Whole-exome sequencing prioritizes candidate genes for hereditary cataract in the Emory mouse mutant

Abstract The Emory cataract (Em) mouse mutant has long been proposed as an animal model for age-related or senile cataract in humans—a leading cause of visual impairment. However, the genetic defect(s) underlying the autosomal dominant Em phenotype remains elusive. Here, we confirmed development of the cataract phenotype in commercially available Em/J mice [but not ancestral Carworth Farms White (CFW) mice] at 6–8 months of age and undertook whole-exome sequencing of candidate genes for Em. Analysis of coding and splice-site variants did not identify any disease-causing/associated mutations in over 450 genes known to underlie inherited and age-related forms of cataract and other lens disorders in humans and mice, including genes for lens crystallins, membrane/cytoskeleton proteins, DNA/RNA-binding proteins, and those associated with syndromic/systemic forms of cataract. However, we identified three cataract/lens-associated genes each with one novel homozygous variant including predicted missense substitutions in Prx (p.R167C) and Adamts10 (p.P761L) and a disruptive in-frame deletion variant (predicted missense) in Abhd12 (p.L30_A32delinsS) that were absent in CFW and over 35 other mouse strains. In silico analysis predicted that the missense substitutions in Prx and Adamts10 were borderline neutral/damaging and neutral, respectively, at the protein function level, whereas, that in Abhd12 was functionally damaging. Both the human counterparts of Adamts10 and Abhd12 are clinically associated with syndromic forms of cataract known as Weil-Marchesani syndrome 1 and polyneuropathy, hearing loss, ataxia, retinitis pigmentosa, and cataract syndrome, respectively. Overall, while we cannot exclude Prx and Adamts10, our data suggest that Abhd12 is a promising candidate gene for cataract in the Em/J mouse.


Introduction
Clouding of the crystalline lens, or cataract(s), is commonly associated with aging and, despite advances in surgical treatment, age-related or senile cataract is a leading cause of visual impairment (low vision and blindness) worldwide (GBD 2019 Blindness and Vision Impairment Collaborators 2021). Cataract may also be inherited either as an isolated phenotype or as part of a multisystem disease typically with an early-onset (birth to 40 years), and over 450 underlying genes have been identified including those for lens crystallins, connexins and other membrane proteins, ocular transcription factors, and RNA-binding proteins (Shiels and Hejtmancik 2017, 2019 (https://cat-map.wustl. edu). Hereditary forms of cataract also afflict many domesticated animals including sheep, cattle, horses/ponies, and over 60 breeds of dogs (Pinard and Basrur 2011;Wilson et al. 2012;Mellersh 2014;Murgiano et al. 2014;Ricketts et al. 2015). In addition, numerous inbred strains of laboratory rats and mice serve as animal models of human cataract including the Shumiya cataract rat and senescence-accelerated mouse strains (Mori et al. 2006;Graw 2019).
The Emory cataract (Gene symbol: Em) mutation arose spontaneously in an inbred colony of Carworth Farms White (CFW) mice, and due to the relatively late onset of autosomal dominant lens opacities at 6-8 months of age, the Em mouse was originally proposed (in 1981) as an animal model for human senile cataract (Kuck et al. 1981;Kuck 1990). In vivo imaging of the Em mouse lens revealed that the cataract progressed slowly with variable severity in four stages starting with irregularly round opacification confined to the central region of the anterior superficial cortex by 2 months of age (stage-1), progressing into the anterior deep cortex by 6 months (stage-2), then to the supranuclear region by 7-8 months (stage-3), and eventually to total lens opacification by 10-12 months (stage-4) (Takizawa and Sasaki 1986). Histopathological studies of Em lenses using light and electron microscopy have detected early ultrastructural changes including acellular regions in the anterior epithelium by 2 weeks of age and swelling of the anterior cortical fibers by 1-2 months of age (Uga et al. 1988). Cataract onset was delayed in Em mice fed a calorie (20-40%) restricted diet (Taylor et al. 1989(Taylor et al. , 1995a(Taylor et al. , 1995bScrofano et al. 1998aScrofano et al. , 1998b, and cataract grading studies have shown that beyond 6 months of age, Em females develop cataract more rapidly (7-11 months) than age-matched males, whereas, the final cataract severity (≥13 months) was indistinguishable between sexes (Shang et al. 2002).
