measured and used to look up a corresponding quality score in a large
table specific to the sequencing machine and chemistry used. This quality score is
logarithmically transformed to obtain a Phred quality score (Illumina inc., 2011). A Phred quality
score of 30 indicates a 1 in 1000 probability for a base call error or a 99.9% base call accuracy.
The average scores in both batches are greater than 36.5 meaning the base calls are at least
99.97% accurate.

SLIT3 Gene Expression by 1,25(OH)2D3 16

Table 1a. Post-alignment QC chart for ESC samples, UPC batch.
Sample
name
Total reads Aligned
Total non-
unique
Non-unique Coverage
Avg.
coverage
depth
Avg. length
Avg.
quality
%GC
Control 1 63,968,788 94.99% 4,437,608 6.94% 6.11% 39.77 54.46 36.56 51.80%
Control 2 49,689,150 94.68% 3,430,470 6.90% 5.68% 33.12 54.44 36.53 51.87%
Treatment 1 64,768,284 95.15% 4,594,332 7.09% 6.81% 36.34 54.47 36.60 51.62%
Treatment 2 41,018,542 94.95% 2,684,669 6.55% 5.53% 27.89 54.47 36.55 51.38%
Uniquely aligned reads (not shown) are only aligned to one location. Non-uniquely aligned reads align to more than one location.

Table 1b. Post-alignment QC chart for ESC samples, HSC batch.
Sample
name
Total reads Aligned
Total non-
unique
Non-unique Coverage
Avg.
coverage
depth
Avg. length
Avg.
quality
%GC
Control 1 41,658,774 94.13% 2,819,849 6.77% 4.88% 29.19 49.52 36.74 51.44%
Control 2 43,181,722 93.82% 2,909,779 6.74% 5.11% 28.82 49.51 36.73 51.52%
Treatment 1 42,371,971 94.32% 2,942,765 6.95% 5.45% 26.76 49.52 36.75 51.24%
Treatment 2 38,194,411 94.10% 2,445,197 6.40% 5.13% 25.26 49.52 36.74 51.02%
Uniquely aligned reads (not shown) are only aligned to one location. Non-uniquely aligned reads align to more than one location.
SLIT3 Gene Expression by 1,25(OH)2D3 17

Principle component analysis (PCA) is an explanatory technique that describes high
dimensional data in a reduced number of uncorrelated dimensions or principle components (PCs)
using linear transformations. Each dot in the plot represents a sample (Figure 1). The first 3 PCs
are shown, which explain approximately 75% of the variation. These plots can be used to view
groupings in the data. The plots of the raw reads show some batch effect. Reads were normalized
using the upper quartile method. In this method, raw count data for all genes in the samples is
divided by the mean upper quartile across all samples after removing genes with zero reads (Lin
et al., 2016). After normalizing reads, the data shows a more dramatic separation between the
treatment and control groups, however, there is still some batch effect seen as HSC samples are
higher than UPC samples on PC2 (Figure 2).
Figure 1. PCA plot of raw reads.

SLIT3 Gene Expression by 1,25(OH)2D3 18

A total of 199,169 transcripts belonging to 59,213 genes were differentially expressed by
1,25(OH2)D3 in this endometriosis stromal cell line. Of this total number, 1627 genes were
statistically significant (p<0.05 & FDR<0.05) and at least two-fold differentially expressed with
886 down-regulated and 741 up-regulated. CYP24A1 was the only strongly up-regulated gene
(369-fold). A total of 49 genes were down-regulated at least 50-fold and 20 of these were down-
regulated at least 100-fold. One of the 20 most down-regulated genes is SLIT3, a secreted
protein involved in cell migration and neuronal guidance. Tables 2 and 3 list the most up- and
down-regulated genes in our ESC samples according to Partek
®
Flow.

Figure 2. PCA plot of normalized reads.

SLIT3 Gene Expression by 1,25(OH)2D3 19

Table 2. Most up-regulated genes in ESC samples.
Gene Fold Change p-value FDR
CYP24A1 368.93 1.59E-07 3.73E-05
G0S2 6.56 1.61E-04 5.08E-04
FENDRR 5.02 1.17E-04 4.14E-04
KRBOX1 4.90 2.84E-04 7.32E-04
PLAU 4.76 3.30E-04 8.11E-04
FOXF1 4.38 1.26E-04 4.34E-04
FHOD3 4.31 2.84E-04 7.32E-04
INHBA 4.29 5.85E-04 1.23E-03
GPR68 4.27 7.24E-04 1.45E-03
KCNH1 4.27 7.19E-04 1.44E-03
Most up-regulated 10 genes in ESC samples with corresponding folds of change, p-values, and
FDR. Analysis was performed with Partek
®
Flow.

Table 3. Most down-regulated genes in ESC samples.
Gene Fold Change p-value FDR
DSC3 -537.54 1.77E-07 3.73E-05
ALDH1A2 -359.19 2.94E-07 4.15E-05
NPY -317.09 2.76E-07 4.08E-05
LRRN4 -313.18 6.10E-08 3.52E-05
CD24 -247.88 8.87E-07 5.34E-05
CADM3 -178.12 1.58E-07 3.73E-05
H19 -147.63 2.15E-07 3.91E-05
SLIT3 -140.19 2.72E-07 4.08E-05
SLITRK5 -139.32 3.68E-07 4.38E-05
BST2 -136.89 1.13E-06 5.82E-05
CACNG4 -134.41 4.30E-06 9.31E-05
CDH3 -131.60 2.32E-07 3.91E-05
FAT3 -126.72 1.22E-08 2.77E-05
TRIM58 -121.01 1.64E-07 3.73E-05
RORB -118.18 4.08E-07 4.48E-05
FAM84B -115.72 1.19E-07 3.73E-05
PARM1 -112.02 5.43E-07 4.68E-05
TFAP2A -109.17 1.73E-08 3.11E-05
KRT8 -105.47 2.35E-08 3.12E-05
SLIT3 Gene Expression by 1,25(OH)2D3 20

FRAS1 -102.32 2.29E-06 7.35E-05
Most down-regulated 20 genes in ESC samples with corresponding folds of change, p-values,
and FDR. Analysis was done with Partek
®
Flow.

Gene Ontology and Pathway Analysis
We used Ingenuity
®
Pathway Analysis to obtain gene ontology (GO) and pathway
analysis results. The original 1627 differentially expressed genes from Partek
®
Flow were
ultimately reduced to 1536 unambiguous, mapped genes; the rest were excluded from this
analysis because their identifiers mapped to multiple genes with no way of determining which
mapping is the correct one.
Figure 3 is a heat map image organizing our genes by the diseases and biological
functions they are involved in. The heat map is divided into 3 hierarchical levels: level 1
represents the largest grouping of disease and function and level 3 the smallest. The size and
color of the squares are sorted by a z-score. This z-score is calculated by IPA
®
to make
predictions on the changes in gene expression and decrease the chance that such predictions are
not due to random error. The size of the squares reflects the predicted magnitude of change in
diseases and functions; larger squares have larger z-scores and more extreme changes in gene
expression. The color of the squares reflects the direction of the change with orange being
associated with a positive z-score and upregulation while blue is associated with a negative z-
score and downregulation. The most significant categories of disease and function by p-value are
located on the left side of the heat map. Organismal injury and abnormalities, cellular movement,
and cancer are the most significant level 1 categories of disease and function as well as the
categories that contain the most up-regulated gene groups. The blue level 2 squares of these 3
SLIT3 Gene Expression by 1,25(OH)2D3 21

groups mostly represent downregulation of cell death and certain cancer pathways. Cell death
and survival was another important level 1 category with most of its level 2 subcategories being
highly down-regulated.
The organismal injury block contains 1333 differentially expressed genes while the
cancer block contains 1324. Within these, we see that the invasion, angiogenesis, and
proliferation level 2 blocks are up-regulated while cell death and apoptosis are down-regulated
(Figure 4). In the cellular movement block, movement, invasion and migration of all kinds of
cells including nervous, cancer, and embryonic are up-regulated (Figure 5). The highlighted level
3 blocks are cell movement, migration of cells, and invasion of cells and they represent the three
of the most statistically significant level 3 blocks in the overall analysis. The cell death and
survival block shows that overall cell death, apoptosis, and necrosis are decreased whereas cell
viability and survival are increased (Figure 6).
Figure 3. Heat map representing disease and function relationships of differentially expressed
genes.

SLIT3 Gene Expression by 1,25(OH)2D3 22

Figure 4. Heat map representing the organismal injury and abnormalities block.

