University of Southern California Dissertations and Theses

Genomic and phenotypic novelties in the Southeast Asian house mouse

(USC Thesis Other)

Genomic and phenotypic novelties in the Southeast Asian house mouse

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Genomic and phenotypic novelties in the Southeast Asian house mouse
by
Noah Malae Simon
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
(BIOLOGY OF AGING)
May 2024
Copyright 2024 Noah Malae Simon

ii
Dedicated to:
Friends and family,
Wherever you may be
Thanks for keeping me sane
While the world goes mad

iii
Acknowledgements
Thank you Rachel, for mentorship, guidance, and lots of patience.
Special thanks to Taekyu Kang: I learned a lot since shadowing you during my
lab rotation. All of my coding and cell culture knowledge is built on that foundation. More
importantly, you were an invaluable drinking buddy while we were penned up during
COVID lockdown.
Thanks to present and past members of the Brem Lab for good scientific
discussion and good vibes, and a legacy of beautiful python scripts for me to cut my
teeth on.
These projects would have been impossible without the expertise in stem cell
biology of James Dutton and his team at the Stem Cell Institute at the University of
Minnesota. Particularly, so much of this hinges on masterful stem cell culture by Yujin
Kim. Maxwell Frier, a recent addition, has already made valuable contributions on this
front and with modeling of axon growth data.,
The SENse lab at UC Berkeley, co-run by Diana Bautista and Ellen Lumpkin,
provided many resources. I learned a lot of neurobiology that I otherwise wouldn’t have.
And this space was like a second scientific home over the past few years. Great science
and, somehow, even better friends.
To my committee members, Jennifer Garrison, Tara Tracy, Bérénice Benayoun
and Matt Dean: thank for your valuable discussions and feedback throughout this
journey.

iv
Chapter 2 is a version of a manuscript titled “Stem cell transcriptional profiles
from mouse subspecies reveal cis-regulatory evolution at translation genes.” We are
currently finalizing revisions for resubmission to Heredity. Chapter 3 is a version of an
in-progress manuscript titled “A stem cell-derived neuronal culture model for mouse
species variation in neurite growth” that will be submitted at a later date. The authors on
both of these are myself (as first author), Yujin Kim, Maxwell Frier, Diana Bautista,
James Dutton, and Rachel Brem. I am very grateful to my fellow authors for granting
permission to use these works here. This work was supported by the National Institutes
of Health grant R01NS116992 awarded to Diana M. Bautista, James R. Dutton, and
Rachel B. Brem.

v
Table of Contents
Dedication ........................................................................................................................ii
Acknowledgements .........................................................................................................iii
List of Tables...................................................................................................................ix
List of Figures.................................................................................................................. x
List of Supplemental Tables ............................................................................................xi
List of Supplemental Figures..........................................................................................xii
Abstract.........................................................................................................................xiii
Chapter 1: Introduction.................................................................................................... 1
The genetic study of natural variation.......................................................................... 1
Mus musculus castaneus as a model of natural variation............................................ 1
Neurite degeneration ................................................................................................... 2
The genetics of neurite regeneration ........................................................................... 4
Translation changes in response to stress and aging.................................................. 7
Translation and neurite extension................................................................................ 9
Chapter 2: Stem cell transcriptional profiles from mouse subspecies reveal
cis-regulatory evolution at translation genes ................................................................. 11
Abstract ..................................................................................................................... 11
Introduction................................................................................................................ 11
Results....................................................................................................................... 13

vi
A screen for directional cis-regulatory change in mouse stem cell pathways......... 13
Unique and conserved alleles in M. m. castaneus in translation gene
upstream regions ................................................................................................... 17
A role for transcription factors in divergently regulated translation genes .............. 19
Discussion ................................................................................................................. 21
Materials and Methods .............................................................................................. 24
RNA-seq data sources ........................................................................................... 24
RNA-seq read mapping.......................................................................................... 25
RNA-seq normalization and quantification ............................................................. 26
Hybrid RNA-seq mapping quality control ............................................................... 26
In silico screen for directional cis-regulatory variation............................................ 27
Induced pluripotent stem cell derivation and culture .............................................. 28
RNA isolation and sequencing ............................................................................... 28
Population genomic analysis.................................................................................. 29
Analysis of transcription factor binding sites .......................................................... 30
Supplemental Figures................................................................................................ 32
Supplemental Tables................................................................................................. 35
Chapter 3: A stem cell-derived neuronal culture model for mouse species variation
in neurite growth............................................................................................................ 53
Abstract ..................................................................................................................... 53

vii
Introduction................................................................................................................ 53
Results....................................................................................................................... 55
Discussion ................................................................................................................. 59
Methods..................................................................................................................... 62
Mouse pluripotent stem cell lines ........................................................................... 62
Stem cell culture..................................................................................................... 63
Motor neuron differentiation ................................................................................... 64
Preparation of mPSC motor neurons for culture or cryopreservation..................... 65
mPSC derived motor neuron culture...................................................................... 65
Microfluidic mPSC motor neuron culture................................................................ 66
RNA isolation and sequencing ............................................................................... 66
Transcriptional profiling .......................................................................................... 67
Immunocytochemistry ............................................................................................ 68
Fluorescence microscopy ...................................................................................... 69
Quantification of neurite area ................................................................................. 69
Machine learning model for quantifying cell bodies................................................ 70
Linear assessment of neurite extension................................................................. 71
Supplemental Figures................................................................................................ 73
Chapter 4: Revamping a genome-wide reciprocal hemizygosity screening method
for use in a mammalian cell culture system................................................................... 79

viii
Abstract ..................................................................................................................... 79
Introduction................................................................................................................ 79
Results....................................................................................................................... 82
Discussion ................................................................................................................. 85
Methods..................................................................................................................... 85
Lentivirus production.............................................................................................. 85
Hybrid cell line........................................................................................................ 87
mPSC Culture ........................................................................................................ 87
Hybrid Mutagenesis ............................................................................................... 88
NGS library preparation and DNA sequencing....................................................... 89
Data analysis.......................................................................................................... 91
Supplemental Figures................................................................................................ 92
Conclusions................................................................................................................... 94
High expression of translation genes in M. m. castaneus stem cells is partly
driven by directional cis-regulatory evolution ............................................................. 94
M. m. castaneus neurons grow longer neurites in the absence of injury ................... 94
References.................................................................................................................... 95

ix
List of Tables
Table 2.1. A screen for directional allele-specific expression in CAST x 129 hybrid
embryonic stem cells..................................................................................................... 15
Table 2.2. Two-factor ANOVA results, Du/Pu enrichment in transcription factor
binding sites upstream of translation GO genes............................................................ 22

x
List of Figures
Figure 2.1: Induction of translation genes in M. m. castaneus ...................................... 16
Figure 2.2: Elevated nucleotide divergence between M. m. castaneus and other
subspecies in regions upstream of translation genes.................................................... 18
Figure 2.3: The transcription factor Eed is a strong candidate as a regulator of
translation genes in M. m. castaneus............................................................................ 20
Figure 3.1: Differentiation of motor neurons from stem cells and transcriptional
profiling.......................................................................................................................... 56
Figure 3.2: Microfluidic assays of neurite extension from mouse stem cell-derived
neurons ......................................................................................................................... 58
Figure 3.3: Inheritance from M. m. castaneus is sufficient for increased neurite
outgrowth in stem cell-derived neurons......................................................................... 60
Figure 4.1: Reciprocal hemizygosity test....................................................................... 80
Figure 4.2: PCR step of library prep amplifies gDNA fragments containing
lentiviral LTR sequence................................................................................................. 81
Figure 4.3: Lentiviral insertions are dispersed across the genome. .............................. 83
Figure 4.4: Comparing read abundances of lentiviral insertions after cell culture
expansion...................................................................................................................... 84

xi
List of Supplemental Tables
Supplemental Table 2.1: A screen for directional allele-specific expression in
CAST x 129 hybrid embryonic stem cells...................................................................... 35
Supplemental Table 2.2. Signatures of selection between mouse subspecies in
transcription factor binding sites upstream of translation genes.................................... 39

xii
List of Supplemental Figures
Supplemental Figure 2.1: Relative expression of translation genes in RNA-seq
datasets containing CAST homozygous or hybrid stem cells........................................ 32
Supplemental Figure 2.2: Ehmt2 knockout weakly induces translation genes .............. 34
Supplemental Figure 3.1: Comparing pluripotent stem cells and stem cell-derived
neuronal transcriptomes across Mus genotypes ........................................................... 73
Supplemental Figure 3.2: Design of neuronal culture experiments. .............................. 74
Supplemental Figure 3.3: Cell body inference in image analysis of stem cellderived neurons. ........................................................................................................... 75
Supplemental Figure 3.4: Quantification of neurite extension measurements............... 76
Supplemental Figure 3.5: Statistical analyses of neurite extension and growth ............ 78
Supplemental Figure 4.1: Amplification of endogenous loci in initial mutagenesis
experiments................................................................................................................... 92
Supplemental Figure 4.2: Primer and linker sequences used for library prep ............... 93

xiii
Abstract
Much of what we know about biology comes from studies of a handful of model
organisms. The diversity across the rest of the taxa that inhabit our planet represents a
treasure trove of information about biological mechanisms that is relatively untouched.
In this thesis, I focus on a little-studied subspecies of the common house mouse that
originates in Southeast Asia, Mus musculus castaneus. Compared to their sister
subspecies which are commonly used in lab studies, M. m. castaneus have a number of
interesting traits, including axon regeneration (Omura et al. 2015), resistance to agerelated hearing loss (Johnson et al. 1997; Johnson et al. 2006), wound healing (HeberKatz et al. 2004), and resistance to DNA damage (French et al. 2015; Chappell et al.
2017). A better understanding of these and other yet-to-be-discovered unique properties
of M. m. castaneus could inform new biomedical applications, while also serving as a
tool to study evolutionary mechanisms. In this thesis, I describe different approaches
that I, along with collaborators, have used to study M. m. castaneus through the lens of
natural variation.
Chapter 2 describes an unbiased screen for directional gene expression changes
between mouse taxa using expression data in hybrid stem cells. Through this screen,
we discovered a novel program of upregulated expression of translation genes in M. m.
castaneus stem cells driven by changes in cis- and trans-acting regulators, attesting to
a change in selective pressure on translation pathway between M. m. castaneus and its
sister species. Complementing this expression data with sequence-based analyses, we
identified signatures of evolutionary selection (elevated sequence divergence) proximal
to the translation genes. These analyses pinpointed evidence for adaptive changes in

xiv
M. m. castaneus in binding sites for the transcriptional regulators Ehmt2, Eed, and
Msx1.
In Chapter 3, I describe a study of species variation in cell-autonomous axon
growth behaviors, using stem cell-derived neurons. In this in vitro system, we found that
M. m. castaneus neurons grew longer extensions than did those from M. m. domesticus
even in the absence of injury. And we found that stem cell-derived neurons from an F1
hybrid between these two species phenocopied those of M. m. castaneus, indicating a
dominant genetic phenotype. In Chapter 4, I lay out advances in development of
methods for an unbiased genetic screening approach in mammalian cells to identify
allelic variation that underlies species-unique traits. The latter represents important
steps towards the eventual goal of finding the genetic basis of M. m. castaneus axon
extension and regeneration phenotypes.

1
Chapter 1: Introduction
The genetic study of natural variation
A fascination with the trait variation in the natural world has captured the human
imagination as far back as the philosophers of ancient Greece (Osborn 1894). Since
Darwin and Mendel laid the scientific foundations in the 1800s (Darwin 1872; Hoßfeld et
al. 2017), the study of evolution and genetics has come a long way. Modern technology
has allowed us to map out our own genomes (Gibbs 2020) and we now have an arsenal
of tools for linking genotype to phenotype. Modern methods such as linkage analysis
(Ott et al. 2015) and genome-wide association studies (Uffelmann et al. 2021) allow us
to explore the genetics underlying natural variation by pairing high-throughput
sequencing and phenotyping in an unbiased manner. These methods have been
utilized in the context of aging biology to search for genes contributing to natural
variation in longevity or age-related diseases (Pasyukova et al. 2000; Ayyadevara et al.
2001; Jeck et al. 2012; Deelen et al. 2019).
Mus musculus castaneus as a model of natural variation
In this thesis, I use house mice as a model for the study of sequence and trait
differences that manifest in the present day but arose long ago. The species Mus
musculus comprises house mice native to Eurasia. M. musculus is split into three main
subspecies: musculus, domesticus, and castaneus (Yang et al. 2007; Yang et al. 2011;
Fujiwara et al. 2021). Genomic analyses suggest that M. m. domesticus and castaneus
diverged ~500,000 years ago (Phifer-Rixey et al. 2020), but still maintain low levels of
gene flow in the wild (Harr et al. 2016). The most commonly studied laboratory mice are

2
inbred strains such as C57BL/6J, which are primarily of M. m. domesticus origin (Yang
et al. 2011). By contrast, M. m. castaneus mice are relatively understudied, and they
provide a valuable resource for studying natural genetic variation and uncovering novel
genes and traits, especially in the context of the rest of the species complex.
The CAST/EiJ strain (Jackson Laboratory Strain #000928) originated from wild
M. m. castaneus caught in Thailand (Chapman and Ruddle 1972). These mice were
maintained and subsequently derived into an inbred strain (Roderick, Thomas 1982)
which is currently maintained and sold by The Jackson Laboratory. Studies using
CAST/EiJ mice and cells have identified genes involved in resistance to age-related
hearing loss (Johnson et al. 1997; Johnson et al. 2006), wound healing (Heber-Katz et
al. 2004), resistance to DNA damage (French et al. 2015; Chappell et al. 2017), bone
density (Beamer et al. 1999; Yu et al. 2007), behaviors related to short-term memory
(Hsiao et al. 2020), and axon regeneration (Omura et al. 2015).
Neurite degeneration
As a motivating trait for Chapter 3 of this thesis, I study the biology of neurons in
mice. The ability of the central nervous system to coordinate information and action at
distal sites relies on the projections of neuronal cells. The subset of these projections
called axons transmit nerve impulses from the cell body of a neuron to other neurons,
muscles, and glands. During development, young neurons in both the peripheral and
central nervous systems (PNS and CNS respectively) must grow long distances to
innervate target cells and tissues; this ability is greatly diminished in the adult CNS
(Fawcett 2020). Additionally, new neurons in adult mammals are only known to be
generated by neural stem cells in three brain regions: in the subventricular zone of the

3
lateral ventral wall, the subgranular zone of the dentate gyrus in the hippocampus, and
the amygdala (Galvan and Jin, 2007; Fernandes et al., 2015; Jhaveri et al., 2018). In
others tissue of the adult CNS, if axons are lost as a consequence of injury or disease,
they cannot be recovered.
One type of axon loss, Wallerian degeneration, was classically described (Waller
and Owen 1850) as a distal pattern of axon degeneration following a more proximal
nerve transection. Since then, Wallerian-like axon degeneration, characterized by a
“dying back” pattern of damage occurring first in the distal sections of the axons
(Beirowski et al. 2005; Conforti et al. 2014; Koliatsos and Alexandris 2019; Coleman
and Höke 2020), has been described in conditions such as traumatic brain injury
(Koliatsos and Alexandris 2019), stroke (Hinman 2014), diabetic neuropathy (Yasuda et
al. 2003), Alzheimer’s disease (Selkoe 2002; Kanaan et al. 2013), amyotrophic lateral
sclerosis (Dadon-Nachum et al. 2011), Charcot-Marie-Tooth disease (Reilly et al. 2011),
and multiple sclerosis (Dutta and Trapp 2007; Singh et al. 2017).
Loss of axons also occurs during normal aging (Pannese 2011). The number of
axons in the optic nerve decreases with age in multiple mammalian species, whereas
the number of cells from which these project is relatively unchanged (Calkins 2013).
Axon numbers in the ear also decline with age, and also at a higher rate than the loss of
corresponding cells (Wu et al. 2019). Likewise, the decrease of axon number and
function occurs with age in other tissues throughout the body and likely contributes to
age-related sensory, motor and cognitive deficits (Verdú et al. 2000; Pannese 2011). In
conditions where axon loss precedes loss of the cell body, targeting axon damage at
early stages may be effective in preventing cell death (Coleman and Höke 2020),

4
opening up a potential for future regeneration-promoting therapies in treating these
disorders.
Decades of research have shown that the inhibition of axonal regeneration in
adult mammals is complex, involving both cell-autonomous factors and extrinsic factors
of the central nervous system environment (Filbin, 2003; Fitch and Silver, 2008; Kadoya
et al., 2009; Sun and He, 2010; Tedeschi and Bradke, 2017b; Fawcett, 2020). In the
CNS, the extracellular matrix contains factors such as NogoA, chondroitin sulphate
proteoglycans (CSPGs), myelin-associated glycoprotein, oligodendrocyte myelin
glycoprotein, and semaphorin 3A, which inhibit neurite growth after injury (Filbin 2003;
Fitch and Silver 2008; Tedeschi and Bradke 2017; Fawcett 2020). CSPGs in particular
are produced by glial cells and are highly induced after injury (Fitch and Silver 2008).
Injury to PNS neurons induces transcriptional programs that promote regeneration
(Tedeschi 2012), whereas these programs are repressed in adult CNS neurons (Cho et
al. 2013). It is thought that inhibiting axon growth in the adult CNS serves to stabilize the
structure of complex neuronal networks in the brain and spinal cord (Sun and He 2010).
The genetics of neurite regeneration
The literature on axonal injury identifies some genes that have altered expression
or activity after axotomy. Axon injury has been shown to have a priming effect, where
gene expression changes promote axon regrowth (Kadoya et al. 2009). Experimental
manipulation of these genes has been shown either to limit axon degeneration following
injury or to promote subsequent regeneration, highlighting some known biochemical
pathways that may be therapeutic targets for neurologic disease.

5
The ‘slow Wallerian degeneration’ (WldS) mutation in mice is a well-studied
example of this. Mice carrying this mutation were characterized as having delayed axon
degeneration after in vivo nerve transection in an early study (Lunn et al. 1989). Since
then, this mutation has been shown to be neuroprotective in mouse models of diseases
such as Parkinson’s disease (Sajadi et al. 2004), Charcot-Marie-Tooth disease
(Samsam et al. 2003), paclitaxel treatment (Wang et al. 2002), glaucoma (Howell et al.
2007), and a tauopathic model of Alzheimer’s disease (Ljungberg et al. 2012). The WldS
gene is the result of a tandem triplication of an ~85kb stretch of DNA on chromosome 4
(Coleman et al. 1998), where the genomic rearrangement resulted in a novel chimeric
gene encoding a protein which is a fusion of 70 N-terminal amino acids from the
ubiquitination Factor E4B (UBE4B) protein and the full-length nicotinamide
mononucleotide adenylyltransferase 1 (NMNAT1) protein (Conforti et al. 2007).
NMNAT1 is responsible for NAD+ synthesis, and activity of the NMNAT1 portion of the
WLDS protein has been shown to play a role in its neuroprotective effect (Gilley and
Coleman 2010; Milde et al. 2013). The related NMNAT2 enzyme has a similar function
and has been shown to be necessary for axon growth and maintenance (Gilley and
Coleman 2010; Coleman and Höke 2020). NMNAT2 has a short half-life and new
protein must be transported to the distal parts of the axon after being synthesized in the
soma (Gilley and Coleman 2010; Conforti et al. 2014). Disruption of axonal transport,
such as from a transection injury, prevents new NMNAT2 from reaching the distal
sections of axon, and NAD+ production is halted (Gilley and Coleman 2010). Presence
of the more stable WLDS produces requisite amounts of NAD+ and is sufficient to
prolong axon survival after a severing injury compared to wildtype (Gilley and Coleman

6
2010; Milde et al. 2013). Two chemotherapeutic drugs (vincristine and bortezomib) have
been shown to induce decreased levels of axonal NMNAT2 in vitro (Geisler et al. 2019).
Another gene, sarm1, plays a role in axon degeneration via NAD+ signaling. Null
mutation or shRNA knockdown of sarm1 leads to prolonged axon survival after
mechanical injury, indicating that this gene has a pro-degeneration effect (Osterloh et al.
2012; Gerdts et al. 2013). Sarm1 was shown to decrease NAD+ concentrations after
axon injury, and has been demonstrated to be necessary and sufficient for this
decrease (Gerdts et al. 2013; Sasaki et al. 2016). Sarm1 knockout prevented
vincristine-induced axon degeneration and behavioral measures of CIPN in mice
(Geisler et al. 2016) and decreased behavioral CIPN symptoms in paclitaxel treated
mice (Turkiew et al. 2017). In another study, Sarm1 knockout prevented NAD+ depletion
and axon degeneration induced by vincristine or bortezomib treatment (Geisler et al.
2019).
High levels of SPRR1A expression and protein were measured after in vivo
axotomy (Bonilla et al. 2002). SPRR1A overexpression was also shown to increase
axon growth in uninjured cells, including when grown on substrates inhibitory to axon
growth (Bonilla et al. 2002). Sprr1a upregulation by the transcription factor SOX11 was
shown to promote axon regeneration in mice (Jing et al. 2012).
There is evidence that the mammalian target of rapamycin (mTOR) pathway, a
classic determinant of aging (Papadopoli et al. 2019), is involved in axon regeneration.
One study reported that deletion of the phosphatase and tensin homolog gene (PTEN),
which downregulates mTOR, resulted in increased axon regeneration following optic
nerve crush (Park et al. 2008). This effect was abolished after treatment with rapamycin,

7
an mTOR inhibitor. PTEN deletion by itself promoted cortical motor neuron regeneration
(Liu et al. 2010), and concurrent deletion of another gene, SOC3, had a larger effect
(Jin et al. 2015). The pro-regenerative effect of mTOR described in mice has also been
demonstrated in in vivo experiments on human retinal ganglion cells (Teotia et al.
2019).
A 2015 study identified a unique axon regeneration phenotype in CAST/EiJ from
a panel of ex vivo dorsal root ganglia neurons from nine mouse strains (Omura et al.
2015). In addition to the increased regeneration in cell culture, the authors also
demonstrated the ability of CAST/EiJ mice to regenerate axons in vivo in the CNS after
dorsal column crush, optic nerve crush, and induced stroke. Activin signaling was
shown to play a role in this regeneration phenotype through higher baseline expression
of Activin receptors (Omura et al. 2015), but other adaptions in this pathway may be
present in M. m. castaneus. Activin supplementation to C57BL/6J neurons did not
induce the same magnitude of neurite growth measured in CAST/EiJ neurons in either
naïve or injury conditions (Omura et al. 2015), which suggests there are additional
evolved mechanisms contributing to axon regeneration in CAST/EiJ.
Translation changes in response to stress and aging
In Chapter 2 of this thesis, I study an ancient divergence between mouse
lineages in the expression of a suite of genes that act in translation. Translation
pathways have been recognized for decades as a linchpin of aging. Rates of protein
translation peak in early adulthood and decline with age in yeast, C. elegans, D.
melanogaster, mice, and humans (Kim and Pickering 2023). Loss of translation during
aging has been linked to decreases in levels of initiation factors eIF2 and eIF5 (Kimball

8
et al. 1992; Luchessi et al. 2008), elongation factor eEF1A (Webster et al. 1981),
ribosome biogenesis genes (Anisimova et al. 2020), and aberrant ribosome kinetics
(Stein et al. 2022), indicating dysfunction at multiple steps of protein synthesis.
And yet inhibiting translation genetically or pharmacologically has resulted in
increased lifespan and healthspan in multiple species (Kapahi et al. 2004; Scheuner et
al. 2005; Curran and Ruvkun 2007; Pan et al. 2007; Syntichaki et al. 2007; Steffen et al.
2008; Selman et al. 2009; Tain et al. 2009; Blagden and Willis 2011; Rogers et al. 2011;
Martin et al. 2014; Takauji et al. 2016; Ren et al. 2019). Additionally, increased
translation rates have been reported in cells cultured from humans with HutchinsonGilford progeria syndrome, a progeria disease (Buchwalter and Hetzer 2017). A
proposed explanation for this paradox is that protein catabolism rates decrease during
normal aging, and the decrease in protein synthesis may be a compensatory change
(Kim and Pickering 2023).
Regulation of translation is a crucial aspect of cellular homeostasis, as
concentrations of thousands of proteins need to be maintained and balanced against
protein catabolism (Abreu et al. 2009). The mTOR pathway plays a major role in
translation regulation. mTOR is activated by signals such as nutrient availability,
oxygen, and growth factors, and promotes pathways involved in cell growth and
proliferation, including protein synthesis, as a part of the mTOR complex 1 (mTORC1)
(Saxton and Sabatini 2017). Inversely, inhibition of mTOR via stress or starvation
inhibits protein synthesis. mTOR signaling regulates various cellular translation
components including translation initiation factors, elongation factors, and ribosome
biogenesis via direct phosphorylation of target proteins or indirect phosphorylation by

9
signal transduction cascades involving other downstream kinases (Wang and Proud
2006; Ma and Blenis 2009).
In eukaryotes, the general response to stress is to inhibit protein synthesis via
the integrated stress response (ISR) signaling pathway (Simpson and Ashe 2012;
Pakos‐Zebrucka et al. 2016; Lacerda et al. 2019). The primary benefit of inhibiting
protein synthesis is thought to be energetic: by clamping down on costly translational
processes, cellular energy can be dedicated instead to stress response (Advani and
Ivanov 2019). A wide variety of cellular stressors, including nutrient deprivation,
oxidative stress, DNA damage, and oncogene activation, trigger the ISR via
phosphorylation of the eukaryotic initiation factor alpha subunit (eIF2α) (Pakos‐
Zebrucka et al. 2016; Lacerda et al. 2019). Phosphorylation of eIF2α inhibits translation
initiation, which is the rate-limiting step in protein synthesis (Jackson et al. 2010).
Although global protein synthesis is inhibited by the ISR, mRNAs coding for stressresponse proteins are preferentially translated (Lacerda et al. 2019), with the repertoire
of these mRNAs varying between stressors (Simpson and Ashe 2012).
Translation and neurite extension
As I show in Chapter 2, M. m. castaneus, the mouse subspecies whose neurons
exhibit unusually high axon regeneration, is also characterized by a phenotype of high
expression of translation genes in stem cells. A central outcome of this thesis is to raise
the possibility that these two properties might be related. In general, in neurons,
translation plays a critical role in axon growth. In neuronal projections, cellular
processes must be coordinated long distances from the cell’s nucleus. The time it takes
to transport cellular materials such mRNAs and proteins back and forth from the

10
nucleus may be prohibitively long in cases where a quick response is necessary. Local
protein synthesis in axon compartments would solve this issue, and experimental
evidence has indeed implicated axonal translation as a key player in neuronal biology.
Local protein translation at the axon terminal allows for quick, and precise responses to
extracellular guidance cues during axon extension (Lin and Holt 2008; Yoon et al. 2009)
and has been shown to be required for synapse formation (Zhang and Poo 2002).
Transcription factors synthesized at the site of injury are transported back to the nucleus
and have been shown to mediate the neuron’s injury response (Hanz et al. 2003; Hanz
and Fainzilber 2006; Cox et al. 2008). Interestingly, decreasing levels of axonal
translation have been correlated with a decrease in axon regeneration using human
stem cell-derived neurons (van Erp et al. 2021). With respect to M. m. castaneus, future
work will focus on translation regulation as a potential mechanism for the neurite
outgrowth and regeneration traits that distinguish this lineage.