Biochemical studies of Em lenses have found a range of abnormalities including increased levels of protein insolubility, Ca 2+ ions, oxidized glutathione, and membrane lipid peroxidation (Kuck 1990). Lens gene expression studies have detected downregulation of several mRNA transcripts for crystallins (e.g. Cryaa) and major intrinsic protein or aquaporin-0 (MIP/AQP0), and up-regulation of adhesion related kinase (Ark) receptor tyrosine kinase associated with cataract in the Em lens (Shi and Bekhor 1992;Sheets et al. 2002). Further, ultrastructural and immunochemical studies have shown that cortical cataract formation in the Em lens was associated with premature proteolytic-cleavage of MIP/AQP0 and gap-junction alpha-8 protein or connexin-50 (GJA8/Cx50) C-termini resulting in abnormal wavy square-array junctions and smaller gap-junctions, respectively, compared to wild-type lens fiber cells (Biswas et al. 2014). Finally, imbalances in the lens crystallin proteome and changes in transfer RNA-derived fragments have been associated with cataract development in the Em mouse (Schmid et al. 2021;Zhang et al. 2022). However, all of these observed changes are likely secondary to the underlying genetic defect(s). Here we have undertaken a whole-exome sequencing approach to analyze variants in candidate genes for Em.

Materials and methods
Mice CFW-Em/J (Em/J, Stock no. 001998) and C57BL/6J (B6J, Stock no. 000664) mice were obtained from The Jackson Laboratory (Bar Harbor, ME, USA). CFW(SW) mice (Strain code 024) were obtained from Charles River Laboratories (Wilmington, MA, USA). Lens imaging of conscious mice was performed using a slit-lamp at 25× magnifications (SL-D7; Topcon, Tokyo, Japan) equipped with a digital camera (D800; Nikon, Tokyo, Japan) under identical image acquisition and processing settings. Anterior chamber imaging of mice was performed by spectral domain-ocular coherence tomography using a 12-mm bore (Bioptigen Envisu, Leica Microsystems, Deerfield, IL, USA). Briefly, mice were anesthetized with ketamine (100 mg/Kg body-weight) and xylazine (10 mg/Kg body-weight) and both eyes immediately hydrated with balanced salt solution (BSS, Alcon Laboratories, Fort Worth, TX, USA). Radial volume scan parameters within the Bioptigen InVivoVueClinic software were set to: 2-mm diameter, 1,000 A-scans/B-scan, 100 B-scans/ volume, 1 frame/B-scan, and 1 volume. Central corneal thickness was measured within the Bioptigen InVivoVueClinic Sofware using vertical angle-locked B-scan calipers. Mice were humanely euthanized according to the American Veterinary Medical Association and the eyes were removed. Whole globes or dissected lenses immersed in phosphate buffered saline (#P4417-100TAB, MilliporeSigma, Burlington, MA, USA) were imaged with a dissecting microscope (Stemi 2000, Zeiss, Thornwood, NY, USA) equipped with a digital camera (Spot Insight, Sterling Heights, MI, USA). All mouse studies were approved by the Institutional Animal Care and Use Committee at Washington University in St. Louis and the University of Iowa in compliance with the Institute for Laboratory Animal Research guidelines.

Exome sequencing and variant analysis
Genomic DNA was isolated from mouse spleen using the Gentra Puregene Kit (Qiagen, Valencia, CA, USA) and quantified by absorbance at 260 nm (NanoDrop 2000, Wilmington, DE, USA).