Figure 5. Heat map representing the cancer block.

SLIT3 Gene Expression by 1,25(OH)2D3 23

Figure 6. Heat map representing the cellular movement block.

Figure 7. Heat map representing the cellular death and survival block.

SLIT3 Gene Expression by 1,25(OH)2D3 24

There are 358 statistically significant pathways that were differentially regulated in our
data. Figure 8 is a bar chart representation of the specific disease and biological function
pathways originally presented in the heat maps and shows the top 10 most statistically significant
pathways ordered by –log(p-value). Again, the color (orange or blue) is representative of the z-
score magnitude and direction. White bars have a z-score close to 0 and grey bars are
undetermined. The floating orange points and connecting lines represent the ratio of the number
of our differentially expressed genes in this pathway to the complete pathway stored in IPA’s
database. Figure 9 breaks the pathways down by percentage of genes up-regulated (red) and
down-regulated (green). The number shown is IPA’s total number of genes in the pathway.

The tables below show the most differentially expressed genes in the VDR/RXR
activation, inhibition of matrix metalloprotease (MMP), and axonal guidance signaling pathways
Figure 8. The 10 most
statistically significant
pathways ordered by –
log(p-value).
SLIT3 Gene Expression by 1,25(OH)2D3 25

(Tables 4a-c). These pathways are all statistically significant and influence cell migration,
differentiation, proliferation, and survival (p=1.42E-04, 1.53E-04, & 2.71E-09 respectively).

VDR/RXR Signaling
Symbol Gene Name Absolute fold change Fold Change FDR p-value
CYP24A1 cytochrome P450 family 24 368.925 368.925 3.73E-05 1.59E-07
HR lysine demethylase, nuclear
receptor corepressor
31.639 -31.639 8.16E-05 3.10E-06
CCL5 C-C motif chemokine ligand
5
25.898 -25.898 4.48E-05 3.99E-07
CD14 CD14 molecule 24.189 -24.189 3.73E-05 1.15E-07
WT1 Wilms tumor 1 17.965 -17.965 3.91E-05 2.45E-07
SEMA3B semaphorin 3B 10.557 -10.557 5.59E-05 1.01E-06
SPP1 secreted phosphoprotein 1 3.793 3.793 6.75E-04 5.93E-04
IGFBP3 IGF binding protein 3 3.278 -3.278 9.47E-05 4.46E-06
KLF4 Kruppel like factor 4 3.213 3.213 2.77E-05 1.25E-08
IGFBP5 IGF binding protein 5 3.045 3.045 7.84E-04 3.14E-04
CYP27B1 cytochrome P450 family 27 2.858 -2.858 1.88E-04 2.50E-05
Figure 9. The 10 most
statistically significant
pathways ordered by –
log(p-value) and
broken down by
percentage of
upregulated and down-
regulated genes.
SLIT3 Gene Expression by 1,25(OH)2D3 26

COL13A1 collagen type XIII alpha 1
chain
2.701 2.701 1.07E-03 4.86E-04
VDR 1,25(OH)2D3 receptor 2.339 2.339 7.41E-04 2.90E-04
IGFBP6 IGF binding protein 6 2.032 -2.032 2.07E-04 3.18E-05
PDGFA platelet derived growth
factor subunit A
2.003 -2.003 1.05E-03 4.78E-04
Table 4a. The 15 most differentially expressed genes in the VDR/RXR activation pathway.

Inhibition of Matrix Metalloproteases
Symbol Gene Name Absolute fold change Fold Change FDR p-value
MMP28 matrix metallopeptidase 28 8.974 -8.974 2.00E-04 2.00E-04
MMP24 matrix metallopeptidase 24 7.745 -7.745 3.91E-05 3.91E-05
MMP15 matrix metallopeptidase 15 6.869 -6.869 4.71E-05 4.71E-05
TFPI2 tissue factor pathway
inhibitor 2
3.41 3.41 1.45E-03 1.45E-03
MMP2 matrix metallopeptidase 2 3.389 3.389 1.18E-03 1.18E-03
THBS2 thrombospondin 2 3.347 -3.347 7.69E-05 7.69E-05
TIMP1 TIMP metallopeptidase
inhibitor 1
3.221 3.221 9.18E-04 9.18E-04
MMP23B matrix metallopeptidase
23B
2.737 -2.737 1.73E-04 1.73E-04
MMP14 matrix metallopeptidase 14 2.175 2.175 6.98E-04 6.98E-04
Table 4b. This pathway had only 9 genes with an absolute fold change greater than 2.0.

Axonal Guidance Signaling
Symbol Gene Name Absolute fold change Fold Change FDR p-value
SLIT3 slit guidance ligand 3 140.191 -140.191 4.08E-05 2.72E-07
UNC5C unc-5 netrin receptor C 58.833 -58.833 3.70E-05 7.39E-08
L1CAM L1 cell adhesion molecule 47.223 -47.223 3.47E-05 5.75E-08
PDGFB platelet derived growth
factor subunit B
36.523 -36.523 5.13E-05 7.27E-07
WNT4 Wnt family member 4 13.901 -13.901 4.08E-05 2.66E-07
FZD3 frizzled class receptor 3 11.561 -11.561 9.16E-05 3.41E-04
FGFR3 FGF receptor 3 11.342 -11.342 7.73E-05 2.54E-06
TUBB2B tubulin beta 2B class IIb 11.226 -11.226 6.97E-05 1.90E-06
EFNA1 ephrin A1 11.051 -11.051 4.61E-05 4.58E-07
WNT2B Wnt family member 2B 10.558 -10.558 4.61E-05 4.65E-07
SEMA3B semaphorin 3B 10.557 -10.557 5.59E-05 1.01E-06
SLIT3 Gene Expression by 1,25(OH)2D3 27

PLXNA2 plexin A2 10.444 -10.444 8.59E-05 3.54E-06
ADAM22 -- 7.617 -7.617 5.73E-04 1.95E-04
EFNA5 ephrin A5 7.296 -7.296 6.73E-05 1.65E-06
SEMA6D semaphorin 6D 7.258 -7.258 9.90E-05 5.08E-06
Table 4c. The 15 most differentially expressed genes in the axonal guidance signaling pathway.

Quantitative RT-PCR of SLIT3 expression in tissue samples
In our previous study, we looked at CYP24A1 expression as it was very strongly and
directly up-regulated by 1,25(OH)2D3 in the vitamin D pathway (Appendix A). CYP24A1
expression was higher in endometriosis samples than in controls (t test, p=0.01). Overall,
expression was elevated by 8-fold in cases compared to controls. Compared to controls, uterine
cavity expression was 9.8-fold higher, 9.2-fold higher in endometriomas, 4.6-fold higher in
bladder lesions, and 6.5-fold higher in ectopic lesions. Pairwise comparisons were only
statistically significant for uterine cavity lesions compared to controls (Bonferroni, p=0.05).
SLIT3 was very strongly down-regulated in cases compared to controls according to our
sequencing results. We looked at SLIT3 expression in the same tissue samples using quantitative
RT-PCR. Raw data is in the form of ΔCT values (Figure 10). Ct (cycle threshold) values are the
number of PCR cycles required for a sample’s fluorescent signal to cross the background
fluorescence level. ΔCt values are the difference between cases and controls.

SLIT3 Gene Expression by 1,25(OH)2D3 28

SLIT3 expression was lower in all cases when compared to controls, however, this
difference was not statistically significant (t-test, p=0.56). In endometriosis, expression was
elevated 1.6-fold compared to controls. When comparing site-specific data using one-way
ANOVA, there was at least one statistically significant difference in mean fold change (p<0.01).
Using Bonferroni’s method, several statistically significant pairwise comparisons were found.
The uterine cavity lesions were the only group with a statistically significant difference in SLIT3
expression compared to controls (p<0.01). However, the pairwise comparisons for intra-uterine
Figure 10. SLIT3 expression in tissue samples. Higher ΔCt values indicate
lower expression.

SLIT3 Gene Expression by 1,25(OH)2D3 29

lesions vs. endometrioma, bladder, and other invasive lesions were all statistically significant
(p’s<0.05). Furthermore, endomtrioma SLIT3 expression was also statistically significantly
different from bladder and other invasive lesions (p’s<0.05).
An interesting grouping in expression is seen in the fold change between intra-uterine and
ovarian lesions compared to the ectopic lesions (Table 5). These ectopic sites show a marked
increase in SLIT3 expression; bladder expression is elevated 3.42-fold while peritoneal, uterine-
sacral, cervical, and rectal lesion expression is elevated 3.07-fold in comparison to controls.
These numbers are 0.65- and 0.15-fold respectively for endometrioma and intra-uterine lesions
compared to controls.