11
Chapter 2: Stem cell transcriptional profiles from mouse subspecies reveal cisregulatory evolution at translation genes
Abstract
A key goal of evolutionary genomics is to harness molecular data to draw
inferences about selective forces that have acted on genomes. The field progresses in
large part through the development and benchmarking of advanced molecular-evolution
analysis methods. Here we evaluated the rigor and performance of a test of directional,
cis-regulatory evolution across genes in pathways, using stem cells from Mus musculus
subspecies as a model. We discovered a unique program of induction of translation
genes in stem cells of the Southeast Asian mouse M. m. castaneus relative to its sister
taxa, driven in part by cis-regulatory variation. As a complement, we used sequence
analyses to find population-genomic signatures of selection in M. m. castaneus, at the
upstream regions of the translation genes. Analysis of mouse species variants in
transcription factor binding sites identified Eed and Ehmt2 as candidate contributors to
divergence in translation gene expression in stem cells. We interpret our data under a
model of changes in lineage-specific pressures across Mus musculus in stem cells with
high translational capacity. Together, our findings underscore the rigor of integrating
expression and sequence-based methods to generate hypotheses about evolutionary
events from long ago.
Introduction
The central objective of molecular-evolution research is to draw inferences about
changing selective pressures between lineages, based on clues from omics data. Since

12
its inception, the field has relied in large part on model-fitting methods using DNA
sequence (Kreitman 2000) and gene expression (Price et al. 2022). As a complement,
empirical genome ranking/bootstrapping approaches have emerged in the more recent
literature (Ferguson and Chang 2020), with development and refinement remaining an
active area of research (Berg et al. 2019; Sohail et al. 2019; Johri et al. 2020; Price et
al. 2022).
One clear-cut empirical molecular-evolution strategy (Bullard et al. 2010; Fraser
et al. 2010; Fraser et al. 2011; Martin et al. 2012; York et al. 2018; Gokhman et al.
2021) takes as input measurements of cis-regulatory variation from expression profiles.
The test identifies cases in which, among the unlinked genes of a pathway subject to
cis-regulatory change, alleles in one taxon tend to drive expression mostly up, or mostly
down, relative to another taxon. This pattern of independent genetic variants with similar
effects at similar genes is unlikely under neutral expectations (Orr 1998). It thus serves
as a suggestive signature of changes in selective pressure on the pathway between
lineages. At this point, rigorous evolutionary inference requires additional follow-up,
including sequence-based tests to distinguish between positive and relaxed selection as
the driver of expression divergence. A number of studies in yeast have made this link
(Fraser et al. 2012; Martin et al. 2012; Roop et al. 2016); in metazoans, the rigor and
utility of expression-based cis-regulatory pathway analyses remain to be fully validated
(though see Mack et al. 2023).
In the current work, we set out to harness the diversity among mouse lineages in
pluripotent stem cell expression programs, to model the integration of expression- and
sequence-based tests for selection in multi-gene pathways. We focused on the mouse

13
Mus musculus castaneus (M. m. castaneus). This subspecies is endemic to southeast
Asia and diverged 0.5-1 MYA from other Mus musculus (Chapman and Ruddle 1972;
Sangster et al. 1993). Previous surveys have established divergence between M. m.
castaneus and laboratory strains in terms of gene expression (Fraser et al. 2011; Xiong
et al. 2014; Chappell et al. 2017; Tkatchenko et al. 2019; Chou et al. 2022) and
phenotype (Johnson et al. 1997; Beamer et al. 1999; Johnson et al. 2006; Yu et al.
2007; Koturbash et al. 2011; French et al. 2015; Omura et al. 2015; Chappell et al.
2017; Hsiao et al. 2020; Chou et al. 2022). Our goal was to use stem cell
transcriptomes to identify pathways subject to directional cis-regulatory change between
M. m. castaneus and laboratory mice. We earmarked one pathway hit, a set of
ribosomal genes at which M. m. castaneus alleles acted in cis to drive uniquely high
expression, for independent validation analyses with sequence data. At these loci,
phylogenetic and population-genomic tests revealed signals of unique evolution in M. m.
castaneus. Together, our results validate the utility of an expression-based molecularevolution test with sequence-based follow-up, in a model mouse tissue.
Results
A screen for directional cis-regulatory change in mouse stem cell pathways
As a first testbed for analyses of pathway cis-regulatory change, we used
transcriptional profiles (Marks et al. 2015) of embryonic stem cells of an F1 hybrid
background from a cross between two homozygous mouse strains: a male M. m.
castaneus (CAST/EiJ, hereafter CAST), and a female of the 129/SvImJ laboratory strain
genotype (hereafter 129), which is of admixed origin (Frazer et al. 2007; Yang et al.
2007; Yang et al. 2011). In the F1, because the two subspecies’ alleles of a given gene

14
are in the same nucleus, any difference in allele-specific expression between them can
be attributed to genetic variation acting in cis, i.e., not mediated through a soluble factor
(Wittkopp et al. 2004). We implemented a pipeline of allele-specific read-mapping taking
account of mapping artifacts (see Methods); and we earmarked all genes exhibiting
significant expression divergence between the species’ alleles (962 genes at a 0.05 pvalue threshold). We tabulated the directional effect for each gene—whether the CAST
allele was more highly expressed than the laboratory-strain allele, or vice versa—using
the log2-transformed fold-change between expression of the 129 allele and CAST allele.
Then, to formulate our test, we used as pathways groups of genes of common function,
each comprised of a Gene Ontology biological process term. For each such group, we
quantified the agreement in the direction of allelic expression differences between
species across the gene members. We evaluated significance based on resampling. Of
the complete survey results, one pathway showed significant signal: a cohort of genes
from the GO term for translation (Table 2.1). This represented a potential case in which
selective pressures on regulation of the respective loci had changed between the M. m.
castaneus and laboratory-strain lineages.
M. m. castaneus cis-regulatory alleles drive high expression of translation genes
in stem cells
Within the translation GO genes, the CAST allele was upregulated relative to that
of 129 in the hybrid at a ~2-fold excess of genes relative to those with higher expression
from the 129 allele (15.3% versus 7.6% respectively of translation genes with differential
allele-specific expression; Supplemental Figure 2.1A and Figure 2.1C). As it
encompassed a relatively small number of loci, we sought to verify this trend in

15
independent experimental data sources. For this purpose, we repeated the culture and
sequencing of CAST x 129 hybrid stem cells and detected an even more robust
directional imbalance in allele-specific expression among translation genes, with the
CAST allele again expressed more highly across the set 2.54-fold more often than the
129 allele (42.6% versus 16.8% of genes respectively; Figure 2.1A and 2.1C). Similarly,
we analyzed hybrid stem cells from a cross between CAST and the admixed C57BL/6J
laboratory strain (hereafter BL6; Werner et al. 2017), and observed a 2.48-fold
imbalance favoring high expression by the CAST allele among translation genes (Figure
2.1C and Supplemental Figure 2.1B). Together, these data suggest that M. m.
castaneus cis-regulatory alleles at translation genes encode a unique activating
program relative to those encoded by 129 and BL6 alleles, in stem cells.
GO term
Sign
statistic
Number
of genes
Resampling
p-value p-adj
translation -46.05 157 0.0000 0
cytoplasmic translation -19.01 19 0.0003 0.018
ribosomal large subunit assembly -13.45 17 0.0043 0.140
ion transport -34.75 188 0.0050 0.140
actin filament organization 21.31 65 0.0057 0.140
Table 2.1. A screen for directional allele-specific expression in CAST x 129
hybrid embryonic stem cells. Each row reports results of a statistical test for
directional allele-specific expression variation in CAST/EiJ male x 129/SvImJ female
F1 hybrid embryonic stem cells (Marks et al., 2015) in the indicated Gene Ontology
term. Top 5 results are shown here; full results table is reported in Supplemental
Table 2.1. GO term, Gene Ontology term analyzed. Sign statistic, sum of log2(129
allele expression / CAST allele expression) across genes of the term reporting which
parental allele was more highly expressed (<0 for higher expression of the CAST
allele, >0 for higher expression of the 129 allele, 0 for no significant expression
difference between alleles). Number of genes, number of analyzable genes in the
GO term. Resampling p-value, resampling-based significance of the enrichment for
high absolute value of the sign statistic in the term. p-adj, p-value after BenjaminiHochberg correction for multiple testing.

16
Figure 2.1: Directional cis-regulatory variation in translation gene expression in
stem cells between mouse subspecies. (A) Each point reports allele-specific
expression of genes from the translation GO term in CASTx129 hybrid stem cells
cultured in-house: the x-axis reports the log2 ratio of expression of the respective
strain alleles, and the y-axis reports the log10 of the significance of the difference.
Point colors report significance of differential allele-specific expression (p-adj,
adjusted p-value). Red and black text inlays report the percentage of translation
genes where expression of the respective parental allele is higher with and without
filtering for significance, respectively. (B) Data are as in (A), but in a comparison
between homozygous CAST and 129 stem cells cultured in-house. (C) Each column
reports results from analyses of allele-specific expression as in (A) in stem cells of an
F1 hybrid between CAST and either C57BL/6 (from Werner et al. 2017) or 129
((Marks et al., 2015) or this study) as indicated. The first and second rows report the
percentage of translation genes in which the CAST allele is expressed higher or
lower, respectively, than the allele of the other mated parent, and the third row
reports the ratio of these quantities. (D) Data are as in (B) except that homozygous
mouse stem cell transcriptomes were used, in comparisons between CAST and
either 129 (this study or (Skelly et al., 2020)), C57BL/6J, A/J, NZO/HILtJ,
NOD/ShiLtJ, WSB/EiJ, CAST/EiJ, or PWD/PhJ (Skelly et al., 2020) as indicated.

17
Homozygous M. m. castaneus stem cells are distinguished by high expression of
translation genes
We expected that, if regulatory divergence between CAST and other lineages at
translation genes had been important for fitness in the organismal and ecological
context, it would be apparent in the context of homozygous strains, which integrate the
effects of genetic factors acting both in cis and in trans (Signor and Nuzhdin 2018).
Consistent with this prediction, translation genes were more highly expressed in CAST
homozygous stem cells relative to those of 129 and BL6, in a published data resource
(Skelly et al. 2020) and in our own culture and sequencing (Figure 2.1B and
Supplemental Figure 2.1C-D). Interestingly, this imbalance was of larger magnitude
(~3.4-10-fold more genes expressed highly in CAST than in the other lineages; Figure
2.1D, Supplemental Figure 2.1C-F) than we had noted in our analyses of cis-regulatory
variation, indicating that the latter was reinforced by a more pervasive effect of
divergence in trans-acting regulators. Given these signatures of directional cis- and
trans-acting variation between Mus subspecies, we considered the translation gene
cohort to be a compelling candidate for molecular-evolution follow-up, to investigate the
selective pressures underlying these changes.
M. m. castaneus harbors unique and conserved alleles in spatial windows of
translation gene upstream regions
We next turned to sequence-based tests of selection to explore the evolutionary
history of translation gene regulation between mouse lineages. We reasoned that
analysis of multiple M. musculus subspecies could help polarize any sequence variation
we would detect at the loci. Toward this end, we developed an analysis pipeline for

18
single-nucleotide variants from wild-caught mice of the Mus genus (Harr et al. 2016), in
which we tabulated divergence between subspecies, normalized by intra-subspecies
polymorphism, in spatial windows of upstream regions for every gene in the genome.
We then used these values as input into resampling tests for enrichment of high values
of this normalized quantity in translation genes relative to a genomic null, representing a
hallmark of positive selection (Kreitman 2000). In a first such approach, we classified
variant sites based on their spatial positions in upstream loci (Figure 2.2A). The results
revealed elevated sequence divergence between M. m. castaneus on the one hand and
Figure 2.2: Elevated nucleotide divergence between M. m. castaneus and other
subspecies in regions upstream of translation genes. (A) Each bar reports a
metagene analysis of nucleotide divergence between a M. m. castaneus population
from India and a M. m. musculus population from Afghanistan in a 5000-bp window
starting at the indicated distance upstream of transcription start sites. The y-axis
reports the ratio of the average number of sites in the respective window exhibiting
divergence between the subspecies (Du) to the average number of polymorphic sites
(Pu) across genes from the Gene Ontology term GO:0006412, translation,
normalized against the analogous quantity for all other genes in the genome. *: p >
0.05 from a one-sided resampling test for elevated Du/Pu in the translation gene
cohort against a genomic null. (B) Summary of the analysis of Du/Pu between the M.
m. castaneus population and 6 M. m. musuclus and M. m. domesticus populations.
Normalized Du/Pu and resampling p-values are reported for the 5kb window from
25kb to 20kb upstream for each comparison.

19
populations of M. m. musculus and M. m. domesticus on the other (Figure 2.2A-B). The
signal was most marked in a window 25kb upstream of translation start among genes of
the translation GO term and fell off gradually on either side of this peak (Figure 2.2A),
as expected if M. m. castaneus had acquired unique alleles in cis-regulatory elements
with positional preferences across this gene cohort. Such a trend was consistent with
the unique expression program in M. m. castaneus that we had noted in stem cells as a
product in part of cis-regulatory change (Figure 2.1A and 2.1C).
A role for transcription factors in divergently regulated translation genes
As a complement to our focus on the spatial positioning of cis-regulatory
sequence changes in translation genes, we developed a separate molecular-evolution
approach classifying variants on the basis of binding sites by transcriptional regulators
(Kolmykov et al. 2021). For the sites bound by a given regulator, we again tested for
elevated divergence between M. m. castaneus and its relatives, relative to intrasubspecies polymorphism, across the genes of our translation cohort (Figure 2.3A-B).
Three regulators (Ehmt2, Eed, and Msx1) exhibited robust signal in this scheme, with
binding sites at translation genes enriched for divergence between M. m. castaneus and
its sister lineages (Table 2.2, Supplemental Table 2.2, and Figure 2.3A-B). This trend
for unique alleles in M. m. castaneus in regulator binding sites mirrored the results of
our spatially resolved analysis of upstream regions (Figure 2.2). None of the divergent
loci in the binding sites for Eed or Ehmt2 were in the 25kb to 20kb upstream window
where we detected the most signal in our spatial analysis, and thus provided an
independent line of evidence for positive selection driving such changes at translation
genes. In Ehmt2 and Eed knockout stem cell transcriptomes (Auclair et al. 2016;

20

21
Weigert et al. 2023), we detected robust evidence for a directional impact of these
factors on translation gene expression (Figure 2.3C and Supplemental Figure 2.2), as
expected if their binding had direct regulatory function at translation genes. Together,
our analyses reveal changes in these sites as a particularly compelling candidate
molecular mechanism for non-neutral expression variation across mouse lineages, and
its manifestation in stem cells.
Discussion
Empirical molecular evolution tests, once developed, must be validated to
establish their rigor and utility for the field. A compelling means toward this end is to
integrate results from an emergent test strategy with those of more classic tools when
they complement each other in support of an evolutionary inference. In the current work,
we have detected sequence signatures of selection between mouse subspecies in a
cohort of translation genes that also exhibits directional, polygenic cis-regulatory
variation in stem cells. Against a backdrop of other case studies of cis-acting expression
change (Bullard et al. 2010; Fraser et al. 2010; Fraser et al. 2011; Martin et al. 2012;
Figure 2.3: Signatures of species divergence and regulatory function of Eed at
translation genes. (A) In a given row, cells report the ratio of inter-subspecies
divergence (Du) to within-subspecies polymorphism (Pu) at Eed binding sites
upstream of translation genes or all other genes of the genome, in a comparison
between M. m. castaneus and the indicated population of a sister taxon. Divergence
is significantly enriched in Eed binding sites upstream of translation genes (see Table
2.2). (B) An Eed binding footprint upstream of the translation gene Eif4e2 as an
example of mouse subspecies variation. Each row reports the sequence of one M.
m. castaneus, domesticus, or musculus individual (M. m. c., M. m. d., or M. m. m.
respectively). (C) Data and symbols are as in Figure 2.1A except that each point
reports a comparison of expression of one gene from the translation GO term
between Eed knockout and wild-type mouse stem cells (Weigert et al 2023), and red
and black text inlays report the percentage of translation genes where expression is
higher in the indicated genotype with and without filtering for significance,
respectively.
Figure 2.3: Signatures of species divergence and regulatory function of Eed at
translation genes. (A) In a given row, cells report the ratio of inter-subspecies
divergence (Du) to within-subspecies polymorphism (Pu) at Eed binding sites
upstream of translation genes or all other genes of the genome, in a comparison
between M. m. castaneus and the indicated population of a sister taxon. Divergence
is significantly enriched in Eed binding sites upstream of translation genes (see Table
2.2). (B) An Eed binding footprint upstream of the translation gene Eif4e2 as an
example of mouse subspecies variation. Each row reports the sequence of one M.
m. castaneus, domesticus, or musculus individual (M. m. c., M. m. d., or M. m. m.
respectively). (C) Data and symbols are as in Figure 2.1A except that each point
reports a comparison of expression of one gene from the translation GO term
between Eed knockout and wild-type mouse stem cells (Weigert et al 2023), and red
and black text inlays report the percentage of translation genes where expression is
higher in the indicated genotype with and without filtering for significance,
respectively.

22
York et al. 2018; Agoglia et al. 2021; Gokhman et al. 2021; Mack et al. 2023; Wang et
al. 2024), our results represent a proof of concept for sequence-based validation of this
approach in mammals.
Our analyses implicate several transcriptional regulators in the divergence
between mouse subspecies in translation genes. The variation we have detected in
translation gene binding sites for the polycomb repressive complex 2 component Eed
(Mozzetta et al. 2014; Obier et al. 2015; Li et al. 2018), as a candidate determinant of
stem cell expression change between mice, dovetails with the known role for Eed in
development and stem cell differentiation (Young 2011; Obier et al. 2015). The
polycomb repressive complex 2 more broadly has been shown to regulate pluripotency
of stem cells through methylation of the DNA encoding ribosomal RNAs (Zhang et al.
2020); our focused study of Eed
suggests that its effects in stem
cells are also mediated by other
translation components, as
reflected in binding at these loci in
stem cells (Li et al. 2018; Oksuz et
al. 2018; Kaaij et al. 2019; Kriz et
al. 2021). Likewise, the evolutionary
and genomic profiles of binding
sites at translation genes for the
methyltransferase Ehmt2, including
their inferred impact on expression
TF anova_F anova_pval BH
Ehmt2 237.78 1.44E-12 8.55E-10
Eed 29.74 2.45E-05 0.007
Msx1 22.19 0.0001 0.026
Table 2.2. Signatures of selection between
mouse subspecies in transcription factor
binding sites upstream of translation genes.
Each row reports results of a two-factor ANOVA
testing for an excess of divergent variants
between M. m. castaneus on the one hand and
M. m. musculus or M. m. domesticus on the
other, in binding sites of the indicated
transcription factor upstream of genes of the
translation GO term (see Figure 2.3A). The
second and third columns report ANOVA F
statistic and nominal p-value, respectively, and
the third reports the p-value after BenjaminiHochberg correction for multiple testing. Only
the TFs with a p-adj < 0.05 are shown here.
The results for all 592 TFs containing data is
reported in Supplemental Table 2.2.
TF anova_F anova_pval BH
Ehmt2 237.78 1.44E-12 8.55E-10
Eed 29.74 2.45E-05 0.007

23
in stem cells, dovetail with the known function of this factor in stem cell maintenance
and differentiation (Ikegami et al. 2007; Leitch et al. 2013; Boroviak et al. 2014; Auclair
et al. 2016; Kim et al. 2020) and direct binding by Ehmt2 to translation genes in stem
cells (Mozzetta et al. 2014). As such, our results suggest that the known roles of Eed
and Ehmt2 in stem cell identity and differentiation may be mediated by regulation of
protein synthesis, and that evolution used these binding sites in part to tune expression
of translation factors between mouse lineages. A similar function may ultimately also
emerge for the development regulator Msx1, whose binding sites were also top-scoring
in our molecular-evolution analyses across Mus. That said, we expect that our genomic
approach affords only partial coverage and power in discovering elements of the
mechanism of mouse subspecies divergence at translation genes, and thus that many
other contributing transcription factors likely remain to be identified.
Our finding of divergence at translation genes between mouse lineages echoes
reports of cis-regulatory change in the translation machinery of other organismal
systems, from yeasts (Tanay et al. 2005; Hogues et al. 2008; Li and Fay 2017; Sorrells
et al. 2018) to animal ancestors distributed over deep time (Brown et al. 2008). This
literature leaves open the question of what ecological forces might drive cis-regulatory
change at translation loci, and the phenotypes that would mediate such effects. Under a
model in which ribosomal protein dosage governs the readiness of a cell to divide
(Polymenis and Aramayo 2015), adaptive cis-regulatory variation at translation genes
may often reflect species-unique logic of cell growth decisions.
In metazoan stem cells in particular, translation plays a critical role in
differentiation. Inducing (Easley et al. 2010) or compromising (Khajuria et al. 2018) stem

24
cell translation can drive qualitative differences in differentiation behavior. According to
current models, high expression of translation genes in stem cells (Sampath et al. 2008)
sets up a poised state to enable rapid protein production in their differentiated progeny
(Gabut et al. 2020). On the basis of this tight link between translation and differentiation,
it is tempting to speculate that the variation we have seen among Mus musculus
lineages in stem-cell translation gene expression was driven by selection in distinct
niches for unique differentiation properties. If so, the loci we study here would act in a
complex genetic architecture of stem cell differentiation, alongside other variants
characterized across mouse lineages (Ortmann et al. 2020; Skelly et al. 2020). Future
work will explore the phenotypic relevance of translation gene divergence across Mus,
and its role in the slew of traits (Johnson et al. 1997; Beamer et al. 1999; Johnson et al.
2006; Yu et al. 2007; Koturbash et al. 2011; French et al. 2015; Chappell et al. 2017;
Hsiao et al. 2020; Chou et al. 2022) that distinguish M. m. castaneus from the rest of its
genus.
Materials and Methods
RNA-seq data sources
For our initial screen for directional cis-regulatory variation in pathways, we used
transcriptional profiles of CAST/EiJ male x 129/SvImJ female F1 hybrid embryonic stem
cells (NCBI [National Center for Biotechnology Information] accession GSE60738,
samples SRR1557132, SRR1557133, SRR1557112, and SRR1557123; Marks et al.
2015). For validation and follow-up we used additional transcriptional profiles from
reciprocal crosses of CAST/EiJ x C57BL/6J hybrid pluripotent stem cells (NCBI
accession GSE90516, samples SRR5054337-5054348 and SRR5054353-5054364;

25
Werner et al. 2017); homozygous embryonic stem cells from a panel of M. musculus
subspecies (EBI [European Bioinformatics Institute; Sarkans et al. 2018] accession EMTAB-7730 (Skelly et al. 2020)); and in-house CAST/EiJ male x 129/SvImJ female F1
hybrid embryonic stem cells, and homozygous induced pluripotent stem cells from inhouse cultures (see below).
For validation of the role of candidate transcription factors in translation
regulation in stem cells, we analyzed published transcriptional data from stem cell
knockouts of Eed (Weigert et al., 2023; NCBI accession GSE198076, samples
SRR18260888, SRR18260892, SRR18260902, and SRR18260903) and Ehmt2 (Auclair
et al., 2016; NCBI accession GSE71500, samples SRR2133432, SRR2133433,
SRR2133434, and SRR2133435), respectively.
RNA-seq read mapping
For analysis of CAST/EiJ x 129/SvImJ and CAST/EiJ x C57BL/6J hybrid stem
cell transcriptomes from (Marks et al. 2015) we downloaded raw reads and mapped
with the STAR aligner (Dobin et al 2013) to a concatenated genome file containing
chromosomal sequences from both parent strains (CAST/EiJ and either 129/SvImJ or
C57BL/6J).
For analysis of transcriptomes of homozygous CAST/EiJ and C57BL/6J stem
cells cultured in-house (see below), we downloaded raw reads and mapped them to the
corresponding reference genome for the respective strain (CAST/EiJ or GRCm38). For
validation of the role of candidate transcription factors in translation regulation in stem
cells, the Ehmt2 knockout dataset was mapped to the GRCm38 reference; the Eed

26
knockout dataset was in cells from the CD-1 strain, and these RNA-seq reads were
mapped to the CD-1 reference genome (NCBI accession GSE209815; Jung et al.,
2023).
RNA-seq normalization and quantification
Reads that mapped ambiguously to more than one locus were discarded. Read
counts were generated during the STAR alignment step using the ‘--quantMode
GeneCounts’ option. For analysis of transcriptomes of homozygous stem cells from
(Skelly et al. 2020) (genotypes C57BL/6J, A/J, 129S1/SvImJ, NZO/HILtJ, NOD/ShiLtJ,
WSB/EiJ, CAST/EiJ, and PWD/PhJ), we downloaded mapped read counts from EBI’s
ArrayExpress database (Sarkans et al. 2018). For each data set in turn, normalized
(TPM, transcripts per million) counts were generated by dividing read counts per gene
by transcript length (using annotations from the Ensembl database, build 102) and then
dividing by library size.
Hybrid RNA-seq mapping quality control
To eliminate potential artifacts from allele-specific mapping errors in hybrid RNAseq analyses, we performed a simulated RNA-seq experiment using the Polyester
package in R (Frazee et al. 2015) as follows. For the CAST x 129 hybrid stem cell
transcriptome from (Marks et al. 2015), we generated two replicates of simulated reads
from the hybrid genome with ~200 reads for each annotated transcript. These simulated
reads were then mapped back to the genome with STAR as above. For a given gene
with called orthologs in the CAST and 129 genomes, for each allele in turn we tabulated
the ratio between the number of successfully mapped simulated reads and the number

27
of simulated reads that went into the mapping, as a report of the extent of artifact-free
mapping. We converted each such ratio to a percentage; we then took the absolute
value of the difference between the ratio for the 129 and CAST allele as a report of the
difference in mapping fidelity between them. We filtered out genes for which the latter
parameter exceeded 5%, a total of 1,852 genes of 10,380 initial homologs in the data
from (Marks et al. 2015). This protocol was repeated separately for analyses of CAST x
BL6 hybrid stem cell transcriptomes from (Werner et al. 2017), where we filtered out we
excluded 2,886 genes out of an initial 8,887 homologs expressed in the RNA-seq data.
In silico screen for directional cis-regulatory variation
For our initial screen of polygenic, directional cis-regulatory variation in pathways,
we harnessed profiles from CAST/EiJ male x 129/SvImJ female F1 hybrid pluripotent
stem cells (Marks et al. 2015). We generated a list of one-to-one orthologous genes
between CAST/EiJ and 129/SvImJ from the Ensembl database (build 102) using
biomaRt (Durinck et al. 2009). At a given gene, we tested for differential expression
between the CAST/EiJ and 129/SvImJ alleles using the reads mapping to each as input
into DESeq2 (Love et al. 2014). We eliminated from further analysis genes with fewer
than 10 total reads across all samples.
For each gene with differential allele-specific expression at adjusted P < 0.05, we
assigned a quantitative sign statistic s, equal to the log2(129 allele expression / CAST
allele expression). We assigned all genes without differential allele-specific expression
to have s = 0. We downloaded gene annotations in Gene Ontology ‘biological process’
terms from the AmiGO database (Carbon et al. 2009). We eliminated from further
testing all terms containing fewer than 10 genes with significant differential allele-