Whole exome capture was achieved using the SureSelect Mouse All Exon (50 Mb) Kit that targets over 221,784 exons from 24,306 genes (Agilent Technologies, Santa Clara, CA, USA) followed by next-generation sequencing (1 lane, paired-end reads 2 × 101 bps) on an Illumina HiSeq 2000 system using the Multiplexing Sample Preparation Oligo-nucleotide Kit and the HiSeq 2000 Paired-End Cluster Generation Kit (Illumina, San Diego, CA, USA) according to the manufacturers' instructions and as briefly described (Mackay et al. 2014(Mackay et al. , 2015. Raw sequence data were aligned to the mouse reference genome (build mm10) by NovoalignMPI (www.novocraft.com), and sequence variants were called using the Sequence Alignment/Map format (SAMtools) and Picard programs (http://samtools.sourceforge. net/). Target coverage and read-depth were reviewed by the Integrated Genomics Viewer (IGV; http://www.broadinstitute. org/igv/). Called variants were reviewed using the SNP and Variation Suite software (SVS 8.9.1, Golden Helix, Bozeman, MT, USA). Variant effects on protein function were predicted using the Sorting Intolerant from Tolerant (SIFT) web server (https:// sift.bii.a-star.edu.sg/) (Sim et al. 2012) and the Protein Variation Effect Analyzer (PROVEAN) web server (http://provean.jcvi.org) (Choi and Chan 2015).

Em/J mouse lens phenotype
Em/J inbred albino mice are reported to be homozygous for an autosomal dominant cataract (Em) phenotype that arose spontaneously in an inbred colony of CFW mice at Emory University, Atlanta, GA, USA (Kuck et al. 1981;Kuck 1990) (https://www.jax. org). CFW mice were derived from so-called Swiss mice and have been maintained as an outbred population for many generations causing reduced linkage disequilibrium between nearby alleles/markers compared to inbred mouse strains-rendering them useful for genetic analysis of complex traits (Lynch 1969;Chia et al. 2005;Yalcin et al. 2010;Parker et al. 2016). First, we performed a gross comparison of postmortem eyes and lenses from Em/J (mutant sub-strain) and CFW (ancestral strain) mice (n = ≥3 animals) before and after obvious cataract formation in the former. While Em/J and CFW eyes were grossly indistinguishable at post-natal day 21 (P21), we observed occasional, unilateral, displacement of the pupil margin in both strains (Fig. 1a). Upon dissection, the otherwise clear lenses from pupil-displaced Em/J eyes were found to be misshapen with an "egg-shaped" appearance (Fig. 1b). Similarly, at 7 months of age, CFW lenses were transparent with an occasional, unilateral, egg-shaped appearance ( Fig. 1c and d). These observations suggested that variable pupil displacement may be due to an adhesion between the iris and lens capsule (suspected posterior synechia) that resulted in asymmetric lens growth. Regardless, occasional pupil displacement with a misshapen lens in the Em/J mutant sub-strain was likely derived from the ancestral CFW strain. In Em/J lenses at 7 months of age, severe, bilateral, cortical, cataract was manifest without signs of lens rupture ( Fig. 1e and f) and by 9 months of age, Em/J lenses appeared slightly shrunken with signs of liquefaction in the outer cortex region surrounding the nucleus (Fig. 1g Fig. 1. Lens phenotype of Em/J mice. Dark-field (a, f, h) and bright-field (b-d, e, g) dissecting microscope images of Em/J eye (a) and lenses (b, e-h) and CFW lenses (c, d) at P21 (a, b), 7 months of age (c-f), and 9 months of age (g, h). Arrow in (a) indicates pupil displacement. Arrows in (h) indicate signs of liquefaction. Scale bar: 500 µm. and h). Slit-lamp examination of conscious Em/J mice (without pupil dilation or anesthesia) at 7 months of age confirmed the presence of bilateral, cortical cataract with mild inter-ocular variability and, in one case, a suspected synechia without other signs of ocular inflammation ( Fig. 2a1 and a3). Otherwise, Em/J (and CFW) eyes had clear corneas and translucent irises with unremarkable central corneal thickness (average CCT = 99.4 µm ± SD 2.51, n = 6 eyes) and anterior chamber dimensions (Supplementary Fig. 1) (Lively et al. 2010).
