The effect of sex on the mouse lens transcriptome

Abstract

The transcriptome of mammalian tissues differs between males and females, and these differences can change across the lifespan, likely regulating known sexual dimorphisms in disease prevalence and severity. Cataract, the most prevalent disease of the ocular lens, occurs at similar rates in young individuals, but its incidence is elevated in older women compared to men of the same age. However, the influence of sex on the lens transcriptome was unknown. RNAseq based transcriptomic profiling of young adult C57BL/6J mouse lens epithelial and fiber cells revealed that few genes are differentially expressed between the sexes. In contrast, lens cells from aged (24 month old) male and female C57BL/6J mice differentially expressed many genes, including several whose expression is lens preferred. . Like cataracts, posterior capsular opacification (PCO), a major sequela of cataract surgery, may also be more prevalent in women. Lens epithelial cells isolated from mouse eyes 24 hours after lens fiber cell removal exhibited numerous transcriptomic differences between the sexes, including genes implicated in complement cascades and extracellular matrix regulation, and these differences are much more pronounced in aged mice than in young mice. These results provide an unbiased basis for future studies on how sex affects the lens response to aging, cataract development, and cataract surgery.

Introduction

Biological sex in vertebrates is commonly recognized as the set of male or female traits necessary for an organism to successfully contribute to sexual reproduction. In mammals, this is conferred primarily by the organism’s sex chromosome complement, with female primary sexual characteristics considered as the “default” state. Males form when genes present on the Y chromosome drive testes differentiation from the genital ridge. The resulting Leydig cell differentiation leads to the fetal testosterone production necessary for the development of the male genitalia (Greenfield, 2015; Zhao and Yao, 2019). This concept was further bolstered by the realization that females, which carry two copies of the X chromosome, randomly inactivate one copy via the action of the long non-coding RNA Xist, yielding males and females who have similar X chromosome gene dosage (Arnold, 2017).

However, sex affects numerous aspects of biology besides those directly relevant to reproduction, as exemplified by the wide variety of human diseases for which disease incidence, severity, or presentation differs greatly between sexes (Mauvais-Jarvis et al., 2020). These differences, along with sexually dimorphic aging mechanisms, likely contribute to the longer life expectancy of women compared to men (Sampathkumar et al., 2020). However, in contrast to this, epidemiological studies have consistently found that post-menopausal women have a higher risk of cataract than age-matched men, although the cataract rates in younger people are similar in both sexes (Zetterberg and Celojevic, 2015). Further, several epidemiological studies have suggested that women have a significantly higher risk of developing posterior capsular opacification (PCO) after cataract surgery than men (Congdon et al., 2008; Fong et al., 2014; Lee et al., 2016), although other studies do not support this correlation (Tokko et al., 2019).

Most experimental investigations into the increased risk of cataract in post-menopausal women have focused on a possible protective effect of estrogen on cataract development (Zetterberg and Celojevic, 2015). The prevalent hypothesis in the literature suggests that the withdrawal of estrogen at menopause results in cataract due to a “rebound” effect on the lens (Zetterberg and Celojevic, 2015). This postulate is supported by epidemiological studies finding a correlation between hormonal replacement therapy of menopausal women and protection from cataract (Noran et al., 2007; Younan et al., 2002), while treatment of breast cancer patients with the anti-estrogen, tamoxifen, increases cataract risk (Gorin et al., 1998; Nelson et al., 2019).

In other tissues, sex confers more differences on cells than can be explained by just differential exposure to sex hormones, including changes in DNA methylation (Hodes et al., 2017) and the expression of numerous autosomal genes (Arnold, 2019), even prior to embryonic sex determination (Lowe et al., 2015). However, no studies have investigated how sex affects global gene expression in the lens, and it is unknown whether any sex differences vary during aging. We used RNAseq to perform transcriptome profiling of lens cells obtained from inbred C57BL/6J strain mice to discover the effect of age on the lens epithelial and fiber cell transcriptomes, and the transcriptional response of lens epithelial cells to a surgery that models lens epithelium injury during cataract surgery (Faranda et al, 2021). As that study was performed on samples from both sexes, here, we reanalyze that dataset to elucidate how the global transcriptome of lens epithelial and fiber cells isolated from young (3 months) and old (24 month) old C57BL/6J strain mice differ by sex and to investigate how sex and age affect the response of lens epithelial cells to lens fiber cell removal performed to model modern cataract surgery. As the National Institutes of Health of the United States has mandated that research studies should include animals from both sexes, and consider the possibility that sex is a relevant biological variable in biomedical research (Woitowich et al., 2020), this study will provide important baseline information for the field of lens and cataract research.

Materials and methods:

Mice

All studies using mice comply with the Association for Research in Vision and Ophthalmology Statement on the Use of Animals in Vision Research and were approved by the University of Delaware Institutional Animal Care and Use Committee. Twenty four month old C57BL/6NIA mice (10 males and 10 females) were obtained from the National Institute on Aging Biological Resources Colony in October of 2018. These animals are derived from C57BL/6J foundation stock obtained from the Jackson Laboratory in 2016. Ten week old C57BL/6J mice (10 males and 10 females, Stock # 000664) were obtained from the Jackson Laboratory in October of 2018. In both cases, animals were housed at the University of Delaware animal facility under a 14/10 hour light-dark cycle for two weeks prior to tissue isolation. The eyes from all mice used in this study were of normal size and did not manifest signs of the sporadic eye defects that have been reported in this strain (Smith et al., 1994).

Mouse cataract surgery model and tissue isolation

Surgical removal of lens fiber cells was performed on adult mice to mimic human cataract surgery as previously described (Desai et al., 2010; Manthey et al., 2014b) following procedures first reported by Call et al. (Call et al., 2004). Briefly, two weeks after arrival at the University of Delaware, mice were anesthetized following iris dilation, a central corneal incision made, and the entire lens fiber cell mass was removed from one eye by a sharp forceps, leaving behind an intact lens capsule. The cornea was sutured and the eye restored to normal shape with balanced saline solution. Twenty four hours later, mice were re- anesthetized, and the surgery repeated on the other eye. Mice were then immediately sacrificed and lens capsular bags isolated by dissection.

For RNA sequencing, lens capsular bags from either 24 hours post cataract surgery (PCS) or zero hours PCS were pooled from five animals to make a single biological replicate, while lens fiber masses from two independent animals were pooled per replicate. Two biological replicates were created for each condition (male and female samples from 3 month old versus 24 month old animals at zero hours PCS, 24 hours PCS, and lens fiber cells), and flash frozen on dry ice. RNA was isolated from these samples, RNA libraries prepared using the SMARTer® Stranded Total RNA-Seq Kit-Pico Input Mammalian (Takara Bio USA, Inc., Mountain View, CA, USA) and sequenced by DNA Link, USA (901 Morena Blvd. Ste 730 San Diego CA 92117, USA) on a NovaSeq 6000 (San Diego, CA, USA). Read pairs (101 basepairs, bidirectional) were aligned to the Ensembl primary assembly of the mouse GRCm38 genome (Yates et al., 2020) using Hisat2 with its default parameters (Kim et al., 2019). Read pairs aligning to genomic features in the Ensembl Mouse version 101 GTF file were quantified as gene level counts, using HTSeq-Count in union mode (Anders et al., 2014). Length normalized abundance estimates (Fragments per Kilobase-Million (FPKM)) were calculated from gene level counts using the total length of all known exons for a given gene, after merging overlapping exons.

Samples were partitioned for TMM (Trimmed Median of Means) scaling (Robinson and Oshlack, 2010) and differential expression analyses were performed based on the objective of a particular contrast. For contrasts evaluating differences between epithelial cells and fiber cells, and sex or age effects in un-injured tissues, all samples collected at 0 hours post-surgery were grouped together. For contrasts evaluating LEC injury responses, all LECs samples were grouped together.

The “exactTest” method from the edgeR statistical package (version 3.30.3) was used to estimate the magnitude and statistical significance of differential gene expression, with robust dispersion estimates (Phipson et al., 2016; Robinson et al., 2010). Genes with at least 10 mapped reads in at least two samples were considered to have “detectable” levels of expression. Genes failing “detectable” criteria were removed prior to running the “exactTest”, using edgeR’s “filterByExpr” function (Chen et al., 2016).

Biologically significant differentially expressed genes (DEGs) were defined as those exhibiting a statistically significant difference in expression using Storey’s Q value to adjust for False Discovery Rate (FDR ≤ 0.05; (Storey and Tibshirani, 2003)), a difference in expression level greater than 2 FPKM between conditions, Fold Change (FC) greater than 2 in either the positive or negative direction, and expressed at a level greater than 2 FPKM. (Manthey et al., 2014a).

Pathway analyses

Pathway analysis was performed on all statistically significant DEGs defined as those exhibiting a fold change ≥ |2| and FDR < 0.05 using iPathwayGuide (Advaita Bioinformatics, Plymouth Michigan, USA). This software package uses Impact Analysis, an approach that considers the directed interactions of a DEG within a given pathway (as defined by the Kyoto Encyclopedia of Genes and Genomes, KEGG, (Kanehisa et al., 2017), Release 96.0+/11-21, Nov 20), and also whether more pathway participants are observed in the DEG list than would be expected by chance (Ahsan and Draghici, 2017; Draghici et al., 2007; Tarca et al., 2009). Gene ontology comparisons were made against the October 14, 2020 release of the Gene Ontology Consortium database

Lens Preferred Genes in the iSyte database:

The Integrated systems tool for eye gene discovery (iSyTe) database (Kakrana et al., 2018) uses the principle of “whole body subtraction” to identify genes with a significant difference in expression between the lens, and the rest of the body. Genes expressed at higher levels in the lens vs non-lens tissues are described as “lens-preferred”, and are expected to contribute to the lens phenotype. In the current study, DEGs identified in sex and age dependent contrasts were compared to the 412 genes that iSyTe considers “highly lens preferred” at postnatal day 56.

Results

We studied how aging affects the lens transcriptome by performing RNAseq analyses on lens cells isolated from 3 and 24 month old inbred C57BL/6J strain mice housed under standard laboratory conditions (Faranda et al., 2021) as this minimized potential variability driven by differences in genetic background and environment (Ackert-Bicknell et al., 2015). During the design of this study, we included equal numbers of animals from both sexes as mandated by NIH guidelines (Woitowich et al., 2020), which led us to reexamine these data (Gene Expression Omnibus Accession number GSE166619) to discover how sex affects the lens transcriptome. Principal component analysis (PCA) and a Spearman’s Rho correlation matrix revealed that cell type (epithelium versus fibers) was the largest influence on transcriptomic differences in this study, followed by the effect of injury on the lens epithelial cell (LEC) transcriptome, followed by age. The greatest variation between replicates was seen in samples from injured lens samples suggesting some sample to sample variability in the injury response at 24 hours post surgery. (Supplemental Figure 1). However, as expected for inbred mice housed under controlled conditions, Spearman’s Rho correlation plots comparing the global levels of gene expression between sex specific biological replicates strongly correlate (Supplemental figure 2) although, sex appears to have much less influence on the global lens transcriptome than cell type, injury status, or age (Supplemental figure 1).

The effect of sex on the transcriptome of the young adult lens epithelium

Consistent with the PCA analysis, comparison between the transcriptome of young adult male and female mouse lens epithelial cells revealed only five differentially expressed genes (DEGs) that met our previously developed criteria for “biologically significant” changes that could plausibly affect lens cells (at least 2 fold change in expression level, FDR corrected p value ≤ 0.05, gene expressed at least 2 FPKM in at least one condition (Manthey et al., 2014a)). Of these, three were genes residing on sex chromosomes, while the other two are predicted genes of unknown function that reside on autosomes. All three genes residing on sex chromosomes exhibit sex-dependent expression in other mouse tissues as well based on comparisons with the Sex associated genes data base (SAGD (Shi et al., 2019)), while neither autosomal gene has been reported to be sex-dependent in SAGD (Table 1). Of these genes, only Xist exhibited “lens preferred” expression in 56 day postnatal LECs based on comparisons with the Integrated systems tool for eye gene discovery (iSyTe) database (Kakrana et al., 2018) which likely reflects the sex of the tissue used to create this resource. Too few genes exhibited significant differential expression between male and female young LECs to perform Adviata ipathway guide analyses (Ahsan and Draghici, 2017) on this gene list.