Discussion
In this study, we report on the key genes that seem to be regulated by 1,25(OH)2D3 in
cells affected by endometriosis. We treated the ESC22B cell line with 1,25-dihydroxyvitamin D3
and looked at differential gene expression using NGS. In our pathway analysis, we found
statistically significant pathways related to vitamin D metabolism, cell invasion and migration,
and neuroangiogenesis. In the vitamin D pathway, VDR binds 1,25(OH)2D3 for transportation
and processing. CYP24A1 hydroxylates the side chain of 1,25(OH)2D3 degrading it for
excretion. As a result, this protein has a negative feedback impact on the vitamin D system.
Samples N ΔCt Mean Std dev Fold-change
All Controls 5 3.91 0.90 -
All Cases 50 4.59 2.55 0.62
Uterine Cavity 23 6.69 1.56 0.15
Endometrioma 7 4.52 1.02 0.65
Bladder 12 2.13 1.40 3.42
Misc. Ectopic Sites 8 2.29 1.75 3.07
Table 5. Mean ΔCts and fold
changes for cases compared to
controls. Cases are also
broken down by location.
SLIT3 Gene Expression by 1,25(OH)2D3 30

Signals from inflammatory states and infiltrating immune cells translationally up-regulate
CYP24A1 (Rubsamen et al., 2014). In our study, CYP24A1 was up-regulated 368-fold while
VDR was up-regulated about 2-fold indicating that the vitamin D system in these cells was
indeed responsive. The inflammatory microenvironment created by endometriosis and treatment
with by 1,25(OH)2D3 may explain the extremely high expression of CYP24A1 with only
moderate upregulation of VDR. In the quantitative RT-PCR experiment, CYP24A1 expression
was greater in uterine lesions compared to ectopic lesions. The locational differences in
expression are consistent with more active and dominant estrogen microenvironments found in
intra-uterine ad ovarian lesions compared to ectopic lesions (Ingles et al., 2017).
Another vitamin D regulated pathway that was detected by our study is the matrix
metalloprotease inhibition pathway. MMPs degrade the extracellular matrix and play roles in cell
invasion and migration. In cancer, MMP activity is increased allowing for metastasis and new
tumors to form. In endometriosis, expression of several different types of MMPs is increased
aiding in the invasion and migration of cells forming new lesions (Kokorine et al., 1997; Chung
et al., 2001). In this study, the expression of several types of MMPs was reduced by
1,25(OH)2D3. These results align with a previous study that showed that vitamin D reduces
expression of MMPs 2 and 9 in ESC samples of patients with endometriosis (Miyashita et al.,
2016). These results provide evidence for the possibility of vitamin D as a treatment to eliminate
the invasive properties of endometriosis cells and prevent the formation of new lesions.
Similar to tumors, endometriosis implants require angiogenesis to establish themselves as
well as continue surviving. Without the constant development of new vasculature, ectopic lesions
begin to shrink (Nap et al., 2004). Along with angiogenesis, neurogenesis is equally as important
SLIT3 Gene Expression by 1,25(OH)2D3 31

for maintaining endometriosis lesions as the two processes are directly linked (Weinstein, 2005).
Major and peripheral nerves are often aligned with blood vessels resulting from similar
molecular patterning mechanism using semaphorins and netrins (Asante & Taylor, 2011).
Semaphorins appear as part of the most statistically significant, differentially regulated
molecules in the axonal guidance signaling pathway. In short, semaphorins are membrane-
associated and secreted ligands that direct axons and vessels to their intended locations. In cases
of endometriosis, one study has shown that semaphorin 3C is up-regulated. Netrins tend to be
proangiogenic and are capable of repulsive and attractive signaling. Through their receptors,
netrins can reduce angiogenesis and cell migration (Nap et al., 2004). In our study, semaphorins
3B, 6D, 3F, 6B, and 5A as well as UNC5 netrin receptors B and C are down-regulated
suggesting that endometriosis cells are responsive to 1,25(OH)2D3.
In the axonal guidance signaling pathway, SLIT3 is the most down-regulated gene. SLITs
are secreted proteins that interact with other proteins through leucine-rich repeats and epidermal
growth factor-like motifs (Itoh, Miyabayashi, Ohno, & Sakano, 1998). They bind to the Robo
family of receptors that mediate a repulsive response that can repel both migrating cells and
axons (Brose & Tessier-Lavigne, 2000; Geutskens et al., 2012). Additionally, SLIT3 has been
shown to stimulate angiogenesis. Ultimately, this protein is capable of repressing nerve growth
and increasing vessel growth (Lin et al., 2016). SLIT3 is expressed in many types of tissues
including skin, colon, kidney, brain, and ovary indicating that its functions may not be restricted
to just axonal guidance and migration. It has been shown to be epigenetically inactivated due to
hypermethylation in tumor cell lines and primary tumors (Dickinson et al., 2004). When
functional, SLIT3 can inhibit migration and invasion of melanoma and thyroid cancer cells
SLIT3 Gene Expression by 1,25(OH)2D3 32

(Guan et al., 2013; Denk, Braig, Schubert, & Bosserhoff, 2011). Along with the other genes in
this pathway, the dysregulation of SLIT3 may result in the invasive nature of endometriosis. Like
with CYP24A1, we observed differences by site that varied by an estrogen dominant
microenvironment for the intra-uterine and ovarian lesions compared to the various ectopic
lesions in our PCR assays. Bladder, uterine-sacral, rectal, cervical, and peritoneal lesions are
ectopic by nature and the cells that have seeded these implants have increased expressions of
SLIT3. Our data indicates that SLIT3 works with semaphorins, netrins, as well as other axonal
guidance genes to make endometriosis implants viable by providing the necessary blood supply
and nerves required for survival. When comparing the 1,25(OH)2D3 treated and untreated
stromal endometriosis cells in the sequencing results, we see a sharp decrease in SLIT3
expression. Overall, the repression of the axonal guidance signaling pathway provides evidence
for the idea that vitamin D is capable of slowing down the constant neuroangiogenesis required
for endometriosis implants.
Conclusion
The activation of the VDR/RXR signaling pathway in ESC22B cells indicates that
1,25(OH)2D3 does have an impact on endometriosis tissues. Additionally, the inhibition of genes
involved in the MMP and axonal guidance pathways strengthens the idea that vitamin D is a
viable treatment against endometriosis. The difference in SLIT3 expression between eutopic and
ectopic endometriosis sites suggests that invasion and neuroangiogenesis are targetable factors of
endometriosis. Previous studies that have focused on the phenotypic and behavioral factors have
shown a link between 1,25(OH)2D3 and endometriosis. In one study, women with endometriosis
were more sensitive to sunlight, more likely to apply sunscreen, and thus more likely to have
SLIT3 Gene Expression by 1,25(OH)2D3 33

lower levels of 1,25(OH)2D3 (Somigliana et al., 2010). Another study linked increases in
benzophenone-based sunscreen use with likelihood of endometriosis diagnosis. However, the
estrogenic properties of benzophenone may be a stronger factor for endometriosis in these cases
(Kunisue et al., 2012). Altogether, such results show that vitamin D may be a viable treatment
for endometriosis, support previous knowledge on certain genes and their involvement in the
disease, and provide new targets for further investigation on endometriosis.
Acknowledgements
I would like to express my sincere gratitude to Dr. Sue Ingles, my advisor throughout my
graduate studies; Drs. Mariana Stern and Lynda McGinnis for guiding me through the thesis
process; the USC UPC Genome & Cytometry Core and USC Epigenome Center for their help
with molecular techniques, Drs. Yibu Chen and Meng Li for support with bioinformatics and
analyses; and finally, my parents, Dr. CY Wang and Dr. YC Liu, for always supporting me and
helping me to realize my own potential.