28
specific expression. For a given remaining term containing n genes, we summed the s
values across the component genes to yield a summary statistic Strue. To evaluate
significance by resampling, we randomly sampled n genes from the total set with
expression data and summed their s values, generating a resampled summary statistic
Sresample. We carried out this calculation 10,000 times and used as a two-sided P value
the proportion of resamples in which |Sresample| ≥ |Strue|. We corrected for multiple testing
with the Benjamini-Hochberg method. P values are reported in Supplemental Table 2.1.
All further expression and molecular-evolution analyses focused on genes in the topscoring term, GO:0006412, translation (Figure 2.1).
Induced pluripotent stem cell derivation and culture
To establish C57BL/6J and CAST/EiJ induced pluripotent stem cell lines, we
used mouse embryonic fibroblasts E13.5 obtained from Jackson Laboratories (Bar
Harbor, ME, USA) as input into a stem cell derivation protocol as previously described
(Terzic et al. 2016). Briefly, we used octamer-binding transcription factor-4 (Oct4), sexdetermining region y-box 2 (Sox2), and Kruppel-like factor-4 (Klf4) as reprogramming
factors, introduced using pMXs retroviral vectors. A 129S6/SvEv embryonic stem cell
line was obtained from Millipore Sigma, (Catalog no. SCR012), Burlington, MA, USA). A
CAST/EiJ x 129/SvImJ mouse embryonic stem cell line was generously provided by
Joost Gribnau (Marks et al. 2015).
Stem cells were cultured on irradiated mouse embryonic fibroblasts (R and D
Systems, Minneapolis, MN, USA) in miPSC medium: knockout DMEM with 4.5 g/ L dglucose (Gibco, Grand Island, NY, USA), 10% knockout serum replacement (KSR)
(Gibco), 10% fetal bovine serum (FBS) (HyClone, Logan, UT, USA), 1× MEM

29
nonessential amino acids (MEM NEAA) (Gibco), 1× GlutaMAX (Gibco), 0.1 mM 2-
mercaptoethanol (BME) (Life Technologies, Grand Island, NY, USA), and 0.02%
ESGRO-LIF (Millipore, Billerica, MA, USA). Cells were incubated at 37°C in 5% CO2.
RNA isolation and sequencing
RNA was extracted from undifferentiated pluripotent stem cell cultures (2 BL6
replicates, 2 CAST replicates) following feeder depletion using the RNAqueous™-Micro
Total RNA Isolation Kit (Thermo Fisher Scientific) and on-column DNase treatment
(QIAGEN, Hilden, Germany). RNA samples were processed into mRNA libraries and
sequenced on an Illumina NovaSeq 6000 Sequencing System, yielding ~20M pairedend 150 bp reads per sample. Read-mapping was as above.
Population genomic analysis
We downloaded resequencing data from wild populations of M. m. domesticus
(from France, Germany and Iran), M. m. musculus (from Afghanistan, Czech Republic
and Kazakhstan), and M. m. castaneus (from northwest India) (Harr et al. 2016). VCF
files were used to map SNPs from each individual mouse onto the C57BL/6J reference
genome (GRCm38) using the ‘consensus’ command from bcftools software (Danecek et
al. 2021). Default options for ‘bcftools consensus’ algorithm were used, meaning that in
the case of the heterozygous loci, the reference GRCm38 allele would be used.
Sequences upstream of the transcription start site for each gene were extracted utilizing
the pybedtools Python package (Quinlan and Hall 2010; Dale et al. 2011). In order to
assess only a single transcript per gene, gffread (Pertea and Pertea 2020) was used to
extract coding sequences (CDS) from each annotated transcript. A custom Python

30
script was used to check whether each CDS contained an in-frame start and stop
codon, signifying a valid open reading frame (ORF). For genes with multiple transcripts
containing a valid ORF, the longest transcript was selected. Genes without an in-frame
ORF were eliminated from further analysis.
We carried out selection tests separately in a comparison of each M. m.
domesticus or M. m. musculus population in turn against the M. m. castaneus
population. In each case, for a given gene, we generated an alignment of the 50kb
region upstream of the gene start site and split it into ten 5kb windows. In each window
of each gene, we tabulated the number of sites harboring polymorphisms across one or
both populations, and the number of sites with alleles fixed in each population and
divergent between them (Pu and Du respectively). For a given window, we took the sum
of Du and, separately, the sum of Pu across the translation GO term genes, and we
calculated the ratio Rtranslation = Du/Pu. To evaluate significance by resampling, for a
given window we randomly sampled 10 genes from the total set with sequence data;
calculated Pu and Du, their sums across genes, and their ratio Rresample as above; and
over 10,000 resamples, we used as a one-sided P value the proportion of resamples in
which Rresample ≥ Rtranslation. Resampling statistics are reported in Figure 2.2B.
Analysis of transcription factor binding sites
In a public resource of chromatin immunoprecipitation sequencing data sets
(Kolmykov et al. 2021), we tabulated binding sites for 680 transcription factors across
the mouse genome. For each factor, Du/Pu was calculated as above for binding sites
within 50kb upstream of genes, and then separately in the remaining non-binding site
regions. These values were then used as input for a two-factor ANOVA, testing for an

31
interaction effect between binding site identity and membership in the translation GO
term for each factor. Du/Pu between the M. m. castaneus population and each M. m.
domesticus or M. m. musculus population were used as replicates in the ANOVA
analysis. ANOVA results are reported in Table 2.2. Multiple sequence alignment was
edited for display using Jalview software (Waterhouse et al. 2009).

32
Supplemental Figures

33
Supplemental Figure 2.1: Directional cis-regulatory variation in translation
gene expression in stem cells between mouse subspecies. For individual plots:
Each point reports allele-specific expression of genes from the translation GO term in
a single RNA-seq dataset. The x-axis reports the log2 ratio of expression of the
respective strain alleles, and the y-axis reports the log10 of the significance of the
difference. Point colors report significance of differential allele-specific expression (padj, adjusted p-value). Red and black text inlays report the percentage of translation
genes where expression of the respective parental allele is higher with and without
filtering for significance, respectively. (A) CAST x 129 hybrid stem cells (Marks et al.,
2015). (B) CAST x BL6 hybrid stem cells (Werner et al., 2017). (C-I) Comparisons
between homozygous CAST stem cells and homozygous stem cells from other
strains (Skelly et al., 2020): (C) CAST and 129; (D) CAST and BL6; (E) CAST and
WSB/EiJ; (F) CAST and PWD/PhJ.

34
Supplemental Figure 2.2: Ehmt2 knockout weakly induces translation genes.
Volcano plot reporting relative expression of genes from the translation GO term after
Ehmt2 knockout in stem cells (Auclair et al. 2016). The dotted line on the x-axis
shows the p-adj cutoff of 0.05 for significance. Red text inlays indicate the
percentage of translation genes where expression of either parental allele is
significantly higher based on differential expression analysis. Black text inlays
indicate the percentage of genes where expression of either parent allele is higher,
including both significantly and non-significantly induced genes.

35
Supplemental Tables
GO term Sign
statistic
Number of
genes in RNAseq data
Resampling
p-value
p-adj
translation -46.05 157 0 0.000
cytoplasmic translation -19.01 19 0.0003 0.018
ribosomal large subunit assembly -13.45 17 0.0043 0.140
ion transport -34.75 188 0.005 0.140
actin filament organization 21.31 65 0.0057 0.140
proteasome-mediated ubiquitindependent protein catabolic process
21.12 97 0.0184 0.377
regulation of cell population
proliferation
20.25 102 0.0287 0.504
meiotic cell cycle 13.90 57 0.0429 0.660
neuron projection development -13.75 76 0.0745 0.809
protein stabilization -14.87 92 0.08 0.809
negative regulation of protein binding -9.33 39 0.0909 0.809
mRNA transport 11.37 61 0.0931 0.809
positive regulation of GTPase activity -11.74 70 0.1042 0.809
rhythmic process -13.70 90 0.1046 0.809
intracellular signal transduction 20.30 205 0.1063 0.809
defense response to virus -11.38 65 0.1074 0.809
ribosomal small subunit biogenesis -4.31 10 0.1193 0.809
cytoskeleton organization 10.24 56 0.1209 0.809
cell migration 15.00 124 0.125 0.809
angiogenesis -14.04 123 0.1439 0.885
transmembrane transport -14.44 139 0.1631 0.909
regulation of apoptotic process 12.12 99 0.1661 0.909
lipid catabolic process -7.65 42 0.1699 0.909
positive regulation of translation -7.32 44 0.1901 0.922
protein dephosphorylation -9.74 75 0.1932 0.922
Supplemental Table 2.1: A screen for directional allele-specific expression in
CAST x 129 hybrid embryonic stem cells. Each row reports results of a statistical
test for directional allele-specific expression variation in CAST/EiJ male x 129/SvImJ
female F1 hybrid embryonic stem cells (Marks et al., 2015) in the indicated Gene
Ontology term. A total of 123 TFs were tested. GO term, Gene Ontology term
analyzed. Sign statistic, sum of log2(129 allele expression / CAST allele expression)
across genes of the term reporting which parental allele was more highly expressed
(<0 for higher expression of the CAST allele, >0 for higher expression of the 129
allele, 0 for no significant expression difference between alleles). Number of genes,
number of analyzable genes in the GO term. Resampling p-value, resampling-based
significance of the enrichment for high absolute value of the sign statistic in the term.
p-adj, p-value after Benjamini-Hochberg correction for multiple testing.

36
cell cycle
-22.40 391 0.1949 0.922
cellular response to hypoxia
-6.27 35 0.2051 0.927
Wnt signaling pathway
-10.98 105 0.2111 0.927
positive regulation of transcription,
DNA
-templated
-19.10 315 0.223 0.935
stem cell population maintenance
-6.80 46 0.2399 0.935
positive regulation of apoptotic process
-13.61 193 0.2711 0.935
positive regulation of NF
-
kappaB transcription factor activity
7.29 64 0.2857 0.935
spermatogenesis 11.89 166 0.2875 0.935
cellular response to DNA damage
stimulus
-16.24 315 0.2978 0.935
protein polyubiquitination 8.01 84 0.3059 0.935
chromosome segregation 6.78 61 0.3108 0.935 G proteincoupled receptor signaling pathway -7.96 88 0.3158 0.935
protein localization to plasma
membrane
-7.30 82 0.3458 0.935
negative regulation of transcription,
DNA
-templated
-13.77 278 0.3467 0.935
positive regulation of cell population
proliferation
-11.85 213 0.3522 0.935
response to drug 6.35 65 0.353 0.935
regulation of translation 7.17 87 0.3595 0.935
rRNA processing
-7.29 88 0.3639 0.935
autophagy
-7.63 99 0.3703 0.935
ubiquitin
-
dependent protein catabolic process
9.68 155 0.3765 0.935
positive regulation of ERK1 and ERK2
cascade
-6.65 79 0.3767 0.935
cell
-cell adhesion 5.81 62 0.3844 0.935
heart development
-8.10 120 0.3851 0.935
DNA repair
-11.27 244 0.4086 0.935
post
-embryonic development
-4.59 46 0.4098 0.935
regulation of transcription by RNA poly
merase II
-11.79 277 0.426 0.935
oxidation
-reduction process
-11.11 250 0.4278 0.935
nervous system development 9.32 186 0.4304 0.935
negative regulation of cell growth 5.51 72 0.4424 0.935
chromatin remodeling
-4.74 58 0.4432 0.935
positive regulation of transcription by
RNA polymerase II
-15.42 556 0.4727 0.935
positive regulation of cell growth
-3.86 46 0.4821 0.935
positive regulation of neuron projection
development
5.06 80 0.492 0.935
response to oxidative stress 4.35 57 0.4921 0.935
RNA splicing 7.40 156 0.4936 0.935
mitotic cell cycle 4.60 72 0.5065 0.935

37
apoptotic process
-10.11 308 0.5149 0.935
in utero embryonic development 7.50 183 0.5264 0.935
carbohydrate metabolic process
-5.16 94 0.5282 0.935
Golgi organization
-4.06 64 0.5342 0.935
regulation of transcription, DNA
-
templated
-11.58 428 0.5353 0.935
negative regulation of transcription by
RNA polymerase II
-10.96 400 0.5439 0.935
intracellular protein transport
-7.05 185 0.5536 0.935
protein transport
-10.29 375 0.556 0.935
vesicle
-mediated transport
-6.28 154 0.5695 0.935
negative regulation of apoptotic
process
8.11 255 0.5752 0.935
axon guidance
-4.04 79 0.5759 0.935
positive regulation of MAPK cascade 3.25 54 0.5865 0.935
endocytosis
-4.72 112 0.6012 0.935
fatty acid metabolic process 3.94 82 0.6029 0.935
transcription, DNA
-templated
-3.36 62 0.6049 0.935
chromatin organization
-6.03 180 0.6083 0.935
response to hypoxia 3.01 53 0.6116 0.935
mRNA processing 6.10 201 0.6221 0.935
actin cytoskeleton organization
-3.97 97 0.6327 0.935
neuron migration 2.85 58 0.6389 0.935
negative regulation of cell population
p
roliferation
-5.50 186 0.6462 0.935
protein autophosphorylation
-3.72 96 0.6472 0.935
transcription by RNA polymerase II
-2.95 68 0.6568 0.935
protein ubiquitination
-4.87 162 0.6591 0.935
negative regulation of neuron apoptotic
process
3.20 77 0.6635 0.935
mRNA splicing, via spliceosome 3.48 89 0.6651 0.935
inflammatory response
-3.30 88 0.6774 0.935
positive regulation of gene expression 5.13 207 0.681 0.935
cellular response to leukemia inhibitory
factor
-3.09 82 0.6845 0.935
positive regulation of IkappaB kinase/NF-kappaB signaling
2.11 47 0.6968 0.942
multicellular organism development 6.50 416 0.7189 0.958
phosphorylation
-5.20 320 0.7405 0.958
positive regulation of protein
phosphorylation
-2.82 106 0.7485 0.958
regulation of cell cycle
-2.24 82 0.771 0.958
innate immune response
-2.69 134 0.7842 0.958
signal transduction 4.27 317 0.7855 0.958
protein phosphorylation
-4.04 291 0.7908 0.958
positive regulation of cell migration 2.40 125 0.7984 0.958

38
immune system process -2.59 157 0.8066 0.958
cell adhesion -2.69 175 0.8138 0.958
regulation of gene expression 2.64 172 0.8162 0.958
cell differentiation 4.06 398 0.8247 0.958
cell division 2.78 239 0.831 0.958
negative regulation of gene expression 2.02 150 0.8511 0.958
lipid metabolic process 2.52 254 0.8525 0.958
proteolysis 2.21 186 0.8551 0.958
negative regulation of neuron projectio
n development
0.84 43 0.8674 0.958
ribosome biogenesis -0.94 61 0.8848 0.958
regulation of cell shape -1.06 75 0.8848 0.958
regulation of alternative mRNA splicing
, via spliceosome
-0.68 37 0.8876 0.958
biological_process -3.52 820 0.8882 0.958
metabolic process 1.10 96 0.8888 0.958
microtubule cytoskeleton organization 0.91 75 0.8925 0.958
lung development 0.68 49 0.8975 0.958
cell population proliferation -0.90 87 0.9033 0.958
methylation -0.65 94 0.9342 0.982
dendrite morphogenesis 0.14 32 0.9548 0.994
protein folding -0.26 56 0.9614 0.994
sensory perception of sound -0.22 60 0.9697 0.994
dephosphorylation 0.10 56 0.9824 0.994
cilium assembly -0.16 115 0.9856 0.994
cell projection organization 0.06 112 0.9945 0.995

39
TF ANOVA F ANOVA pval p-adj
Ehmt2 237.78 1.44E-12 8.55E-10
Eed 29.74 2.45E-05 0.007
Msx1 22.19 0.0001 0.026
Fbxl19 14.36 0.001 0.156
Mllt3 13.91 0.001 0.156
Rfx6 13.06 0.002 0.171
Ell 12.11 0.002 0.200
Nfatc1 10.98 0.003 0.256
Ubn1 10.47 0.004 0.273
Lhx6 9.29 0.006 0.376
Nr5a1 8.16 0.010 0.494
Bhlhe41 8.09 0.010 0.494
Vsx2 7.28 0.014 0.588
Nelfb 7.27 0.014 0.588
ND5 7.08 0.015 0.592
Pou2f3 6.36 0.020 0.708
Nr1h4 6.35 0.020 0.708
Aire 6.20 0.022 0.713
Chaf1b 5.97 0.024 0.747
Bcor 5.38 0.031 0.887
Ets2 5.28 0.032 0.887
Myt1l 5.25 0.033 0.887
Ctr9 5.07 0.036 0.919
Foxo1 4.95 0.038 0.925
Nr1i2 4.77 0.041 0.925
Spin1 4.68 0.043 0.925
Crtc2 4.50 0.047 0.925
Kdm6b 4.49 0.047 0.925
Brd3 4.43 0.048 0.925
Ell2 4.43 0.048 0.925
Phf19 4.35 0.050 0.925
Ruvbl2 4.35 0.050 0.925
Pou3f1 4.28 0.052 0.927
Supplemental Table 2.2. Signatures of selection between mouse subspecies in
transcription factor binding sites upstream of translation genes. Each row
reports results of a two-factor ANOVA testing for an excess of divergent variants
between M. m. castaneus on the one hand and M. m. musculus or M. m. domesticus
on the other, in binding sites of the indicated transcription factor upstream of genes
of the translation GO term (see Figure 2.3A). The second and third columns report
ANOVA F statistic and nominal p-value, respectively, and the third reports the pvalue after Benjamini-Hochberg correction for multiple testing. 592 TFs containing
data were analyzed.
Supplemental Table 2.2. Signatures of selection between mouse subspecies in
transcription factor binding sites upstream of translation genes. Each row
reports results of a two-factor ANOVA testing for an excess of divergent variants
between M. m. castaneus on the one hand and M. m. musculus or M. m. domesticus
on the other, in binding sites of the indicated transcription factor upstream of genes
of the translation GO term (see Figure 2.3A). The second and third columns report
ANOVA F statistic and nominal p-value, respectively, and the third reports the pvalue after Benjamini-Hochberg correction for multiple testing. 592 TFs containing
data were analyzed.

40
Mafa 4.05 0.058 0.981
Npas3 3.93 0.061 0.981
Nfe2 3.91 0.062 0.981
E2f4 3.88 0.063 0.981
Ing1 3.88 0.063 0.981
Maf 3.73 0.068 0.999
Znf296 3.65 0.070 0.999
Sp5 3.59 0.073 0.999
Brwd1 3.57 0.073 0.999
Zmiz1 3.52 0.075 0.999
Mef2d 3.50 0.076 0.999
Ubn2 3.45 0.078 0.999
Gli2 3.42 0.079 0.999
Pcgf2 3.35 0.082 0.999
Foxd3 3.30 0.084 0.999
Utp3 3.12 0.093 0.999
Foxp1 3.08 0.095 0.999
Bmi1 3.02 0.098 0.999
Pwwp2b 3.00 0.099 0.999
Hdac4 2.93 0.102 0.999
Nlrc5 2.89 0.105 0.999
Nucks1 2.85 0.107 0.999
Zfat 2.82 0.108 0.999
Zmynd8 2.81 0.109 0.999
Elob 2.80 0.110 0.999
Anp32e 2.70 0.116 0.999
Nfil3 2.67 0.118 0.999
Wdr43 2.64 0.120 0.999
Spen 2.61 0.122 0.999
Sfmbt1 2.58 0.124 0.999
Usf1 2.49 0.130 0.999
Pou3f2 2.45 0.133 0.999
Tlx1 2.41 0.136 0.999
Nfyb 2.39 0.138 0.999
Jph2 2.37 0.139 0.999
Irf2 2.37 0.139 0.999
Dpf2 2.33 0.143 0.999
Hoxb5 2.28 0.146 0.999
Gabpa 2.25 0.149 0.999
Htatsf1 2.23 0.151 0.999
Snai1 2.22 0.152 0.999
Taf9b 2.20 0.154 0.999
Fezf2 2.18 0.155 0.999
Morc3 2.11 0.162 0.999

41
Cebpd 2.09 0.164 0.999
Baz1a 2.07 0.166 0.999
Mfsd11 2.06 0.167 0.999
Cbfb 2.06 0.167 0.999
Zfp217 2.05 0.167 0.999
Arnt2 2.00 0.173 0.999
Ncoa3 1.95 0.178 0.999
Supt5h 1.90 0.183 0.999
Lin28a 1.80 0.195 0.999
Tead2 1.79 0.196 0.999
Prdm5 1.76 0.200 0.999
Cat 1.72 0.204 0.999
Cbx2 1.72 0.204 0.999
Kdm5a 1.72 0.205 0.999
Lyl1 1.71 0.206 0.999
Smarcc1 1.65 0.214 0.999
Smad1 1.62 0.217 0.999
Pax3 1.58 0.223 0.999
Atf4 1.56 0.226 0.999
Sox30 1.52 0.232 0.999
Rybp 1.52 0.232 0.999
Supt16h 1.52 0.232 0.999
Nkx2
-
5 1.52 0.232 0.999
Nelfe 1.52 0.233 0.999
Znf24 1.51 0.233 0.999
Spib 1.51 0.234 0.999
Prdm14 1.50 0.235 0.999
Erg 1.47 0.239 0.999
Znf652 1.45 0.243 0.999
Pex2 1.44 0.244 0.999
Stra8 1.44 0.244 0.999
Nacc1 1.44 0.244 0.999
Ascl2 1.44 0.245 0.999
Kdm3a 1.43 0.245 0.999
Cbx3 1.43 0.245 0.999
Ubtf 1.41 0.249 0.999
Insm1 1.41 0.249 0.999
Sap18 1.37 0.256 0.999
Batf3 1.37 0.256 0.999
Hoxa13 1.36 0.257 0.999
Stag1 1.33 0.263 0.999
Mafb 1.32 0.264 0.999
Rbbp4 1.31 0.265 0.999
Thap1 1.31 0.265 0.999

42
Hnrnpu 1.30 0.268 0.999
Nrf1 1.30 0.268 0.999
Pitx1 1.28 0.271 0.999
Gli1 1.26 0.274 0.999
Nrdc 1.26 0.275 0.999
Mta1 1.26 0.275 0.999
Zc3h11a 1.20 0.287 0.999
Tbpl1 1.18 0.289 0.999
Dux 1.16 0.295 0.999
Tbx19 1.15 0.295 0.999
Myo1c 1.13 0.300 0.999
Zfp809 1.12 0.302 0.999
Zfp449 1.11 0.304 0.999
Hoxd11 1.11 0.304 0.999
Cebpe 1.10 0.306 0.999
Kmt2d 1.09 0.309 0.999
Foxh1 1.07 0.312 0.999
Foxa3 1.07 0.313 0.999
Magi1 1.07 0.314 0.999
Tbl1x 1.06 0.315 0.999
Pou2f1 1.06 0.316 0.999
Nfkb1 1.05 0.317 0.999
Ccny 1.01 0.326 0.999
Tet3 1.01 0.328 0.999
Dmrtb1 1.00 0.329 0.999
Xbp1 0.99 0.332 0.999
PcgF3 0.99 0.332 0.999
Rcor1 0.98 0.333 0.999
Lhx1 0.96 0.338 0.999
Hsf1 0.96 0.339 0.999
Zfx 0.95 0.342 0.999
Ccl5 0.93 0.346 0.999
Sin3b 0.92 0.349 0.999
Setd1a 0.90 0.354 0.999
Chd8 0.89 0.356 0.999
Lmnb1 0.89 0.357 0.999
Foxg1 0.88 0.360 0.999
Egr2 0.87 0.362 0.999
Foxp3 0.87 0.362 0.999
Zbtb2 0.87 0.363 0.999
Rora 0.87 0.363 0.999
Na 0.85 0.367 0.999
Kat5 0.85 0.367 0.999
Znf319 0.85 0.368 0.999

43
Brd9 0.85 0.368 0.999
Foxk1 0.84 0.369 0.999
Bach1 0.84 0.370 0.999
Jarid2 0.84 0.371 0.999
Mxi1 0.84 0.371 0.999
Kmt2b 0.83 0.373 0.999
Relb 0.82 0.377 0.999
Gfi1b 0.81 0.378 0.999
Cbx7 0.78 0.386 0.999
Rbpjl 0.78 0.388 0.999
Brd2 0.76 0.393 0.999
Sox3 0.75 0.395 0.999
Smarca5 0.75 0.396 0.999
Tfap4 0.75 0.396 0.999
Ncaph2 0.74 0.400 0.999
Klf1 0.73 0.402 0.999
Hey2 0.72 0.407 0.999
Smyd3 0.71 0.408 0.999
Onecut1 0.71 0.410 0.999
Rara 0.69 0.415 0.999
Chd7 0.69 0.415 0.999
Dpep2 0.69 0.417 0.999
Usf2 0.68 0.418 0.999
Meis1 0.68 0.419 0.999
Obox1 0.68 0.420 0.999
Rag2 0.67 0.423 0.999
Gsk3b 0.66 0.427 0.999
Kdm4c 0.66 0.427 0.999
Mcrs1 0.65 0.428 0.999
Elf5 0.65 0.428 0.999
Mitf 0.65 0.430 0.999
Chd4 0.65 0.430 0.999
Grhl2 0.65 0.431 0.999
Foxo3 0.64 0.435 0.999
Creld1 0.63 0.435 0.999
Arid3a 0.63 0.436 0.999
Top2b 0.63 0.436 0.999
Hif1a 0.63 0.436 0.999
Atrx 0.63 0.437 0.999
Stat2 0.63 0.438 0.999
Nipbl 0.63 0.438 0.999
Osr1 0.62 0.440 0.999
Nfya 0.62 0.441 0.999
E2f3 0.61 0.443 0.999

44
Pax6 0.61 0.443 0.999
Six2 0.61 0.445 0.999
Etv5 0.60 0.446 0.999
Smarcc2 0.59 0.450 0.999
Aurkb 0.59 0.452 0.999
Kdm6a 0.59 0.452 0.999
Kdm2b 0.58 0.453 0.999
Rsf1 0.58 0.455 0.999
Lef1 0.58 0.457 0.999
Cdk6 0.57 0.457 0.999
Tcf7 0.57 0.461 0.999
Kdm1a 0.56 0.462 0.999
Tbx21 0.56 0.462 0.999
Nr1h2 0.55 0.468 0.999
Dr1 0.55 0.469 0.999
Men1 0.54 0.469 0.999
Taf10 0.54 0.469 0.999
Myb 0.54 0.470 0.999
Prdm15 0.54 0.471 0.999
Prox1 0.54 0.472 0.999
Tp53 0.54 0.473 0.999
Tead1 0.53 0.473 0.999
Ikzf2 0.52 0.479 0.999
Nfatc2 0.52 0.479 0.999
Ebf2 0.52 0.481 0.999
Arid1a 0.51 0.482 0.999
Ikzf1 0.51 0.483 0.999
Crx 0.51 0.483 0.999
Chd2 0.51 0.484 0.999
Sin3a 0.51 0.484 0.999
Rfx1 0.50 0.487 0.999
Npas4 0.50 0.487 0.999
Irf5 0.50 0.489 0.999
Kat2a 0.49 0.490 0.999
Dlx5 0.48 0.496 0.999
Usp16 0.48 0.496 0.999
Myf5 0.48 0.498 0.999
Tex10 0.47 0.501 0.999
Zbtb17 0.47 0.502 0.999
Esr2 0.46 0.506 0.999
Pgr 0.45 0.509 0.999
Ncoa2 0.45 0.511 0.999
Sp9 0.44 0.514 0.999
Sumo1 0.44 0.517 0.999