Table 1.

DEGs between young male and female lens epithelial cells. Bold denotes genes that have been previously reported to exhibit sex dependent expression in adult mouse tissues based on comparisons with SAGD (Shi et al., 2019). Italics denotes genes that exhibit lens specific expression in postnatal day 56 mouse lenses according to the iSyTe database (Kakrana et al., 2018).

SYMBOL	DESCRIPTION	Chromosome	Fold Change	FDR	Male FPKM	Female FPKM
Xist	inactive X specific transcripts	X	2756.3	1.7E-91	0.01	28.3
Ddx3y	DEAD (Asp-Glu-Ala-Asp) box polypeptide 3, Y-linked	Y	−8455.0	2.8E-89	6.3	0.0
Eif2s3y	eukaryotic translation initiation factor 2, subunit 3, Y-linked	Y	−4692.1	2.2E-25	4.4	0.0
Gm10800	predicted gene 10800	2	2114.4	3.1E-8	0.00	12.4
Gm42972	predicted gene 42972	3	−2.8	3.5E-2	4.9	1.8

The effect of sex on the response of young adult lens epithelial cells to lens fiber cell removal

Comparison between the transcriptome of young adult male and female mouse lens epithelial cells isolated 24 hours after lens fiber cell removal to model cataract surgery (Jiang et al., 2018; Manthey et al., 2014b) revealed only ten biologically significant DEGs; six are located on sex chromosomes, four on autosomes. Of these, only Xist exhibited lens-preferred expression in adults based on comparisons with the iSyTe database and only five genes (all located on sex chromosomes) exhibit sexually dimorphic gene expression in adult mice in other tissues based on SAGD (Table 2). We were unable to perform iPathway guide analyses of these DEGs as too few genes were identified for the underlying algorithm to detect enriched pathways.

Table 2.

DEGs between young male and female lens epithelial cells at 24 hours PCS. Bold denotes genes that have been previously reported to exhibit sex dependent expression in adult mouse tissues based on comparisons with SAGD (Shi et al., 2019). Italics denotes genes that exhibit lens preferred expression in postnatal day 56 mouse lenses according to the iSyTe database (Kakrana et al., 2018).

SYMBOL	DESCRIPTION	Chromosome	Fold Change	FDR	Male FPKM	Female FPKM
Xist	inactive X specific transcripts	X	1672.4	2.63E-3	0.0	49.5
Gm13375	predicted gene 13375	2	16.2	4.91E-2	0.2	2.5
Eif2s3x	eukaryotic translation initiation factor 2, subunit 3, X-linked [	X	2.7	4.47E-3	9.9	27.2
Serping1	serine (or cysteine) peptidase inhibitor, clade G, member 1	2	2.4	4.38E-2	7.8	18.7
G0s2	G0/G1 switch gene 2	1	−3.1	4.38E-2	35.6	11.6
Ccdc88b	coiled-coil domain containing 88B	19	−3.6	1.01E-2	8.1	2.2
Uty	ubiquitously transcribed tetratricopeptide repeat gene, Y chromosome	Y	−3158.4	2.27E-34	2.1	0.0
Eif2s3y	eukaryotic translation initiation factor 2, subunit 3, Y-linked	Y	−3449.3	2.17E-38	5.4	0.0
Kdm5d	lysine (K)-specific demethylase 5D	Y	−5243.9	7.43E-44	2.2	0.0
Ddx3y	DEAD (Asp-Glu-Ala-Asp) box polypeptide 3, Y-linked	Y	−6043.4	2.11E-59	7.5	0.0

The effect of sex on the young adult lens fiber cell transcriptome

Comparison between the transcriptome of young adult male and female mouse lens fiber cells revealed that 49 genes met the criteria for “biologically significant” differential expression between the sexes, with 28 genes expressed at higher levels in females and 21 expressed at higher levels in males. Of these, five DEGs reside on sex chromosomes while the remainder are autosomal (Table 3). Advaita iPathway guide (Ahsan and Draghici, 2017) analysis of the 191 DEGs meeting statistical criteria for differential expression (FDR ≤ 0.05; absolute fold change > 2; 191 genes total) did not detect any ontology terms or KEGG pathways suggesting common regulated mechanisms.

Table 3.

DEGs between young male and female lens fiber cells. Bold denotes genes that have been previously reported to exhibit sex dependent expression in adult mouse tissues based on comparisons with SAGD (Shi et al., 2019). Italics denotes genes that exhibit lens preferred expression in postnatal day 56 mouse lenses according to the iSyTe database (Kakrana et al., 2018).

SYMBOL	DESCRIPTION	Chromosome	Fold Change	FDR	Male FPKM	Female FPKM
Xist	inactive X specific transcripts	X	312.7	9.72E-48	0.0	10.1
Adrb2	adrenergic receptor, beta 2	18	173.0	7.14E-3	0.0	2.3
Mrpl22	mitochondrial ribosomal protein L22	11	16.5	9.66E-3	0.2	4.1
Tm7sf3	transmembrane 7 superfamily member 3	6	13.1	3.63E-8	0.3	4.3
Selenon	selenoprotein N	4	6.3	8.76E-7	0.7	4.3
Gm10557	predicted gene 10557	18	5.9	6.90E-3	1.9	11.8
Sdhc	succinate dehydrogenase complex, subunit C	1	5.8	5.00E-4	0.6	3.6
Mfsd1	major facilitator superfamily domain containing 1	3	5.7	5.01E-3	0.5	2.9
Olfml3	olfactomedin-like 3	3	5.0	1.31E-5	0.7	3.3
Ppp3cc	protein phosphatase 3, catalytic subunit, gamma isoform	14	4.7	4.52E-2	0.6	2.7
Rab39b	RAB39B, member RAS oncogene family	X	4.1	2.43E-2	0.8	3.4
Rassf7	Ras association (RalGDS/AF-6) domain family (N-terminal) 7	7	3.8	3.79E-2	1.2	4.8
Samd1	sterile alpha motif domain containing 1	8	3.7	1.70E-3	1.6	5.9
Ahcy	S-adenosylhomocysteine hydrolase	2	3.5	1.14E-2	2.4	8.4
Armc10	armadillo repeat containing 10	5	3.3	2.62E-2	1.2	3.9
Bnip2	BCL2/adenovirus E1B interacting protein 2	9	2.8	2.48E-2	1.3	3.6
Aldh3a1	aldehyde dehydrogenase family 3, subfamily A1	11	2.8	7.19E-4	5.5	15.6
Mapkapk2	MAP kinase-activated protein kinase 2	1	2.7	2.01E-2	1.2	3.4
Rbm7	RNA binding motif protein 7	9	2.6	2.35E-2	1.4	3.6
Ptpn2	protein tyrosine phosphatase, non-receptor type 2	18	2.6	2.69E-2	2.6	6.5
Epg5	ectopic P-granules autophagy protein 5 homolog (C. elegans)	18	2.5	1.29E-3	1.4	3.5
Retreg1	reticulophagy regulator 1	15	2.5	5.94E-3	2.4	6.0
Flcn	folliculin	11	2.3	2.62E-2	1.8	4.2
Acat1	acetyl-Coenzyme A acetyltransferase 1	9	2.3	4.15E-2	2.3	5.4
Gstm1	glutathione S-transferase, mu 1	3	2.3	6.70E-4	15.1	34.7
2410004B18Rik	RIKEN cDNA 2410004B18 gene	3	2.2	6.39E-3	8.1	18.2
Zc3h4	zinc finger CCCH-type containing 4	7	2.1	5.86E-4	2.3	4.9
Exoc3	exocyst complex component 3	13	2.0	1.14E-2	3.7	7.5
Arpc5	actin related protein 2/3 complex, subunit 5	1	−2.0	3.02E-2	15.3	7.6
Hmgb3	high mobility group box 3	X	−2.1	5.94E-3	25.3	12.3
Cxxc1	CXXC finger 1 (PHD domain)	18	−2.1	1.81E-2	8.4	4.0
Mtmr4	myotubularin related protein 4	11	−2.2	3.79E-2	5.3	2.4
Arhgef19	Rho guanine nucleotide exchange factor (GEF) 19	4	−2.2	2.46E-2	3.7	1.7
Ahdc1	AT hook, DNA binding motif, containing 1	4	−2.3	5.53E-4	4.1	1.8
Oplah	5-oxoprolinase (ATP-hydrolysing)	15	−2.3	1.14E-2	3.8	1.7
Tm9sf1	transmembrane 9 superfamily member 1	14	−2.7	1.79E-2	3.5	1.3
Zdhhc4	zinc finger, DHHC domain containing 4	5	−2.7	3.85E-2	7.1	2.6
Susd4	sushi domain containing 4	1	−2.9	2.83E-2	5.0	1.7
Spsb3	splA/ryanodine receptor domain and SOCS box containing 3	17	−2.9	4.52E-2	6.3	2.1
Pygm	muscle glycogen phosphorylase	19	−3.1	7.34E-3	4.6	1.5
Gm14322	predicted gene 14322	2	−3.2	3.71E-2	13.3	4.1
Map2k1	mitogen-activated protein kinase kinase 1	9	−3.3	1.15E-4	6.2	1.9
Cntfr	ciliary neurotrophic factor receptor	4	−3.8	4.15E-2	3.2	0.8
Yipf2	Yip1 domain family, member 2	9	−6.4	5.09E-3	4.6	0.7
S100a1	S100 calcium binding protein A1	3	−17.3	1.02E-2	6.7	0.3
Tmub1	transmembrane and ubiquitin-like domain containing 1	5	−18.5	1.04E-3	2.6	0.1
A930036I15Rik	RIKEN cDNA A930036I15 gene	3	−81.7	3.91E-2	2.2	0.0
Ddx3y	DEAD (Asp-Glu-Ala-Asp) box polypeptide 3, Y-linked	Y	−335.4	1.68E-43	9.6	0.0
Eif2s3y	eukaryotic translation initiation factor 2, subunit 3, Y-linked	Y	−570.3	7.04E-11	3.9	0.0

Analysis of these DEGs using iSyte (Kakrana et al., 2018) revealed that only three genes differentially expressed between male and female young lens fibers (Gstm1, Aldh3a1, and Pygm) exhibited lens preferred expression in 56-day old whole lenses. Four DEGs located on sex chromosomes and five autosomal DEGs were previously found to exhibit sexually dimorphic expression in other mouse tissues based on comparisons with SAGD (Table 3).

The effect of sex on the aged lens epithelial cell transcriptome

Comparison between the transcriptome of aged male and female mouse LECs revealed 37 DEGs between the sexes, with 24 genes expressed at higher levels in females and 13 expressed at higher levels in males. Of these, three DEGs reside on sex chromosomes while the remainder are autosomal (Table 4). Three DEGs located on sex chromosomes and nine autosomal DEGs were previously found to exhibit sex dependent expression in other mouse tissues based on comparisons with SAGD (Table 4). Analysis of all DEGs with a statistically significant difference in expression (FDR < 0.05, absolute fold change ≥ 2, 74 genes total) for impacted pathways or enriched gene ontology (GO) terms using iPathway guide (Ahsan and Draghici, 2017) revealed that “lens development in camera type eye” (FDR corrected p value 5 X10⁻⁷) and structural constituent of the eye lens (FDR 2.3 X10⁻¹¹) as enriched GO terms.-As in young LECs, KEGG Pathway analysis was not informative.

Table 4.

DEGs between aged male and female lens epithelial cells. Bold denotes genes that have been previously reported to exhibit sex dependent expression in adult mouse tissues based on comparisons with SAGD (Shi et al., 2019). Italics denotes genes that exhibit lens preferred expression in postnatal day 56 mouse lenses according to the iSyTe database (Kakrana et al., 2018).