SLIT3 Gene Expression by 1,25(OH)2D3 34

References
Kennedy, S., Bergqvist, A., Chapron, C., D'Hooghe, T., Dunselman, G., Greb, R., . . . ESHRE
Special Interest Group for Endometriosis and Endometrium Guideline Development
Group. (2005). ESHRE guideline for the diagnosis and treatment of
endometriosis. Human Reproduction (Oxford, England), 20(10), 2698-2704.
doi:10.1093/humrep/dei135
Ingles, S. A., Wu, L., Liu, B. T., Chen, Y., Wang, C., Templeman, C., & Brueggmann, D.
(2017). Differential gene expression by 1,25(OH)2D3 in an endometriosis stromal cell
line. The Journal of Steroid Biochemistry and Molecular
Biology, doi:10.1016/j.jsbmb.2017.01.011
Nezhat, C., Nezhat, F., & Nezhat, C. (2012). Endometriosis: Ancient disease, ancient
treatments. Fertility and Sterility, 98(6 Suppl), S1-S62.
doi:10.1016/j.fertnstert.2012.08.001
Bulletti, C., Coccia, M. E., Battistoni, S., & Borini, A. (2010). Endometriosis and
infertility. Journal of Assisted Reproduction and Genetics, 27(8), 441-447.
doi:10.1007/s10815-010-9436-1
Vercellini, P., Viganò, P., Somigliana, E., & Fedele, L. (2014). Endometriosis: Pathogenesis and
treatment. Nature Reviews. Endocrinology, 10(5), 261. doi:10.1038/nrendo.2013.255
Sasson, I. E., & Taylor, H. S. (2008). Stem cells and the pathogenesis of endometriosis. Annals
of the New York Academy of Sciences, 1127(1), 106-115. doi:10.1196/annals.1434.014
Schrodt, G. R., Alcorn, M. O., & Ibanez, J. (1980). Endometriosis of the male urinary system: A
case report. The Journal of Urology, 124(5), 722-723.
Simpson, J. L., Elias, S., Malinak, L. R., & Buttram, J., V C. (1980). Heritable aspects of
endometriosis. I. genetic studies. American Journal of Obstetrics and
Gynecology, 137(3), 327.
Stefansson, H., Geirsson, R. T., Steinthorsdottir, V., Jonsson, H., Manolescu, A., Kong,
A.,...Stefansson, K. (2002). Genetic factors contribute to the risk of developing
endometriosis. Human Reproduction (Oxford, England), 17(3), 555-559.
doi:10.1093/humrep/17.3.555
Nyholt, D. R., Low, S., Anderson, C. A., Painter, J. N., Uno, S., Morris, A. P., ...Montgomery, G.
W. (2012). Genome-wide association meta-analysis identifies new endometriosis risk
loci. Nature Genetics, 44(12), 1355-1359. doi:10.1038/ng.2445
SLIT3 Gene Expression by 1,25(OH)2D3 35

Pagliardini, L., Gentilini, D., Vigano', P., Panina-Bordignon, P., Busacca, M., Candiani, M., &
Di Blasio, A. M. (2013). An italian association study and meta-analysis with previous
GWAS confirm WNT4, CDKN2BAS and FN1 as the first identified susceptibility loci
for endometriosis. Journal of Medical Genetics, 50(1), 43. doi:10.1136/jmedgenet-2012-
101257
Pellegrini, C., Gori, I., Achtari, C., Hornung, D., Chardonnens, E., Wunder, D., ...Canny, G. O.
(2012). The expression of estrogen receptors as well as GREB1, c-MYC, and cyclin D1,
estrogen-regulated genes implicated in proliferation, is increased in peritoneal
endometriosis. Fertility and Sterility, 98(5), 1200. doi:10.1016/j.fertnstert.2012.06.056
Dassen, H., Punyadeera, C., Kamps, R., Delvoux, B., Van Langendonckt, A., Donnez, J.,
...Groothuis, P. (2007). Estrogen metabolizing enzymes in endometrium and
endometriosis. Human Reproduction, 22(12), 3148-3158. doi:10.1093/humrep/dem310
Deligdisch, L. (2000). Hormonal pathology of the endometrium. Modern Pathology, 13(3), 285-
294. doi:10.1038/modpathol.3880050
Bukulmez, O., Hardy, D. B., Carr, B. R., Word, R. A., & Mendelson, C. R. (2008). Inflammatory
status influences aromatase and steroid receptor expression in
endometriosis. Endocrinology, 149(3), 1190-1204. doi:10.1210/en.2007-0665
Bulun, S. E., Cheng, Y., Pavone, M. E., Xue, Q., Attar, E., Trukhacheva, E., ...Kim, J. J. (2010).
Estrogen receptor-beta, estrogen receptor-alpha, and progesterone resistance in
endometriosis. Seminars in Reproductive Medicine, 28(1), 36.
Bulun, S. E., Cheng, Y., Yin, P., Imir, G., Utsunomiya, H., Attar, E., ...Julie Kim, J. (2006).
Progesterone resistance in endometriosis: Link to failure to metabolize
estradiol. Molecular and Cellular Endocrinology, 248(1), 94-103.
doi:10.1016/j.mce.2005.11.041
Matsuzaki, S., Fukaya, T., Uehara, S., Murakami, T., Sasano, H., & Yajima, A. (2000).
Characterization of messenger RNA expression of estrogen receptor-α and -β in patients
with ovarian endometriosis. Fertility and Sterility, 73(6), 1219-1225. doi:10.1016/S0015-
0282(00)00527-6
Giangrande, P. H., & McDonnell, D. P. (1999). The A and B isoforms of the human
progesterone receptor: Two functionally different transcription factors encoded by a
single gene. Recent Progress in Hormone Research, 54, 291.
Zeitoun, K., Takayama, K., Sasano, H., Suzuki, T., Moghrabi, N., Andersson, S., ...Bulun, S. E.
(1998). Deficient 17β-hydroxysteroid dehydrogenase type 2 expression in endometriosis:
SLIT3 Gene Expression by 1,25(OH)2D3 36

Failure to metabolize 17β-Estradiol1. The Journal of Clinical Endocrinology &
Metabolism, 83(12), 4474-4480. doi:10.1210/jcem.83.12.5301
Attia, G. R., Zeitoun, K., Edwards, D., Johns, A., Carr, B. R., & Bulun, S. E. (2000).
Progesterone receptor isoform A but not B is expressed in endometriosis. The Journal of
Clinical Endocrinology and Metabolism, 85(8), 2897-2902. doi:10.1210/jc.85.8.2897
Ross, A. C., & Institute of Medicine (U. S.). Committee to Review Dietary Reference Intakes for
Vitamin D and Calcium. (2011). Dietary reference intakes: Calcium, vitamin D.
Washington, DC: National Academies Press.
Ma, Y., Johnson, C. S., & Trump, D. L. (2016). Mechanistic insights of vitamin D anticancer
effects. Vitamins and Hormones, 100, 395.
Hewison, M., & Adams, J. S. (2008). Unexpected actions of vitamin D: New perspectives on the
regulation of innate and adaptive immunity. Nature Clinical Practice Endocrinology &
Metabolism, 4(2), 80-90. doi:10.1038/ncpendmet0716
Kachkache, M., Rebut-Bonneton, C., Demignon, J., Cynober, E., & Garabédian, M. (1993).
Uterine cells other than stromal decidual cells are required for 1,25‐dihydroxyvitamin D3
production during early human pregnancy. FEBS Letters, 333(1-2), 83-88.
doi:10.1016/0014-5793(93)80379-9
Barrera, D., Avila, E., Hernández, G., Halhali, A., Biruete, B., Larrea, F., & Díaz, L. (2007).
Estradiol and progesterone synthesis in human placenta is stimulated by
calcitriol. Journal of Steroid Biochemistry and Molecular Biology, 103(3), 529-532.
doi:10.1016/j.jsbmb.2006.12.097
Kinuta, K., Tanaka, H., Moriwake, T., Aya, K., Kato, S., & Seino, Y. (2000). Vitamin D is an
important factor in estrogen biosynthesis of both female and male
gonads. Endocrinology, 141(4), 1317-1324. doi:10.1210/en.141.4.1317
Agic, A., Xu, H., Altgassen, C., Noack, F., Wolfler, M. M., Diedrich, K., ...Hornung, D. (2007).
Relative expression of 1,25-dihydroxyvitamin D3 receptor, vitamin D 1α-hydroxylase,
vitamin D 24-hydroxylase, and vitamin D 25-hydroxylase in endometriosis and
gynecologic cancers. Reproductive Sciences, 14(5), 486-497.
doi:10.1177/1933719107304565
Miyashita, M., Koga, K., Izumi, G., Sue, F., Makabe, T., Taguchi, A., ...Osuga, Y. (2016).
Effects of 1,25-dihydroxy vitamin D3 on endometriosis. The Journal of Clinical
Endocrinology & Metabolism, 101(6), 2371-2379. doi:10.1210/jc.2016-1515
SLIT3 Gene Expression by 1,25(OH)2D3 37