45
Tcf7l1 0.42 0.522 0.999
Snai2 0.42 0.523 0.999
Hdac2 0.42 0.523 0.999
Maz 0.42 0.524 0.999
Rad23b 0.41 0.527 0.999
Nkx3
-
1 0.41 0.530 0.999
Msl2 0.41 0.530 0.999
Kdm5c 0.41 0.531 0.999
Dnmt3a 0.41 0.531 0.999
Mta2 0.40 0.535 0.999
Hey1 0.40 0.535 0.999
Nfib 0.40 0.536 0.999
Shox2 0.39 0.537 0.999
Smc5 0.39 0.540 0.999
Dppa4 0.39 0.540 0.999
Arntl 0.38 0.542 0.999
Ep300 0.38 0.543 0.999
Pcgf6 0.37 0.550 0.999
Wiz 0.37 0.552 0.999
Lhx2 0.36 0.554 0.999
Yy1 0.36 0.555 0.999
Sox9 0.35 0.561 0.999
Rfx2 0.34 0.564 0.999
Gata6 0.34 0.565 0.999
Dpep3 0.34 0.568 0.999
Tfap2a 0.34 0.569 0.999
Ep400 0.33 0.572 0.999
Mtf2 0.33 0.572 0.999
Auts2 0.33 0.572 0.999
Sox6 0.33 0.573 0.999
Rfx3 0.32 0.576 0.999
Tcf7l2 0.32 0.576 0.999
Gata2 0.32 0.578 0.999
Utp6 0.32 0.579 0.999
Elf1 0.32 0.579 0.999
Sox4 0.32 0.580 0.999
Purb 0.31 0.581 0.999
Maff 0.31 0.584 0.999
Hnf4a 0.31 0.584 0.999
Gfi1 0.31 0.584 0.999
Creb1 0.31 0.586 0.999
Irf4 0.30 0.591 0.999
Nkx3
-
2 0.30 0.593 0.999
Med1 0.29 0.593 0.999

46
Ehf 0.29 0.593 0.999
Sall4 0.29 0.594 0.999
Zfpm1 0.29 0.595 0.999
Rarb 0.29 0.596 0.999
Kdm2a 0.29 0.598 0.999
Yap1 0.29 0.598 0.999
Foxn1 0.28 0.601 0.999
Sap130 0.28 0.603 0.999
Egr1 0.28 0.603 0.999
Znf281 0.28 0.605 0.999
Chd1 0.27 0.608 0.999
Smarcad1 0.27 0.610 0.999
Klf5 0.27 0.611 0.999
Mef2a 0.27 0.611 0.999
Ruvbl1 0.26 0.615 0.999
Nfe2l2 0.26 0.615 0.999
Etv6 0.26 0.616 0.999
Max 0.26 0.617 0.999
L3mbtl2 0.26 0.618 0.999
Kmt2c 0.25 0.619 0.999
Znf143 0.25 0.623 0.999
E2f1 0.24 0.627 0.999
Satb1 0.23 0.633 0.999
Rnf2 0.23 0.634 0.999
Cxxc1 0.23 0.635 0.999
Hcfc1 0.23 0.637 0.999
Leo1 0.23 0.638 0.999
Rad51 0.23 0.638 0.999
Mafk 0.23 0.638 0.999
Sirt1 0.23 0.639 0.999
Hes1 0.23 0.640 0.999
Fosb 0.22 0.641 0.999
Otx2 0.22 0.641 0.999
Gata4 0.22 0.641 0.999
Setdb1 0.22 0.642 0.999
Tbx6 0.22 0.643 0.999
Ahr 0.22 0.646 0.999
Epop 0.22 0.647 0.999
Smad4 0.21 0.650 0.999
Stag2 0.21 0.651 0.999
Nbn 0.21 0.652 0.999
Gata3 0.21 0.653 0.999
Zfp57 0.21 0.654 0.999
Hoxc9 0.21 0.654 0.999

47
Srf 0.21 0.655 0.999
Myod1 0.20 0.659 0.999
Nup153 0.19 0.665 0.999
Fosl2 0.19 0.665 0.999
Hnf4g 0.19 0.666 0.999
Rorc 0.19 0.666 0.999
Ppargc1a 0.19 0.667 0.999
Ldb1 0.19 0.668 0.999
Batf 0.19 0.671 0.999
Acss2 0.19 0.671 0.999
Rpa2 0.18 0.672 0.999
Taf12 0.18 0.673 0.999
Pcgf1 0.18 0.674 0.999
Pax5 0.18 0.675 0.999
Zbtb16 0.18 0.677 0.999
Taf2 0.18 0.678 0.999
Kmt2a 0.18 0.679 0.999
Crebbp 0.17 0.682 0.999
Mef2c 0.17 0.683 0.999
Smad2 0.17 0.684 0.999
Elk4 0.17 0.684 0.999
Foxa2 0.17 0.685 0.999
Ar 0.17 0.687 0.999
Plagl1 0.17 0.688 0.999
Nup98 0.16 0.693 0.999
Six4 0.16 0.694 0.999
Tcf3 0.15 0.698 0.999
Thap11 0.15 0.699 0.999
Esrrg 0.15 0.699 0.999
Tbr1 0.15 0.701 0.999
Ezh1 0.15 0.702 0.999
Prdm16 0.15 0.702 0.999
Hoxa9 0.15 0.703 0.999
Stfa1 0.15 0.706 0.999
Sumo2 0.15 0.706 0.999
Rtf1 0.15 0.707 0.999
Pax7 0.15 0.707 0.999
Msl1 0.14 0.708 0.999
Taf1 0.14 0.710 0.999
Onecut2 0.14 0.711 0.999
Brca1 0.14 0.712 0.999
Pbrm1 0.14 0.715 0.999
Wt1 0.14 0.716 0.999
Ikzf3 0.14 0.717 0.999

48
Bcl11b 0.13 0.719 0.999
Ascl1 0.13 0.719 0.999
Tbx5 0.13 0.721 0.999
Hnf1a 0.13 0.724 0.999
Myc 0.12 0.728 0.999
Cebpg 0.12 0.728 0.999
Ezh2 0.12 0.728 0.999
Gtf2b 0.12 0.731 0.999
Kat8 0.12 0.732 0.999
Thra 0.12 0.737 0.999
Smc3 0.11 0.739 0.999
Hand1 0.11 0.740 0.999
Cdx2 0.11 0.748 0.999
Tfcp2l1 0.11 0.749 0.999
Sp1 0.10 0.750 0.999
Cbx8 0.10 0.750 0.999
Ncor1 0.10 0.751 0.999
Nelfa 0.10 0.751 0.999
Dmap1 0.10 0.752 0.999
Nkx6
-
1 0.10 0.752 0.999
Srsf2 0.10 0.752 0.999
Tal1 0.10 0.754 0.999
Klf3 0.10 0.754 0.999
Klf6 0.10 0.758 0.999
Sirt6 0.10 0.760 0.999
Hoxc12 0.09 0.762 0.999
Dpy30 0.09 0.763 0.999
Junb 0.09 0.763 0.999
Tcf12 0.09 0.768 0.999
Sall1 0.09 0.770 0.999
Ssrp1 0.09 0.773 0.999
Pbx1 0.09 0.773 0.999
Kansl3 0.08 0.775 0.999
Zic2 0.08 0.775 0.999
Suz12 0.08 0.775 0.999
Zeb1 0.08 0.778 0.999
Hnf1b 0.08 0.780 0.999
Trim33 0.08 0.782 0.999
Notch1 0.08 0.785 0.999
Nr1d2 0.07 0.789 0.999
Trim28 0.07 0.790 0.999
Nsd3 0.07 0.790 0.999
Bhlhe40 0.07 0.791 0.999
Fosl1 0.07 0.794 0.999

49
Rbfox2 0.07 0.795 0.999
Dnmt3b 0.07 0.796 0.999
Supt6h 0.07 0.796 0.999
Zkscan1 0.07 0.797 0.999
Irf7 0.07 0.798 0.999
Tbp 0.07 0.800 0.999
Sox2 0.06 0.802 0.999
Nr3c1 0.06 0.803 0.999
Med26 0.06 0.803 0.999
Elf4 0.06 0.807 0.999
Mybl1 0.06 0.809 0.999
Runx1 0.06 0.813 0.999
Neurod2 0.06 0.814 0.999
Nkx2
-
1 0.06 0.816 0.999
Spi1 0.06 0.817 0.999
Prop1 0.06 0.817 0.999
Brd4 0.06 0.817 0.999
Mafg 0.05 0.819 0.999
Stat5a 0.05 0.819 0.999
Nr5a2 0.05 0.820 0.999
Bach2 0.05 0.821 0.999
Irf1 0.05 0.824 0.999
Kdm4a 0.05 0.824 0.999
Zic3 0.05 0.827 0.999
Kdm5b 0.05 0.829 0.999
Neurog2 0.05 0.829 0.999
Smarca4 0.05 0.830 0.999
Ebf1 0.05 0.831 0.999
Ash2l 0.05 0.831 0.999
Ino80 0.05 0.831 0.999
Myog 0.05 0.834 0.999
Hivep3 0.04 0.835 0.999
Zic1 0.04 0.835 0.999
Hand2 0.04 0.836 0.999
Atf2 0.04 0.837 0.999
Olig2 0.04 0.837 0.999
Dppa2 0.04 0.838 0.999
Nr4a1 0.04 0.839 0.999
Pou5f1 0.04 0.840 0.999
Cdk8 0.04 0.841 0.999
Capg 0.04 0.844 0.999
Erf 0.04 0.847 0.999
Sox17 0.04 0.849 0.999
Rxra 0.04 0.849 0.999

50
Nr6a1 0.04 0.849 0.999
Fli1 0.04 0.850 0.999
Btaf1 0.04 0.850 0.999
Foxl2 0.04 0.852 0.999
Phrf1 0.04 0.852 0.999
Cux2 0.03 0.857 0.999
Rag1 0.03 0.857 0.999
Stat1 0.03 0.857 0.999
Tfe3 0.03 0.860 0.999
Isx 0.03 0.860 0.999
Esr1 0.03 0.865 0.999
Esrrb 0.03 0.866 0.999
Nfyc 0.03 0.868 0.999
Tet2 0.03 0.868 0.999
Phf5a 0.03 0.868 0.999
Bcl6 0.03 0.871 0.999
Hdac1 0.03 0.873 0.999
Esrra 0.03 0.876 0.999
Runx3 0.02 0.879 0.999
Hoxb4 0.02 0.881 0.999
Rcor2 0.02 0.882 0.999
Stat4 0.02 0.884 0.999
Ptf1a 0.02 0.884 0.999
Clock 0.02 0.886 0.999
Srebf1 0.02 0.891 0.999
Aff3 0.02 0.892 0.999
Tp63 0.02 0.894 0.999
Jund 0.02 0.894 0.999
Sinhcaf 0.02 0.895 0.999
Prdm9 0.02 0.895 0.999
Rai1 0.02 0.896 0.999
Ets1 0.02 0.897 0.999
Stat5b 0.02 0.899 0.999
Cebpa 0.02 0.899 0.999
Atf7 0.02 0.900 0.999
Tead4 0.02 0.904 0.999
Foxa1 0.01 0.905 0.999
Ctnnb1 0.01 0.905 0.999
Lhx3 0.01 0.909 0.999
Fos 0.01 0.909 0.999
Dmc1 0.01 0.910 0.999
Nkx2
-
2 0.01 0.910 0.999
Cebpb 0.01 0.913 0.999
Prdm13 0.01 0.913 0.999

51
Phf6 0.01 0.915 0.999
Ctcf 0.01 0.915 0.999
Nr1h3 0.01 0.916 0.999
Stat3 0.01 0.916 0.999
Med23 0.01 0.918 0.999
Ctcfl 0.01 0.918 0.999
Ncor2 0.01 0.921 0.999
Pias2 0.01 0.921 0.999
Chaf1a 0.01 0.922 0.999
Mycn 0.01 0.923 0.999
Atoh1 0.01 0.926 0.999
Isl1 0.01 0.926 0.999
Rad21 0.01 0.927 0.999
Baz1b 0.01 0.928 0.999
Ndufs2 0.01 0.928 0.999
Irf3 0.01 0.930 0.999
Etv4 0.01 0.932 0.999
Prdm1 0.01 0.934 0.999
Gata1 0.01 0.937 0.999
Tgif1 0.01 0.937 0.999
Abcc9 0.01 0.938 0.999
Dlx3 0.01 0.940 0.999
Rest 0.01 0.941 0.999
Foxf1 0.01 0.941 0.999
Asxl1 0.01 0.941 0.999
Ogt 0.01 0.942 0.999
Nfe2l1 0.00 0.945 0.999
Med12 0.00 0.947 0.999
Gps2 0.00 0.948 0.999
Twist2 0.00 0.950 0.999
Rbpj 0.00 0.950 0.999
Zfp384 0.00 0.950 0.999
Nanog 0.00 0.951 0.999
Neurod1 0.00 0.952 0.999
Mecp2 0.00 0.953 0.999
Ppara 0.00 0.953 0.999
Anpep 0.00 0.955 0.999
Klf4 0.00 0.958 0.999
Kansl1 0.00 0.962 0.999
Stat6 0.00 0.962 0.999
Tfap2c 0.00 0.963 0.999
Hes5 0.00 0.966 0.999
Tox 0.00 0.966 0.999
Smad3 0.00 0.966 0.999

52
Gtf3c1 0.00 0.970 0.999
Lmo2 0.00 0.971 0.999
Pcna 0.00 0.972 0.999
Cdk9 0.00 0.976 0.999
Sigmar1 0.00 0.977 0.999
Rela 0.00 0.978 0.999
Dmrt1 0.00 0.978 0.999
Hdac3 0.00 0.979 0.999
Cdk7 0.00 0.981 0.999
Terf2ip 0.00 0.982 0.999
Sp7 0.00 0.982 0.999
Pknox1 0.00 0.983 0.999
Runx2 0.00 0.983 0.999
Tet1 0.00 0.984 0.999
Ncapd3 0.00 0.985 0.999
Smc1a 0.00 0.985 0.999
Vdr 0.00 0.986 0.999
Bhlha15 0.00 0.987 0.999
T 0.00 0.988 0.999
Irf8 0.00 0.989 0.999
Atf3 0.00 0.990 0.999
Pparg 0.00 0.992 0.999
Nfia 0.00 0.993 0.999
Elk1 0.00 0.994 0.999
Jun 0.00 0.994 0.999
Tbx3 0.00 0.995 0.999
Nr1d1 0.00 0.995 0.999
Eomes 0.00 0.997 0.999
Ell3 0.00 0.997 0.999
Utf1 0.00 0.998 0.999
Sim2 0.00 0.999 0.999

53
Chapter 3: A stem cell-derived neuronal culture model for mouse species
variation in neurite growth
Abstract
Understanding the molecular control of neurite extension by neurons represents
a key goal of cellular neuroscience. Neurons of the Southeast Asian mouse Mus
musculus castaneus exhibit a unique phenotype of neurite growth and regeneration
whose mechanism is poorly understood. In this work, we pioneered the use of stem cellderived neurons for in vitro analyses of neurite phenotypes in M. m. castaneus and its
relatives. We differentiated neurons from stem cells of a panel of Mus genotypes and
assayed their cellular properties. Cells of the M. m. castaneus genotype extended more
and longer neurites than did neurons from standard laboratory mice. In stem cellderived neurons from F1 interspecific crosses using M. m. castaneus, the avid neurite
growth phenotype followed a dominant genetic model. Thus, M. m. castaneus neurons
are hard-wired for neurite extension even outside of an organismal context, and the
underlying genetic factors are sufficient to confer the trait on other backgrounds. This
study establishes the power of stem cell models for mechanistic dissection of neuronal
biology in non-model mice.
Introduction
In the mammalian central nervous system, once axons are laid down during
development, they do not regenerate efficiently, especially after injury. The failure to
regrow axons is the principal reason why traumatic brain or spinal cord damage does
not heal in adult mammals (Tedeschi et al. 2017). Years of research have revealed
mutations and drugs that promote axon growth and regeneration, including elegant

54
screening approaches in laboratory strains of mice (Nix et al. 2014; Tedeschi et al.
2017; Mahar and Cavalli 2018; Sekine et al. 2018; Lindborg et al. 2021; Lear and Moore
2023). Against this backdrop, natural genetic variation across genomes from the wild
represents another class of perturbation that can be harnessed for discoveries of the
molecular mechanisms of neurite growth. In a now-classic survey of axonal injury and
repair in genetically distinct mice, a little-studied Southeast Asian mouse subspecies,
Mus musculus castaneus, emerged with unique phenotypes: its neurons could
regenerate axons far beyond those of any other known mouse system, after injury in
vivo and ex vivo, in which transcriptional and genetic analyses implicated a potential
role for Activin signaling (Omura et al. 2015) and control of Ascl1 expression (Lisi et al.
2017). This was received by the field as a landmark (Mahar and Cavalli 2018), with the
implication that there are no fundamental biological roadblocks that limit axonal
regeneration in the mammalian central nervous system. But exactly what M. m.
castaneus neurons do differently than those of other species, and what the underlying
molecular and genetic mechanisms might be, remains largely unknown.
Stem cell-derived neurons represent a powerful experimental system for the
study of axon growth and regeneration, cutting cost and technical challenges (Harper et
al. 2004; Lu et al. 2014; Pelkonen et al. 2021). We reasoned that development of an in
vitro culture model could serve as a foundational step for the genetic dissection of M. m.
castaneus neuronal phenotypes. We thus set out to establish a comparative biology
framework with a stem cell-based approach. We developed a strategy using a panel of
pluripotent stem cells of homozygous and interspecies F1 hybrid mouse genotypes with
directed neuron differentiation and image analysis tools. The resulting comparison

55
workflow serves as an efficient and powerful jumping-off point for studying the
mechanism of mouse species variation in neurite extension properties.
Results
To pursue comparative analyses of neurite growth across Mus with an in vitro
setup, we chose an approach that used a panel of pluripotent stem cell lines of different
genetic backgrounds. These included induced pluripotent stem cell lines derived from
M. m. castaneus strain CAST/Ei/J (hereafter CAST) and from the laboratory strain
C57BL/6 (BL6; mostly of M. domesticus origin); and embryonic stem cell lines from the
laboratory strain 129S6/SvEv (129; also largely of M. domesticus origin) and from
reciprocal F1 hybrid crosses of CAST x 129S6/SvEv. We then utilized a two-step
directed differentiation protocol (Wu et al. 2012) to yield stem cell-derived motor
neurons of each genotype (Figure 3.1A, B). To assess differentiation outcomes across
the panel, we cultured stem cells and neurons of each genotype and isolated RNA for
sequencing. Analysis of the resulting reads identified 9784 genes differentially
expressed between neurons and stem cells across genotypes (Figure 3.1C and
Supplemental Fig 1A). Genes induced in neuron cultures were enriched for annotations
in synapse function, axon biology, and membrane potential regulation (Figure 3.1D),
and included genes related to motor neuron identity (Figure 3.1E). Likewise, in a
principal component analysis (Supplemental Figure 3.1), our stem cell transcriptomes
grouped with previously published RNA-seq data from C57BL/6, CAST, and 129
embryonic stem cell lines (Skelly et al. 2020). These results indicated that the reagents
used here were suitable models for informative biological analyses: our stem cell lines
were transcriptionally similar to pluripotent stem cells cultured and sequenced

56

57
independently, and our differentiation protocol generated neurons with high expression
of motor neuron-specific genes.
Next, we used our stem cell-derived neurons to test for genotype effects on
neurite outgrowth. We used our five focal genotypes (CAST, 129, BL6, and two CAST x
129 hybrids) as input into quantitative assays of neurite extension using a microfluidic
approach. For this purpose, we cultured neurons of each genotype in the Xona
Innsbruck PDMS chip, a Campenot-style device with cell bodies cultured in a central
chamber extending neurites through microcombs into flanking chambers (Figure 3.2AB). We used an experimental design that included replicate differentiations and imaging
multiple devices for each discrete incubation period (Supplemental Figure 3.2). After
fixation, we used immunohistochemistry to assay neuronal outgrowth in each device,
and from the resulting images we adapted an image analysis strategy to quantify cell
bodies in the central chamber and neurite outgrowth in the flanking chambers (Figure
3.2A). Neurite density in the outer chambers was measured as the total area of bright
Figure 3.1: Differentiation of motor neurons from stem cells and transcriptional
profiling. (A) Representative images from cultures at the indicated timepoints; at
day 14, shown are results of immunohistochemistry assays for expression of the
indicated neuronal markers. Scale bars: 100 µm. (B) Protocol for directed
differentiation of mouse pluripotent stem cells to motor neurons; neural identity was
induced in pluripotent stem cell aggregates that were then patterned for motor
neuron generation. Schematized at top is the timeline of motor neuron differentiation
and phenotyping (d, days). Media and supplements, substrates and treatment details
are described in the supplemental methods. (C-E) Transcriptome analysis of stem
cells and stem cell-derived motor neurons. (C) Each row reports transcripts per
million, as a mean across replicates and median-centered at 0, for one gene
differentially expressed between neurons and stem cells. Each column reports
results from the indicated genotype, in transcriptomes from neurons or pluripotent
stem cells (PSC) from this study or (Skelly et al. 2020). (D) Top Gene Ontology
terms enriched among genes upregulated in neurons relative to stem cells. In a
given row, the dot size reports the proportion of all genes upregulated in neurons that
fall into the term on the y-axis. The x-axis reports the number of genes upregulated
in neurons in the term. (E) Data are as in (C) except that only motor neuron identity
marker genes are shown.

58
pixels after a uniformly applied thresholding algorithm. To estimate cell body density for
each culture, we trained a machine learning image classification model to identify
regions as cell body, neurite, or background (Figure 3.2Aiii) and measured the total area
from pixels classified as cell body. In analyses focused on linear neurite extension
(Jocher et al. 2018), we found that neurons derived from stem cells of M. m. castaneus
Figure 3.2: Microfluidic assays of neurite extension from mouse stem cellderived neurons. (A) A custom image analysis workflow was used for neurite
detection and to train a machine learning classification tool to identify cell bodies for
normalization. (i) Representative image of III tubulin immunohistochemistry of stem
cell-derived neurons, with cell bodies cultured in the central chamber extending
neurites through the microgroove comb (500 μm length) into outer chambers at left
and right. (ii) Central chamber image segmented by classification model. Colors
represent pixels classified as either cell bodies (red), axons (green), or background
(magenta). (iii) Processed outer chamber image. Neurite outgrowth areas were
quantified from regions above an algorithmically defined brightness threshold (white).
(B) Each panel is a representative tiled image of III tubulin immunohistochemistry of
stem cell-derived neurons of the indicated mouse genotype after culture for three
days in the device. Scale bar: 500 μm.

59
origin (both the CAST and CAST x 129 hybrids) extended more neurites early in the
time course, and that these neurites reached longer across the device, relative to
laboratory-strain neurons (Figure 3.3A and Supplemental Figure 3.4). Separately, we
quantified total neurite outgrowth for each genotype and again observed a more
dramatic growth phenotype among neurons with inheritance from M. m. castaneus
(Figure 3.3B). Together, these data make clear that stem cell-derived CAST neurons
exhibit more avid neurite growth than those of laboratory mice and that the phenotype
follows a dominant genetic model. Thus, neurite growth and regeneration phenotypes
previously reported in vivo in M. m. castaneus can be recapitulated in an in vitro culture
system even in the absence of injury, in which alleles from M. m. castaneus are
sufficient to drive enhanced neurite extension.
Discussion
Traits that vary naturally in the wild can serve as a rich data source for
mechanistic insights of evolutionary and applied interest. Wild-derived members of the
genus Mus represent a particularly useful system for this framework, given the
applicability of tools from decades of work in their laboratory cousins. In this study, we
have established a stem cell-derived CNS neuronal model for comparative biology
across Mus. We focused on neurite growth differences between the neurons of M. m.
castaneus and those of laboratory mouse backgrounds, based on previous studies in
vivo after injury and in short-term culture ex vivo (Omura et al. 2015; Lisi et al. 2017).
Our findings reveal that the species divergence in neurite phenotypes can be modeled
with stem cell-derived neurons in vitro. The methods we have developed set a
precedent for stem cell-based evolutionary and genetic dissection of neuronal

60
Figure 3.3: Inheritance from M. m. castaneus is sufficient for increased neurite
outgrowth in stem cell-derived neurons. (A) Each panel reports a comparison,
across mouse genotypes, of the neurite outgrowth of stem cell-derived neurons at
the indicated day of culture. Each trace reports, for cultures of neurons of the
indicated genotype, the number of neurites (y-axis) that extended to each distance
(x-axis) from the inner boundary of the neurite chamber (the outer boundary of the
comb; see Figure 3.2A). (B) Shown is a comparison across mouse genotypes of total
neurite density in the neurite chamber across the culture time course. On a given day
(x-axis), each bar reports the mean (solid black), upper and lower quartiles (bar
borders), and individual replicate measurements (points) of neurite density,
normalized by cell body count, for neuron cultures of the indicated genotype. In each
panel, day and time were significant predictors of the respective phenotype by twoway ANOVA (p < 0.05).