SYMBOL	DESCRIPTION	Chromosome	Fold Change	FDR	Male FPKM	Female FPKM
Xist	inactive X specific transcripts	X	1932.6	3.61E-92	0.0	34.6
Crygd	crystallin, gamma D	1	60.3	2.36E-2	0.0	2.8
Gm49958	predicted gene, 49958	17	7.9	6.62E-5	0.3	2.5
Rgs4	regulator of G-protein signaling 4	1	5.2	1.29E-2	0.6	3.3
Lgsn	lengsin, lens protein with glutamine synthetase domain	1	4.5	4.79E-3	19.9	88.6
Pip5kl1	phosphatidylinositol-4-phosphate 5-kinase-like 1	2	3.6	1.18E-3	1.1	3.8
Gm42808	predicted gene 42808	5	3.5	5.55E-3	10.6	36.9
Fabp5	fatty acid binding protein 5, epidermal	3	3.3	1.18E-2	14.8	48.1
Acta2	actin, alpha 2, smooth muscle, aorta	19	3.3	4.76E-9	10.8	35.1
Cryba4	crystallin, beta A4	5	3.2	5.90E-3	179.0	581.3
Rny1	RNA, Y1 small cytoplasmic, Ro-associated	6	3.2	3.16E-2	10.8	34.7
Pstpip2	proline-serine-threonine phosphatase-interacting protein 2	18	2.8	1.11E-3	1.3	3.8
Mboat1	membrane bound O-acyltransferase domain containing 1	13	2.8	2.73E-9	3.2	8.9
Crybb1	crystallin, beta B1	5	2.4	1.50E-2	173.1	407.3
Cryba1	crystallin, beta A1	11	2.3	8.08E-3	356.8	830.6
Crygs	crystallin, gamma S	16	2.3	8.58E-3	567.8	1321.6
Lctl	lactase-like	9	2.3	2.66E-2	7.5	17.0
Smco3	single-pass membrane protein with coiled-coil domains 3	6	2.3	5.34E-4	2.4	5.4
Basp1	brain abundant, membrane attached signal protein 1	15	2.2	8.58E-3	39.9	87.0
Snhg12	small nucleolar RNA host gene 12	4	2.2	1.06E-2	1.8	3.9
Dpf3	D4, zinc and double PHD fingers, family 3	12	2.2	8.58E-3	1.9	4.1
Tmem40	transmembrane protein 40	6	2.1	8.96E-3	5.0	10.6
Arhgap31	Rho GTPase activating protein 31	16	2.1	5.34E-4	2.1	4.4
Cryba2	crystallin, beta A2	1	2.1	4.99E-3	661.3	1371.1
Efemp1	epidermal growth factor-containing fibulin-like extracellular matrix protein 1	11	−2.0	2.36E-2	33.6	16.7
Slc6a6	solute carrier family 6 (neurotransmitter transporter, taurine), member 6	6	−2.0	1.83E-4	128.6	62.7
Ntrk2	neurotrophic tyrosine kinase, receptor, type 2	13	−2.3	9.69E-3	4.6	2.0
F5	coagulation factor V	1	−2.3	9.80E-3	15.2	6.5
Gpx3	glutathione peroxidase 3	11	−2.4	9.69E-5	229.3	97.4
Matn2	matrilin 2	15	−2.6	3.26E-2	5.7	2.2
Plin4	perilipin 4	17	−3.2	1.76E-2	3.7	1.2
Col9a1	collagen, type IX, alpha 1	1	−3.5	2.00E-6	11.9	3.4
Crhbp	corticotropin releasing hormone binding protein	13	−3.5	1.18E-3	21.4	6.2
Fbn1	fibrillin 1	2	−3.7	3.58E-10	32.3	8.6
Krt12	keratin 12	11	−27.8	1.18E-3	3.5	0.1
Eif2s3y	eukaryotic translation initiation factor 2, subunit 3, Y-linked	Y	−592.9	5.01E-21	2.6	0.0
Ddx3y	DEAD (Asp-Glu-Ala-Asp) box polypeptide 3, Y-linked	Y	−856.2	5.65E-93	7.8	0.0

Analysis of these DEGs using iSyte (Kakrana et al., 2018) revealed that 14 DEGs exhibited lens preferred expression in adult lenses. Comparison of genes differentially expressed in male and female aged LECs with those found to differentially expressed between young versus old LECs when sex was not considered as a variable (Faranda et al 2021), revealed that 10 of the 37 genes differentially expressed by sex in aged LECs also significantly change expression level during aging. While the direction of change in males and females was the same in all cases, the extent of these changes was sexually dimorphic (Table 5).

Table 5.

The genes whose expression changes significantly with age in both sexes, and are differentially expressed between lens epithelial cells from aged female mice and male mice.

SYMBOL	DESCRIPTION	aged vs young FC	aged vs young FDR	young FPKM	aged FPKM	aged female vs male FC	aged female vs male FDR	Male aged FPKM	Female aged FPKM
Cryba1	crystallin, beta A1	−3.05	3.33E-05	1807.54	590.89	2.33	8.08E-03	356.79	830.56
Crybb1	crystallin, beta B1	−2.66	1.08E-03	769.04	288.84	2.35	1.50E-02	173.11	407.3
Slc6a6	solute carrier family 6 (neurotransmitter transporter, taurine), member 6	2.17	1.33E-03	43.78	95.29	−2.06	1.83E-04	128.59	62.73
Crygs	crystallin, gamma S	−2.39	3.03E-03	2255.19	940.23	2.33	8.58E-03	567.75	1321.57
Fabp5	fatty acid binding protein 5, epidermal	−3.32	4.14E-03	104.03	31.26	3.25	1.18E-02	14.75	48.07
Cryba2	crystallin, beta A2	−2.01	6.83E-03	2030.8	1011.47	2.07	4.99E-03	661.25	1371.15
Crygd	crystallin, gamma D	−85.04	1.16E-02	119.19	1.4	60.13	2.36E-02	0.04	2.77
Cryba4	crystallin, beta A4	−2.87	1.29E-02	1087.55	378.32	3.25	5.90E-03	178.99	581.28
Gpx3	glutathione peroxidase 3	2.66	4.68E-02	61.17	162.73	−2.36	9.69E-05	229.28	97.36
Matn2	matrilin 2	2.31	4.93E-02	1.71	3.95	−2.62	3.26E-02	5.75	2.19

The effect of sex on the response of aged lens epithelial cells to lens fiber cell removal

Comparison between the transcriptome of aged male and female mouse LECs at 24 hours PCS revealed 430 biologically significant DEGs between the sexes, with 267 genes expressed at higher levels in females and 163 expressed at higher levels males. Of these, 17 DEGs reside on sex chromosomes, 8 are encoded on the mitochondrial genome, while the remainder are autosomal (Table 6). Six DEGs located on sex chromosomes, one on the mitochondrial genome, and 90 autosomal DEGs were previously found to exhibit sex dependent expression in other mouse tissues based on comparisons with SAGD (Table 6). Comparison of this list with the iSyTE database revealed that 30 DEGs between male and female LECs at 24 hours PCS exhibit lens preferred expression in adult lenses.

Table 6.

DEGs between aged male and female lens epithelial cells at 24 hours PCS Bold denotes genes that have been previously reported to exhibit sex dependent expression in adult mouse tissues based on comparisons with SAGD (Shi et al., 2019).. Italics denotes genes that exhibit lens preferred expression in postnatal day 56 mouse lenses according to the iSyTe database (Kakrana et al., 2018).