Somigliana, E., Panina-Bordignon, P., Murone, S., Di Lucia, P., Vercellini, P., & Vigano, P.
(2007). Vitamin D reserve is higher in women with endometriosis. Human
Reproduction, 22(8), 2273-2278. doi:10.1093/humrep/dem142
Harris, H. R., Chavarro, J. E., Malspeis, S., Willett, W. C., & Missmer, S. A. (2013). Dairy-food,
calcium, magnesium, and vitamin D intake and endometriosis: A prospective cohort
study. American Journal of Epidemiology, 177(5), 420-430. doi:10.1093/aje/kws247
Hartwell, D., Rødbro, P., Jensen, S. B., & Christiansen, C. (1990). Vitamin D metabolites -
relation to age, menopause and endometriosis. Scandinavian Journal of Clinical &
Laboratory Investigation, 50(2), 115-121. doi:10.3109/00365519009089142
Dobin, A., Davis, C. A., Schlesinger, F., Drenkow, J., Zaleski, C., Jha, S., ...Gingeras, T. R.
(2013). STAR: Ultrafast universal RNA-seq aligner. Bioinformatics (Oxford,
England), 29(1), 15-21. doi:10.1093/bioinformatics/bts635
Illumina, Inc. (2011). Quality Scores for Next-Generation Sequencing: Assessing Sequencing
Accuracy Using Phred Quality Scoring. Retrieved from
https://www.illumina.com/documents/products/technotes/technote_Q-Scores.pdf
Lin, Y., Golovnina, K., Chen, Z., Lee, H. N., Negron, Y. L. S., Sultana, H., ...Harbison, S. T.
(2016). Comparison of normalization and differential expression analyses using RNA-seq
data from 726 individual drosophila melanogaster. BMC Genomics, 17, 28.
doi:10.1186/s12864-015-2353-z
Greaves, E., Collins, F., Esnal-Zufiaurre, A., Giakoumelou, S., Horne, A. W., & Saunders, P. T.
K. (2014). Estrogen receptor (ER) agonists differentially regulate neuroangiogenesis in
peritoneal endometriosis via the repellent factor SLIT3. Endocrinology, 155(10), 4015-
4026. doi:10.1210/en.2014-1086
Rübsamen, D., Kunze, M. M., Buderus, V., Brauß, T. F., Bajer, M. M., Brüne, B., & Schmid, T.
(2014). Inflammatory conditions induce IRES-dependent translation of cyp24a1. PloS
One, 9(1), e85314. doi:10.1371/journal.pone.0085314
Kokorine, I., Nisolle, M., Donnez, J., Eeckhout, Y., Courtoy, P. J., & Marbaix, E. (1997).
Expression of interstitial collagenase (matrix metalloproteinase-1) is related to the
activity of human endometriotic lesions. Fertility and Sterility, 68(2), 246-251.
doi:10.1016/S0015-0282(97)81510-5\
Chung, H., Wen, Y., Chun, S., Nezhat, C., Woo, B., & Lake Polan, M. (2001). Matrix
metalloproteinase-9 and tissue inhibitor of metalloproteinase-3 mRNA expression in
ectopic and eutopic endometrium in women with endometriosis: A rationale for
SLIT3 Gene Expression by 1,25(OH)2D3 38

endometriotic invasiveness. Fertility and Sterility, 75(1), 152-159. doi:10.1016/S0015-
0282(00)01670-8
Nap, A. W., Griffioen, A. W., Dunselman, G. A. J., Bouma-Ter Steege, Jessica C. A, Thijssen,
Victor L. J. L, Evers, J. L. H., & Groothuis, P. G. (2004). Antiangiogenesis therapy for
endometriosis. The Journal of Clinical Endocrinology & Metabolism, 89(3), 1089-1095.
doi:10.1210/jc.2003-031406
Weinstein, B. M. (2005). Vessels and nerves: Marching to the same tune. Cell, 120(4), 299.
doi: http://dx.doi.org/10.1016/j.cell.2005.01.010
Asante, A., & Taylor, R. N. (2011). Endometriosis: The role of neuroangiogenesis. Annual
Review of Physiology, 73(1), 163-182. doi:10.1146/annurev-physiol-012110-142158
Itoh, A., Miyabayashi, T., Ohno, M., & Sakano, S. (1998). Cloning and expressions of three
mammalian homologues of drosophila slit suggest possible roles for slit in the formation
and maintenance of the nervous system. Molecular Brain Research, 62(2), 175-186.
doi:10.1016/S0169-328X(98)00224-1
Brose, K., & Tessier-Lavigne, M. (2000). Slit proteins: Key regulators of axon guidance, axonal
branching, and cell migration. ENGLAND: Elsevier Ltd. doi:10.1016/S0959-
4388(99)00066-5
Geutskens, S. B., Andrews, W. D., van Stalborch, A. D., Brussen, K., Holtrop-de Haan, S. E.,
Parnavelas, J. G., ...van Hennik, P. B. (2012). Control of human hematopoietic
stem/progenitor cell migration by the extracellular matrix protein Slit3. Laboratory
Investigation; a Journal of Technical Methods and Pathology, 92(8), 1129.
doi:10.1038/labinvest.2012.81
Dickinson, R. E., Dallol, A., Bieche, I., Krex, D., Morton, D., Maher, E. R., & Latif, F. (2004).
Epigenetic inactivation of SLIT3 and SLIT1 genes in human cancers. British Journal of
Cancer, 91(12), 2071-2078. doi:10.1038/sj.bjc.6602222
Guan, H., Wei, G., Wu, J., Fang, D., Liao, Z., Xiao, H., . . . Li, Y. (2013). Down-regulation of
miR-218-2 and its host gene SLIT3 cooperate to promote invasion and progression of
thyroid cancer. The Journal of Clinical Endocrinology & Metabolism, 98(8), E1334-
E1344. doi:10.1210/jc.2013-1053
Denk, A., Braig, S., Schubert, T., & Bosserhoff, A. (2011). Slit3 inhibits activator protein 1-
mediated migration of malignant melanoma cells. International Journal of Molecular
Medicine, 28(5), 721-726.
SLIT3 Gene Expression by 1,25(OH)2D3 39

Somigliana, E., Viganò, P., Abbiati, A., Gentilini, D., Parazzini, F., Benaglia, L., ...Fedele, L.
(2010). 'Here comes the sun': Pigmentary traits and sun habits in women with
endometriosis. Human Reproduction, 25(3), 728-733. doi:10.1093/humrep/dep453
Kunisue, T., Chen, Z., Buck Louis, G. M., Sundaram, R., Hediger, M. L., Sun, L., & Kannan, K.
(2012). Urinary concentrations of benzophenone-type UV filters in US women and their
association with endometriosis. Environmental science & technology, 46(8), 4624-4632.

SLIT3 Gene Expression by 1,25(OH)2D3 40

Appendix A

Differential Gene Expression by 1,25(OH)2D3 in an Endometriosis Stromal Cell Line

Sue Ann Ingles
a,c
, Liang Wu
a
, Benjamin T. Liu
a
, Yibu Chen
b
, Chun-Yeh Wang
c
, Claire
Templeman
c
, Doerthe Bruggmann
c,d

a
University of Southern California, Department of Preventive Medicine, Los Angeles CA, U.S.A.
90089
b
University of Southern California, Department of Health Science Libraries, Bioinformatics
Service, Los Angeles, CA, U.S.A. 90089
c
University of Southern California, Department of Obstetrics/Gynecology, Los Angeles CA,
U.S.A. 90089
d
Present address: Institute of Occupational, Social and Environmental Medicine, Goethe
University, 60329 Frankfurt/Main, Germany

Note: L. Wu and B.T. Liu contributed equally to this manuscript.

E-mail addresses: ingles@usc.edu (S. Ingles), lwu441@usc.edu (L. Wu), liubenja@usc.edu (B.
Liu), yibuchen@usc.edu (Y. Chen), Doerthe.Brueggmann@med.usc.edu (D. Bruggmann)

Corresponding author: Sue Ann Ingles ingles@usc.edu

Abstract

Endometriosis is a common female reproductive disease characterized by invasion of
endometrial cells into other organs, frequently causing pelvic pain and infertility. Alterations
of the vitamin D system have been linked to endometriosis incidence and severity. To shed
light on the potential mechanism for these associations, we examined the effects of
1,25(OH)2D3 on gene expression in endometriosis cells. Stromal cell lines derived from
endometriosis tissue were treated with 1,25(OH)2D3, and RNA-seq was used to identify genes
differentially expressed between treated and untreated cells. Gene ontology and pathway
analyses were carried out using Partek Flow and Ingenuity software suites, respectively. We
identified 1627 genes that were differentially expressed (886 down-regulated and 741 up-
regulated) by 1,25(OH)2D3. Only one gene, CYP24A1, was strongly up-regulated (369-fold).
Many genes were strongly down-regulated. 1,25(OH)2D3 treatment down-regulated several
genetic pathways related to neuroangiogenesis, cellular motility, and invasion, including
pathways for axonal guidance, Rho GDP signaling, and matrix metalloprotease
inhibition. These findings support a role for vitamin D in the pathophysiology of
endometriosis, and provide new targets for investigation into possible causes and
treatments.