61
phenotypes in non-model mouse species, and our proof-of-concept experiments with
these tools shed new light on the mechanisms of the M. m. castaneus phenotype.
Our work builds on recent investments in stem cell resources for non-model mice
that have laid the foundation for in vitro approaches to comparative biology across these
backgrounds. The field to date has emphasized the phenotypes of stem cells
themselves, including their variation across mouse strains and species and the nuclear
and mitochondrial genetic basis (Kelly et al. 2013; Garbutt et al. 2018; Lazzarano et al.
2018; Ortmann et al. 2020; Skelly et al. 2020; Byers et al. 2022; Aydin et al. 2023). By
contrast, comparative studies of stem cell-derived tissues across mouse backgrounds
remain at a premium in the current literature (although see (Kelly et al. 2013)). Our work
represents a case study for the latter paradigm in neurons. By modeling M. m.
castaneus neurite outgrowth in stem cell-derived CNS neurons, we have demonstrated
the feasibility and power of the in vitro approach.
Indeed, our results, against the backdrop of previous work from M. m. castaneus
animals, have important implications for models of the cellular mechanism of the neurite
growth trait. By virtue of its appearance in stem cell-derived neurons, we infer that the
M. m. castaneus phenotype does not depend on contact with other tissue types, in
development or in the adult. Furthermore, although the M. m. castaneus trait was
originally described for neurons isolated from dorsal root ganglia (Omura et al. 2015;
Lisi et al. 2017), our recapitulation in stem cell derived motor neurons supports
arguments against a role for programs of particular neuronal types, consistent with the
observation of enhanced M. m. castaneus CNS regeneration observed in vivo after
injury conditioning (Omura et al. 2015). Our in vitro data from neurons cultured for up to

62
seven days indicates that M. m. castaneus neurites grow longer and faster than those of
other mouse backgrounds, even in the absence of injury. We propose that M. m.
castaneus neurons are hard-wired for avid neurite outgrowth regardless of their
environment or developmental lineage. If so, this property could be linked causally to
the unique response by M. m. castaneus neurons to injury (Omura et al. 2015; Lisi et al.
2017), or alternatively it could represent an independent behavior of neurons of this
background.
Additionally, our analysis of neurons from interspecies crosses has revealed a
dominant genetic mechanism for the M. m. castaneus neurite growth trait. This echoes
previous reports in other systems of dominant negative alleles in repressors of neurite
outgrowth (Tomoda et al. 1999; Yang et al. 2010; Catlett et al. 2021). In principle,
molecular factors underlying the M. m. castaneus phenotype could have such a mode of
action, or they could instead act as gains of function in pro-growth pathways. In either
case, it is tempting to speculate that these determinants, once discovered in M. m.
castaneus, could be sufficient to confer the trait in other backgrounds. Ultimately, of
relevance for regeneration applications, with the continued implementation of stem cellbased experimental designs, expedient analyses to identify the molecular basis of M. m.
castaneus neuronal phenotypes will come within reach.
Methods
Mouse pluripotent stem cell lines
Mouse induced pluripotent stem cell (miPSC) lines strains were derived from
CAST/EiJ and C57BL/6 embryonic fibroblasts obtained from Jackson labs at the UMN
Stem Cell Institute, as previously described (Greder et al. 2012). A 129S6/SvEv ESC

63
line was obtained from Millipore Sigma, (Catalog no. SCR012), Burlington, MA, USA).
CAST/EiJ x 129/SvImJ and 129/SvImJ x CAST/EiJ mESC lines were provided by Joost
Gribnau.
Stem cell culture
Mouse pluripotent stem cell (PSC) culture was performed essentially as
described previously (Greder et al. 2012). PSCs were cultured on a feeder layer of
irradiated mouse embryonic fibroblasts (R&D Systems, Minneapolis, MN, USA) on 0.1%
gelatin (Millipore, Burlington, MA, USA) -coated 6-well tissue culture plates (Thermo
Fisher Scientific, Waltham, MA, USA). iMEFs were thawed and maintained in high
glucose Dulbecco’s Modified Eagle Medium (DMEM) (Sigma-Aldrich, St. Louis, MO,
USA), 10% fetal bovine serum (FBS) (HyClone, Logan, UT, USA), 1x MEM nonessential amino acids (NEAA) (Sigma-Aldrich), 1x penicillin/streptomycin (Corning,
Corning, NY, USA), and 1x sodium pyruvate (Sigma-Aldrich). C57BL/6 miPSCs were
cultured in miPSC medium (knockout DMEM (Thermo Fisher Scientific), 10% knockout
serum replacement (KSR) (Thermo Fisher Scientific), 10% FBS, 1x glutaMAX (Thermo
Fisher Scientific), and 1x MEM NEAA, supplemented with 200 units/ml LIF (Millipore)
and 110 μM 2-mercaptoethanol (Thermo Fisher Scientific). CAST/EiJ miPSCs,
129S6/SvEv mESCs and the hybrid ESCs were cultured in miPSC media medium
supplemented with 3 μM CHIR99021 (Sigma-Adrich) and 1 μM PD0325901 (SigmaAldrich). All cell lines were incubated at 37oC in 5% CO2 with daily media changes.
mPSCs were passaged every 2 to 3 days depending on colony size and density. For
passaging, cells were washed with phosphate buffered saline (PBS) +/+ (Thermo Fisher
Scientific), detached with 0.25% trypsin (Sigma-Aldrich), incubated at 37oC for 3 to 4

64
minutes, then neutralized with an equal volume of miPSC medium without supplements.
Cells were further gently dissociated with a P1000 micropipette and collected by
centrifugation at 300g for 4 minutes. The supernatant was aspirated, and the cells
resuspended in complete medium. The resuspended cells were seeded onto a fresh
feeder layer for maintenance culture.
Motor neuron differentiation
mPSCs were differentiated into motor neurons using a protocol adapted from Wu
Whye et al., (2012). After cell passaging, iMEFs were depleted by incubation in a T75
flask (Thermo Fisher Scientific) at 37oC for 40 minutes in miPSC media. mPSCs were
then collected by centrifugation at 329g for 4 minutes and resuspended in Neural
Induction medium (DFK5 medium supplemented with 50 ng/mL Noggin (Stemcell
Technologies, Vancouver, Canada), 20 ng/mL FGF2 (Thermo Fisher Scientific), and 20
ng/mL FGF8 (Thermo Fisher Scientific)). DKF5 medium is 1:1 v/v miPSC medium
without supplements and DMEM/F12 (Sigma-Aldrich), supplemented with 5% FBS, 5%
KSR, 1% NEAA, 1% glutaMAX, 40 μM 2-mercaptoethanol, and 0.05%v/v InsulinTransferrin-Selenium (ITS-G) (Thermo Fisher Scientific). The resuspended cells were
cultured in 6-well ultra-low attachment plates (Corning) with orbital shaking at 65 rpm
(SHEX1619DG shaker, OHAUS, Parsippany, NJ, USA). The embryoid body (EB) cell
aggregates were cultured in Neural Induction medium for 2 days, with a half media
change on the second day. For specification into motor neurons. EBs from the Neural
Induction media were collected by centrifugation at 228g for 1 minute and resuspended
in Motor Neuron specification medium (DFK5 medium supplemented with 1 μM retinoic
acid (RA) (Millipore) and 1 μM smoothened agonist (SAG) (Millipore)). The resuspended

65
EBs were returned to the 6 well ultra-low attachment plates, with a feeding volume of 4
mL/ well and cultured with shaking for 5 days, with half media changes each day.
Preparation of mPSC motor neurons for culture or cryopreservation
Differentiated cell aggregates were collected and washed once with PBS +/+.
After allowing the aggregates to settle, the PBS +/+ was aspirated and replaced with
Accutase (Millipore) and incubated in a 37oC water bath for 8 minutes, with gentle
tapping of the tube at around 4 minutes. After incubation further mechanical dissociation
was conducted using a P1000 micropipette. The Accutase was then diluted with at least
an equal volume of DFK5 medium and the cells collected by centrifugation at 329g for 4
minutes before resuspension for cell culture or cryopreservation. For cryopreservation,
the cells were resuspended in Freezing media (90% FBS, 10% DMSO) at room
temperature, pipetted into cryogenic storage vials (Thermo Fisher Scientific) and cooled
in a Mr. Frosty Freezing Container (Thermo Fisher Scientific) at -80oC for 24 hours
before being moved to long-term storage in liquid nitrogen.
mPSC derived motor neuron culture
One day prior to plating neural cells, culture plates were coated with 0.01% polyL-ornithine (Millipore) overnight. The following day, the wells were washed twice with
PBS +/+, coated with 5 μg/mL laminin (StemCell Technologies) diluted in PBS +/+ and
incubated for at least two hours. For neuron culture, disaggregated cells or cells thawed
from cryopreservation were resuspended in DFK5 medium supplemented with 10 ng/mL
each of BDNF (Thermo Fisher Scientific), GDNF (GenScript, Piscataway, NJ, USA),
CNTF (BioLegend, San Diego, CA, USA), and NT3 (PeproTech, Cranbury, NJ, USA)
and cultured at 37oC, 5% CO2. 10μM Y-27632 (VWR International, Radnor, PA, USA)

66
was supplemented into the culture medium for 24 hours post thawing. 10 μM AraC
(Sigma-Aldrich) was added to the culture medium for the first 48 hours and then the
medium was changed. Half volume media changes were then performed daily.
Microfluidic mPSC motor neuron culture
For axon quantification dissociated mPSC neurons were cultured in microfluidic
Innsbruck devices (Catalog #IND500 Xona Microfluidics, Research Triangle Park, NC,
USA) mounted on 0.13-0.16mm thick Marienfeld Superior™ Cover Glass (Electron
Microscopy Sciences, Hatfield, PA, USA). The cover glass with the device mounted on
top was coated with 0.01% poly-L-ornithine two days prior to cell plating and incubated
overnight. The following day, it was washed twice with PBS+/+ and coated with 5μg/mL
rhLaminin-521 in PBS+/+ overnight. The extended period of rhLaminin-521 coating was
to ensure full coating of the microgrooves of the device. 250,000 live, disaggregated
cells were then seeded into the central chamber of each microfluidic device and
cultured at 37oC, 5% CO2. 10 μM AraC (Sigma-Aldrich) was added to the culture
medium for the first 48 hours. Full media changes for the cell chambers and half volume
media changes for the axon chambers were performed daily. Neurons were fixed using
10% buffered formalin after either 3, 5, or 7 days of culture for axon outgrowth
quantification.
RNA isolation and sequencing
RNA was extracted from undifferentiated PSC cultures following feeder depletion
and from PSC-derived motor neurons following 7 days of culture. For PSC samples the
RNAqueous™-Micro Total RNA Isolation Kit (Thermo Fisher Scientific) was used. For
neuron samples, we used the RNAeasy-Micro Total RNA Isolation Kit (Thermo Fisher

67
Scientific). In each case on-column DNase treatment (QIAGEN, Hilden, Germany) was
performed. RNA samples were processed into mRNA libraries and sequenced on an
Illumina NovaSeq 6000 Sequencing System, yielding ~20M paired-end 150 bp reads
per sample. Reads were mapped to reference genomes for the appropriate strains
using the STAR aligner (Dobin et al. 2013). Reference genomes and annotations were
downloaded from Ensembl database (Cunningham et al. 2022) build 102. BL6 samples
were mapped to the GRCm38 reference genome, 129 samples were mapped to the
129S1/SvImJ reference, and CAST samples to the CAST/EiJ reference genome. CAST
x 129 hybrid samples were mapped to a concatenated genome of the two strains. As a
reference point for the transcriptional identity of the various cell lines used in this study,
we also included published RNA-seq data from mESCs of various strains (Skelly et al.
2020). Raw read files were downloaded from the European Nucleotide Archive
(Sarkans et al. 2018) (accession E-MTAB-7730) and mapped using STAR.
Transcriptional profiling
The R package DESeq2 (Love et al. 2014) was used to test for differential
expression between the PSC and neuron RNA-seq samples generated in this study.
GO enrichment analysis on genes significantly upregulated (adjusted p-value < 0.05) in
neuronal samples was performed in R using the package clusterProfiler (Wu et al.
2021), testing only GO terms in the ‘biological processes’ ontology. Additionally, a set of
genes expressed highly or specifically in motor neurons was selected from the literature
(Ho et al. 2016; Blum et al. 2021; Solomon et al. 2021; Guerra San Juan et al. 2022).
Heatmaps were generated using the pheatmap package in R using log2(TPM+1) values
as input and the row-wise scaling option. For principal component analysis (PCA),

68
expression data from both stem cell and neuron samples was filtered to remove noncoding RNAs and mitochondrially encoded genes. 315 genes that were dysregulated in
hybrids (with a log2 fold-change < -10 between hybrid and parental strains) were also
filtered out. For the remaining genes, log2(TPM+1) values were calculated and input into
the prcomp module in R. In Supplemental Figure 3.1, background ellipses were drawn
for each cell type using the stat_ellipse function. Published RNA-seq data from (Skelly
et al. 2020) was included in Figure 3.1 and Supplemental Figure 3.1 for comparison.
Immunocytochemistry
Neuron cultures were fixed using 10% buffered formalin (Thermo Fisher
Scientific) for 10 minutes at room temperature and washed once with PBS -/- (Thermo
Fisher Scientific). Cells were permeabilized using 0.2% Triton X-100 (Sigma-Aldrich) in
PBS -/- for 10 minutes at room temperature and then incubated in blocking buffer (0.1%
Tween 20 (Sigma-Aldrich) and 1% bovine serum albumin (BSA) (Sigma-Aldrich) in PBS
-/-) for 2 hours at room temperature. The cells were then incubated overnight at 4oC in
primary antibody diluted in blocking buffer. The following day the cells were washed
three times with PBS -/- and incubated in secondary antibody diluted in blocking buffer
for 30 minutes at room temperature. The secondary antibody was replaced with DAPI
(Thermo Fisher Scientific, catalog #D3571) diluted in blocking buffer then incubated at
room temperature for 5 minutes. The cells were washed once with PBS -/- and
incubated in PBS -/- at 4oC until imaging. (bIII tubulin primary antibody, Millipore,
catalog #MAB1637, dilution 1:1000, ISL1 primary antibody, Millipore, catalog #AB4326,
dilution 1:200, MNX1 primary antibody, DSHB (Iowa City, IA, USA), catalog #81.5C10-s,
dilution 1:25, MAP2 primary antibody, Abcam (Cambridge, UK), catalog # ab32454,

69
dilution 1:200, Alexa Fluor 488 Donkey anti-Mouse secondary antibody, Thermo Fisher
Scientific, catalog #A21202, dilution 1:500, Alexa Fluor 555 Donkey anti-Rabbit
secondary antibody, Thermo Fisher Scientific, catalog #A31572, dilution 1:500).
Fluorescence microscopy
Fluorescence imaging was performed using a DMI6000 B microscope (Leica,
Wetzlar, Germany). Images were captured using images were captured using the Leica
Application Suite X (LAS X) software. Tile scans were acquired based on individual 10X
images. The full length of the central chamber was imaged in 2-3 tile scans.
Quantification of neurite area
Tiled images were stitched together manually using GIMP image processing
software to obtain a full view of the areas imaged. For each device, the image was
cropped into three sections: the central chamber where the cell bodies were seeded,
and the two outer chambers containing only neurite extensions. Outer chamber images
were cropped to a width of 100µm past the maximum neurite extension for the image
(see Linear assessment of neurite extension below). Cropping was done in ImageJ
by drawing a line between the first and last visible microgroove in the device and using
a custom ImageJ macro to draw a rectangle with a set width: specific maximum
extension for the outer chambers, 2000µm for the middle chamber. To measure axon
area in the outer chambers, images were first processed with ImageJ’s Rolling Ball
Background Subtraction module with a rolling ball radius of 50 µm to correct for
differences in local background fluorescence. Since the orientation of the chambers is
slightly rotated in the images, the cropped images contained blank space outside of the
bounding rectangle. To avoid the darker blank space skewing the background

70
subtraction at the edges of the chamber, pixels with a gray value of 0 were set to the
median gray value for that image prior to background subtraction. Images were then
converted into a binary mask using the triangle algorithm option in ImageJ’s Auto
Threshold module to set a brightness threshold in an unbiased manner. The total area
of pixels above the threshold was then measured and counted as the total area of axon
growth in that chamber.
Machine learning model for quantifying cell bodies
For the central chamber, we used the Trainable Weka Segmentation module in
ImageJ to measure the total cell body area in an unbiased manner. For machine
learning training, we cropped out a 600x600 pixel panel from the central chamber of
each image in our dataset to generate training data. The model was trained to segment
images into three object classes: cell body, axon, or background. To minimize the
computational power required, we selected a subset of the available training features
(various image filters utilized by the software) empirically. An initial classification model
was trained on a small subset of the data with all possible features selected. This model
was then loaded into the Weka Explorer to identify which features were most predictive.
Feature selection was done using the options of ‘CfsSubsetEval -P 1 -E 1’ for Attribute
Evaluator and ‘BestFirst -D 1 -N 5’ for Search Method, respectively. For the final version
of the mode, the following feature options were used: Gaussian blur, Bilateral, Lipschitz,
Kuwahara, Gabor, Sobel filter, Hessian, Minimum, Membrane projections, and
Difference of gaussians. A set of training data was generated by drawing regions of
interest (ROIs) in the Weka software interface for each object class. The model was
trained in an iterative method, where a model was trained with an initial set of ROIs. The

71
labeled images were then visually evaluated and additional ROIs were added to
mislabeled regions and the model was retrained; this was repeated until the panel of
segmented images were accurately labeled based on visual assessment. For each
culture, we took the sum of the total axon area measured from the two outer chambers
and normalized this to the cell body area in the central chamber from our machine
learning segmentation model. To assess the model’s performance, central chamber
images were processed with ImageJ’s Rolling Ball Background Subtraction and AutoThreshold modules (see Quantification of neurite area above) and the total bright
pixel area was measured from each image. We took the ratio of model-segmented cell
body area divided by the total bright pixel area for each sample. We identified one
sample where the segmentation model was severely undercounting cell bodies
(Supplemental Figure 3.3). This outlier was excluded from further analyses. A two-way
ANOVA was performed on normalized neurite outgrowth to test for effects of strain and
days in culture on normalized neurite outgrowth. We used a post-hoc Tukey’s HSD test
to perform multiple pairwise comparisons between strains across all timepoints.
Linear assessment of neurite extension
Image analysis was performed on fluorescent images to quantify neurite
extension into the Xona device axon chambers at each time point for each species. This
was accomplished using a pair of Matlab scripts adapted from similar analysis
described by(Jocher et al. 2018). Fluorescent images were rotated using a Hough
transform to align Xona devices with the image axes. Left and right axon chambers
were manually identified from the comb outward to the edge of each image. Images
were cropped to these regions to generate a pair of rectangles corresponding to the left

72
and right chambers. Cropped images were then thresholded and filtered to remove
noise and identify pixels positive for antibody-stained neurites. Counts were obtained by
quantifying the number of intersections parallel to the comb at 50µm intervals outward
into each chamber. Two consecutive intervals with no neurite crossings defined the
outer boundary of extension in each cropped rectangle. In Figure 3.3A, the sum of
neurite intersections for a given device at a given interval was normalized by the cell
body count (see Machine learning model for quantifying cell bodies above).

73
Supplemental Figures
Supplemental Figure 3.1: Comparing pluripotent stem cells and stem cellderived neuronal transcriptomes across Mus genotypes. Principal component
analysis of transcriptomes from this study and from (Skelly et al. 2020). Each point
reports the transcriptome from a stem cell (PSC) or neuronal culture of the indicated
genotype, in terms of the value of the first (x-axis) or second (x-axis) principal
component. Axis labels report variance across the data set explained by the
respective components.

74
Supplemental Figure 3.2: Design of neuronal culture experiments. Each row
reports the number of replicate differentiations from pluripotent stem cells to neurons
(first column), and the number of replicate cultures in devices imaged for each
timepoint per batch (second column), for the indicated genotype.

75
Supplemental Figure 3.3: Cell body inference in image analysis of stem cellderived neurons. (A) Shown is a histogram, across all images of stem cell-derived
neurons in microfluidic devices in this study, of the proportion of the bright pixels of
the central chamber of a given image (see Figure 2A of the main text) that were
labeled as cell bodies by our segmentation model (see Methods). The red bar
reports an image eliminated for further analysis based on quality control filtering. (B)
Imaging of nuclear staining (DAPI) and III tubulin immunohistochemistry overlaid
with the pixels labeled as cell bodies by the model, on a representative image.

76

77
Supplemental Figure 3.4: Quantification of neurite extension measurements.
(A) Data are as in Figure 3A of the main text except that in a given trace, the point
reports the mean across replicates and error bars report standard error. (B) Data are
as in Figure 3B of the main text except that measurements analyzed were the
maximum neurite length on a given device.

78
Supplemental Figure 3.5: Statistical analyses of neurite extension and growth.
Each row reports the difference in the distance corresponding to neurite extension of
50% of the neuron population (A) or cell body-normalized overall neurite growth (B)
and its significance, after multiple testing correction, between the indicated pair of
genotypes.

79
Chapter 4: Revamping a genome-wide reciprocal hemizygosity screening method
for use in a mammalian cell culture system
Abstract
In the study of natural genetic variation, linking genotype to phenotype using
unbiased screening techniques is a staple of the modern literature. Classic approaches
make use of recombinants in panels of meiotic progeny from crosses, and are not
applicable in systems of distantly related, reproductively isolated species. Previously,
the Brem lab pioneered a high-throughput reciprocal hemizygosity screening method to
dissect the genetics underlying trait variation between partially isolated yeast species. In
this chapter, we established foundational elements for this tool to be used in
mammalian cell culture. We used viral insertional mutagenesis to generate pools of
hemizygote clones in an interspecies hybrid mouse stem cell line, and we catalogued
and quantified growth phenotypes of the viral insertion mutants en masse by
sequencing. These results pave the way towards screens of trait variation across
species boundaries.
Introduction
Geneticists have long sought to understand what drives the trait diversity we see
in the natural world. Powerful linkage and association mapping approaches have been
developed to identify natural DNA sequence variants underlying traits of interest by
tracing co-inheritance across meiotic recombinants from crosses, either in the lab or in
the wild. These are useful for genetic analyses within a species but cannot as a rule be

80
applied between reproductively isolated species (Morgan 1910; Hale et al. 1993;
Sampath et al. 2008; Bertram and Tanzi 2009; Weiss and Brem 2019).
One genetic method that can bridge the gap across species is the reciprocal
hemizygosity (RH) test (Steinmetz et al. 2002; Farahani et al. 2004; Stern 2014). The
classic application (Figure 4.1) was a test of the impact of allelic variation, in a particular
candidate gene, on a trait of interest as it differed between genetic backgrounds. The
latter were mated to generate an F1 hybrid, and the allele of the candidate gene from
each parent in turn was disrupted to create two hemizygotes, each harboring only one
functional copy of the locus. Because the hemizygotes are genetically identical
everywhere else in the genome, any phenotypic difference between them can be
uniquely ascribed to variation at the disrupted gene. Importantly, the RH test can be
used to compare species whose
hybrid offspring are sterile.
The Brem lab previously
developed a genome-wide screening
technique based on the RH test
called RH-seq and successfully
utilized it in yeast (Weiss et al. 2018).
Here insertional transposon
mutagenesis, applied to an
interspecies F1 hybrid, generated a
large pool of hemizygote mutants,
which were then screened in a
Figure 4.1: Reciprocal hemizygosity test. In
this classic genetic test, hemizygote mutants
are generated from an inter-species hybrid. For
a candidate gene, the allele from either parent
(shown in blue and orange) is disrupted in turn
to create a pair of hemizygote mutants. Any
phenotypic differences between the reciprocal
hemizygotes can be attributed to the betweenspecies variation at the disrupted locus.

81
functional assay designed around a growth phenotype of interest. Transposon insertion
sequencing catalogued the mutants and quantified the abundance of reads mapping to
each insertion as a proxy for the abundance of hemizygote mutants, and thus their
fitness in the growth assay. At each gene, these quantitative fitness measurements
were then used to perform the RH test as a comparison between hemizygotes in each
species’ allele in turn.
Figure 4.2: PCR step of library prep amplifies gDNA fragments containing
lentiviral LTR sequence. (A) DNA fragment containing endogenous genomic
sequence (orange), the lentiviral LTR sequence (grey). Custom linkers (blue) are
ligated to the ends after sonication. The linkers contain nonhomologous short (light
blue) and long (dark blue) arms. The forward PCR primer contains a 3’ region
specific to the viral LTR sequence (dark grey) and a 5’ overhang containing the
Illumina P5 adapter sequence. (B) The 3’ region of the reverse primer is specific to
the reverse complement of the long arm portion of the linker. Therefore, the reverse
primer will only prime to PCR product amplified from of the forward primer. This
increases the specificity of the reaction and ensures that only DNA fragments
containing the lentiviral LTR sequence are amplified. The reverse primer also
contains the Illumina P7 adapter sequence (dark green) as a 5’ overhang. (C) The
final PCR product contains the Illumina adapter sequences required for NGS
(green), the 3’ end of the viral LTR sequence (grey) and the gDNA adjacent to the
end of the LTR sequence (orange).
Figure 4.2: PCR step of library prep amplifies gDNA fragments containing
lentiviral LTR sequence. (A) DNA fragment containing endogenous genomic
sequence (orange), the lentiviral LTR sequence (grey). Custom linkers (blue) are
ligated to the ends after sonication. The linkers contain nonhomologous short (light
blue) and long (dark blue) arms. The forward PCR primer contains a 3’ region

82
RH-seq was initially developed for the discovery of genes underlying a
thermotolerance difference between yeast species (Weiss et al. 2018; Abrams et al.
2022). It has also been utilized in C. elegans (Wang et al. 2022), but a full-scale screen
has yet to be carried out in a mammalian system.
Some first steps in adapting RH-seq for mammalian cell culture were taken in the
thesis work of a previous graduate student in the Brem lab (Kang 2022) using a
lentiviral vector to mutagenize hybrid mouse fibroblasts. Here, we expand upon these
methods to advance the RH-seq technique in hybrid mouse stem cells. Ultimately, our
approach may be of use in dissecting the genetic basis of the neurite outgrowth
phenotype of M. m. castaneus neurons we investigated in stem cell-derived models in
Chapter 3.
Results
As a pilot, we established methods to infect an induced pluripotent stem cell line
from a cross between M. m. castaneus CAST Ei/J and the standard laboratory mouse
strain 129S6/SvEv (of mixed M. m. musculus and M. m. domesticus origin), with a HIVbased, puromycin- and BFP-labeled virus. With this approach we generated a pilot set
of 12 independent mutant pools, each involving co-inoculation of the wild-type hybrid
with virus, selection, and collection of ~106 virally infected cells per sample. We isolated
DNA from each for sequencing of the junctions between viral insertions and the
genome: we fragmented the DNA, ligated the fragments to adapters, and used them as
input into a PCR using primers recognizing the virus and the adapter (Figure 4.2). The
results revealed that the majority of the sequencing reads mapped to identical positions
in the genome and were shared across all 12 replicates (Supplemental Figure 4.1A).

83
Further investigation of these sites revealed that the forward primer used in the PCR
step of this initial library prep protocol was annealing to endogenous genomic
sequences in the mouse genome rather than the insertions of our marked HIV virus.
Based on this, we designed and tested an alternative primer for the viral
insertions (Supplemental Figure 4.1B). We re-sequenced DNAs from our 12 pilot
mutagenesis cultures of hybrid stem cells and detected no amplification of endogenous
genomic sequences in these profiles; rather, our sequencing output successfully
Figure 4.3: Lentiviral insertions are dispersed across the genome. From a
small-scale lentiviral mutagenesis and sequencing experiment, we detected 263
unique insertions. Of these, 124 mapped to the M. m. musculus genome, with the
other 139 mapping to the M. m. castaneus genome. Insertions with only a single
corresponding read were omitted.

84
reported viral insertions across the hybrid genome (Figure 4.3) with no reported
insertions common across the independent infections, as expected (data not shown).
We next investigated the extent to which our culture and sequencing could
provide reproducible estimates of hemizygote abundance in pools. Focusing on one
pool of ~1000 hemizygote mutants in the hybrid background, we expanded this pool for
seven days, split it into two cultures, and sequenced the viral insertions from each.
Since these were split from the same sample and cultured under identical conditions,
Figure 4.4: Comparing read abundances of lentiviral insertions after cell
culture expansion. An initial pool of mutagenized stem cells was split between two
cell culture plates and expanded further for 7 days. Sequencing libraries were
generated from gDNA extracted from each sample. Read counts were tabulated for
each unique insertion and normalized to the total mapped reads for each sample.
173 insertions which were shared between both samples are plotted. Insertions
close to the diagonal line have consistent abundance between samples.