SYMBOL	DESCRIPTION	Chromosome	Fold Change	FDR	Male FPKM	Female FPKM
Endog	endonuclease G	2	400.3	1.52E-5	0.0	2.7
A730094K22Rik	RIKEN cDNA A730094K22 gene	9	361.3	2.06E-4	0.0	3.5
Gm24644	predicted gene, 24644	2	337.4	8.11E-4	0.0	17.5
Gm18840	predicted gene, 18840	6	264.2	2.65E-2	0.0	3.7
Xist	inactive X specific transcripts	X	107.5	1.47E-2	0.5	54.8
Lipt2	lipoyl(octanoyl) transferase 2 (putative)	7	21.3	2.66E-4	0.1	2.5
Rpl17-ps8	ribosomal protein L17, pseudogene 8	X	13.8	3.63E-3	0.4	5.6
Steap4	STEAP family member 4	5	13.8	3.13E-2	2.3	31.4
Efemp1	epidermal growth factor-containing fibulin-like extracellular matrix protein 1	11	13.4	1.35E-8	4.2	57.2
Foxi3	forkhead box I3	6	12.1	2.48E-3	0.4	4.4
Kcnj13	potassium inwardly-rectifying channel, subfamily J, member 13	1	12.1	9.93E-3	0.3	3.8
Lgi2	leucine-rich repeat LGI family, member 2	5	11.3	5.49E-4	0.5	5.5
Myl6b	myosin, light polypeptide 6B]	10	11.1	4.34E-3	0.3	3.0
C1ra	complement component 1, r subcomponent A	6	10.8	2.99E-5	0.6	6.5
Slc6a13	solute carrier family 6 (neurotransmitter transporter, GABA), member 13	6	10.4	3.98E-4	0.6	6.8
Rida	reactive intermediate imine deaminase A homolog	15	10.3	1.71E-3	0.3	3.5
Chil1	chitinase-like 1	1	9.9	2.76E-2	7.9	78.4
Dio3	deiodinase, iodothyronine type III	12	9.8	2.50E-2	1.0	10.1
Klhdc7a	kelch domain containing 7A	4	9.6	2.10E-6	0.3	3.2
Npvf	neuropeptide VF precursor	6	9.5	1.33E-2	1.3	12.2
Six6	sine oculis-related homeobox 6	12	9.5	2.33E-4	0.4	3.6
Tyrp1	tyrosinase-related protein 1	4	9.1	4.52E-4	0.4	4.1
Mfrp	membrane frizzled-related protein	9	8.9	1.06E-3	0.5	4.6
A2m	alpha-2-macroglobulin	6	8.8	3.65E-3	1.9	16.5
Bmp4	bone morphogenetic protein 4	14	8.7	1.52E-3	0.7	6.5
Chrm1	cholinergic receptor, muscarinic 1, CNS	19	8.7	2.39E-4	0.4	3.5
Slc7a4	solute carrier family 7 (cationic amino acid transporter, y+ system), member 4	16	8.6	5.18E-3	0.5	4.0
Sod3	superoxide dismutase 3, extracellular	5	8.5	4.27E-2	18.5	156.7
Slc16a2	solute carrier family 16 (monocarboxylic acid transporters), 2	X	8.5	1.25E-2	2.1	18.1
Rab3b	RAB3B, member RAS oncogene family	4	7.8	2.41E-2	0.4	3.3
Rtl8c	retrotransposon Gag like 8C	X	7.8	1.67E-2	0.9	7.5
Macrod1	mono-ADP ribosylhydrolase 1	19	7.8	1.90E-2	0.3	2.3
Cdh11	cadherin 11	8	7.7	7.28E-4	0.3	2.5
Mme	membrane metallo endopeptidase	3	7.5	3.56E-2	1.7	13.0
Prkcq	protein kinase C, theta	2	7.4	5.13E-3	0.5	4.0
Wfdc1	WAP four-disulfide core domain 1	8	7.2	4.73E-3	0.8	5.9
Otx1	orthodenticle homeobox 1	11	7.0	4.85E-2	0.5	3.4
Sord	sorbitol dehydrogenase	2	7.0	1.49E-2	1.0	6.8
Cldn2	claudin 2	X	6.9	6.83E-4	0.6	4.1
Ajap1	adherens junction associated protein 1	4	6.4	1.14E-2	0.4	2.8
Cavin2	caveolae associated 2	1	6.4	9.88E-3	2.8	17.8
Map3k21	mitogen-activated protein kinase kinase kinase 21	8	6.4	7.27E-3	0.6	3.9
Zmat5	zinc finger, matrin type 5	11	6.3	1.33E-2	0.6	3.6
Rab3il1	RAB3A interacting protein (rabin3)-like 1	19	6.3	2.11E-2	0.4	2.5
Zic1	zinc finger protein of the cerebellum 1	9	6.2	2.35E-3	1.2	7.2
Gm14322	predicted gene 14322	2	6.2	9.97E-3	0.5	3.2
Ephb2	Eph receptor B2	4	6.1	2.02E-5	0.4	2.4
Selenop	selenoprotein P	15	6.0	2.10E-2	1.6	9.9
Pdgfra	platelet derived growth factor receptor, alpha polypeptide	5	6.0	4.97E-2	0.4	2.5
Slc13a4	solute carrier family 13 (sodium/sulfate symporters), member 4	6	5.8	9.49E-4	1.5	8.9
Apcdd1	adenomatosis polyposis coli down-regulated 1	18	5.8	9.60E-4	0.5	2.8
Serping1	serine (or cysteine) peptidase inhibitor, clade G, member 1	2	5.7	1.13E-12	4.8	27.6
Aass	aminoadipate-semialdehyde synthase	6	5.6	1.62E-2	0.5	2.9
Mrgprf	MAS-related GPR, member F	7	5.6	1.64E-2	0.5	2.9
Zic2	zinc finger protein of the cerebellum 2	14	5.6	2.19E-4	1.9	10.7
Mab21l2	mab-21-like 2	3	5.5	6.22E-6	1.9	10.6
Lrp2	low density lipoprotein receptor-related protein 2	2	5.5	6.15E-4	1.0	5.3
Nr2f1	nuclear receptor subfamily 2, group F, member 1	13	5.4	3.42E-2	0.9	5.1
Cp	ceruloplasmin	3	5.4	3.08E-6	21.7	117.4
Tmem176b	transmembrane protein 176B	6	5.4	1.24E-7	7.0	37.9
Adamts5	a disintegrin-like and metallopeptidase (reprolysin type) with thrombospondin type 1 motif, 5 (aggrecanase-2)	16	5.3	5.58E-10	1.2	6.2
C1qb	complement component 1, q subcomponent, beta polypeptide	4	5.3	9.88E-3	1.5	8.3
Gm14325	predicted gene 14325	2	5.1	2.51E-2	0.8	4.4
Lsr	lipolysis stimulated lipoprotein receptor	7	5.1	4.12E-2	0.8	4.0
F5	coagulation factor V	1	5.0	4.02E-5	2.8	14.3
Bcam	basal cell adhesion molecule	7	5.0	1.27E-4	1.4	7.2
Nid2	nidogen 2	14	4.8	1.20E-2	1.9	9.1
Snapc5	small nuclear RNA activating complex, polypeptide 5	9	4.8	2.02E-2	0.9	4.4
Rhoj	ras homolog family member J	12	4.6	1.12E-2	2.7	12.4
Reln	reelin	5	4.6	2.88E-6	4.0	18.6
Fmod	fibromodulin	1	4.6	1.89E-5	3.7	17.0
Colgalt2	collagen beta(1-O)galactosyltransferase 2	1	4.6	3.08E-6	1.1	5.1
Pcyt1b	phosphate cytidylyltransferase 1, choline, beta isoform	X	4.6	2.41E-3	1.9	8.6
Olfml2a	olfactomedin-like 2A	2	4.6	2.62E-7	4.7	21.7
Vat1l	vesicle amine transport protein 1 like	8	4.6	3.45E-4	1.0	4.7
Hopx	HOP homeobox	5	4.6	6.42E-3	1.4	6.6
S1pr3	sphingosine-1-phosphate receptor 3	13	4.5	4.38E-9	6.6	29.6
Jam2	junction adhesion molecule 2	16	4.3	1.65E-2	1.6	6.8
Tmem176a	transmembrane protein 176 A	6	4.3	2.79E-4	2.6	11.4
Aqp4	aquaporin 4	18	4.3	8.57E-8	3.2	13.7
Rtn4rl1	reticulon 4 receptor-like 1	11	4.3	1.26E-3	1.9	8.2
Gpm6a	glycoprotein m6a	8	4.3	8.36E-4	2.0	8.6
Il1r1	interleukin 1 receptor, type I	1	4.3	3.56E-2	1.2	5.2
Enpp5	ectonucleotide pyrophosphatase/phosphodiesterase 5	17	4.2	1.89E-3	1.4	6.1
Trp53i11	transformation related protein 53 inducible protein 11	2	4.2	2.72E-3	1.0	4.3
Sema3d	sema domain, immunoglobulin domain (Ig), short basic domain, secreted, (semaphorin) 3D	5	4.2	2.67E-2	1.4	5.7
Fbn1	fibrillin 1	2	4.2	2.04E-3	8.1	33.8
C1s1	complement component 1, s subcomponent 1	6	4.2	2.43E-2	1.3	5.3
Mturn	maturin, neural progenitor differentiation regulator homolog	6	4.1	2.20E-2	1.9	7.8
Ogn	osteoglycin	13	4.1	2.92E-2	3.7	15.2
Klf15	Kruppel-like factor 15	6	4.1	1.74E-2	0.9	3.6
Rarres2	retinoic acid receptor responder (tazarotene induced) 2	6	4.1	4.42E-4	1.4	5.6
Napepld	N-acyl phosphatidylethanolamine phospholipase D	5	4.0	4.10E-2	1.4	5.4
Cdh4	cadherin 4	2	4.0	9.60E-4	1.5	6.0
Lrrc75b	leucine rich repeat containing 75B	10	4.0	1.90E-2	1.0	3.9
Lypd6	LY 6/PLAUR domain containing 6	2	4.0	4.40E-2	1.2	4.7
Flrt1	fibronectin leucine rich transmembrane protein 1	19	4.0	8.28E-5	1.5	5.8
Pcx	pyruvate carboxylase	19	3.9	4.63E-5	1.0	4.0
Gpx3	glutathione peroxidase 3	11	3.9	3.04E-3	58.8	230.7
Zfp982	zinc finger protein 982	4	3.9	1.02E-2	1.2	4.6
Col18a1	collagen, type XVIII, alpha 1	10	3.9	3.32E-2	23.3	91.0
Serpinb9	serine (or cysteine) peptidase inhibitor, clade B, member 9	13	3.9	4.39E-2	0.8	3.3
C4b	complement component 4B (Chido blood group)	17	3.8	2.46E-4	3.7	14.3
Amph	amphiphysin	13	3.8	1.50E-5	5.3	20.2
Mrc2	mannose receptor, C type 2	11	3.8	2.70E-2	1.2	4.6
Alad	aminolevulinate, delta-, dehydratase	4	3.8	3.07E-2	7.3	27.7
Col1a2	collagen, type I, alpha 2	6	3.8	9.39E-3	0.9	3.5
Prelp	proline arginine-rich end leucine-rich repeat	1	3.8	2.63E-3	5.2	19.8
Optc	opticin	1	3.8	7.73E-5	21.5	81.0
Otulinl	OTU deubiquitinase with linear linkage specificity like	15	3.7	4.30E-2	0.8	3.0
Sorcs1	sortilin-related VPS10 domain containing receptor 1	19	3.7	2.34E-2	0.8	2.9
Ephx2	epoxide hydrolase 2, cytoplasmic	14	3.7	1.01E-3	3.0	10.9
Csad	cysteine sulfinic acid decarboxylase	15	3.6	4.33E-2	3.9	13.9
Ntrk2	neurotrophic tyrosine kinase, receptor, type 2	13	3.6	6.51E-3	1.0	3.4
Fbln1	fibulin 1	15	3.5	3.09E-4	15.3	54.3
Smarcd3	SWI/SNF related, matrix associated, actin dependent regulator of chromatin, subfamily d, member 3	5	3.5	2.97E-6	1.8	6.5
Mtus1	mitochondrial tumor suppressor 1	8	3.5	2.89E-4	2.0	7.0
Pla2g5	phospholipase A2, group V	4	3.5	1.71E-3	4.7	16.5
Copz2	coatomer protein complex, subunit zeta 2	11	3.5	2.54E-2	1.3	4.4
Abat	4-aminobutyrate aminotransferase	16	3.5	9.37E-3	2.1	7.3
Ipp	IAP promoted placental gene	4	3.5	4.84E-2	0.8	2.9
Atp2b4	ATPase, Ca++ transporting, plasma membrane 4	1	3.5	2.18E-4	2.0	6.9
Fundc1	FUN14 domain containing 1	X	3.5	1.10E-2	3.0	10.4
Mgst1	microsomal glutathione S-transferase 1	6	3.4	9.57E-6	3.6	12.6
Ccn5	cellular communication network factor 5	2	3.4	1.23E-2	3.0	10.4
Hacd4	3-hydroxyacyl-CoA dehydratase 4	4	3.4	4.17E-2	1.0	3.3
Galnt12	polypeptide N-acetylgalactosaminyltransferase 12	4	3.4	3.17E-2	1.5	5.2
Boc	biregional cell adhesion molecule-related/down-regulated by oncogenes (Cdon) binding protein	16	3.4	1.84E-3	1.6	5.4
Atp1a2	ATPase, Na+/K+ transporting, alpha 2 polypeptide	1	3.4	6.34E-3	29.1	99.1
Pafah2	platelet-activating factor acetylhydrolase 2	4	3.4	2.93E-3	1.2	4.2
Cdon	cell adhesion molecule-related/down-regulated by oncogenes	9	3.4	2.66E-5	4.3	14.7
Crhbp	corticotropin releasing hormone binding protein	13	3.4	4.54E-2	5.7	19.4
Lhx2	LIM homeobox protein 2	2	3.4	2.22E-2	1.6	5.3
Oplah	5-oxoprolinase (ATP-hydrolysing)	15	3.3	7.22E-4	1.5	4.9
St8sia1	ST8 alpha-N-acetyl-neuraminide alpha-2,8-sialyltransferase 1	6	3.3	5.52E-3	1.2	4.1
Fcgrt	Fc receptor, IgG, alpha chain transporter	7	3.3	3.65E-3	2.3	7.5
Oat	ornithine aminotransferase	7	3.3	1.15E-2	15.7	52.0
Vcan	versican	13	3.3	7.61E-3	4.0	13.3
Pdk2	pyruvate dehydrogenase kinase, isoenzyme 2	11	3.3	9.84E-3	4.2	13.7
Col23a1	collagen, type XXIII, alpha 1	11	3.3	3.96E-2	8.1	26.7
Anapc13	anaphase promoting complex subunit 13	9	3.3	4.80E-2	2.3	7.5
Scx	scleraxis	15	3.2	1.45E-3	1.9	6.2