Subsequently, in an effort to genetically map the Em phenotype to a mouse chromosome, we began an F2 intercross with the C57BL/6J (B6J) genome reference strain (Sarsani et al. 2019). However, F1 progeny failed to develop the characteristic cataract phenotype by 9-12 months of age suggesting that heterozygous Em mice developed cataract much later than homozygotes and/ or that the cataract exhibited reduced penetrance on the B6J genetic background-making further genetic mapping of Em challenging.

Exome metrics and variant analysis
For both Em/J and CFW exome samples, over 86% of total paired-end reads were mapped to the C57BL/6J mouse reference genome assembly (Genome Reference Consortium Mouse Build 38 or GRCm38/mm10) and over 86% of mapped reads were present in the captured exome sequences (Supplementary Table 2). Over 98% of each exome achieved a read depth of ≥10 × coverage and over 93% of each achieved 25× coverage. Unexpected gaps in coverage were found on chromosome-4 of CFW (32 exons) and on chromosome Y of Em/J (140 exons). However, close inspection of these regions of low coverage using the IGV (https://software. broadinstitute.org/software/igv/) did not reveal credible candidate genes for ocular abnormalities.
Since Em/J (and CFW) mice are albinos with pink eyes, we first sought to confirm the presence of an underlying mutation in the tyrosinase gene (Tyr) on chromosome-7 (Beermann et al. 2004). IGV confirmed that both strains were homozygous for the classical albino missense mutation (c.G308C, p.C103S) in exon-1 of Tyr (Supplementary Table 3). In addition, we found that both Em/J and CFW mice were homozygous for a nonsense mutation (p.Y347X) in exon-7 of the gene for phosphodiesterase 6 betasubunit (Pde6b) on chromosome-5 (Supplementary Table 3) that underlies an autosomal recessive form of retinal (rodphotoreceptor) degeneration 1 (rd1, rodless)-consistent with prior reports that CFW mice undergo early-onset retinal degeneration, (Kuck et al. 1981;Serfilippi et al. 2004;Han et al. 2013).
Fourth, we proceeded to filter these "novel" variants in Cat-Map genes against other public databases including the University of California, Santa Cruz (UCSC) Genome Browser (Lee et al. 2022) and the Ensembl Genome Browser (https://ensembl.org) (Cunningham et al. 2022) in order to access later versions of dbSNP and genomic variants in other mouse strains. Variant coordinates were input to the UCSC Genome Browser and analyzed using the Annotated SNPs from Mouse Strain Comparison Analysis followed by the Ensemble Browser (release 106). Of the 43 first-round variants, 20 were found to have rs identifiers and occurred in other mouse strains (Table 2). A further 19 variants were present in other mouse strains and/or were reference sequence in Em-J, leaving four Cat-Map genes each with a novel variant comprising predicted missense SNVs in Fktn, Prx, and Adamts10, and a disruptive in-frame deletion variant in Abhd12 (Table 2). In order to validate these novel variants, we undertook PCR-Sanger sequencing and/or allele-specific PCR amplification. The Prx and Adamts10 missense SNVs were confirmed to be heterozygous in CFW and homozygous in Em/J mice and both were consistent with predicted p.R167C and p.P761L amino-acid substitutions, respectively ( Fig. 3a and b). However, the predicted missense  Supplementary Fig. 2). Close inspection of this variant using IGV revealed that it exhibited very low coverage (12.5% of reads) further suggesting that it was a sequencing artefact. Additionally, we found that the disruptive in-frame 6-bp deletion (c.89_94delTGGACG) in Abhd12, predicted to result in a missense substitution (p.L30-A32delinsP), was discordant with PCR-Sanger sequencing. Figure 3c revealed that while the Abhd12 variant was absent in CFW mice and homozygous in Em/J mice, it was more consistent with a 7-bp deletion and 1-bp insertion (c. 88_94delCTGGACGinsT) predicted to result in the missense substitution p.L30_A32delinsS. Alternatively, we cannot exclude the possibility that a 5-bp deletion (c.90_94delGGACG) and a 1-bp deletion (c.88delC), or some other rearrangement, occurred to generate the p.L30_A32delinsS substitution.