85
we expected the individual mutants to grow and divide at similar rates – which should
be reflected in the sequencing data as comparable read abundances between
replicates. After analyzing the data, we found 173 mapped insertions that were shared
between replicates with reasonable agreement between abundances (Figure 4.4).
These data provide a first validation of the mutagenesis, culture, and sequencing
techniques established here and serve as a foundation for continued methods
development beyond the scope of this thesis.
Discussion
Genetic variation between species holds a wealth of knowledge that is largely
inaccessible with conventional unbiased screening techniques. RH-seq provides us a
tool to tap into this potential. Here, we expand on previous work from the lab towards
the goal of adapting this technique for use in mammalian cell culture.
The results presented here demonstrate that we have a system for culturing and
mutagenizing stem cells that is viable for performing RH-seq. This opens the door for
unbiased screening of potential traits in stem cells related to the regulatory program in
translation genes discussed in Chapter 2. With the possibility of differentiating
hemizygote stem cells into other cell types, this also paves the way for performing RHseq screens in post-mitotic cell types, such as neurons.
Methods
Lentivirus production
Production of our lentiviral vector was performed as described by Kang (2022).
Lentivirus was produced by calcium phosphate transfection of HEK 293T cells cultured

86
in DMEM (Gibco cat. #21-063-029) supplemented with 10% FBS (Genesee Scientific
cat. #25-514H) and 1% pen-strep (Corning cat. #30002CI). The lentiviral backbone was
a third-generation modified HIV vector containing BFP and puromycin resistance
markers driven by a Ef1α promoter, and standard central polypurine tract-central
termination sequences (cPPT-CTS). The pMDLg/pRRE packaging and pCMV-VSV-G
envelope vectors were delivered through separate plasmids. The lentiviral plasmids
were provided by Dr. Marius Walter at the Buck Institute. The day before transfection,
HEK 293T cells were seeded in 15 cm cell culture dishes at an appropriate density to
reach 50-70% confluence on the day of transfection. A 2x HeBS solution was prepared
by dissolving NaCl (Sigma-Aldrich cat. #S7653), HEPES (Sigma-Aldrich cat. #H7523),
and Na2HPO4 (Sigma- Aldrich cat. #S7907) into water, and adjusting the pH to 7.0 with
NaOH (Thomas Scientific cat. #C753N15) solution. Prior to transfection, the cell culture
media was replaced with media supplemented with 25 μM chloroquine (Thermo
Scientific cat. #C868U90). 40 μg of the lentiviral backbone plasmid, 15 μg of the
envelope plasmid, and 20 μg of the packaging vector were mixed into 500 μL of water
and added to 500 μL of 0.5 M CaCl2 (Sigma-Aldrich cat. #C5080). This solution was
then mixed dropwise into 1 mL of 2x HeBS while vortexing and incubated at room
temperature for 20-30 minutes. An hour after chloroquine supplementation, the resulting
precipitate solution was added dropwise onto HEK 293T cells. The cells were then
placed in a 37°C humidified incubator with 10% CO2 for 8-10 hours before changing to
fresh media. Two days after transfection, the viral supernatant was collected and
passed through a 0.45 μm filter before being flash-frozen in dry ice. The frozen
supernatant was stored at -80°C.

87
Hybrid cell line
The hybrid stem cell line CAS129 was provided by Joost Gribnau’s lab. This cell
line was derived previously from a crossing of a male 129S6/SvEv mouse and a female
CAST/EiJ mouse (Marks et al. 2015).
mPSC Culture
The mPSCs were cultured essentially as described previously (Greder et al.
2012; Sajini et al. 2012; Wu et al. 2012). Cells were cultured on a feeder layer of
irradiated mouse embryonic fibroblasts (iMEFs) (R&D Systems, Minneapolis, MN,
USA). iMEFs were plated on 0.1% gelatin (Millipore, Burlington, MA, USA) -coated 6-
well tissue culture plates (Thermo Fisher Scientific, Waltham, MA, USA) at least one
day prior to plating stem cells. iMEFs were thawed, and maintained in iMEF medium:
High Glucose Dulbecco’s Modified Eagle Medium (DMEM) (Sigma-Aldrich, St. Louis,
MO, USA), 10% fetal bovine serum (FBS) (HyClone, Logan, UT, USA), 1x MEM nonessential amino acids (NEAA) (Sigma-Aldrich), 1x penicillin/streptomycin (Corning,
Corning, NY, USA), and 1x sodium pyruvate (Sigma-Aldrich). miPSC -/- medium
consisted of knockout DMEM (Thermo Fisher Scientific), 10% knockout serum
replacement (KSR) (Thermo Fisher Scientific), 10% FBS, 1x glutaMAX (Thermo Fisher
Scientific), and 1x MEM NEAA. miPSC +/+ medium was made as miPSC -/-
supplemented with 0.02% LIF (Millipore) and 1x 2-mercaptoethanol (BME) (Thermo
Fisher Scientific). Hybrid Cas129 were cultured in 2i medium (miPSC +/+ medium
supplemented with 3μM CHIR99021 (Sigma-Adrich) and 1μM PD0325901 (SigmaAldrich)). Stem cells were incubated at 37°C in 5% CO2 with daily media changes.

88
The mPSCs were passaged every 2 to 3 days depending on colony size and
density. For passaging, cells were washed with phosphate buffered saline (PBS) +/+
(Thermo Fisher Scientific), detached with 0.25% trypsin (Sigma-Aldrich), incubated at
37°C for 3 to 4 minutes, then neutralized with an equal volume of miPSC -/- medium.
Cells were further dissociated with a P1000 micropipette, after which they were
collected by centrifugation at 329 × g for 4 minutes. The supernatant was aspirated, and
the cells resuspended in complete medium. The resuspended cells were seeded onto a
fresh feeder layer for colony formation.
Hybrid Mutagenesis
The mutagenesis protocol was adapted from Kang (2022). The hybrid CAS129
mESCs were subjected to untargeted viral mutagenesis. mESCs were passaged onto
puromycin-resistant iMEFs (Thermo Fisher Scientific), which were plated a day prior in
iMEF medium. mESCs were passaged in 2i medium supplemented with varying
concentrations of our lentivirus and 4μg/ml polybrene (Sigma-Aldrich). The day of the
passage was denoted day 1. On day 2 the medium was replaced with 2i medium with
1μg/ml puromycin (Millipore), allowing for selection. mESCs were cultured in puromycin
medium until day 7 with daily media changes. Surviving colonies on day 7 were further
expanded for cryopreservation or differentiation. To limit the number of viral insertion
events per cell, varying concentrations of the virus were tested to find the minimum
concentration that enables puromycin selection.
An initial mutagenesis experiment was conducted by treating 12x wells of hybrid
stem cells with our lentivirus. These cells were expanded under puromycin selection for
7 days before gDNA was extracted for library prep.

89
In a follow-up experiment, we performed a small-scale mutagenesis and
sequencing experiment to assess consistency at the final data analysis step between
technical replicates. An initial mutant pool was generated from hybrid stem cells and
expanded for 7 days under puromycin selection. This pool was split into three aliquots
and frozen to simulate how cells would be treated under a full-scale screen. One of
these aliquots was thawed and cultured for 7 days in one well of a 6-well plate. Cells
were then passaged equally across two 6-well plates and expanded for an additional 7
days. Then, cells from each plate were pooled and processed as two replicates for
library prep and sequences.
NGS library preparation and DNA sequencing
Library preparation was adapted from (Serrao et al. 2016) Cells were lysed in
digestion buffer (aqueous solution containing 100 mM NaCl l (Sigma-Aldrich cat.
#S7653), 10 mM Tris (Corning cat. #45031CM), 25 mM EDTA (Corning cat. #43034C),
0.5% SDS S (Bio-Rad cat. #1610416), 0.1 mg/mL proteinase K (NEB cat. #P8107S) at
a volume of 1 mL per 1 million cells, incubated at 50°C overnight. Genomic DNA
(gDNA) was extracted using an equal volume of phenol-chloroform-isoamyl alcohol
(Sigma-Aldrich cat. #P2069) followed by ethanol precipitation. gDNA was then
fragmented into ~300 bp pieces by sonication (Covaris S2 Ultrasonicator; settings: Peak
power = 175, Duty factor = 20, Cycles/burst = 200, treatment = 80s).
Sonicated gDNA was then purified with a two-step (right-side size selection)
magnetic bead purification using Ampure XP beads (Beckman-Coulter Product No:
A63881). DNA was first treated using a 0.85x volume of beads. Beads were pulled
down using a magnetic tube rack and supernatant was transferred to a fresh tube.

90
Supernatant was then treated with 0.55x volume (relative to initial sample volume) of
beads. After magnetic pulldown, supernatant was discarded, the pellet was washed with
80% ethanol, and purified DNA was eluted with ultrapure water.
Fragmented DNA was then treated with the NEBNext® End Repair Module (N
EB Catalog # E6050), yielding blunt-end DNA. DNA was purified with a one-step (leftside size selection) bead cleanup using a 1.4x volume of Ampure XP beads. Custom
adapters (Supplemental Table 4.1) were then ligated to the DNA fragments using T4
DNA ligase (NEB cat. #M0202S) at 12°C overnight, followed by one-step purification
with 1.0x volume of Ampure XP beads.
To generate the sequencing library, the adapter-ligated DNA fragments were then
used as template in a PCR reaction. Forward primers were designed to anneal to the
lentiviral long terminal repeat (LTR) sequence, while the reverse primer contained a
sequence complementary to the custom adapter sequence (see Supplemental Table
4.1 for primer sequences, and Figure 4.2 for a schematic). Both primers contained the
standard Illumina P5 and P7 priming sites and chip adapter sequences. PCR was run
using Advantage® 2 Polymerase Mix (Takara Bio cat. # 639201), final concentrations of
1.5 μM LTR primer, 0.3 μM linker primer, and 50 μM dNTPs. PCR cycle settings were
94°C for 2 minutes, followed by 25 cycles of 94°C for 15s, 55°C for 30s, 68°C for 45s,
followed by a final 68°C for 10 minutes.
PCR product was purified with two one-step bead cleanups using 0.9x followed by
0.85x volumes of Ampure XP beads. The purified sample was sequenced by UC
Berkeley QB3 Genomics core on the Illumina MiSeq with the v2 reagent kit, generating
1 million 300bp single-end reads per run.

91
DNA at each step of the library prep was quantified using a QubitTM 4 fluorometer
(Thermo Fisher cat. #Q33238) with the QubitTM 1x dsDNA High Sensitivity kit (Thermo
Fisher cat. #Q33231). Quality and fragment size of the sonicated gDNA and final
libraries were assessed using the Agilent 2100 Bioanalyzer Instrument by the UC
Berkeley QB3 core.
Data analysis
Sequencing data was analyzed using custom Python scripts. First, reads were
filtered to select only those containing the last 20 bp at the 3’ end of the lentiviral LTR.
One mismatch to the LTR sequence was allowed, to account for sequencing errors. The
LTR sequence was then trimmed off, leaving only the adjacent genomic sequence. The
trimmed reads were then mapped to a concatenation of the 129S1/SvImJ and CAST/EiJ
reference genomes using pblat v2.5 (Wang and Kong 2019) with the flags "-minIdentity
=100", "-tileSize=12", and "-out=blast8". Mapped reads were filtered to include only
those mapping to a single locus. Finally, the number of reads per unique insertion was
tabulated, reflecting the abundance of cells with the corresponding genotype in the
initial sequencing pool.

92
Supplemental Figures
Supplemental Figure 4.1: Amplification of endogenous loci in initial
mutagenesis experiments. Sequencing libraries were generated from 12
independent pools of mutagenized hybrid stem cells. (A) The 10 most abundant
putative across the dataset. Nine of these inserts contained reads in all 12 replicates.
Reads mapping to these 10 loci comprised an average of ~51% of the total mapped
reads. Read counts were normalized to the total number of mapped reads for each
library, and displayed as a percentage of the total reads per library. (B) Sequence
alignments of putative insertion sites with the lentiviral LTR sequence. Multiple
sequence alignment (ClustalW) of the 100bp directly upstream of three putative
insertion sites with the LTR sequence of HIV used in our lentiviral vector. The
complementary region from the PCR primer ‘thiv4’ used in library prep is included for
reference. High homology of the primer to the genomic sequences suggests nonspecific priming to these regions.
Supplemental Figure 4.1: Amplification of endogenous loci in initial
mutagenesis experiments. Sequencing libraries were generated from 12
independent pools of mutagenized hybrid stem cells. (A) The 10 most abundant
A
B

93
Supplemental Figure 4.2: Primer and linker sequences used for library prep.
Linker sequences are ligated onto DNA fragments after sonication. LTR primers
contain regions complementary to the HIV LTR region used in our lentiviral vector
(red) and the p5 Illumina adapter (light green) required for sequencing. Black Ns are
replaced with custom 6-bp indices for multiplexing. The linker primer contains an
identical region (blue) to the linker long arm, and will therefore only prime to the
reverse complement. Linker primer also contains the p7 Illumina adapter (green).
Supplemental Figure 4.2: Primer and linker sequences used for library prep.
Linker sequences are ligated onto DNA fragments after sonication. LTR primers
contain regions complementary to the HIV LTR region used in our lentiviral vector
(red) and the p5 Illumina adapter (light green) required for sequencing. Black Ns are
replaced with custom 6-bp indices for multiplexing. The linker primer contains an
identical region (blue) to the linker long arm, and will therefore only prime to the
reverse complement. Linker primer also contains the p7 Illumina adapter (green).

94
Conclusions
High expression of translation genes in M. m. castaneus stem cells is partly
driven by directional cis-regulatory evolution
We detected sequence signatures of selection between mouse subspecies in a
cohort of translation genes that also exhibits directional, polygenic cis-regulatory
variation in stem cells. These results represent a proof of concept for the validation of
this approach in mammals. A sequence-based analysis using population genomics data
identified signatures of selection in the binding sites of candidate transcription factors
targeting these genes. Expression profiles from knockouts of two of these candidates,
Eed and Ehmt2, provide further evidence that these regulate expression of translation
genes in stem cells. Future work will explore the phenotypic relevance of translation
gene divergence across Mus, and its role in the slew of traits that distinguish M. m.
castaneus from the rest of its genus.
M. m. castaneus neurons grow longer neurites in the absence of injury
We expand on the previously known phenotype of axon regeneration in M. m.
castaneus neurons (Omura et al. 2015), reporting higher neurite growth than their sister
species without any injury stimulus. The cell culture system we utilized is completely in
vivo through the use of stem cell-derived neurons and will serve for further study of this
trait. Our finding of a dominant axon growth phenotype in M. m. castaneus hybrid
neurons shows strong potential that the underlying mechanisms could harnessed for
therapeutic use. The methods developed here set the stage for exciting future studies in
these traits.

95
References
Abrams MB, Chuong JN, AlZaben F, Dubin CA, Skerker JM, Brem RB. 2022. Barcoded
reciprocal hemizygosity analysis via sequencing illuminates the complex genetic basis of
yeast thermotolerance. G3 (Bethesda). 12(2):jkab412. doi:10.1093/g3journal/jkab412.
Abreu R de S, Penalva LO, Marcotte EM, Vogel C. 2009. Global signatures of protein and
mRNA expression levels. Mol Biosyst. 5(12):1512–1526. doi:10.1039/b908315d.
Advani VM, Ivanov P. 2019. Translational control under stress: reshaping the translatome.
Bioessays. 41(5):e1900009. doi:10.1002/bies.201900009.
Agoglia RM, Sun D, Birey F, Yoon S-J, Miura Y, Sabatini K, Pașca SP, Fraser HB. 2021.
Primate cell fusion disentangles gene regulatory divergence in neurodevelopment.
Nature. 592(7854):421–427. doi:10.1038/s41586-021-03343-3.
Anisimova AS, Meerson MB, Gerashchenko MV, Kulakovskiy IV, Dmitriev SE, Gladyshev VN.
2020. Multifaceted deregulation of gene expression and protein synthesis with age.
Proceedings of the National Academy of Sciences. 117(27):15581–15590.
doi:10.1073/pnas.2001788117.
Auclair G, Borgel J, Sanz LA, Vallet J, Guibert S, Dumas M, Cavelier P, Girardot M, Forné T,
Feil R, et al. 2016. EHMT2 directs DNA methylation for efficient gene silencing in mouse
embryos. Genome Res. 26(2):192–202. doi:10.1101/gr.198291.115.
Aydin S, Pham DT, Zhang T, Keele GR, Skelly DA, Paulo JA, Pankratz M, Choi T, Gygi SP,
Reinholdt LG, et al. 2023. Genetic dissection of the pluripotent proteome through multiomics data integration. Cell Genom. 3(4):100283.
doi:https://doi.org/10.1016/j.xgen.2023.100283.
Ayyadevara S, Ayyadevara R, Hou S, Thaden JJ, Reis RJS. 2001. Genetic Mapping of
Quantitative Trait Loci Governing Longevity of Caenorhabditis elegans in RecombinantInbred Progeny of a Bergerac-BO × RC301 Interstrain Cross. Genetics. 157(2):655–666.
doi:10.1093/genetics/157.2.655.
Beamer WG, Shultz KL, Churchill GA, Frankel WN, Baylink DJ, Rosen CJ, Donahue LR. 1999.
Quantitative trait loci for bone density in C57BL/6J and CAST/EiJ inbred mice.
Mammalian Genome. 10(11):1043–1049. doi:10.1007/s003359901159.
Beirowski B, Adalbert R, Wagner D, Grumme DS, Addicks K, Ribchester RR, Coleman MP.
2005. The progressive nature of Wallerian degeneration in wild-type and slow Wallerian
degeneration (WldS) nerves. BMC Neuroscience. 6(1):6. doi:10.1186/1471-2202-6-6.
Berg JJ, Harpak A, Sinnott-Armstrong N, Joergensen AM, Mostafavi H, Field Y, Boyle EA,
Zhang X, Racimo F, Pritchard JK, et al. 2019. Reduced signal for polygenic adaptation
of height in UK Biobank. Elife. 8:e39725. doi:10.7554/eLife.39725.
Bertram L, Tanzi RE. 2009. Genome-wide association studies in Alzheimer’s disease. Human
Molecular Genetics. 18(R2):R137–R145. doi:10.1093/hmg/ddp406.

96
Blagden SP, Willis AE. 2011. The biological and therapeutic relevance of mRNA translation in
cancer. Nat Rev Clin Oncol. 8(5):280–291. doi:10.1038/nrclinonc.2011.16.
Blum JA, Klemm S, Shadrach JL, Guttenplan KA, Nakayama L, Kathiria A, Hoang PT, Gautier
O, Kaltschmidt JA, Greenleaf WJ, et al. 2021. Single-cell transcriptomic analysis of the
adult mouse spinal cord reveals molecular diversity of autonomic and skeletal motor
neurons. Nat Neurosci. 24(4):572–583. doi:10.1038/s41593-020-00795-0.
Bonilla IE, Tanabe K, Strittmatter SM. 2002. Small Proline-Rich Repeat Protein 1A Is Expressed
by Axotomized Neurons and Promotes Axonal Outgrowth. J Neurosci. 22(4):1303–1315.
doi:10.1523/JNEUROSCI.22-04-01303.2002.
Boroviak T, Loos R, Bertone P, Smith A, Nichols J. 2014. The ability of inner-cell-mass cells to
self-renew as embryonic stem cells is acquired following epiblast specification. Nat Cell
Biol. 16(6):513–525. doi:10.1038/ncb2965.
Brown SJ, Cole MD, Erives AJ. 2008. Evolution of the holozoan ribosome biogenesis regulon.
BMC Genomics. 9(1):442. doi:10.1186/1471-2164-9-442.
Buchwalter A, Hetzer MW. 2017. Nucleolar expansion and elevated protein translation in
premature aging. Nat Commun. 8(1):328. doi:10.1038/s41467-017-00322-z.
Bullard JH, Mostovoy Y, Dudoit S, Brem RB. 2010. Polygenic and directional regulatory
evolution across pathways in Saccharomyces. Proceedings of the National Academy of
Sciences. 107(11):5058–5063. doi:10.1073/pnas.0912959107.
Byers C, Spruce C, Fortin HJ, Hartig EI, Czechanski A, Munger SC, Reinholdt LG, Skelly DA,
Baker CL. 2022. Genetic control of the pluripotency epigenome determines
differentiation bias in mouse embryonic stem cells. EMBO J. 41(2):e109445.
doi:10.15252/embj.2021109445.
Calkins DJ. 2013. Age-Related Changes in the Visual Pathways: Blame It on the Axon. Invest
Ophthalmol Vis Sci. 54(14):ORSF37–ORSF41. doi:10.1167/iovs.13-12784.
Carbon S, Ireland A, Mungall CJ, Shu S, Marshall B, Lewis S, the AmiGO Hub, the Web
Presence Working Group. 2009. AmiGO: online access to ontology and annotation data.
Bioinformatics. 25(2):288–289. doi:10.1093/bioinformatics/btn615.
Catlett TS, Onesto MM, McCann AJ, Rempel SK, Glass J, Franz DN, Gómez TM. 2021. RHOA
signaling defects result in impaired axon guidance in iPSC-derived neurons from
patients with tuberous sclerosis complex. Nat Commun. 12(1):2589.
doi:10.1038/s41467-021-22770-4.
Chapman VM, Ruddle FH. 1972. Glutamate ocaloacetate transaminase (GOT) genetics in the
mouse: polymorphism of GOT-1. Genetics. 70(2):299–305.
doi:10.1093/genetics/70.2.299.
Chappell GA, Israel JW, Simon JM, Pott S, Safi A, Eklund K, Sexton KG, Bodnar W, Lieb JD,
Crawford GE, et al. 2017. Variation in DNA-Damage Responses to an Inhalational
Carcinogen (1,3-Butadiene) in Relation to Strain-Specific Differences in Chromatin

97
Accessibility and Gene Transcription Profiles in C57BL/6J and CAST/EiJ Mice.
Environmental Health Perspectives. 125(10):107006. doi:10.1289/EHP1937.
Cho Y, Sloutsky R, Naegle KM, Cavalli V. 2013. Injury-induced HDAC5 nuclear export is
essential for axon regeneration. Cell. 155(4):894–908. doi:10.1016/j.cell.2013.10.004.
Chou M-Y, Appan D, Chang K-W, Chou C-H, Lin C-Y, Gau SS-F, Huang H-S. 2022. Mouse
hybrid genome mediates diverse brain phenotypes with the specificity of reciprocal
crosses. FASEB J. 36(3):e22232. doi:10.1096/fj.202101624R.
Coleman MP, Conforti L, Buckmaster EA, Tarlton A, Ewing RM, Brown MC, Lyon MF, Perry VH.
1998. An 85-kb tandem triplication in the slow Wallerian degeneration (Wlds) mouse.
Proc Natl Acad Sci U S A. 95(17):9985–9990.
Coleman MP, Höke A. 2020. Programmed axon degeneration: from mouse to mechanism to
medicine. Nature Reviews Neuroscience. 21(4):183–196. doi:10.1038/s41583-020-
0269-3.
Conforti L, Fang G, Beirowski B, Wang MS, Sorci L, Asress S, Adalbert R, Silva A, Bridge K,
Huang XP, et al. 2007. NAD(+) and axon degeneration revisited: Nmnat1 cannot
substitute for Wld(S) to delay Wallerian degeneration. Cell Death Differ. 14(1):116–127.
doi:10.1038/sj.cdd.4401944.
Conforti L, Gilley J, Coleman MP. 2014. Wallerian degeneration: an emerging axon death
pathway linking injury and disease. Nature Reviews Neuroscience. 15(6):394–409.
doi:10.1038/nrn3680.
Cox LJ, Hengst U, Gurskaya Nadya, Lukyanov KA, Jaffrey SR. 2008. Intra-axonal translation
and retrograde trafficking of CREB promotes neuronal survival. Nat Cell Biol. 10(2):149–
159. doi:10.1038/ncb1677.
Cunningham F, Allen JE, Allen J, Alvarez-Jarreta J, Amode MR, Armean IM, Austine-Orimoloye
O, Azov AG, Barnes I, Bennett R, et al. 2022. Ensembl 2022. Nucleic Acids Research.
50(D1):D988–D995. doi:10.1093/nar/gkab1049.
Curran SP, Ruvkun G. 2007. Lifespan regulation by evolutionarily conserved genes essential for
viability. PLoS Genet. 3(4):e56. doi:10.1371/journal.pgen.0030056.
Dadon-Nachum M, Melamed E, Offen D. 2011. The “Dying-Back” Phenomenon of Motor
Neurons in ALS. J Mol Neurosci. 43(3):470–477. doi:10.1007/s12031-010-9467-1.
Dale RK, Pedersen BS, Quinlan AR. 2011. Pybedtools: a flexible Python library for manipulating
genomic datasets and annotations. Bioinformatics. 27(24):3423–3424.
doi:10.1093/bioinformatics/btr539.
Danecek P, Bonfield JK, Liddle J, Marshall J, Ohan V, Pollard MO, Whitwham A, Keane T,
McCarthy SA, Davies RM, et al. 2021. Twelve years of SAMtools and BCFtools.
GigaScience. 10(2):giab008. doi:10.1093/gigascience/giab008.
Darwin C. 1872. On the Origin of Species, 6th Edition. [accessed 2024 Feb 20].
http://libarch.nmu.org.ua/handle/GenofondUA/5143.

98
Deelen J, Evans DS, Arking DE, Tesi N, Nygaard M, Liu X, Wojczynski MK, Biggs ML, van der
Spek A, Atzmon G, et al. 2019. A meta-analysis of genome-wide association studies
identifies multiple longevity genes. Nat Commun. 10(1):3669. doi:10.1038/s41467-019-
11558-2.
Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, Batut P, Chaisson M, Gingeras
TR. 2013. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 29(1):15–21.
doi:10.1093/bioinformatics/bts635.
Durinck S, Spellman PT, Birney E, Huber W. 2009. Mapping identifiers for the integration of
genomic datasets with the R/Bioconductor package biomaRt. Nat Protoc. 4(8):1184–
1191. doi:10.1038/nprot.2009.97.
Dutta R, Trapp BD. 2007. Pathogenesis of axonal and neuronal damage in multiple sclerosis.
Neurology. 68(Issue 22, Supplement 3):S22–S31.
doi:10.1212/01.wnl.0000275229.13012.32.
Easley CA, Ben-Yehudah A, Redinger CJ, Oliver SL, Varum ST, Eisinger VM, Carlisle DL,
Donovan PJ, Schatten GP. 2010. mTOR-Mediated Activation of p70 S6K Induces
Differentiation of Pluripotent Human Embryonic Stem Cells. Cellular Reprogramming.
12(3):263–273. doi:10.1089/cell.2010.0011.
van Erp S, van Berkel AA, Feenstra EM, Sahoo PK, Wagstaff LJ, Twiss JL, Fawcett JW, Eva R,
ffrench-Constant C. 2021. Age-related loss of axonal regeneration is reflected by the
level of local translation. Exp Neurol. 339:113594. doi:10.1016/j.expneurol.2020.113594.
Farahani P, Fisler JS, Wong H, Diament AL, Yi N, Warden CH. 2004. Reciprocal Hemizygosity
Analysis of Mouse Hepatic Lipase Reveals Influence on Obesity. Obesity Research.
12(2):292–305. doi:10.1038/oby.2004.37.
Fawcett JW. 2020. The Struggle to Make CNS Axons Regenerate: Why Has It Been so
Difficult? Neurochem Res. 45(1):144–158. doi:10.1007/s11064-019-02844-y.
Ferguson J, Chang J. 2020. An empirical Bayesian ranking method, with applications to high
throughput biology. Bioinformatics. 36(1):177–185. doi:10.1093/bioinformatics/btz471.
Filbin MT. 2003. Myelin-associated inhibitors of axonal regeneration in the adult mammalian
CNS. Nat Rev Neurosci. 4(9):703–713. doi:10.1038/nrn1195.
Fitch MT, Silver J. 2008. CNS injury, glial scars, and inflammation: Inhibitory extracellular
matrices and regeneration failure. Experimental Neurology. 209(2):294–301.
doi:10.1016/j.expneurol.2007.05.014.
Fraser HB, Babak T, Tsang J, Zhou Y, Zhang B, Mehrabian M, Schadt EE. 2011. Systematic
detection of polygenic cis-regulatory evolution. PLoS Genet. 7(3):e1002023.
doi:10.1371/journal.pgen.1002023.
Fraser HB, Levy S, Chavan A, Shah HB, Perez JC, Zhou Y, Siegal ML, Sinha H. 2012.
Polygenic cis-regulatory adaptation in the evolution of yeast pathogenicity. Genome Res.
22(10):1930–1939. doi:10.1101/gr.134080.111.