Cyp39a1	cytochrome P450, family 39, subfamily a, polypeptide 1	17	3.2	1.19E-2	1.2	3.9
Pdcd4	programmed cell death 4	19	3.2	2.11E-2	17.1	54.7
Paxx	non-homologous end joining factor	2	3.2	2.74E-2	2.0	6.3
Pcolce	procollagen C-endopeptidase enhancer protein	5	3.2	5.69E-4	3.9	12.2
Ildr2	immunoglobulin-like domain containing receptor 2	1	3.2	6.81E-7	5.6	17.5
Gm7694	predicted gene 7694	1	3.2	3.27E-4	3.3	10.4
Mrpl34	mitochondrial ribosomal protein L34	8	3.2	4.23E-2	3.5	11.3
Jup	junction plakoglobin	11	3.2	1.62E-3	1.5	4.6
Gas1	growth arrest specific 1	13	3.2	7.41E-6	7.6	24.0
9130023H24Rik	RIKEN cDNA 9130023H24 gene	7	3.1	1.44E-2	1.0	3.1
Zfp185	zinc finger protein 185	X	3.1	4.76E-5	3.3	10.2
Oxct1	3-oxoacid CoA transferase 1	15	3.1	1.03E-4	3.1	9.5
Slc1a3	solute carrier family 1 (glial high affinity glutamate transporter), 3	15	3.1	5.18E-3	1.2	3.7
Prxl2a	peroxiredoxin like 2A	14	3.1	7.30E-6	7.4	22.7
Rcan2	regulator of calcineurin 2	17	3.1	1.02E-2	3.3	10.2
Isyna1	myo-inositol 1-phosphate synthase A1	8	3.1	9.03E-3	6.2	19.0
Epm2a	epilepsy, progressive myoclonic epilepsy, type 2 gene alpha	10	3.1	2.37E-2	1.0	3.2
Atp1b3	ATPase, Na+/K+ transporting, beta 3 polypeptide	9	3.0	2.43E-2	53.6	163.6
Pnmal2	PNMA-like 2	7	3.0	1.86E-2	5.5	16.6
Rorc	RAR-related orphan receptor gamma	3	3.0	1.91E-2	1.7	5.1
Epas1	endothelial PAS domain protein 1	17	3.0	1.90E-2	5.5	16.6
Ndufb4	NADH:ubiquinone oxidoreductase subunit B4	16	3.0	8.21E-3	4.2	12.6
Ech1	enoyl coenzyme A hydratase 1, peroxisomal	7	3.0	7.41E-4	2.4	7.0
Ndrg1	N-myc downstream regulated gene 1	15	3.0	7.59E-3	8.0	23.7
Acadsb	acyl-Coenzyme A dehydrogenase, short/branched chain	7	3.0	1.52E-3	2.4	7.1
Scara3	scavenger receptor class A, member 3	14	3.0	1.02E-2	1.5	4.4
Lix1l	Lix1-like	3	3.0	1.34E-4	3.2	9.3
Abi3bp	ABI gene family, member 3 (NESH) binding protein	16	2.9	3.12E-3	4.4	12.9
Stra6	stimulated by retinoic acid gene 6	9	2.9	2.66E-2	1.3	4.0
Sntb1	syntrophin, basic 1	15	2.9	9.84E-3	3.3	9.7
Ephb6	Eph receptor B6	6	2.9	7.79E-3	1.3	3.7
Loxl1	lysyl oxidase-like 1	9	2.9	2.29E-5	11.3	33.1
Fat4	FAT atypical cadherin 4	3	2.9	3.62E-3	6.1	17.9
Itgb8	integrin beta 8	12	2.9	2.06E-2	1.8	5.2
Sesn1	sestrin 1	10	2.9	1.94E-2	3.3	9.6
Irs1	insulin receptor substrate 1	1	2.9	3.46E-3	2.0	5.8
Gldc	glycine decarboxylase	19	2.9	1.62E-3	5.0	14.3
Porcn	porcupine O-acyltransferase	X	2.9	1.12E-3	3.4	9.7
Gulp1	GULP, engulfment adaptor PTB domain containing 1	1	2.9	1.52E-2	1.3	3.8
Sqor	sulfide quinone oxidoreductase	2	2.9	3.69E-3	1.1	3.1
Fbxo10	F-box protein 10	4	2.9	1.23E-3	1.4	4.1
Dll 1	delta like canonical Notch ligand 1	17	2.9	1.92E-2	2.4	6.7
Plaat3	phospholipase A and acyltransferase 3	19	2.9	3.56E-2	1.8	5.1
Ptgds	prostaglandin D2 synthase (brain)	2	2.8	1.37E-3	116.1	329.2
Dync2li 1	dynein cytoplasmic 2 light intermediate chain 1	17	2.8	4.17E-2	1.5	4.2
Ndufa4	Ndufa4, mitochondrial complex associated	6	2.8	4.72E-2	2.2	6.2
Zhx1	zinc fingers and homeoboxes 1	15	2.8	1.95E-2	2.2	6.3
Alg14	asparagine-linked glycosylation 14	3	2.8	1.71E-2	1.3	3.6
Chchd10	coiled-coil-helix-coiled-coil-helix domain containing 10	10	2.8	4.27E-4	7.0	19.4
Fads2	fatty acid desaturase 2	19	2.8	2.70E-4	6.9	19.2
Surf1	surfeit gene 1	2	2.7	6.80E-3	3.3	9.1
Pdpr	pyruvate dehydrogenase phosphatase regulatory subunit	8	2.7	5.96E-3	1.3	3.5
Rbms2	RNA binding motif, single stranded interacting protein 2	10	2.7	5.19E-3	1.7	4.7
Idh1	isocitrate dehydrogenase 1 (NADP+), soluble	1	2.7	3.60E-3	3.3	8.9
Rnf43	ring finger protein 43]	11	2.7	7.45E-3	1.3	3.6
Aldh6a1	aldehyde dehydrogenase family 6, subfamily A1	12	2.7	2.89E-2	1.6	4.2
Parm1	prostate androgen-regulated mucin-like protein 1	5	2.6	1.26E-2	3.4	9.1
Igfbp4	insulin-like growth factor binding protein 4	11	2.6	2.02E-2	6.9	18.2
Shisa2	shisa family member 2	14	2.6	4.95E-2	1.6	4.2
Fam107a	family with sequence similarity 107, member A	14	2.6	2.76E-2	2.5	6.5
Hltf	helicase-like transcription factor	3	2.6	5.11E-3	1.7	4.3
Gm44250	predicted gene, 44250	6	2.6	1.54E-2	1.3	3.3
Slc4a5	solute carrier family 4, sodium bicarbonate cotransporter, member 5	6	2.6	2.86E-2	3.0	7.7
Scp2	sterol carrier protein 2, liver [	4	2.6	9.03E-3	9.4	24.1
Antxr1	anthrax toxin receptor 1	6	2.6	5.58E-3	5.3	13.5
Rdh10	retinol dehydrogenase 10 (all-trans)	1	2.5	1.03E-2	8.9	22.7
Slc22a17	solute carrier family 22 (organic cation transporter), member 17	14	2.5	1.06E-2	2.8	7.2
Apoe	apolipoprotein E	7	2.5	1.18E-2	238.9	605.0
Acsl3	acyl-CoA synthetase long-chain family member 3	1	2.5	7.41E-4	6.2	15.7
Ace	angiotensin I converting enzyme (peptidyl-dipeptidase A) 1	11	2.5	1.03E-2	1.5	3.9
Scd1	stearoyl-Coenzyme A desaturase 1	19	2.5	4.13E-2	3.1	7.8
Ldhb	lactate dehydrogenase B	6	2.5	1.15E-2	8.9	22.4
Crb2	crumbs family member 2	2	2.5	3.34E-2	3.2	8.1
Scel	sciellin	14	2.5	2.03E-2	12.3	30.8
Tsc22d1	TSC22 domain family, member 1	14	2.5	2.93E-3	30.2	74.8
Sod2	superoxide dismutase 2, mitochondrial	17	2.5	1.61E-3	3.0	7.4
Kcnk1	potassium channel, subfamily K, member 1	8	2.4	1.18E-3	7.4	18.2
Tmem106c	transmembrane protein 106C	15	2.4	1.63E-2	3.4	8.3
Tenm4	teneurin transmembrane protein 4	7	2.4	1.61E-3	3.1	7.5
Gas6	growth arrest specific 6	8	2.4	4.05E-3	15.5	37.7
Tns2	tensin 2	15	2.4	2.76E-2	3.9	9.4
Cox7b	cytochrome c oxidase subunit 7B	X	2.4	1.90E-2	11.0	26.2
Vps41	VPS41 HOPS complex subunit	13	2.4	4.16E-3	2.7	6.3
Ddah2	dimethylarginine dimethylaminohydrolase 2	17	2.4	1.47E-2	5.2	12.3
Sort1	sortilin 1	3	2.4	1.25E-3	4.5	10.5
Cd99l2	CD99 antigen-like 2	X	2.4	1.13E-2	4.6	10.9
Gpc6	glypican 6	14	2.3	2.96E-2	1.8	4.3
Rfxap	regulatory factor X-associated protein	3	2.3	4.49E-2	3.7	8.6
Ids	iduronate 2-sulfatase	X	2.3	1.57E-2	2.8	6.4
Tspan7	tetraspanin 7	X	2.3	7.47E-3	2.5	5.8
Efna5	ephrin A5	17	2.3	2.69E-3	5.0	11.6
Phactr2	phosphatase and actin regulator 2	10	2.3	9.39E-3	4.1	9.3
Nfia	nuclear factor I/A	4	2.3	8.76E-3	2.7	6.1
Tmem94	transmembrane protein 94	11	2.3	2.35E-2	3.6	8.3
Yipf4	Yip1 domain family, member 4	17	2.3	8.43E-3	4.3	9.8
Ube2d-ps	ubiquitin-conjugating enzyme E2D, pseudogene	11	2.3	3.94E-2	2.1	4.8
Cxxc5	CXXC finger 5	18	2.3	2.85E-3	5.6	12.8
Gpc4	glypican 4	X	2.3	5.44E-3	8.0	18.3
Mmut	methylmalonyl-Coenzyme A mutase	17	2.3	4.60E-2	2.3	5.3
Zfp52	zinc finger protein 52	17	2.3	3.90E-2	2.1	4.7
Nek7	NIMA (never in mitosis gene a)-related expressed kinase 7	1	2.3	3.42E-2	2.4	5.5
Ptpn13	protein tyrosine phosphatase, non-receptor type 13	5	2.2	3.13E-4	2.3	5.1
Cst3	cystatin C	2	2.2	1.02E-2	64.4	144.5
Uqcr11	ubiquinol-cytochrome c reductase, complex III subunit XI	10	2.2	3.00E-2	8.7	19.5
Cadm2	cell adhesion molecule 2	16	2.2	2.19E-2	4.2	9.3
Phyh	phytanoyl-CoA hydroxylase	2	2.2	7.59E-3	5.7	12.7
Insig1	insulin induced gene i	5	2.2	2.55E-2	6.7	14.8
Bcl6	B cell leukemia/lymphoma 6	16	2.2	3.22E-2	2.6	5.8
Fads1	fatty acid desaturase 1	19	2.2	9.88E-3	9.5	21.0
Gprc5c	G protein-coupled receptor, family C, group 5, member C	11	2.2	3.13E-2	3.5	7.6
Pygb	brain glycogen phosphorylase	2	2.2	9.93E-3	5.5	12.0
Gpalpp1	GPALPP motifs containing 1	14	2.1	3.35E-2	2.6	5.5
Etv5	ets variant 5	16	2.1	4.17E-2	5.9	12.5
Laptm4a	lysosomal-associated protein transmembrane 4A	12	2.1	2.37E-2	25.8	54.1
Il11ra1	interleukin 11 receptor, alpha chain 1	4	2.1	3.52E-2	3.3	7.0
Dusp3	dual specificity phosphatase 3	11	2.1	1.35E-2	3.1	6.4
Slc7a11	solute carrier family 7 (cationic amino acid transporter, y+ system), member 11	3	2.1	1.33E-2	13.7	28.5
Igsf8	immunoglobulin superfamily, member 8	1	2.1	4.85E-2	3.4	7.1
Scrn1	secernin 1	6	2.1	4.56E-2	4.0	8.4
Prickle2	prickle planar cell polarity protein 2	6	2.1	1.23E-2	2.4	4.9
Axin2	axin 2	11	2.1	2.76E-2	3.3	6.9
Fdft1	farnesyl diphosphate farnesyl transferase 1	14	2.1	1.02E-2	5.0	10.3
Creg1	cellular repressor of E1A-stimulated genes 1	1	2.1	3.23E-2	5.9	12.1
Pdhx	pyruvate dehydrogenase complex, component X	2	2.0	3.18E-2	3.1	6.4
Ndufb10	NADH:ubiquinone oxidoreductase subunit B10	17	2.0	4.16E-2	16.8	34.2
Arf6	ADP-ribosylation factor 6	12	−2.0	2.59E-2	66.5	32.9
Pprc1	peroxisome proliferative activated receptor, gamma, coactivator-related 1	19	−2.0	3.52E-2	18.2	9.0
Eif6	eukaryotic translation initiation factor 6	2	−2.0	4.24E-2	40.2	19.8
Sema4d	sema domain, immunoglobulin domain (Ig), transmembrane domain(TM) and short cytoplasmic domain, (semaphorin) 4D	13	−2.0	6.13E-3	10.6	5.2
Noc2l	NOC2 like nucleolar associated transcriptional repressor	4	−2.0	3.11E-2	47.6	23.4
Bcar1	breast cancer anti-estrogen resistance 1	8	−2.1	3.42E-2	33.2	16.2
Bach2	BTB and CNC homology, basic leucine zipper transcription factor 2	4	−2.1	1.66E-2	4.8	2.3
Ero1l	ERO1-like (S. cerevisiae)	14	−2.1	1.47E-2	8.6	4.2
Cttnbp2nl	CTTNBP2 N-terminal like	3	−2.1	4.67E-2	16.7	8.0
Abl2	v-abl Abelson murine leukemia viral oncogene 2 (arg, Abelson-related gene)	1	−2.1	8.19E-3	10.5	5.0
Trim25	tripartite motif-containing 25	11	−2.1	1.79E-2	14.3	6.8
Rapgef3	Rap guanine nucleotide exchange factor (GEF) 3	15	−2.1	6.66E-3	14.1	6.7
Card10	caspase recruitment domain family, member 10	15	−2.1	4.40E-2	17.6	8.4
Ccdc137	coiled-coil domain containing 137	11	−2.1	3.74E-2	8.1	3.8
Pttg 1ip	pituitary tumor-transforming 1 interacting protein	10	−2.1	4.69E-2	26.7	12.6
Lrrfip2	leucine rich repeat (in FLII) interacting protein 2	9	−2.2	2.60E-2	26.0	12.1
Prtg	protogenin	9	−2.2	1.35E-2	4.3	2.0
Tjap1	tight junction associated protein 1	17	−2.2	1.48E-2	12.5	5.8
Ppp1r15a	protein phosphatase 1, regulatory subunit 15A	7	−2.2	9.93E-3	28.1	12.9
Ehd1	EH-domain containing 1	19	−2.2	1.57E-2	58.3	26.7