SLIT3 Gene Expression by 1,25(OH)2D3 41

Keywords: endometriosis, vitamin D, CYP24A1, differential gene expression, next generation
sequencing

1. Introduction

Endometriosis is one of the most common female health disorders, affecting 10-15% of all
women of reproductive age. Frequently, the condition causes chronic pelvic pain and infertility,
impacting the patient’s quality of life. It also places a considerable economic burden on society
with approximately $20 billion disease-related costs spent in the United States per year [1].
Endometriosis is characterized by growth of endometrial tissue outside of the uterine cavity.
Functional epithelial and stromal elements can be differentiated into peritoneal, ovarian and deep
infiltrating lesions according to their location [2].

Exact causes and pathogenetic pathways of this heterogenetic disease are not fully
understood. In line with the widely accepted implantation theory [3] endometrial cells enter the
abdominal cavity and attach, proliferate and invade local structures to form ectopic foci.
Estrogen dependency and changes of cellular characteristics, e.g. epithelial-mesenchymal
transition, are a prerequisite for the establishment and maintenance of lesions. It is notable that
endometriosis shares several characteristics with cancer, including excessive proliferation,
invasion into other organs, peripheral metastasis, inflammation, and (for some cancers) estrogen
dependence. During their lifetime, up to 30% of endometriosis patients will also suffer from
ovarian cancer; the risk to develop low-grade serous or endometrioid cancers is increased
approximately 2-fold, and for clear cell subtypes by 3-fold [4].

The Vitamin D system is important for numerous crucial processes in the human reproductive
system. The endometrium is not only a target for 1,25 dihydroxyvitamin D3 (1,25(OH)2D3), but
also constitutes a site of 1,25(OH)2D3 synthesis. Specifically, stromal endometrial cells were
shown to express Vitamin D receptors and 1-alpha hydroxylase, which activates 25OHD [5].
Serum levels of 25(OH)D have been reported to be both higher [6] and lower [7,8] in
endometriosis patients than in healthy individuals. Compared to controls, endometriosis patients
have been reported to have higher serum 1,25(OH)2D3 [9] and higher urinary DBP levels [10].
Also, expression of 1-alpha hydroxylase has been reported to be higher in endometriosis tissues
compared to normal endometrium [11]. Further, vitamin D has been widely studied as a possible
cancer preventive agent. Anti-tumor effects of 1,25(OH)2D3 and its analogues are mediated by
multiple pathways in a cell type and tissue specific manner. Mechanisms include inhibition of
proliferation, cell motility, invasion, metastasis, and inflammation (reviewed by Ma [12]).

Since findings on endometriosis and the vitamin D system remain inconclusive and this
complex relationship needs further investigation, the objectives of this study were (1) to identify
genes and genetic pathways that are regulated by vitamin D in endometriosis cells, and (2) to
determine whether the vitamin D system is more active in endometrial tissue and ectopic lesions
SLIT3 Gene Expression by 1,25(OH)2D3 42

than in normal endometrium. We treated stromal endometriosis cell lines with 1,25(OH)2D3 and
examined global gene expression using next-generation sequencing. We also examined
expression levels of one gene that is strongly regulated by vitamin D, CYP24A1, in tissue from
endometriosis lesions and normal endometrium.

2. Materials and Methods

2.1 Cell Culture

The endometriosis cell line (ESC22B) used in this study was of stromal origin, and derived
from peritoneal endometriosis lesions [13-14]. Cells were cultured for 25 days in media
composed of Dulbeco’s Modified Eagle’s Medium/F12 (DMEM/F12), 10% fetal bovine serum
(FBS), and 5% L-glutamine with antibiotics Penicillin-Streptomycin in humidified 5% CO2 and
95% air at 37
o
C. Cells were passaged twice, divided into 6 tissue culture dishes (3 treatment and
3 control), and allowed to proliferate to approximately 60 - 80% confluency 24 hours prior to
treatment. At this point, cells were placed into the same medium described, except that no FBS
was added, to eliminate potential influence of 1,25(OH)2D3 present in FBS. In the treatment
group, 0.1 μM 1,25(OH)2D3, which represents a supra-physiologic concentration, was dissolved
in DSMO and added into each dish. Cells were moved back into the incubator to culture for
another 4 hours before RNA extraction using RNeasy
®
Mini Kits (Qiagen, Valencia, CA, USA).

2.2 Library Construction and Next-gen Sequencing

cDNA library construction and next-generation sequencing were carried out at the USC
Epigenome Center. RNA libraries were generated using Illumina TruSeq
®
RNA-Sample
Preparation Kits (Illumina, La Jolla, CA, USA) according to Illumina’s protocols, with starting
material of 1 µg of total RNA. The Illumina Hi-seq
®
2000 platform (Illumina, La Jolla, CA,
USA) was used to generate 50 base pair paired-end reads. Libraries were applied to an Illumina
version 3 flow cell at a concentration of 16 pM. We obtained an average of 41 million reads per
sample (range 38-44).

2.3 Bioinformatic Analysis.

Genome alignment, quality control and differential gene expression was analyzed with Partek
®

Flow and Partek
®
Genomic Suite (Partek, St. Louis, MO, USA). Trimming of raw reads (both
ends) was based on minimum quality score of 20 and minimum read length of 25. Trimmed
reads were aligned to the human genome hg38 using Star2.4.1d [15] with the guidance of
Gencode v24. Aligned reads were quantified to Gencode v24 using Partek E/M
method. Normalized counts (upper quartile normalized with offset 1) were analyzed for
differential expression with the Gene-specific analysis (GSA) method. Genes were considered to
be differentially expressed if the false discovery rate (FDR) was less than 0.05 and there was at
SLIT3 Gene Expression by 1,25(OH)2D3 43

least a 2-fold difference in expression. Further function and pathway analysis was done by
Ingenuity
®
Pathway Analysis (IPA) (Ingenuity, Redwood City, CA, USA).

2.4 Tissue samples

A total of 43 tissue samples (38 endometriosis and 5 controls) were collected from women of
reproductive age. Of the 38 endometriosis samples, 6 were from endometriomas, 6 were from
bladder lesions, 5 were from other extra-uterine sites (1 cervical, 1 peritoneum, 2 rectal, 1 uterine
sacral) and 21 were from within the uterine lining. Pathology reports were used to confirm
endometriosis diagnoses and to exclude endometrial pathology in control samples. All subjects
were between 23 and 49 years of age, reported regular menstrual cycles, were not pregnant, and
had not been on hormonal treatment. Phase of menstrual cycle was recorded per the patient’s
verbal history as follicular (cycle day 5–14) versus luteal phase (day 15–30). Endometrial
pathology was excluded in healthy controls. The study was approved by the University of
Southern California Institutional Review Board (#HS 032005).
2.5 Quantitative RT-PCR

Total RNA was extracted using Qiagen Mini Kits (Qiagen, Valencia CA). cDNA was
generated using the Qiagen RT2 First Strand kit (Qiagen, Valencia CA) and then stored at -40
o
C.
Forward and reverse primers for CYP24A1 (gene of interest) and GADPH (housekeeping gene)
were designed by Qiagen. GADPH expression was found to be relatively unchanged between
case and control cell lines allowing for its use in the comparative Ct method. SYBR Green with
ROX passive reference, HotStart DNA Taq polymerase, and nucleotides were all also provided
by Qiagen in an optimized real-time PCR buffer. Quantitative PCR was performed on an
Applied Biosystems 7900HT Fast Real-Time PCR System at the USC Epigenome Center. All
samples were run in triplicate for the gene of interest and housekeeping gene.