99
Fraser HB, Moses AM, Schadt EE. 2010. Evidence for widespread adaptive evolution of gene
expression in budding yeast. Proceedings of the National Academy of Sciences.
107(7):2977–2982. doi:10.1073/pnas.0912245107.
Frazee AC, Jaffe AE, Langmead B, Leek JT. 2015. Polyester: simulating RNA-seq datasets with
differential transcript expression. Bioinformatics. 31(17):2778–2784.
doi:10.1093/bioinformatics/btv272.
Frazer KA, Eskin E, Kang HM, Bogue MA, Hinds DA, Beilharz EJ, Gupta RV, Montgomery J,
Morenzoni MM, Nilsen GB, et al. 2007. A sequence-based variation map of 8.27 million
SNPs in inbred mouse strains. Nature. 448(7157):1050–1053. doi:10.1038/nature06067.
French JE, Gatti DM, Morgan DL, Kissling GE, Shockley KR, Knudsen GA, Shepard KG, Price
HC, King D, Witt KL, et al. 2015. Diversity Outbred Mice Identify Population-Based
Exposure Thresholds and Genetic Factors that Influence Benzene-Induced Genotoxicity.
Environ Health Perspect. 123(3):237–245. doi:10.1289/ehp.1408202.
Fujiwara K, Kawai Y, Moriwaki K, Takada T, Shiroishi T, Saitou N, Suzuki H, Osada N. 2021.
Insights into Mus musculus population structure across Eurasia revealed by wholegenome analysis. Genomics. [accessed 2024 Feb 20].
http://biorxiv.org/lookup/doi/10.1101/2021.02.05.429881.
Gabut M, Bourdelais F, Durand S. 2020. Ribosome and Translational Control in Stem Cells.
Cells. 9(2):497. doi:10.3390/cells9020497.
Garbutt TA, Konneker TI, Konganti K, Hillhouse AE, Swift-Haire F, Jones A, Phelps D, Aylor DL,
Threadgill DW. 2018. Permissiveness to form pluripotent stem cells may be an
evolutionarily derived characteristic in Mus musculus. Sci Rep. 8(1):14706.
doi:10.1038/s41598-018-32116-8.
Geisler S, Doan RA, Cheng GC, Cetinkaya-Fisgin A, Huang SX, Höke A, Milbrandt J, DiAntonio
A. 2019. Vincristine and bortezomib use distinct upstream mechanisms to activate a
common SARM1-dependent axon degeneration program. JCI Insight. 4(17):e129920,
129920. doi:10.1172/jci.insight.129920.
Geisler S, Doan RA, Strickland A, Huang X, Milbrandt J, DiAntonio A. 2016. Prevention of
vincristine-induced peripheral neuropathy by genetic deletion of SARM1 in mice. Brain.
139(12):3092–3108. doi:10.1093/brain/aww251.
Gerdts J, Summers DW, Sasaki Y, DiAntonio A, Milbrandt J. 2013. Sarm1-mediated axon
degeneration requires both SAM and TIR interactions. J Neurosci. 33(33):13569–13580.
doi:10.1523/JNEUROSCI.1197-13.2013.
Gibbs RA. 2020. The Human Genome Project changed everything. Nat Rev Genet. 21(10):575–
576. doi:10.1038/s41576-020-0275-3.
Gilley J, Coleman MP. 2010. Endogenous Nmnat2 is an essential survival factor for
maintenance of healthy axons. PLoS Biol. 8(1):e1000300.
doi:10.1371/journal.pbio.1000300.

100
Gokhman D, Agoglia RM, Kinnebrew M, Gordon W, Sun D, Bajpai VK, Naqvi S, Chen Coral,
Chan A, Chen Chider, et al. 2021. Human-chimpanzee fused cells reveal cis-regulatory
divergence underlying skeletal evolution. Nat Genet. 53(4):467–476.
doi:10.1038/s41588-021-00804-3.
Greder LV, Gupta S, Li S, Abedin MJ, Sajini A, Segal Y, Slack JMW, Dutton JR. 2012. Analysis
of endogenous Oct4 activation during induced pluripotent stem cell reprogramming using
an inducible Oct4 lineage label. Stem Cells. 30(11):2596–2601. doi:10.1002/stem.1216.
Guerra San Juan I, Nash LA, Smith KS, Leyton-Jaimes MF, Qian M, Klim JR, Limone F, Dorr
AB, Couto A, Pintacuda G, et al. 2022. Loss of mouse Stmn2 function causes motor
neuropathy. Neuron. 110(10):1671-1688.e6. doi:10.1016/j.neuron.2022.02.011.
Hale DW, Washburn LL, Eicher EM. 1993. Meiotic abnormalities in hybrid mice of the C57BL/6J
x Mus spretus cross suggest a cytogenetic basis for Haldane’s rule of hybrid sterility.
Cytogenet Cell Genet. 63(4):221–234. doi:10.1159/000133539.
Hanz S, Fainzilber M. 2006. Retrograde signaling in injured nerve – the axon reaction revisited.
Journal of Neurochemistry. 99(1):13–19. doi:10.1111/j.1471-4159.2006.04089.x.
Hanz S, Perlson E, Willis D, Zheng J-Q, Massarwa R, Huerta JJ, Koltzenburg M, Kohler M, vanMinnen J, Twiss JL, et al. 2003. Axoplasmic importins enable retrograde injury signaling
in lesioned nerve. Neuron. 40(6):1095–1104. doi:10.1016/s0896-6273(03)00770-0.
Harper JM, Krishnan C, Darman JS, Deshpande DM, Peck S, Shats I, Backovic S, Rothstein
JD, Kerr DA. 2004. Axonal growth of embryonic stem cell-derived motoneurons in vitro
and in motoneuron-injured adult rats. Proc Natl Acad Sci U S A. 101(18):7123–7128.
doi:10.1073/pnas.0401103101.
Harr B, Karakoc E, Neme R, Teschke M, Pfeifle C, Pezer Ž, Babiker H, Linnenbrink M, Montero
I, Scavetta R, et al. 2016. Genomic resources for wild populations of the house mouse,
Mus musculus and its close relative Mus spretus. Sci Data. 3(1):160075.
doi:10.1038/sdata.2016.75.
Heber-Katz E, Chen P, Dvm LC, Zhang X-M, Troutman S, Blankenhorn EP. 2004. Regeneration
in MRL mice: further genetic loci controlling the ear hole closure trait using MRL and
M.m. Castaneus mice. Wound Repair and Regeneration. 12(3):384–392.
doi:10.1111/j.1067-1927.2004.012308.x.
Hinman JD. 2014. The back and forth of axonal injury and repair after stroke. Curr Opin Neurol.
27(6):615–623. doi:10.1097/WCO.0000000000000149.
Ho R, Sances S, Gowing G, Amoroso MW, O’Rourke JG, Sahabian A, Wichterle H, Baloh RH,
Sareen D, Svendsen CN. 2016. ALS disrupts spinal motor neuron maturation and aging
pathways within gene co-expression networks. Nat Neurosci. 19(9):1256–1267.
doi:10.1038/nn.4345.
Hogues H, Lavoie H, Sellam A, Mangos M, Roemer T, Purisima E, Nantel A, Whiteway M.
2008. Transcription Factor Substitution during the Evolution of Fungal Ribosome
Regulation. Mol Cell. 29(5):552–562. doi:10.1016/j.molcel.2008.02.006.

101
Hoßfeld U, Jacobsen H-J, Plass C, Brors B, Wackernagel W. 2017. 150 years of Johann Gregor
Mendel’s “Versuche über Pflanzen-Hybriden.” Mol Genet Genomics. 292(1):1–3.
doi:10.1007/s00438-016-1254-4.
Howell GR, Libby RT, Jakobs TC, Smith RS, Phalan FC, Barter JW, Barbay JM, Marchant JK,
Mahesh N, Porciatti V, et al. 2007. Axons of retinal ganglion cells are insulted in the optic
nerve early in DBA/2J glaucoma. J Cell Biol. 179(7):1523–1537.
doi:10.1083/jcb.200706181.
Hsiao K, Noble C, Pitman W, Yadav N, Kumar S, Keele GR, Terceros A, Kanke M, Conniff T,
Cheleuitte-Nieves C, et al. 2020. A Thalamic Orphan Receptor Drives Variability in
Short-Term Memory. Cell. 183(2):522-536.e19. doi:10.1016/j.cell.2020.09.011.
Ikegami K, Iwatani M, Suzuki M, Tachibana M, Shinkai Y, Tanaka S, Greally JM, Yagi S, Hattori
N, Shiota K. 2007. Genome-wide and locus-specific DNA hypomethylation in G9a
deficient mouse embryonic stem cells. Genes to Cells. 12(1):1–11. doi:10.1111/j.1365-
2443.2006.01029.x.
Jackson RJ, Hellen CUT, Pestova TV. 2010. THE MECHANISM OF EUKARYOTIC
TRANSLATION INITIATION AND PRINCIPLES OF ITS REGULATION. Nat Rev Mol
Cell Biol. 11(2):113–127. doi:10.1038/nrm2838.
Jeck WR, Siebold AP, Sharpless NE. 2012. Review: a meta-analysis of GWAS and ageassociated diseases. Aging Cell. 11(5):727–731. doi:10.1111/j.1474-9726.2012.00871.x.
Jin D, Liu Y, Sun F, Wang X, Liu X, He Z. 2015. Restoration of skilled locomotion by sprouting
corticospinal axons induced by co-deletion of PTEN and SOCS3. Nature
Communications. 6(1):8074. doi:10.1038/ncomms9074.
Jing X, Wang T, Huang S, Glorioso JC, Albers KM. 2012. The transcription factor Sox11
promotes nerve regeneration through activation of the regeneration-associated gene
Sprr1a. Experimental Neurology. 233(1):221–232. doi:10.1016/j.expneurol.2011.10.005.
Jocher G, Mannschatz SH, Offterdinger M, Schweigreiter R. 2018. Microfluidics of SmallPopulation Neurons Allows for a Precise Quantification of the Peripheral Axonal Growth
State. Front Cell Neurosci. 12. doi:10.3389/fncel.2018.00166. [accessed 2024 Mar 24].
https://www.frontiersin.org/articles/10.3389/fncel.2018.00166.
Johnson KR, Erway LC, Cook SA, Willott JF, Zheng QY. 1997. A major gene affecting agerelated hearing loss in C57BL/6J mice. Hearing Research. 114(1):83–92.
doi:10.1016/S0378-5955(97)00155-X.
Johnson KR, Zheng QY, Noben-Trauth K. 2006. Strain background effects and genetic
modifiers of hearing in mice. Brain Research. 1091(1):79–88.
doi:10.1016/j.brainres.2006.02.021.
Johri P, Charlesworth B, Jensen JD. 2020. Toward an Evolutionarily Appropriate Null Model:
Jointly Inferring Demography and Purifying Selection. Genetics. 215(1):173–192.
doi:10.1534/genetics.119.303002.

102
Jung YH, Wang H-LV, Ali S, Corces VG, Kremsky I. 2023. Characterization of a strain-specific
CD-1 reference genome reveals potential inter- and intra-strain functional variability.
BMC Genomics. 24(1):437. doi:10.1186/s12864-023-09523-x.
Kaaij LJT, Mohn F, van der Weide RH, de Wit E, Bühler M. 2019. The ChAHP Complex
Counteracts Chromatin Looping at CTCF Sites that Emerged from SINE Expansions in
Mouse. Cell. 178(6):1437-1451.e14. doi:10.1016/j.cell.2019.08.007.
Kadoya K, Tsukada S, Lu P, Coppola G, Geschwind D, Filbin MT, Blesch A, Tuszynski MH.
2009. Combined Intrinsic and Extrinsic Neuronal Mechanisms Facilitate Bridging Axonal
Regeneration One Year after Spinal Cord Injury. Neuron. 64(2):165–172.
doi:10.1016/j.neuron.2009.09.016.
Kanaan NM, Pigino GF, Brady ST, Lazarov O, Binder LI, Morfini GA. 2013. Axonal degeneration
in Alzheimer’s disease: When signaling abnormalities meet the axonal transport system.
Experimental Neurology. 246:44–53. doi:10.1016/j.expneurol.2012.06.003.
Kang T. 2022. Natural divergence of traits in species of mice reveals novel molecular
mechanisms of cellular senescence [PhD]. [Los Angeles, CA]: University of Southern
California.
Kapahi P, Zid BM, Harper T, Koslover D, Sapin V, Benzer S. 2004. Regulation of lifespan in
Drosophila by modulation of genes in the TOR signaling pathway. Curr Biol. 14(10):885–
890. doi:10.1016/j.cub.2004.03.059.
Kelly RDW, Rodda AE, Dickinson A, Mahmud A, Nefzger CM, Lee W, Forsythe JS, Polo JM,
Trounce IA, McKenzie M, et al. 2013. Mitochondrial DNA Haplotypes Define Gene
Expression Patterns in Pluripotent and Differentiating Embryonic Stem Cells. STEM
CELLS. 31(4):703–716. doi:10.1002/stem.1313.
Khajuria RK, Munschauer M, Ulirsch JC, Fiorini C, Ludwig LS, McFarland SK, Abdulhay NJ,
Specht H, Keshishian H, Mani DR, et al. 2018. Ribosome Levels Selectively Regulate
Translation and Lineage Commitment in Human Hematopoiesis. Cell. 173(1):90-
103.e19. doi:10.1016/j.cell.2018.02.036.
Kim HS, Pickering AM. 2023. Protein translation paradox: Implications in translational regulation
of aging. Frontiers in Cell and Developmental Biology. 11. [accessed 2024 Feb 20].
https://www.frontiersin.org/articles/10.3389/fcell.2023.1129281.
Kim K, Ryu TY, Ryu J-W, Han T-S, Jung C-R, Son M-Y, Kim D-S, Cho H-S. 2020. RNA-seq
based transcriptome analysis of EHMT2 functions in breast cancer. Biochemical and
Biophysical Research Communications. 524(3):672–676.
doi:10.1016/j.bbrc.2020.01.128.
Kimball SR, Vary TC, Jefferson LS. 1992. Age-dependent decrease in the amount of eukaryotic
initiation factor 2 in various rat tissues. Biochem J. 286 ( Pt 1)(Pt 1):263–268.
doi:10.1042/bj2860263.
Koliatsos VE, Alexandris AS. 2019. Wallerian degeneration as a therapeutic target in traumatic
brain injury. Current Opinion in Neurology. 32(6):786–795.
doi:10.1097/WCO.0000000000000763.

103
Kolmykov S, Yevshin I, Kulyashov M, Sharipov R, Kondrakhin Y, Makeev VJ, Kulakovskiy IV,
Kel A, Kolpakov F. 2021. GTRD: an integrated view of transcription regulation. Nucleic
Acids Research. 49(D1):D104–D111. doi:10.1093/nar/gkaa1057.
Koturbash I, Scherhag A, Sorrentino J, Sexton K, Bodnar W, Swenberg JA, Beland FA, PardoManuel deVillena F, Rusyn I, Pogribny IP. 2011. Epigenetic Mechanisms of Mouse
Interstrain Variability in Genotoxicity of the Environmental Toxicant 1,3-Butadiene.
Toxicological Sciences. 122(2):448–456. doi:10.1093/toxsci/kfr133.
Kreitman M. 2000. Methods to detect selection in populations with applications to the human.
Annu Rev Genomics Hum Genet. 1:539–559. doi:10.1146/annurev.genom.1.1.539.
Kriz AJ, Colognori D, Sunwoo H, Nabet B, Lee JT. 2021. Balancing cohesin eviction and
retention prevents aberrant chromosomal interactions, Polycomb-mediated repression,
and X-inactivation. Mol Cell. 81(9):1970-1987.e9. doi:10.1016/j.molcel.2021.02.031.
Lacerda R, Menezes J, Candeias MM. 2019. Alternative Mechanisms of mRNA Translation
Initiation in Cellular Stress Response and Cancer. In: Romão L, editor. The mRNA
Metabolism in Human Disease. Cham: Springer International Publishing. (Advances in
Experimental Medicine and Biology). p. 117–132. [accessed 2024 Feb 20].
https://doi.org/10.1007/978-3-030-19966-1_6.
Lazzarano S, Kučka M, Castro JPL, Naumann R, Medina P, Fletcher MNC, Wombacher R,
Gribnau J, Hochepied T, Van Montagu M, et al. 2018. Genetic mapping of species
differences via in vitro crosses in mouse embryonic stem cells. Proceedings of the
National Academy of Sciences. 115(14):3680–3685. doi:10.1073/pnas.1717474115.
Lear BP, Moore DL. 2023. Moving CNS axon growth and regeneration research into human
model systems. Front Neurosci. 17. doi:10.3389/fnins.2023.1198041. [accessed 2024
Mar 22].
https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2023.1198041/fu
ll.
Leitch HG, McEwen KR, Turp A, Encheva V, Carroll T, Grabole N, Mansfield W, Nashun B,
Knezovich JG, Smith A, et al. 2013. Naive pluripotency is associated with global DNA
hypomethylation. Nat Struct Mol Biol. 20(3):311–316. doi:10.1038/nsmb.2510.
Li XC, Fay JC. 2017. Cis-Regulatory Divergence in Gene Expression between Two Thermally
Divergent Yeast Species. Genome Biol Evol. 9(5):1120–1129. doi:10.1093/gbe/evx072.
Li Y, Zheng H, Wang Q, Zhou C, Wei L, Liu X, Zhang W, Zhang Y, Du Z, Wang X, et al. 2018.
Genome-wide analyses reveal a role of Polycomb in promoting hypomethylation of DNA
methylation valleys. Genome Biol. 19(1):18. doi:10.1186/s13059-018-1390-8.
Lin AC, Holt CE. 2008. Function and regulation of local axonal translation. Curr Opin Neurobiol.
18(1):60–68. doi:10.1016/j.conb.2008.05.004.
Lindborg JA, Tran NM, Chenette DM, DeLuca K, Foli Y, Kannan R, Sekine Y, Wang X, Wollan
M, Kim I-J, et al. 2021. Optic nerve regeneration screen identifies multiple genes
restricting adult neural repair. Cell Rep. 34(9):108777. doi:10.1016/j.celrep.2021.108777.

104
Lisi V, Singh B, Giroux M, Guzman E, Painter MW, Cheng Y-C, Huebner E, Coppola G,
Costigan M, Woolf CJ, et al. 2017. Enhanced Neuronal Regeneration in the CAST/Ei
Mouse Strain Is Linked to Expression of Differentiation Markers after Injury. Cell
Reports. 20(5):1136–1147. doi:10.1016/j.celrep.2017.07.010.
Liu K, Lu Y, Lee JK, Samara R, Willenberg R, Sears-Kraxberger I, Tedeschi A, Park KK, Jin D,
Cai B, et al. 2010. PTEN deletion enhances the regenerative ability of adult corticospinal
neurons. Nature Neuroscience. 13(9):1075–1081. doi:10.1038/nn.2603.
Ljungberg MC, Ali YO, Zhu J, Wu C-S, Oka K, Zhai RG, Lu H-C. 2012. CREB-activity and
nmnat2 transcription are down-regulated prior to neurodegeneration, while NMNAT2
over-expression is neuroprotective, in a mouse model of human tauopathy. Hum Mol
Genet. 21(2):251–267. doi:10.1093/hmg/ddr492.
Love MI, Huber W, Anders S. 2014. Moderated estimation of fold change and dispersion for
RNA-seq data with DESeq2. Genome Biology. 15(12):550. doi:10.1186/s13059-014-
0550-8.
Lu P, Woodruff G, Wang Y, Graham L, Hunt M, Wu D, Boehle E, Ahmad R, Poplawski G, Brock
J, et al. 2014. Long-distance axonal growth from human induced pluripotent stem cells
after spinal cord injury. Neuron. 83(4):789–796. doi:10.1016/j.neuron.2014.07.014.
Luchessi AD, Cambiaghi TD, Alves AS, Parreiras-E-Silva LT, Britto LRG, Costa-Neto CM, Curi
R. 2008. Insights on eukaryotic translation initiation factor 5A (eIF5A) in the brain and
aging. Brain Res. 1228:6–13. doi:10.1016/j.brainres.2008.06.057.
Lunn ER, Perry VH, Brown MC, Rosen H, Gordon S. 1989. Absence of Wallerian Degeneration
does not Hinder Regeneration in Peripheral Nerve. European Journal of Neuroscience.
1(1):27–33. doi:https://doi.org/10.1111/j.1460-9568.1989.tb00771.x.
Ma XM, Blenis J. 2009. Molecular mechanisms of mTOR-mediated translational control. Nat
Rev Mol Cell Biol. 10(5):307–318. doi:10.1038/nrm2672.
Mack KL, Square TA, Zhao B, Miller CT, Fraser HB. 2023. Evolution of Spatial and Temporal
cis-Regulatory Divergence in Sticklebacks. Mol Biol Evol. 40(3):msad034.
doi:10.1093/molbev/msad034.
Mahar M, Cavalli V. 2018. Intrinsic mechanisms of neuronal axon regeneration. Nat Rev
Neurosci. 19(6):323–337. doi:10.1038/s41583-018-0001-8.
Marks H, Kerstens HHD, Barakat TS, Splinter E, Dirks RAM, van Mierlo G, Joshi O, Wang S-Y,
Babak T, Albers CA, et al. 2015. Dynamics of gene silencing during X inactivation using
allele-specific RNA-seq. Genome Biology. 16(1):149. doi:10.1186/s13059-015-0698-x.
Martin HC, Roop JI, Schraiber JG, Hsu TY, Brem RB. 2012. Evolution of a Membrane Protein
Regulon in Saccharomyces. Molecular Biology and Evolution. 29(7):1747–1756.
doi:10.1093/molbev/mss017.
Martin I, Kim JW, Lee BD, Kang HC, Xu J-C, Jia H, Stankowski J, Kim M-S, Zhong J, Kumar M,
et al. 2014. Ribosomal protein s15 phosphorylation mediates LRRK2 neurodegeneration
in Parkinson’s disease. Cell. 157(2):472–485. doi:10.1016/j.cell.2014.01.064.

105
Milde S, Gilley J, Coleman MP. 2013. Subcellular Localization Determines the Stability and
Axon Protective Capacity of Axon Survival Factor Nmnat2. PLOS Biology.
11(4):e1001539. doi:10.1371/journal.pbio.1001539.
Morgan TH. 1910. Sex Limited Inheritance in Drosophila. Science. 32(812):120–122.
doi:10.1126/science.32.812.120.
Mozzetta C, Pontis J, Fritsch L, Robin P, Portoso M, Proux C, Margueron R, Ait-Si-Ali S. 2014.
The histone H3 lysine 9 methyltransferases G9a and GLP regulate polycomb repressive
complex 2-mediated gene silencing. Mol Cell. 53(2):277–289.
doi:10.1016/j.molcel.2013.12.005.
Nix P, Hammarlund M, Hauth L, Lachnit M, Jorgensen EM, Bastiani M. 2014. Axon regeneration
genes identified by RNAi screening in C. elegans. J Neurosci. 34(2):629–645.
doi:10.1523/JNEUROSCI.3859-13.2014.
Obier N, Lin Q, Cauchy P, Hornich V, Zenke M, Becker M, Müller AM. 2015. Polycomb Protein
EED is Required for Silencing of Pluripotency Genes upon ESC Differentiation. Stem
Cell Rev and Rep. 11(1):50–61. doi:10.1007/s12015-014-9550-z.
Oksuz O, Narendra V, Lee C-H, Descostes N, LeRoy G, Raviram R, Blumenberg L, Karch K,
Rocha PP, Garcia BA, et al. 2018. Capturing the Onset of PRC2-Mediated Repressive
Domain Formation. Mol Cell. 70(6):1149-1162.e5. doi:10.1016/j.molcel.2018.05.023.
Omura T, Omura K, Tedeschi A, Riva P, Painter MW, Rojas L, Martin J, Lisi V, Huebner EA,
Latremoliere A, et al. 2015. Robust Axonal Regeneration Occurs in the Injured CAST/Ei
Mouse CNS. Neuron. 86(5):1215–1227. doi:10.1016/j.neuron.2015.05.005.
Orr HA. 1998. Testing natural selection vs. genetic drift in phenotypic evolution using
quantitative trait locus data. Genetics. 149(4):2099–2104.
doi:10.1093/genetics/149.4.2099.
Ortmann D, Brown S, Czechanski A, Aydin S, Muraro D, Huang Y, Tomaz RA, Osnato A, Canu
G, Wesley BT, et al. 2020. Naive Pluripotent Stem Cells Exhibit Phenotypic Variability
that Is Driven by Genetic Variation. Cell Stem Cell. 27(3):470-481.e6.
doi:10.1016/j.stem.2020.07.019.
Osborn HF. 1894. From the Greeks to Darwin: An Outline of the Development of the Evolution
Idea. Macmillan.
Osterloh JM, Yang J, Rooney TM, Fox AN, Adalbert R, Powell EH, Sheehan AE, Avery MA,
Hackett R, Logan MA, et al. 2012. dSarm/Sarm1 is required for activation of an injuryinduced axon death pathway. Science. 337(6093):481–484.
doi:10.1126/science.1223899.
Ott J, Wang J, Leal SM. 2015. Genetic linkage analysis in the age of whole-genome
sequencing. Nat Rev Genet. 16(5):275–284. doi:10.1038/nrg3908.
Pakos‐Zebrucka K, Koryga I, Mnich K, Ljujic M, Samali A, Gorman AM. 2016. The integrated
stress response. EMBO Reports. 17(10):1374–1395. doi:10.15252/embr.201642195.