Vasp	vasodilator-stimulated phosphoprotein	7	−2.2	4.84E-2	60.0	27.5
Erc2	ELKS/RAB6-interacting/CAST family member 2	14	−2.2	4.80E-2	8.3	3.8
Shb	src homology 2 domain-containing transforming protein B	4	−2.2	1.02E-2	21.0	9.6
Vps37b	vacuolar protein sorting 37B	5	−2.2	1.69E-2	15.1	6.9
Rnf217	ring finger protein 217	10	−2.2	2.27E-2	5.0	2.3
Sart3	squamous cell carcinoma antigen recognized by T cells 3	5	−2.2	5.32E-3	20.4	9.2
Eaf1	ELL associated factor 1	14	−2.2	3.56E-2	20.2	9.1
Ccdc9b	coiled-coil domain containing 9B	2	−2.2	8.44E-3	33.4	15.1
Rhbdf2	rhomboid 5 homolog 2	11	−2.2	1.18E-2	12.0	5.4
Ftsj3	FtsJ RNA methyltransferase homolog 3 (E. coli)	11	−2.2	9.26E-3	24.7	11.1
Cdc42ep3	CDC42 effector protein (Rho GTPase binding) 3	17	−2.2	4.99E-2	8.7	3.9
Ctu2	cytosolic thiouridylase subunit 2	8	−2.3	4.81E-2	5.1	2.3
Litaf	LPS-induced TN factor	16	−2.3	1.44E-2	15.0	6.6
Surf6	surfeit gene 6	2	−2.3	1.05E-3	13.9	6.1
Phldb1	pleckstrin homology like domain, family B, member 1	9	−2.3	2.42E-3	44.3	19.4
Igsf9b	immunoglobulin superfamily, member 9B	9	−2.3	4.40E-2	4.9	2.1
Klhdc4	kelch domain containing 4	8	−2.3	2.33E-3	18.1	7.9
Tatdn2	TatD DNase domain containing 2	6	−2.3	1.60E-3	41.9	18.3
mt-Nd2	mitochondrially encoded NADH dehydrogenase 2	MT	−2.3	4.41E-2	3117.5	1361.6
Nol10	nucleolar protein 10	12	−2.3	1.96E-2	22.2	9.6
Srf	serum response factor	17	−2.3	1.06E-2	32.5	14.0
Urb2	URB2 ribosome biogenesis 2 homolog (S. cerevisiae)	8	−2.3	3.10E-2	5.0	2.1
Urb1	URB1 ribosome biogenesis 1 homolog (S. cerevisiae)	16	−2.3	3.88E-2	4.5	1.9
Smad7	SMAD family member 7	18	−2.3	1.18E-2	33.8	14.5
Trim47	tripartite motif-containing 47	11	−2.3	1.68E-2	24.3	10.4
Gprc5a	G protein-coupled receptor, family C, group 5, member A	6	−2.3	3.13E-2	7.5	3.2
Epha2	Eph receptor A2	4	−2.4	3.62E-3	42.1	17.9
Zswim4	zinc finger SWIM-type containing 4	8	−2.4	1.02E-2	19.3	8.1
Mical2	microtubule associated monooxygenase, calponin and LIM domain containing 2	7	−2.4	7.81E-3	13.0	5.5
Tnfaip3	tumor necrosis factor, alpha-induced protein 3	10	−2.4	1.25E-3	20.7	8.7
Sh3bp2	SH3-domain binding protein 2	5	−2.4	4.84E-2	8.6	3.6
Kcnd3	potassium voltage-gated channel, Shal-related family, member 3	3	−2.4	1.30E-2	5.3	2.2
Kif7	kinesin family member 7	7	−2.4	4.84E-2	3.9	1.6
5430416N02Rik	RIKEN cDNA 5430416N02 gene	5	−2.4	4.17E-2	4.8	2.0
Rnf213	ring finger protein 213	11	−2.4	1.10E-2	29.3	12.0
Rrp9	RRP9, small subunit (SSU) processome component, homolog (yeast)	9	−2.4	1.18E-2	16.3	6.7
mt-Nd4	mitochondrially encoded NADH dehydrogenase 4	MT	−2.4	2.05E-2	1988.1	815.8
mt-Cytb	mitochondrially encoded cytochrome b	MT	−2.4	2.34E-2	7413.0	3027.4
Tubb2b	tubulin, beta 2B class IIB	13	−2.5	2.29E-2	18.9	7.6
Gm24187	predicted gene, 24187	13	−2.5	3.56E-2	3979.7	1607.5
Oasl2	2’-5’ oligoadenylate synthetase-like 2	5	−2.5	1.70E-3	18.5	7.5
mt-Nd6	mitochondrially encoded NADH dehydrogenase 6	MT	−2.5	3.10E-2	851.8	342.4
Emilin2	elastin microfibril interfacer 2	17	−2.5	7.19E-3	25.2	10.0
Nbeal2	neurobeachin-like 2	9	−2.5	9.16E-3	4.9	2.0
Plec	plectin	15	−2.5	4.68E-3	466.1	184.4
Snhg17	small nucleolar RNA host gene 17	2	−2.5	2.43E-2	4.0	1.6
Helz2	helicase with zinc finger 2, transcriptional coactivator	2	−2.5	4.29E-2	8.6	3.4
Klf16	Kruppel-like factor 16	10	−2.5	1.68E-2	7.9	3.1
Card19	caspase recruitment domain family, member 19	13	−2.5	3.75E-2	36.2	14.2
mt-Co1	mitochondrially encoded cytochrome c oxidase I	MT	−2.6	2.85E-2	6096.8	2387.9
Tnk2	tyrosine kinase, non-receptor, 2	16	−2.6	1.01E-3	25.7	9.9
Cd274	CD274 antigen	19	−2.6	4.85E-2	12.2	4.7
Gm8995	predicted gene 8995	7	−2.6	2.19E-2	5.0	1.9
Ier5	immediate early response 5	1	−2.7	8.19E-3	41.0	15.5
Ccdc88b	coiled-coil domain containing 88B	19	−2.7	1.68E-2	15.2	5.7
Fmnl1	formin-like 1	11	−2.7	4.57E-2	30.6	11.5
Oaf	out at first homolog	9	−2.7	9.13E-4	40.8	15.3
Pik3ap1	phosphoinositide-3-kinase adaptor protein 1	19	−2.7	5.32E-3	17.2	6.4
mt-Rnr2	mitochondrially encoded 16S rRNA	MT	−2.7	3.72E-2	111335.5	41317.1
Tcirg1	T cell, immune regulator 1, ATPase, H+ transporting, lysosomal V0 protein A3	19	−2.7	2.05E-2	24.3	8.9
Plk3	polo like kinase 3	4	−2.7	2.94E-2	17.3	6.3
mt-Rnr1	mitochondrially encoded 12S rRNA	MT	−2.8	3.49E-2	93896.8	33926.5
Il1rn	interleukin 1 receptor antagonist	2	−2.8	3.15E-2	31.0	11.2
Slfn5	schlafen 5	11	−2.8	3.07E-2	6.6	2.4
Mtmr10	myotubularin related protein 10	7	−2.8	2.73E-3	15.6	5.6
Mertk	MER proto-oncogene tyrosine kinase	2	−2.8	3.04E-3	10.2	3.7
Dot1l	DOT1-like, histone H3 methyltransferase (S. cerevisiae)	10	−2.8	3.35E-4	28.6	10.2
Smox	spermine oxidase	2	−2.8	1.09E-2	23.3	8.3
Pim1	proviral integration site 1	17	−2.8	2.04E-2	56.7	19.9
Syk	spleen tyrosine kinase	13	−2.8	2.03E-2	11.0	3.9
Gm10925	predicted gene 10925	1	−2.9	4.84E-2	31.0	10.8
Rnf19b	ring finger protein 19B	4	−2.9	3.42E-2	52.9	18.4
Lrrfip1	leucine rich repeat (in FLII) interacting protein 1	1	−3.0	1.16E-3	97.8	32.4
Dusp7	dual specificity phosphatase 7	9	−3.0	2.05E-3	38.0	12.5
Arl4d	ADP-ribosylation factor-like 4D	11	−3.0	2.33E-3	36.9	12.1
Dimt1	DIM1 dimethyladenosine transferase 1-like (S. cerevisiae)	13	−3.1	7.97E-4	8.4	2.7
Gm48958	predicted gene, 48958	14	−3.1	2.86E-2	7.7	2.5
Cmtr2	cap methyltransferase 2	8	−3.1	1.24E-2	3.7	1.2
Nrg1	neuregulin 1	8	−3.1	4.11E-2	3.7	1.2
Fjx1	four jointed box 1	2	−3.1	3.42E-2	10.4	3.3
Thoc6	THO complex 6	17	−3.1	2.70E-2	9.9	3.2
Cdt1	chromatin licensing and DNA replication factor 1	8	−3.1	4.32E-2	10.3	3.3
Slc16a3	solute carrier family 16 (monocarboxylic acid transporters), 3	11	−3.2	1.30E-3	15.0	4.7
Elf3	E74-like factor 3 [	1	−3.2	3.34E-2	9.7	3.0
Acod1	aconitate decarboxylase 1	14	−3.2	1.84E-2	24.6	7.6
Arhgap22	Rho GTPase activating protein 22	14	−3.3	1.34E-4	18.0	5.5
G0s2	G0/G1 switch gene 2	1	−3.3	1.13E-3	83.1	25.5
Csf3r	colony stimulating factor 3 receptor (granulocyte)	4	−3.3	3.99E-3	51.9	15.7
Baiap2	brain-specific angiogenesis inhibitor 1-associated protein 2	11	−3.3	1.22E-2	11.3	3.4
Sp100	nuclear antigen Sp100	1	−3.3	2.74E-2	3.8	1.1
Sfn	stratifin	4	−3.3	3.56E-2	14.4	4.3
Atf3	activating transcription factor 3	1	−3.4	3.50E-2	31.4	9.3
Srxn1	sulfiredoxin 1 homolog (S. cerevisiae)	2	−3.4	1.32E-2	67.9	20.0
Phlda1	pleckstrin homology like domain, family A, member 1	10	−3.4	3.60E-3	77.3	22.8
Chka	choline kinase alpha	19	−3.5	1.41E-5	15.6	4.5
Tsc22d2	TSC22 domain family, member 2	3	−3.5	9.57E-6	27.5	7.9
Gm44899	predicted gene 44899	7	−3.5	4.35E-2	6.1	1.7
Ifrd1	interferon-related developmental regulator 1	12	−3.6	1.91E-3	22.2	6.1
Mxd1	MAX dimerization protein 1	6	−3.7	6.51E-3	37.5	10.2
Spata5l1	spermatogenesis associated 5-like 1	2	−3.7	4.45E-2	11.3	3.1
Pxdc1	PX domain containing 1	13	−3.8	4.39E-2	7.0	1.9
Dusp4	dual specificity phosphatase 4	8	−3.8	3.17E-2	20.4	5.4
Nefm	neurofilament, medium polypeptide	14	−3.9	1.02E-2	12.4	3.2
Maff	v-maf musculoaponeurotic fibrosarcoma oncogene family, protein F	15	−4.0	5.19E-4	20.6	5.2
H2-Q4	histocompatibility 2, Q region locus 4	17	−4.1	1.32E-2	36.4	8.9
Borcs6	BLOC-1 related complex subunit 6	11	−4.2	2.92E-2	4.8	1.1
Gadd45g	growth arrest and DNA-damage-inducible 45 gamma	13	−4.3	2.45E-2	140.7	32.4
Kdelr3	KDEL endoplasmic reticulum protein retention receptor 3	15	−4.4	2.42E-3	10.0	2.3
Neat1	nuclear paraspeckle assembly transcript 1 (non-protein coding)	19	−4.4	9.13E-4	94.8	21.7
Gm36940	predicted gene, 36940	9	−4.7	4.29E-2	3.0	0.6
Ngf	nerve growth factor	3	−4.8	3.94E-2	42.0	8.8
Ifi204	interferon activated gene 204	1	−5.0	1.71E-2	14.2	2.8
mt-Atp6	mitochondrially encoded ATP synthase 6	MT	−5.0	1.23E-2	37.7	7.5
Rn7s2	7S RNA 2	12	−5.0	3.12E-3	101.4	20.2
Rbp4	retinol binding protein 4, plasma	19	−5.1	3.18E-2	6.5	1.3
Flnc	filamin C, gamma	6	−5.8	3.35E-2	178.3	30.6
Cyp26a1	cytochrome P450, family 26, subfamily a, polypeptide 1	19	−6.0	1.01E-3	67.8	11.3
Mirt2	myocardial infraction associated transcript 2	15	−6.6	3.13E-2	4.3	0.7
Gm23833	predicted gene, 23833	17	−6.6	4.33E-3	29.7	4.5
Trpc6	transient receptor potential cation channel, subfamily C, member 6	9	−6.6	3.61E-3	4.3	0.7
Rsad2	radical S-adenosyl methionine domain containing 2	12	−7.1	2.01E-2	22.9	3.2
Gm26532	predicted gene, 26532	8	−7.8	3.35E-2	5.4	0.7
Tnfsf14	tumor necrosis factor (ligand) superfamily, member 14	17	−8.7	4.19E-2	3.2	0.4
Emp1	epithelial membrane protein 1	6	−9.2	4.95E-2	59.6	6.5
Oasl1	2’-5’ oligoadenylate synthetase-like 1	5	−9.7	3.62E-3	13.1	1.3
Mir3109	microRNA 3109	9	−9.9	2.92E-2	58.0	5.8
Muc5ac	mucin 5, subtypes A and C, tracheobronchial/gastric	7	−10.2	2.97E-6	5.4	0.5
Cfap58	cilia and flagella associated protein 58	19	−11.3	6.99E-3	3.4	0.3
Foxn1	forkhead box N1	11	−11.7	2.93E-3	3.5	0.3
Asprv1	aspartic peptidase, retroviral-like 1	6	−13.9	6.66E-3	14.8	1.1
Klk9	kallikrein related-peptidase 9	7	−14.0	2.11E-2	3.7	0.3
Gm44275	predicted gene, 44275	6	−14.3	4.44E-2	3.2	0.2
Krt16	keratin 16	11	−16.8	4.29E-2	16.8	1.0
Krt17	keratin 17	11	−23.4	2.33E-3	18.3	0.8
Muc4	mucin 4	16	−25.4	1.74E-2	4.4	0.2
Gm9573	predicted gene 9573	17	−45.5	3.47E-2	3.1	0.1
Irx1	Iroquois homeobox 1	13	−69.9	2.76E-3	2.2	0.0
Krt12	keratin 12	11	−85.9	1.88E-4	6.9	0.1
C2cd4b	C2 calcium-dependent domain containing 4B	9	−92.6	4.27E-4	7.0	0.1
Gm19219	predicted gene, 19219	9	−378.7	6.02E-3	3.8	0.0
Eif2s3y	eukaryotic translation initiation factor 2, subunit 3, Y-linked	Y	−3251.2	1.75E-37	5.6	0.0
Kdm5d	lysine (K)-specific demethylase 5D	Y	−4566.7	7.48E-42	2.2	0.0
Ddx3y	DEAD (Asp-Glu-Ala-Asp) box polypeptide 3, Y-linked	Y	−7263.9	7.91E-63	10.1	0.0