Triplicate Ct readings were averaged and delta Ct values were calculated for each sample as
the difference between CYP24A1 and GADPH Ct values. Mean delta Ct values were compared
between endometrial and control tissues using Student’s t test. One-way analysis of variance
was used to compare tissue sites (uterine cavity, endometrioma, bladder, miscellaneous ectopic
sites, control). Post-hoc comparisons were carried out using the Bonferroni method. Relative
CYP24 expression (endometrial vs. control) was calculated as difference in delta Ct values.

3. Results and Discussion

3.1 Differential gene expression by 1,25(OH)2D3 in endometriosis stromal cells
There were 1627 genes that were at least two-fold differentially expressed (886 down-
regulated and 741 up-regulated) by 1,25(OH)2D3. Only one gene, CYP24A1, was strongly up-
regulated (369-fold). The other 740 up-regulated genes had less than 7-fold change in
expression. Many genes were strongly down-regulated by 1,25(OH)2D3. Among the 886 down-
SLIT3 Gene Expression by 1,25(OH)2D3 44

regulated genes, differential expression was more than 100-fold for 20 genes, more than 50-fold
for 49 genes, and more than 10-fold for 240 genes. The most strongly up-regulated and down-
regulated genes are listed in Tables 1a and 1b.

Several pathways were identified as being strongly regulated by 1,25(OH)2D3. Genes
contributing most strongly to these pathways are shown in Table 2.
The most strongly regulated pathway was the axonal guidance pathway, with the SLIT3 gene
being down-regulated 140-fold. The axonal guidance pathway regulates neuroangiogenesis in
normal development. In endometriosis, neuroangiogenesis plays a crucial role in the
establishment and progression of ectopic endometrial implants and may contribute to pain
associated with the disease (reviewed in [16]). The SLIT3 gene has been found to be
overexpressed in peritoneal endometriosis lesions compared to normal peritoneum
[17]. Overexpression of SLIT3 has also been proposed as a biomarker of recurrence in ovarian
endometriomas [18].

The RhoDGI Signalling pathway was also strongly regulated by 1,25(OH)2D3 in the
endometriosis stromal cells. Rho GDP-Dissociation Inhibitors (RhoGDIs) are important
regulators of the Rho family of GTPases. Rho family GTPase signaling is induced by binding of
SLIT ligands to their receptors, and influences actin dynamics and organization of the
cytoskeleten. Several genes in the cadherin family were strongly down-regulated, contributing to
the significance of this pathway.

Another pathway significantly regulated by 1,25(OH)2D3 was the Matrix Metalloproteinase
Inhibition pathway. Matrix metalloproteinases (MMPs) degrade the extracellular matrix and are
involved in remodeling of the endometrium during normal menstrual cycling. MMPs have been
hypothesized to play a role in invasion of the peritoneum in endometriosis. Altered MMP
expression has also been observed in endometrial tissues [19]. A previous study found that
1,25(OH)2D3 treatment reduced MMP2 and MMP9 expression in endometrioma-derived stromal
cells [7].

Other vitamin D pathway genes that were differentially regulated, in addition to CYP24A1
(369-fold up-regulated), were VDR, the vitamin D receptor gene (2.3-fold up-regulated), and
CYP27B1, responsible for conversion of 25(OH)D to 1,25(OH)2D3 (2.9-fold down-
regulated). This result not only underlines the successful treatment of our cells with vitamin D
and the alteration of related pathways, it also aligns with the finding of Vigano et al [5], who
documented the expression of VDR in stromal cells of the endometrium. Based on the theory of
retrograde menstruation, endometrial cells of endometriosis patients may therefore not lose the
capability to express VDR upon vitamin D stimulation when exposed to an ectopic
microenvironment.

3.2 Quantitative RT-PCR of CYP24 expression in tissue samples

SLIT3 Gene Expression by 1,25(OH)2D3 45

Because the vitamin D pathway gene, CYP24A1, is strongly and directly up-regulated by
1,25(OH)2D3, its expression can be used as an indicator of local 1,25(OH)2D3 signaling
activity. Hence, we examined CYP24A1 expression in tissue samples from endometrium of
healthy subjects and from endometriosis lesions.

We found that CYP24A1 expression was higher in endometriosis samples than in control
samples (t test, p=0.01). Therefore, the local 1,25(OH)2D3 signaling seems to be increased in
eutopic and ectopic tissue of endometriosis patients due to the disease. There was no difference
in expression by luteal vs. follicular cycle phase (p= 0.55). Thus there was no evidence of
dependency on the menstrual cycle. Expression was highest among endometriosis samples
located on the ovary (endometriomas) or within the uterine cavity. Expression was intermediate
among other ectopic lesions (bladder and other sites), and lowest among control samples (Figure
1).

Among all endometriosis lesions, CYP24A1 expression was elevated 8-fold compared to
healthy controls, indicating intense vitamin D metabolism in endometriosis tissues. Expression
was elevated 9.8-fold among lesions in the uterine cavity, 9.2-fold among endometriomas, 4.6-
fold among bladder lesions, and 6.5-fold among other ectopic lesions. Pairwise comparisons
were only statistically significant for intra-uterine lesions vs. controls (Bonferroni p =
0.05). However, the observed differences by site are consistent with an estrogen dominant
microenvironment for intra-uterine and ovarian lesions in comparison to ectopic lesions. Thus,
this finding may indicate a site-specific upregulation of CYP24A1 expression in endometriosis
lesions due to high local levels of estrogen.

We acknowledge that elevated CYP24A1 levels in the endometriosis tissues is not proof of
increased vitamin D activity, since CYP24A1 mRNA levels could also be elevated due to
increased copy numbers. Although we did not investigate the expression level of other vitamin
D pathway genes in the tissue samples, in the endometriosis cell line we did observe significant
changes in expression of the VDR and CYP27B1 genes in response to 1,25(OH)2D treatment (see
section 3.1). These genes and others found to be regulated by 1,25(OH)2D in the endometriosis
stromal cell line are good targets for future investigation.

Finally, certain limitations of this study have to be addressed. We analyzed tissue samples that
were biopsied upon visual identification of endometriosis lesions. Although they were verified as
such by the pathological report, they did not undergo micro-dissection. Hence, the demonstrated
changes in gene expression cannot be attributed to a specific cell population - stromal versus
epithelia cells - within the endometriosis lesion. Also, a minimal contamination of the sample
with local tissue, e.g. with peritoneal cells, cannot be excluded. Further, we based the menstrual
cycle phase on verbal history and did not obtain hormone levels in the patients’ blood to verify
the exact cycle phase. Also, blood vitamin D levels were not measured. This made it impossible
to relate gene expression levels to vitamin D levels of every individual patient and to exclude
potential alterations such as an existing vitamin deficiency. We also acknowledge the current
lack of any functional studies, which investigate proliferation, migration or invasion of the cells
SLIT3 Gene Expression by 1,25(OH)2D3 46

in the presence of 1,25(OH)2D3. As a natural progression of this research, we will perform these
assays in future studies.

3.3 Conclusions:
The vitamin D system appears to be activated in both eutopic and ectopic endometriosis tissue
relative to normal endometrium, as indicated by increased CYP24A1 expression. Genetic
pathways that may be altered by 1,25(OH)2D3 in endometriosis tissues were identified by analysis
of global gene expression. Treatment of endometriosis stromal cells with 1,25(OH)2D3 altered
expression of genetic pathways involved in neuroangiogenesis, cellular motility and invasion.
These results support a role for vitamin D in endometriosis pathogenesis and its transformation
to ovarian cancer and provide new targets for investigation into possible causes and treatments.

Conflict of Interest

The authors declare to have no conflict of interest.

Acknowledgements

We are grateful to Prof. Dr. A. Starzinski-Powitz (Goethe University, Frankurt/Maine) for
providing the cell line used in this study, to Dr. Daniel Weisenberger for expert advice on RT-
PCR, and Dr. Meng Li for bioinformatics assistance. The bioinformatics software and computing
resources used in the analysis are funded by the USC Office of Research and the Norris Medical
Library.

Funding

This research did not receive any specific grant from funding agencies in the public,
commercial, or not-for-profit sectors.