106
Pan KZ, Palter JE, Rogers AN, Olsen A, Chen D, Lithgow GJ, Kapahi P. 2007. Inhibition of
mRNA translation extends lifespan in Caenorhabditis elegans. Aging Cell. 6(1):111–119.
doi:10.1111/j.1474-9726.2006.00266.x.
Pannese E. 2011. Morphological changes in nerve cells during normal aging. Brain Struct
Funct. 216(2):85–89. doi:10.1007/s00429-011-0308-y.
Papadopoli D, Boulay K, Kazak L, Pollak M, Mallette FA, Topisirovic I, Hulea L. 2019. mTOR as
a central regulator of lifespan and aging. F1000Res. 8:F1000 Faculty Rev-998.
doi:10.12688/f1000research.17196.1.
Park KK, Liu K, Hu Y, Smith PD, Wang C, Cai B, Xu B, Connolly L, Kramvis I, Sahin M, et al.
2008. Promoting axon regeneration in the adult CNS by modulation of the PTEN/mTOR
pathway. Science. 322(5903):963–966. doi:10.1126/science.1161566.
Pasyukova EG, Vieira C, Mackay TFC. 2000. Deficiency Mapping of Quantitative Trait Loci
Affecting Longevity in Drosophila melanogaster. Genetics. 156(3):1129–1146.
doi:10.1093/genetics/156.3.1129.
Pelkonen A, Pistono C, Klecki P, Gómez-Budia M, Dougalis A, Konttinen H, Stanová I,
Fagerlund I, Leinonen V, Korhonen P, et al. 2021. Functional Characterization of Human
Pluripotent Stem Cell-Derived Models of the Brain with Microelectrode Arrays. Cells.
11(1):106. doi:10.3390/cells11010106.
Pertea G, Pertea M. 2020. GFF Utilities: GffRead and GffCompare. F1000Res. 9:ISCB Comm
J-304. doi:10.12688/f1000research.23297.2.
Phifer-Rixey M, Harr B, Hey J. 2020. Further resolution of the house mouse (Mus musculus)
phylogeny by integration over isolation-with-migration histories. BMC Evolutionary
Biology. 20(1):120. doi:10.1186/s12862-020-01666-9.
Polymenis M, Aramayo R. 2015. Translate to divide: сontrol of the cell cycle by protein
synthesis. Microbial Cell. 2(4):94–104. doi:10.15698/mic2015.04.198.
Price PD, Palmer Droguett DH, Taylor JA, Kim DW, Place ES, Rogers TF, Mank JE, Cooney
CR, Wright AE. 2022. Detecting signatures of selection on gene expression. Nat Ecol
Evol. 6(7):1035–1045. doi:10.1038/s41559-022-01761-8.
Quinlan AR, Hall IM. 2010. BEDTools: a flexible suite of utilities for comparing genomic
features. Bioinformatics. 26(6):841–842. doi:10.1093/bioinformatics/btq033.
Reilly MM, Murphy SM, Laurá M. 2011. Charcot-Marie-Tooth disease. Journal of the Peripheral
Nervous System. 16(1):1–14. doi:https://doi.org/10.1111/j.1529-8027.2011.00324.x.
Ren X, Hu B, Song M, Ding Z, Dang Y, Liu Z, Zhang W, Ji Q, Ren R, Ding J, et al. 2019.
Maintenance of Nucleolar Homeostasis by CBX4 Alleviates Senescence and
Osteoarthritis. Cell Rep. 26(13):3643-3656.e7. doi:10.1016/j.celrep.2019.02.088.
Roderick, Thomas. 1982. New Inbred Strains: MOLC/Rk, MOLD/Rk, MOLE/Rk, CASA/Rk,
CASB/Rk, CAST/Ei. Mouse News Letter. 67:22–3.

107
Rogers AN, Chen D, McColl G, Czerwieniec G, Felkey K, Gibson BW, Hubbard A, Melov S,
Lithgow GJ, Kapahi P. 2011. Life span extension via eIF4G inhibition is mediated by
posttranscriptional remodeling of stress response gene expression in C. elegans. Cell
Metab. 14(1):55–66. doi:10.1016/j.cmet.2011.05.010.
Roop JI, Chang KC, Brem RB. 2016. Polygenic evolution of a sugar specialization trade-off in
yeast. Nature. 530(7590):336–339. doi:10.1038/nature16938.
Sajadi A, Schneider BL, Aebischer P. 2004. Wlds-mediated protection of dopaminergic fibers in
an animal model of Parkinson disease. Curr Biol. 14(4):326–330.
doi:10.1016/j.cub.2004.01.053.
Sajini AA, Greder LV, Dutton JR, Slack JMW. 2012. Loss of Oct4 expression during the
development of murine embryoid bodies. Dev Biol. 371(2):170–179.
doi:10.1016/j.ydbio.2012.08.008.
Sampath P, Pritchard DK, Pabon L, Reinecke H, Schwartz SM, Morris DR, Murry CE. 2008. A
Hierarchical Network Controls Protein Translation during Murine Embryonic Stem Cell
Self-Renewal and Differentiation. Cell Stem Cell. 2(5):448–460.
doi:10.1016/j.stem.2008.03.013.
Samsam M, Mi W, Wessig C, Zielasek J, Toyka KV, Coleman MP, Martini R. 2003. The Wlds
mutation delays robust loss of motor and sensory axons in a genetic model for myelinrelated axonopathy. J Neurosci. 23(7):2833–2839.
Sangster MY, Heliams DB, MacKenzie JS, Shellam GR. 1993. Genetic studies of flavivirus
resistance in inbred strains derived from wild mice: evidence for a new resistance allele
at the flavivirus resistance locus (Flv). Journal of Virology. 67(1):340–347.
doi:10.1128/jvi.67.1.340-347.1993.
Sarkans U, Gostev M, Athar A, Behrangi E, Melnichuk O, Ali A, Minguet J, Rada JC, Snow C,
Tikhonov A, et al. 2018. The BioStudies database—one stop shop for all data supporting
a life sciences study. Nucleic Acids Research. 46(D1):D1266–D1270.
doi:10.1093/nar/gkx965.
Sasaki Y, Nakagawa T, Mao X, DiAntonio A, Milbrandt J. 2016. NMNAT1 inhibits axon
degeneration via blockade of SARM1-mediated NAD+ depletion. Chao MV, editor. eLife.
5:e19749. doi:10.7554/eLife.19749.
Saxton RA, Sabatini DM. 2017. mTOR Signaling in Growth, Metabolism, and Disease. Cell.
168(6):960–976. doi:10.1016/j.cell.2017.02.004.
Scheuner D, Vander Mierde D, Song B, Flamez D, Creemers JWM, Tsukamoto K, Ribick M,
Schuit FC, Kaufman RJ. 2005. Control of mRNA translation preserves endoplasmic
reticulum function in beta cells and maintains glucose homeostasis. Nat Med. 11(7):757–
764. doi:10.1038/nm1259.
Sekine Y, Lin-Moore A, Chenette DM, Wang X, Jiang Z, Cafferty WB, Hammarlund M,
Strittmatter SM. 2018. Functional Genome-wide Screen Identifies Pathways Restricting
Central Nervous System Axonal Regeneration. Cell Rep. 23(2):415–428.
doi:10.1016/j.celrep.2018.03.058.

108
Selkoe DJ. 2002. Alzheimer’s Disease Is a Synaptic Failure. Science. 298(5594):789–791.
doi:10.1126/science.1074069.
Selman C, Tullet JMA, Wieser D, Irvine E, Lingard SJ, Choudhury AI, Claret M, Al-Qassab H,
Carmignac D, Ramadani F, et al. 2009. Ribosomal protein S6 kinase 1 signaling
regulates mammalian life span. Science. 326(5949):140–144.
doi:10.1126/science.1177221.
Serrao E, Cherepanov P, Engelman AN. 2016. Amplification, Next-generation Sequencing, and
Genomic DNA Mapping of Retroviral Integration Sites. J Vis Exp.(109):53840.
doi:10.3791/53840.
Signor SA, Nuzhdin SV. 2018. The Evolution of Gene Expression in cis and trans. Trends
Genet. 34(7):532–544. doi:10.1016/j.tig.2018.03.007.
Simpson CE, Ashe MP. 2012. Adaptation to stress in yeast: to translate or not? Biochemical
Society Transactions. 40(4):794–799. doi:10.1042/BST20120078.
Singh S, Dallenga T, Winkler A, Roemer S, Maruschak B, Siebert H, Brück W, Stadelmann C.
2017. Relationship of acute axonal damage, Wallerian degeneration, and clinical
disability in multiple sclerosis. Journal of Neuroinflammation. 14(1):57.
doi:10.1186/s12974-017-0831-8.
Skelly DA, Czechanski A, Byers C, Aydin S, Spruce C, Olivier C, Choi K, Gatti DM, Raghupathy
N, Keele GR, et al. 2020. Mapping the Effects of Genetic Variation on Chromatin State
and Gene Expression Reveals Loci That Control Ground State Pluripotency. Cell Stem
Cell. 27(3):459-469.e8. doi:10.1016/j.stem.2020.07.005.
Sohail M, Maier RM, Ganna A, Bloemendal A, Martin AR, Turchin MC, Chiang CW, Hirschhorn
J, Daly MJ, Patterson N, et al. 2019. Polygenic adaptation on height is overestimated
due to uncorrected stratification in genome-wide association studies. Elife. 8:e39702.
doi:10.7554/eLife.39702.
Solomon E, Davis-Anderson K, Hovde B, Micheva-Viteva S, Harris JF, Twary S, Iyer R. 2021.
Global transcriptome profile of the developmental principles of in vitro iPSC-to-motor
neuron differentiation. BMC Molecular and Cell Biology. 22(1):13. doi:10.1186/s12860-
021-00343-z.
Sorrells TR, Johnson AN, Howard CJ, Britton CS, Fowler KR, Feigerle JT, Weil PA, Johnson
AD. 2018. Intrinsic cooperativity potentiates parallel cis-regulatory evolution. eLife.
7:e37563. doi:10.7554/eLife.37563.
Steffen KK, MacKay VL, Kerr EO, Tsuchiya M, Hu D, Fox LA, Dang N, Johnston ED, Oakes JA,
Tchao BN, et al. 2008. Yeast life span extension by depletion of 60s ribosomal subunits
is mediated by Gcn4. Cell. 133(2):292–302. doi:10.1016/j.cell.2008.02.037.
Stein KC, Morales-Polanco F, van der Lienden J, Rainbolt TK, Frydman J. 2022. Ageing
exacerbates ribosome pausing to disrupt cotranslational proteostasis. Nature.
601(7894):637–642. doi:10.1038/s41586-021-04295-4.

109
Steinmetz LM, Sinha H, Richards DR, Spiegelman JI, Oefner PJ, McCusker JH, Davis RW.
2002. Dissecting the architecture of a quantitative trait locus in yeast. Nature.
416(6878):326–330. doi:10.1038/416326a.
Stern DL. 2014. Identification of loci that cause phenotypic variation in diverse species with the
reciprocal hemizygosity test. Trends in Genetics. 30(12):547–554.
doi:10.1016/j.tig.2014.09.006.
Sun F, He Z. 2010. Neuronal intrinsic barriers for axon regeneration in the adult CNS. Current
Opinion in Neurobiology. 20(4):510–518. doi:10.1016/j.conb.2010.03.013.
Syntichaki P, Troulinaki K, Tavernarakis N. 2007. eIF4E function in somatic cells modulates
ageing in Caenorhabditis elegans. Nature. 445(7130):922–926.
doi:10.1038/nature05603.
Tain LS, Mortiboys H, Tao RN, Ziviani E, Bandmann O, Whitworth AJ. 2009. Rapamycin
activation of 4E-BP prevents parkinsonian dopaminergic neuron loss. Nat Neurosci.
12(9):1129–1135. doi:10.1038/nn.2372.
Takauji Y, Wada T, Takeda A, Kudo I, Miki K, Fujii M, Ayusawa D. 2016. Restriction of protein
synthesis abolishes senescence features at cellular and organismal levels. Sci Rep.
6:18722. doi:10.1038/srep18722.
Tanay A, Regev A, Shamir R. 2005. Conservation and evolvability in regulatory networks: The
evolution of ribosomal regulation in yeast. Proc Natl Acad Sci U S A. 102(20):7203–
7208. doi:10.1073/pnas.0502521102.
Tedeschi A. 2012. Tuning the Orchestra: Transcriptional Pathways Controlling Axon
Regeneration. Frontiers in Molecular Neuroscience. 4. [accessed 2024 Feb 20].
https://www.frontiersin.org/articles/10.3389/fnmol.2011.00060.
Tedeschi A, Bradke F. 2017. Spatial and temporal arrangement of neuronal intrinsic and
extrinsic mechanisms controlling axon regeneration. Current Opinion in Neurobiology.
42:118–127. doi:10.1016/j.conb.2016.12.005.
Tedeschi A, Omura T, Costigan M. 2017. CNS repair and axon regeneration: Using genetic
variation to determine mechanisms. Exp Neurol. 287(Pt 3):409–422.
doi:10.1016/j.expneurol.2016.05.004.
Teotia P, Hook MJV, Fischer D, Ahmad I. 2019. Human retinal ganglion cell axon regeneration
by recapitulating developmental mechanisms: effects of recruitment of the mTOR
pathway. Development. 146(13). doi:10.1242/dev.178012. [accessed 2021 Mar 16].
https://dev.biologists.org/content/146/13/dev178012.
Terzic D, Maxon JR, Krevitt L, Dibartolomeo C, Goyal T, Low WC, Dutton JR, Parr AM. 2016.
Directed Differentiation of Oligodendrocyte Progenitor Cells from Mouse Induced
Pluripotent Stem Cells. Cell Transplant. 25(2):411–424.
doi:10.3727/096368915X688137.
Tkatchenko TV, Shah RL, Nagasaki T, Tkatchenko AV. 2019. Analysis of genetic networks
regulating refractive eye development in collaborative cross progenitor strain mice

110
reveals new genes and pathways underlying human myopia. BMC Medical Genomics.
12(1):113. doi:10.1186/s12920-019-0560-1.
Tomoda T, Bhatt RS, Kuroyanagi H, Shirasawa T, Hatten ME. 1999. A mouse serine/threonine
kinase homologous to C. elegans UNC51 functions in parallel fiber formation of
cerebellar granule neurons. Neuron. 24(4):833–846. doi:10.1016/s0896-6273(00)81031-
4.
Turkiew E, Falconer D, Reed N, Höke A. 2017. Deletion of Sarm1 gene is neuroprotective in
two models of peripheral neuropathy. J Peripher Nerv Syst. 22(3):162–171.
doi:10.1111/jns.12219.
Uffelmann E, Huang QQ, Munung NS, de Vries J, Okada Y, Martin AR, Martin HC, Lappalainen
T, Posthuma D. 2021. Genome-wide association studies. Nat Rev Methods Primers.
1(1):1–21. doi:10.1038/s43586-021-00056-9.
Verdú E, Ceballos D, Vilches JJ, Navarro X. 2000. Influence of aging on peripheral nerve
function and regeneration. Journal of the Peripheral Nervous System. 5(4):191–208.
doi:https://doi.org/10.1111/j.1529-8027.2000.00026.x.
Waller AV, Owen R. 1850. XX. Experiments on the section of the glossopharyngeal and
hypoglossal nerves of the frog, and observations of the alterations produced thereby in
the structure of their primitive fibres. Philosophical Transactions of the Royal Society of
London. 140:423–429. doi:10.1098/rstl.1850.0021.
Wang B, Starr AL, Fraser HB. 2024. Cell-type-specific cis-regulatory divergence in gene
expression and chromatin accessibility revealed by human-chimpanzee hybrid cells.
Elife. 12:RP89594. doi:10.7554/eLife.89594.
Wang M, Kong L. 2019. pblat: a multithread blat algorithm speeding up aligning sequences to
genomes. BMC Bioinformatics. 20(1):28. doi:10.1186/s12859-019-2597-8.
Wang MS, Davis AA, Culver DG, Glass JD. 2002. WldS mice are resistant to paclitaxel (taxol)
neuropathy. Ann Neurol. 52(4):442–447. doi:10.1002/ana.10300.
Wang W, Flury AG, Rodriguez AT, Garrison JL, Brem RB. 2022. A role for worm cutl-24 in
background- and parent-of-origin-dependent ER stress resistance. BMC Genomics.
23(1):842. doi:10.1186/s12864-022-09063-w.
Wang X, Proud CG. 2006. The mTOR Pathway in the Control of Protein Synthesis. Physiology.
21(5):362–369. doi:10.1152/physiol.00024.2006.
Waterhouse AM, Procter JB, Martin DMA, Clamp M, Barton GJ. 2009. Jalview Version 2—a
multiple sequence alignment editor and analysis workbench. Bioinformatics. 25(9):1189–
1191. doi:10.1093/bioinformatics/btp033.
Webster GC, Webster SL, Landis WA. 1981. The effect of age on the initiation of protein
synthesis in Drosophila melanogaster. Mech Ageing Dev. 16(1):71–79.
doi:10.1016/0047-6374(81)90034-8.

111
Weigert R, Hetzel S, Bailly N, Haggerty C, Ilik IA, Yung PYK, Navarro C, Bolondi A, Kumar AS,
Anania C, et al. 2023. Dynamic antagonism between key repressive pathways maintains
the placental epigenome. Nat Cell Biol. 25(4):579–591. doi:10.1038/s41556-023-01114-
y.
Weiss CV, Brem RB. 2019. Dissecting Trait Variation across Species Barriers. Trends in
Ecology & Evolution. 34(12):1131–1136. doi:10.1016/j.tree.2019.07.013.
Weiss CV, Roop JI, Hackley RK, Chuong JN, Grigoriev IV, Arkin AP, Skerker JM, Brem RB.
2018. Genetic dissection of interspecific differences in yeast thermotolerance. Nat
Genet. 50(11):1501–1504. doi:10.1038/s41588-018-0243-4.
Werner RJ, Schultz BM, Huhn JM, Jelinek J, Madzo J, Engel N. 2017. Sex chromosomes drive
gene expression and regulatory dimorphisms in mouse embryonic stem cells. Biol Sex
Differ. 8:28. doi:10.1186/s13293-017-0150-x.
Wittkopp PJ, Haerum BK, Clark AG. 2004. Evolutionary changes in cis and trans gene
regulation. Nature. 430(6995):85–88. doi:10.1038/nature02698.
Wu C-Y, Whye D, Mason RW, Wang W. 2012. Efficient Differentiation of Mouse Embryonic
Stem Cells into Motor Neurons. JoVE.(64):3813. doi:10.3791/3813.
Wu PZ, Liberman LD, Bennett K, de Gruttola V, O’Malley JT, Liberman MC. 2019. Primary
Neural Degeneration in the Human Cochlea: Evidence for Hidden Hearing Loss in the
Aging Ear. Neuroscience. 407:8–20. doi:10.1016/j.neuroscience.2018.07.053.
Wu T, Hu E, Xu S, Chen M, Guo P, Dai Z, Feng T, Zhou L, Tang W, Zhan L, et al. 2021.
clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. The
Innovation. 2(3):100141. doi:10.1016/j.xinn.2021.100141.
Xiong H, Morrison J, Ferris MT, Gralinski LE, Whitmore AC, Green R, Thomas MJ, Tisoncik-Go
J, Schroth GP, Pardo-Manuel de Villena F, et al. 2014. Genomic Profiling of
Collaborative Cross Founder Mice Infected with Respiratory Viruses Reveals Novel
Transcripts and Infection-Related Strain-Specific Gene and Isoform Expression. G3
Genes|Genomes|Genetics. 4(8):1429–1444. doi:10.1534/g3.114.011759.
Yang H, Bell TA, Churchill GA, Pardo-Manuel de Villena F. 2007. On the subspecific origin of
the laboratory mouse. Nat Genet. 39(9):1100–1107. doi:10.1038/ng2087.
Yang H, Wang JR, Didion JP, Buus RJ, Bell TA, Welsh CE, Bonhomme F, Yu AH-T, Nachman
MW, Pialek J, Tucker P, Boursot P, McMillan L, Churchill GA, Villena FP-M. 2011.
Subspecific origin and haplotype diversity in the laboratory mouse. Nat Genet.
43(7):648–655. doi:10.1038/ng.847.
Yang P, Wen H, Zhang J. 2010. Expression of a dominant-negative Rho-kinase promotes
neurite outgrowth in a microenvironment mimicking injured central nervous system. Acta
Pharmacol Sin. 31(5):531–539. doi:10.1038/aps.2010.35.
Yasuda H, Terada M, Maeda K, Kogawa S, Sanada M, Haneda M, Kashiwagi A, Kikkawa R.
2003. Diabetic neuropathy and nerve regeneration. Progress in Neurobiology.
69(4):229–285. doi:10.1016/S0301-0082(03)00034-0.

112
Yoon BC, Zivraj KH, Holt CE. 2009. Local Translation and mRNA Trafficking in Axon
Pathfinding. Results Probl Cell Differ. 48:269–288. doi:10.1007/400_2009_5.
York RA, Patil C, Abdilleh K, Johnson ZV, Conte MA, Genner MJ, McGrath PT, Fraser HB,
Fernald RD, Streelman JT. 2018. Behavior-dependent cis regulation reveals genes and
pathways associated with bower building in cichlid fishes. Proc Natl Acad Sci U S A.
115(47):E11081–E11090. doi:10.1073/pnas.1810140115.
Young RA. 2011. Control of Embryonic Stem Cell State. Cell. 144(6):940–954.
doi:10.1016/j.cell.2011.01.032.
Yu H, Mohan S, Edderkaoui B, Masinde GL, Davidson HM, Wergedal JE, Beamer WG, Baylink
DJ. 2007. Detecting Novel Bone Density and Bone Size Quantitative Trait Loci Using a
Cross of MRL/MpJ and CAST/EiJ Inbred Mice. Calcif Tissue Int. 80(2):103–110.
doi:10.1007/s00223-006-0187-z.
Zhang H, Wu Z, Lu JY, Huang B, Zhou H, Xie W, Wang J, Shen X. 2020. DEAD-Box Helicase
18 Counteracts PRC2 to Safeguard Ribosomal DNA in Pluripotency Regulation. Cell
Reports. 30(1):81-97.e7. doi:10.1016/j.celrep.2019.12.021.
Zhang X hui, Poo M. 2002. Localized synaptic potentiation by BDNF requires local protein
synthesis in the developing axon. Neuron. 36(4):675–688. doi:10.1016/s0896-
6273(02)01023-1.

Abstract (if available)

Abstract Much of what we know about biology comes from studies of a handful of model organisms. The diversity across the rest of the taxa that inhabit our planet represents a treasure trove of information about biological mechanisms that is relatively untouched. In this thesis, I focus on a little-studied subspecies of the common house mouse that originates in Southeast Asia, Mus musculus castaneus. Compared to their sister subspecies which are commonly used in lab studies, M. m. castaneus have a number of interesting traits, including axon regeneration (Omura et al. 2015), resistance to age-related hearing loss (Johnson et al. 1997; Johnson et al. 2006), wound healing (Heber-Katz et al. 2004), and resistance to DNA damage (French et al. 2015; Chappell et al. 2017). A better understanding of these and other yet-to-be-discovered unique properties of M. m. castaneus could inform new biomedical applications, while also serving as a tool to study evolutionary mechanisms. In this thesis, I describe different approaches that I, along with collaborators, have used to study M. m. castaneus through the lens of natural variation.

Chapter 2 describes an unbiased screen for directional gene expression changes between mouse taxa using expression data in hybrid stem cells. Through this screen, we discovered a novel program of upregulated expression of translation genes in M. m. castaneus stem cells driven by changes in cis- and trans-acting regulators, attesting to a change in selective pressure on translation pathway between M. m. castaneus and its sister species. Complementing this expression data with sequence-based analyses, we identified signatures of evolutionary selection (elevated sequence divergence) proximal to the translation genes. These analyses pinpointed evidence for adaptive changes in M. m. castaneus in binding sites for the transcriptional regulators Ehmt2, Eed, and Msx1.

In Chapter 3, I describe a study of species variation in cell-autonomous axon growth behaviors, using stem cell-derived neurons. In this in vitro system, we found that M. m. castaneus neurons grew longer extensions than did those from M. m. domesticus even in the absence of injury. And we found that stem cell-derived neurons from an F1 hybrid between these two species phenocopied those of M. m. castaneus, indicating a dominant genetic phenotype. In Chapter 4, I lay out advances in development of methods for an unbiased genetic screening approach in mammalian cells to identify allelic variation that underlies species-unique traits. The latter represents important steps towards the eventual goal of finding the genetic basis of M. m. castaneus axon extension and regeneration phenotypes.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Natural divergence of traits in species of mice reveals novel molecular mechanisms of cellular senescence

PDF

Longitudinal assessment of neural stem-cell aging

PDF

Identifying and measuring cell-intrinsic and cell-extrinsic factors influencing aging

PDF

Defining the functional roles of neurotransmitters and neuropeptides in neural circuits

PDF

Phenotypic and multi-omic characterization of novel C. elegans models of Alzheimer's disease

PDF

Computational approaches to identify genetic regulators of aging and late-life mortality

PDF

Signaling mechanisms governing intestinal regeneration and gut-glia cross-talk in Drosophila

PDF

Investigating brain aging and neurodegenerative diseases through omics data

PDF

Characterization of senescent cell heterogeneity using cell culture models

PDF

A single cell time course of senescence uncovers discrete cell trajectories and transcriptional heterogeneity

PDF

Mediating the homeostatic response during light & cytotoxic induced retinal damage

PDF

Investigating the role of ploidy on genome stability in fission yeast

PDF

TLR4-mediated innate immune response and neuroinflammation: focus on APOE and obesity

PDF

Identifying allele-specific DNA methylation in mammalian genomes

PDF

Genetics and the environment: evaluating the role of noncoding RNA in autism spectrum disorder

PDF

Genetic and molecular insights into the genotype-phenotype relationship

PDF

The overlap between mTOR signaling, rapamycin and cellular senescence

PDF

Modeling neurodegenerative diseases using induced pluripotent stem cells and identifying therapeutic targets

PDF

Variation in Galr1 expression determines the susceptibility to excitotoxin-induced neuronal death in mice

PDF

Natural variation of Arabidopsis thaliana methylome and its impact on genome evolution

Asset Metadata

Creator Simon, Noah Malae (author)

Core Title Genomic and phenotypic novelties in the Southeast Asian house mouse

School Leonard Davis School of Gerontology

Degree Doctor of Philosophy

Degree Program Biology of Aging

Degree Conferral Date 2024-05

Publication Date 04/08/2024

Defense Date 03/18/2024

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

Tag evolutionary biology,genomics,neurons,translation

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Benayoun, Berenice (committee chair), Brem, Rachel (committee member), Dean, Matt (committee member), Garrison, Jennifer (committee member), Tracy, Tara (committee member)

Creator Email nmsimon@usc.edu,noahmalaesimon@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113871263

Unique identifier UC113871263

Identifier etd-SimonNoahM-12772.pdf (filename)

Legacy Identifier etd-SimonNoahM-12772

Document Type Dissertation

Format theses (aat)

Rights Simon, Noah Malae

Internet Media Type application/pdf

Type texts

Source 20240408-usctheses-batch-1137 (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.

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