Analysis of pathways differentially impacted between male and female aged LECs at 24 hours PCS revealed that “response to Staphylococcus aureus infection”, metabolic pathways, and complement pathways were the most impacted (all at FDR≤ 0.04), while the DEGs mapped to the GO terms “cell adhesion” (FDR≤ 6 X10⁻⁵) and “extracellular matrix structural constituent” (FDR≤ 3 X10⁻⁹).

Comparison of genes differentially expressed between male and female aged LECs at 24 hours PCS, with those differentially expressed between young versus old LECs at 24 hours PCS when sex was not considered as a variable (Faranda et al 2021), revealed that five of the 430 genes differentially expressed by sex in aged LECs at 24 hours PCS were also expressed at significantly different levels in aged LECs at 24 hours PCS compared to young. In all cases, these five genes (Pik3ap1, Acod1, Ptafr, Hcar2, and Rpe65) are upregulated in both males and females, but the levels of all of these genes, except for Rpe65, are higher in aged male LECs than female at 24 hours PCS.

The effect of sex on the aged lens fiber cell transcriptome

Comparison between the lens fiber transcriptome of aged male and female mice revealed 41 biologically significant DEGs between the sexes, with 28 genes expressed at higher levels in females and 13 expressed at higher levels in males. Of these, five DEGs reside on sex chromosomes and 36 reside on autosomes (Table 7). Five DEGs located on sex chromosomes and 15 autosomal DEGs were previously found to exhibit sex dependent expression in other mouse tissues based on comparisons with SAGD (Table 7). Eight of the 41 biologically significant DEGs between male and female aged lens fibers exhibit lens preferred expression based on comparisons with the iSyTE database (Table 7).

Table 7.

DEGs between aged male and female lens fiber cells. Bold font denotes genes that have been previously reported to exhibit sex dependent expression in adult mouse tissues based on comparisons with SAGD (Shi et al., 2019). Italics font denotes genes that exhibit lens preferred expression in postnatal day 56 mouse lenses according to the iSyTe database (Kakrana et al., 2018).

SYMBOL	DESCRIPTION	Chromosome	Fold Difference	FDR	Male FPKM	Female FPKM
Xist	inactive X specific transcripts	X	2931.0	8.13E-82	0.0	23.5
Prelp	proline arginine-rich end leucine-rich repeat	1	25.0	4.15E-17	0.1	2.4
Crygd	crystallin, gamma D	1	23.1	4.48E-2	40.2	929.7
Rbp3	retinol binding protein 3, interstitial	14	11.6	7.30E-3	0.4	4.4
H2bc3	H2B clustered histone 3	13	7.4	4.58E-2	0.4	3.4
Ccnd2	cyclin D2	6	5.9	4.21E-3	0.8	5.0
Gm4865	predicted gene 4865	5	5.6	4.60E-2	0.6	3.2
Cxcl14	chemokine (C-X-C motif) ligand 14	13	5.2	3.23E-5	0.9	5.1
Ndufa4l2	Ndufa4, mitochondrial complex associated like 2	10	3.6	2.04E-2	2.9	10.6
Inhba	inhibin beta-A	13	3.5	3.20E-2	1.7	5.8
S1pr3	sphingosine-1-phosphate receptor 3	13	3.1	1.44E-8	1.6	5.2
Chrdl1	chordin-like 1	X	3.1	5.19E-8	1.4	4.3
Pdpn	podoplanin	4	2.9	1.40E-9	7.8	22.4
Aldh3a1	aldehyde dehydrogenase family 3, subfamily A1	11	2.7	1.71E-5	53.1	143.0
Gpsm2	G-protein signalling modulator 2 (AGS3-like, C. elegans)	3	2.6	4.42E-2	2.6	6.9
Serping1	serine (or cysteine) peptidase inhibitor, clade G, member 1	2	2.5	2.80E-3	1.6	4.1
Slc7a8	solute carrier family 7 (cationic amino acid transporter, y+ system), 8	14	2.4	2.41E-4	5.9	14.4
Scarf2	scavenger receptor class F, member 2	16	2.4	2.35E-2	2.0	4.8
Notch2	notch 2	3	2.3	1.46E-2	2.2	5.1
Olfml2a	olfactomedin-like 2A	2	2.3	1.63E-3	1.6	3.6
Kcnk1	potassium channel, subfamily K, member 1	8	2.3	1.06E-2	3.0	6.8
Fez1	fasciculation and elongation protein zeta 1 (zygin I)	9	2.2	5.47E-3	1.7	3.8
F2r	coagulation factor II (thrombin) receptor	13	2.2	3.71E-2	1.6	3.6
Msx1	msh homeobox 1	5	2.2	4.92E-2	2.4	5.3
Susd2	sushi domain containing 2	10	2.1	7.67E-3	1.9	4.1
Glul	glutamate-ammonia ligase (glutamine synthetase)	1	2.1	6.38E-4	4.6	9.8
Mme	membrane metallo endopeptidase	3	2.1	3.08E-3	2.8	5.7
Htra3	HtrA serine peptidase 3	5	2.0	1.57E-2	4.7	9.5
Bpgm	2,3-bisphosphoglycerate mutase	6	−2.1	7.15E-3	5.6	2.7
Guk1	guanylate kinase 1	11	−2.3	4.93E-2	8.6	3.7
Zfyve21	zinc finger, FYVE domain containing 21	12	−2.8	1.30E-2	7.7	2.8
Dmac1	distal membrane arm assembly complex 1	4	−3.4	3.23E-2	7.6	2.1
Igfbp2	insulin-like growth factor binding protein 2	1	−11.4	6.82E-4	4.2	0.4
Hba-a2	hemoglobin alpha, adult chain 2	11	−15.5	2.33E-4	63.2	4.1
Col8a2	collagen, type VIII, alpha 2	4	−23.7	7.32E-12	2.6	0.1
Hbb-bs	hemoglobin, beta adult s chain	7	−31.9	2.25E-3	28.7	0.9
Eif2s3y	eukaryotic translation initiation factor 2, subunit 3, Y-linked	Y	−31.9	6.15E-10	4.2	0.2
Alas2	aminolevulinic acid synthase 2, erythroid	X	−43.2	2.55E-5	5.5	0.1
Krt12	keratin 12	11	−44.1	1.33E-3	3.6	0.1
Car3	carbonic anhydrase 3	3	−173.1	1.60E-3	6.4	0.0
Ddx3y	DEAD (Asp-Glu-Ala-Asp) box polypeptide 3, Y-linked	Y	−408.7	1.56E-60	7.5	0.0

Analysis of all DEGs that exhibited FDR corrected statistically significant levels of expression (233 genes total) for impacted pathways or enriched GO terms using iPathway guide (Ahsan and Draghici, 2017) revealed that “eye development” (19 genes total; FDR corrected p value ≤ 0.003) and extracellular matrix structural constituent (12 genes total; FDR ≤2.1 X10⁻⁴) were among the most enriched GO terms while the pathway “protein digestion and absorption was considered impacted at FDR ≤ 0.002.