References

[1] X. Gao, J. Outley, M. Botteman, J. Spalding, J.A. Simon, C.L. Pashos, Economic burden of
endometriosis, Fertil Steril, 86 (2006) 1561-72. DOI: 10.1016/j.fertnstert.2006.06.015
[2] M. Nisolle, J. Donnez, Peritoneal endometriosis, ovarian endometriosis, and adenomyotic
nodules of the rectovaginal septum are three different entities, Fertil Steril 68 (1997) 585-596.
[3] A. Sampson, Peritoneal endometriosis due to menstrual dissemination of endometrial tissue
into the peritoneal cavity, Am J Obstet Gynecol 14 (1927) 42-469.
[4] S. Ogawa, T. Kaku, S. Amada, H. Kobayashi, T. Hirakawa, K. Ariyoshi, T. Kamura, H.
Nakano, Ovarian endometriosis associated with ovarian carcinoma: a clinicopathological
SLIT3 Gene Expression by 1,25(OH)2D3 47

andimmunohistochemical study, Gynecol Oncol 77 (2000), 298-304. DOI:
10.1006/gyno.2000.5765
[5] P. Vigano, D. Lattuada, S. Mangioni, L. Ermellino, M. Vignali, E. Caporizzo, P. Panina-
Bordignon, M. Besozzi, A.M. Di Blasio, Cycling and early pregnant endometrium as a site of
regulated expression of the vitamin D system, J Mol Endocrinol 36 (2006) 415–24. DOI:
10.1677/jme.1.01946
[6 ] E. Somigliana, P. Panina-Bordignon, S. Murone, P. di Lucia, P. Vercellini, P. Vigano,
Vitamin D Reserve is Higher in Women with Endometriosis, Hum Reprod 22 (2007) 2273–
2278. DOI: 10.1093/humrep/dem142
[7] M. Miyashita, K, Koga, G. Izumi, F. Sue, T. Makabe, A. Taguchi, M. Nagai, Y. Urata, M.
Takamura, M. Harada, T. Hirata, Y. Hirota, O Wada-Hiraike, T Fujii, Y. Osuga, Effects of 1,25-
dihydroxy Vitamin D3 on Endometriosis, J. Clin. Endocinol. Metabolism, 101 (2016) 2371-
2379. DOI: 10.1210/jc.2016-1515
[8] H.R. Harris, J.E. Chavarro, S. Malspeis, W.C. Willet, S.A. Missmer, Dairy-Food, Calcium,
Magnesium, and Vitamin D Intake and Endometriosis: A Prospective Cohort Study, Am. J.
Epidemiol. 177 (2012) 420-430. DOI: 10.1093/aje/kws247
[9] D. Hartwell, P. Rødbro, S.B. Jensen, K. Thomsen, C. Christiansen, Vitamin D Metabolites -
Relation to Age, Menopause, and Endometriosis, Scand. J. Clin. Lab. Invest. 50 (1990) 115-121.
[10] S. Cho, Y.S. Choi, S.Y. Yim, H.I. Yang, Y.E. Jeon, K.E. Lee, HY. Kim, S.K. Seo, B.S.
Lee, Urinary Vitamin D-Binding Protein is elevated in patients with endometriosis, Human
Reproduction 27 (2012) 515-522. DOI: 10.1093/humrep/der345
[11] A. Agic, H. Xu, C. Altgassen, F. Noack, M.M. Wolfler, K. Diedrich, M. Friedrich, R.N.
Taylor, D. Hornung, Relative Expression of 1,25-Dihydroxyvitamin D3 Receptor, Vitamin D 1
alpha Hydroxylase, Vitamin D 24-Hydroxylase, and Vitamin D 25-Hydroxylase in
Endometriosis and Gynecologic Cancers, Reprod. Sci. 14 (2007) 486-497. DOI:
10.1177/1933719107304565
[12] Y. Ma, C.S. Johnson, D.L. Trump, Mechanistic Insights of Vitamin D Anticancer Effects,
Vitamins and Hormones 100 (2016) 395-431. DOI: 10.1016/bs.vh.2015.11.003
[13] A. Zeitvogel, R. Baumann, A. Starzinski-Powitz, Identification of an invasive, N-cadherin-
expressing epithelial cell type in endometriosis using a new cell culture model, 159 (2001) Am J
Pathol 1839–1852. DOI: 10.1016/S0002-9440(10)63030-1
[14] S.K. Banu, J.H. Lee, A. Starzinski-Powitz, J.A. Arosh, Gene expression profiles and
functional characterization of human immortalized endometriotic epithelial and stromal cells,
Fertil. Steril. 90(2008), 972-987. DOI: 10.1016/j.fertnstert.2007.07.1358
[15] A. Dobin, C.A. Davis, F. Schlesinger, J. Drenkow, C. Zaleski, S. Jha, P. Batut, M. Chaisson, T.R.
Gingeras, STAR: ultrafast universal RNA-seq aligner, Bioinformatics 29(2013) 15-21. DOI:
10.1093/bioinformatics/bts635
[16] A. Asante, R.N. Taylor, Endometriosis: the role of neuroangiogenesis, Annu. Rev. Physiol.
73 (2011) 163-182. DOI: 10.1146/annurev-physiol-012110-142158.
[17] E. Greaves, F. Collins, A. Esnal-Zufiaurre, S. Giakoumelou, A.W. Horone, P.T.K.
Saunders, Estrogen Receptor (ER) Agonists Differentially Regulate Neuroangiogenesis in
Peritoneal Endometriosis vis the Repellent Factor SLIT3, Endocrinol. 155 (2014) 4015-
4026. DOI: 10.1210/en.2014-1086
SLIT3 Gene Expression by 1,25(OH)2D3 48

[18] F. Shen, X. Liu, J.-G. Geng, S.-W. Guo, Increased Immunoreactivity to SLIT/ROBO1 in
Ovarian Endometriomas. A Likely Constituent Biomarker for Recurrence, Am. J. Pathol. 175
(2009) 479-488. DOI: 10.2353/ajpath.2009.090024
[19] H.W. Chung, J.Y. Lee, H.-S. Moon, S. E. Hur, M.H. Park, Y. Wen, M.L. Pollan, Matrix
metalloproteinase-2, membranous type 1 matrix metalloproteinase, and tissue inhibitor of
metalloproteinase-2 expression in ectopic and eutopic endometrium, Fertility and Sterility 78
(2002) 787-795.

Abstract (if available)

Abstract Endometriosis is a common and sometimes severe health disorder in women. This disease is characterized by the invasion of endometrial cells into other areas of the body such as the peritoneum, ovaries, bladder, and more. Symptoms of endometriosis include but are not limited to severe dysmenorrhea, deep dyspareunia, chronic pelvic pain, ovulation pain, menstrual pain, irregular flow, premenstrual spotting, infertility, chronic fatigue, or any combination of the list. Vitamin D3 is a secosteroid hormone that plays an important role not only in bone development and maintenance but also in the cell proliferation, differentiation, and apoptosis. Calcitriol (1,25(OH)₂D₃) is the biologically active metabolite of vitamin D3. Several studies have shown that a vitamin D system is active in certain tissues of the female reproductive system. In this study, we examine the effects of 1,25(OH)₂D₃ on gene expression in endometrial tissue obtained from patients with endometriosis. Stromal cell lines derived endometriosis patients were treated with 1,25(OH)₂D₃. Using RNA-Seq, differential gene expression was compared between treated and untreated groups. 1627 genes were at least two-fold differentially expressed. SLIT3, part of the axonal guidance signaling pathway and one of the most down-regulated genes, plays an important role in inhibiting cell migration and invasion and stimulating neuroangiogenesis. Using quantitative RT-PCR, the more invasive endometriosis lesions were determined to have higher SLIT3 expression compared to lesions in the uterine cavity. These results provide evidence for SLIT3 as a gene target in endometriosis and vitamin D3 as a treatment for this disease.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Identification of gene expression regulated by 1,25(OH)₂D3 in human endometriosis cell lines with next-generation sequencing

Asset Metadata

Creator Liu, Benjamin (author)

Core Title SLIT3 gene expression by 1,25(OH)₂D₃ in an endometriosis stromal cell line

School Keck School of Medicine

Degree Master of Science

Degree Program Molecular Epidemiology

Publication Date 07/19/2017

Defense Date 07/05/2017

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

Tag 1,25(OH)₂D3,differential gene expression,endometriosis,next-generation sequencing,OAI-PMH Harvest,quantitative RT-PCR,RNA-seq,SLIT3,vitamin D

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Ingles, Sue (committee chair), McGinnis, Lynda (committee member), Stern, Mariana (committee member)

Creator Email benjaminthomasliu@gmail.com,liubenja@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-403842

Unique identifier UC11264471

Identifier etd-LiuBenjami-5551.pdf (filename),usctheses-c40-403842 (legacy record id)

Legacy Identifier etd-LiuBenjami-5551.pdf

Dmrecord 403842

Document Type Thesis

Rights Liu, Benjamin

Type texts

Source 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 a...

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