Comparison between genes differentially expressed at biologically significant levels during aging in male lens fiber cells with those differentially expressed in aging female lens fibers revealed that 87 genes were upregulated in both sexes, while 324 were downregulated in both sexes, with no genes changing in opposite directions as a function of sex (See Supplemental Table 1). However, in many cases, the extent of age-related changes in gene expression was dimorphic between the sexes. Most notably, the age related decreases in γ-crystallin and other lens fiber cell marker mRNAs in lens fibers were generally more robust in males than females, while the age related upregulation of Ndufa4l2, a gene that encodes a protein that slows electron transport in damaged mitochondria (Li et al., 2017), is more robust in female lens fibers than male (Table 7; Supplemental Table 1).

Discussion

Aging is a biological process during which organisms undergo numerous physiological changes due to cellular senescence, along with age-related cell and tissue damage due to oxidative stress and other environmental insults (da Costa et al., 2016; Slawinska and Krupa, 2021; Xu et al., 2018). Individuals, and even tissues, age at different rates, which appears to be a function of both differential exposure to environmental insults and differing genetic and/or epigenetic landscape (Kenyon, 2010; Wood et al., 2015). The most robust known genetic influence on aging is sex chromosome complement, leading to sex-specific differences between the incidence/clinical presentation of numerous age-related diseases (Marquez et al., 2020; Mauvais-Jarvis et al., 2020; Sampathkumar et al., 2020).

In a companion study, we investigated the effect of age on the lens epithelial and fiber transcriptomes, as well as the response of LECs to cataract surgery, using C57BL/6J mice as the near genetic identity of inbred mice and uniform housing conditions reduces variability (Faranda et al., 2021). The young (12 week) old mice studied are sexually mature adults, while the aged (24 month old) mice appear biologically similar to 70 year old humans based on numerous physiological parameters (Marquez et al., 2020), including the observation that females have entered reproductive senescence based on parameters consistent with human menopause including elevations in serum follicle stimulating hormone (Collins et al., 1981; Urzua et al., 2017). As the prior study included equal numbers of male and female animals, here we were able to determine how sex influences the aging lens transcriptome.

Young LECs exhibit few sex differences in their transcriptomes although young lens fibers exhibit greater sexual dimorphism

Twelve week old mice are sexually mature, have transparent lenses, and have established their adult crystallin complement (Ueda et al., 2002). Transcriptome profiling of LECs from 12 week old mice isolated immediately after lens fiber cell removal revealed only five genes that exhibit at least a two-fold difference between sexes and were expressed at levels plausibly high enough to affect lens biology based on published criteria (Manthey et al., 2014a). One of these is Xist, a non-coding RNA critical for the inactivation of the second X chromosome in females (Brockdorff et al., 2020), while Ddx3y and Eif2s3y are neighboring genes found in the male specific region of the Y chromosome that have the potential to have diverse post transcriptional effects on gene expression (Deschepper, 2020). Neither autosomal DEG has a known function although one is predicted to be a non-coding RNA while the other may be protein coding. Similarly, young male and female LECs also remodel their transcriptome almost identically after a lens injury resembling cataract surgery, although three intriguing sex differences were also seen in aged LECs after injury. Ccdc88b encodes Gipie, a binding partner of GRP78 that can protect cells from the unfolded protein response (Matsushita et al., 2011) and G0s2, a regulator of lipid homeostasis (Zhang et al., 2017), are expressed at higher levels in young male LECs after injury. In contrast, Serping1, which encodes a serine protease inhibitor that regulates innate immunity (Davis et al., 2010), is expressed at higher levels in injured female LECs than male.

Young lens fibers have more transcriptomic differences between sexes than LECs, however, the detected fold changes tend to be modest, and bioinformatics analysis did not reveal significant potential biological connections between these DEGs. Interestingly though, two genes, Aldh3a1, which encodes an aldehyde dehydrogenase that detoxifies lipid peroxides and is known to protect the lens from stress induced opacity (Lassen et al., 2007), as well as Gstm1, which encodes glutathione S transferase mu, a xenobiotic detoxification enzyme which may be protective against cataract in humans (Saadat and Farvardin-Jahromi, 2006), are both expressed at higher levels in young female, compared to male lens fibers.

The lens cell transcriptome of aged mice exhibits greatly increased sexual dimorphism compared to young adults

We reported in the companion paper (Faranda et al., 2021) that both LECs and fibers downregulate the expression of numerous lens preferred genes, most notably the gamma crystallins, with aging, consistent with a prior study (Treton et al., 1988). Here we find that several of the crystallins that downregulate in lens cells with age appear to downregulate less dramatically in female lens cells than male, which correlates with male LECs expressing lower levels of lactase-like/klotho-gamma, a potential activator of FGF receptor activity in the lens (Audette et al., 2016). In contrast, aged male LECs express higher levels of Gpx3 mRNA than females. Notably, Gpx3 encodes a secreted glutathione peroxidase that detoxifies hydrogen peroxide in biological fluids (Baez-Duarte et al., 2014; Olson et al., 2010), suggesting that aged male lenses could be better protected against the oxidative damage than females. This proposal is supported by the observation that aged female lens fibers express higher levels of Ndufa4l2, a marker of mitochondrial damage (Li et al., 2017), than males. These intriguing results could suggest a mechanism for the epidemiological observation that post menopausal women are more prone to developing cataract than men of similar age (Zetterberg and Celojevic, 2015).

The most dramatic sexual dimorphism in lens gene expression detected in this study was for LECs isolated from aged eyes 24 hours after they underwent a surgery that models cataract removal (Manthey et al., 2014b). Pathway analysis of these DEGs revealed that genes encoding several ECM molecules as well components of the complement cascade were differentially expressed in female LECs at 24 hours PCS compared to males. Interestingly, a recent study suggested that Xist, which is a highly female preferred transcript in all lens cells, can increase ECM mRNA levels in injured skin cells by sponging microRNAs that mediate ECM mRNA degradation (Cao and Feng, 2019), an effect that may enhance fibrotic disease progression (Wang et al., 2017). Further, sex differences in complement pathway regulation have been previously reported in C57BL/6 mice (Kotimaa et al., 2016) and normal humans (Gaya da Costa et al., 2018; Troldborg et al., 2017), and such dimorphisms are implicated in differences in disease prevalence, severity, and outcome between sexes (Mauvais-Jarvis et al., 2020; Ward et al., 2018). While it is currently unknown how early transcriptome remodeling of LECs after lens fiber cell removal (Jiang et al., 2018) mechanistically relates to the pathogenesis of posterior capsular opacification (PCO), it is tempting to speculate that sex differences in complement activation and ECM production explain the increased PCO prevalence in women observed in some epidemiological studies (Congdon et al., 2008; Fong et al., 2014; Lee et al., 2016).

Limitations of this study

While the present study and the companion paper (Faranda et al, 2021) revealed that the lens transcriptome differs as a function of sex and age, the limitations of the study design also indicate the need to explore how sex affects lens biology in more depth using both unbiased approaches and directed investigation. First, it is possible, if not likely, that sex has a greater effect on the lens transcriptome than suggested in this study. The limited availability of 24 month old inbred mice housed under controlled conditions led to the present study design where only two biological replicates could be analyzed for sex (male versus female) under each condition (young versus old epithelium, fibers, and injured epithelium). The resulting low statistical power to detect differences in gene expression could mean that there are additional sex-dependent effects on the lens transcriptome, although this issue is likely minimized due to the study of inbred mice housed in a controlled environment. Second, while RNAseq is very reproducible, we were unable to obtain aged animals for validation studies due to their limited availability. Third, it will be important to replicate this work in humans to ensure that these differences have biological relevance to human cataract. Fourth, the broader effects of sex on PCO development are still open questions. The epidemiological evidence for differential PCO rates between sexes is currently mixed (Congdon et al., 2008; Fong et al., 2014; Lee et al., 2016; Tokko et al., 2019) , while many studies on PCO incidence either do not consider sex dependence (Haripriya et al., 2017) or the study design controls for sex (Chen et al., 2019), so sex differences are obscured. Fifth, the present study investigated the acute inflammatory response of mouse LECs to fiber cell removal, at a time 1-2 days prior to the phenotypic transition of LECs to myofibroblasts (Shihan et al., 2020), so the transcriptomic differences by sex detected here may not reflect sex differences in the LEC fibrotic response. Sixth, the animal model used here does not strictly reproduce the surgical technique (including intraocular lens placement) used in human cataract surgery. Thus, future studies are needed to investigate both whether sex influences how robust the later fibrotic response of LECs to lens injury is, the long term survival of these myofibroblasts, and their migratory potential. The latter two points are critical to translate this work to humans as clinical PCO usually occurs months or years after cataract surgery when lens cells initially trapped at the capsular bag periphery by IOL interactions with the lens capsule (Bisevac et al., 2020) escape their containment, and migrate into the optical axis.

Implications for lens biology

The risk of developing age-related cataract in humans is influenced by sex and age (Chang et al., 2011; Hugosson and Ekstrom, 2020), with epidemiological studies routinely finding that post-menopausal females have elevated risk of cataract development compared to aged matched males. While it has been hypothesized that estrogen withdrawal at menopause has direct negative effects on the lens (Zetterberg and Celojevic, 2015), the present study did not find appreciable expression of canonical estrogen receptor mRNA in adult mouse lens cells. However, aged lens cells did exhibit sexually dimorphic expression of numerous genes, several of which have the potential to modify cataract risk. It will be interesting to determine whether such age-related sex differences in gene expression also exist in humans, the mechanisms by which this sexual dimorphism is established, their functional relevance to lens biology, and ultimate risk of cataract development.

Implications for experimental design in lens and cataract research

Similar to the lens, others have found that tissue transcriptomes commonly differ between sexes, with numerous autosomal genes exhibiting sex-preferred expression. However, autosomal genes often only exhibit sexually dimorphic expression in a few, or even a single tissue, demonstrating that chromosomal sex has tissue specific effects (Arnold, 2019; Kassam et al., 2019; Lu and Mar, 2020; Mukaibo et al., 2019; Oliva et al., 2020; Shi et al., 2019). This observation also is valid for the lens, as few autosomal genes exhibiting sexually dimorphic expression in the lens exhibit sexually dimorphic expression in other mouse tissues based on comparisons with the SAGD database (Shi et al., 2019).

These observations, along with increasing recognition that sex confers differences in disease risk/pathogenesis and drug metabolism, has led the National Institutes of Health (NIH) of the United States to mandate that sex be considered as a biological variable in NIH supported biomedical research (Woitowich et al., 2020). Overall, the present study supports the need to both include animals of both sexes in lens research, and to be aware of the potential for sex to modify lens biology, particularly during investigations of lens aging and its influence on cataract.

Highlights:

• Young lens epithelial cells exhibit few sex dependent differences in gene expression

• Young lens epithelial cells exhibit few sex dependent differences in their response to lens injury

• Young lens fibers exhibit numerous sex dependent differences in gene expression

• Aged lens epithelial cells exhibit modest numbers of sexually dimorphic genes

• Aged lens epithelial cells exhibit profound sex specific responses to lens injury

• Aged lens fiber cells exhibit increased numbers of genes that exhibit sexually dimorphic gene expression