Fifth, we used the SIFT web server (https://sift.bii.a-star.edu.sg/ ) to predict the effects of novel missense SNVs on protein function (Sim et al. 2012). A SIFT score of <0.05 was predicted to be intolerant, whereas, a score of ≥0.05 was predicted to be tolerated. The Prx substitution (p.R167C) was predicted to be borderline intolerant (score = 0.05), whereas, the Adamts10 substitution (p.P761L) was tolerated (score = 0.33). Since SIFT did not generate a score for the Abhd12 indel substitution, we used the PROVEAN web server (http://provean.jcvi.org) that predicts the functional effects of single and multiple amino-acid substitutions, insertions, and deletions (Choi and Chan 2015). A PROVEAN score less than −2.5 was predicted to be deleterious, whereas, a score greater than −2.5 was predicted to be neutral. The Prx substitution (p.R167C) was predicted to be borderline neutral (score = −2.31), whereas, the Adamts10 substitution (p.P761L) was predicted to be neutral (score = 0.96). By contrast, the Abhd12 indel substitution (p.L30-A32delinsS) was predicted to be functionally deleterious (score = −20.74).
Finally, we compared the transcript expression profile of Prx, Adamts10, and Abhd12 in mouse eye tissues using the BioGPS gene portal (http://biogps.org) (Wu et al. 2016). Figure 4 shows that all three genes are expressed in the lens at higher levels than in adjacent eye tissues including the iris, ciliary body, and retina. Overall, while we cannot totally exclude Prx and Adamts10, our variant analysis data suggest that Abhd12 was the most plausible candidate gene for Em.

Discussion
The Em mouse inherits a spontaneous, uncharacterized, mutation underlying a reported autosomal dominant form of latemanifest (6-8 months of age), progressive, cortical cataract that has long been proposed as an animal model for age-related cataract in humans (Kuck 1990). We confirmed the cataract phenotype in commercially available homozygous Em/J mutant mice at 6-8 months of age but not in age-matched Em/J heterozygotes or CFW ancestral mice. Using whole-exome sequencing and variant analysis of Em/J and CFW mice, we excluded over 450 known genes for inherited and age-related forms of cataracts-with or without other ocular and/or systemic abnormalities-including those for lens crystallins, membrane and cytoskeletal proteins, DNA or RNA-binding proteins, along with many genes for syndromic/systemic forms of cataract (https:// cat-map.wustl.edu). Our variant analysis culminated in the prioritization of three genes Prx, Adamts10, and Abhd12 each with one predicted missense variant that were not identified in any public genomic database suggesting that they were possible candidates for Em (Table 2). It is noteworthy that the human genes for Adamts10 (ADAMTS10) and Abhd12 (ABHD12) are included in commercially available panels of genes that are routinely sequenced in a clinical setting for the molecular diagnosis of inherited eye diseases including cataract and other lens disorders, anterior segment dysgenesis, and retinal degenerations (e.g. blueprintgenetics.com).
Mutations in the human gene for Prx (PRX, MIM no. 605725) on chromosome 19q13.2 have been associated with autosomal recessive Charcot Marie Tooth disease type 4F (CMT4F, MIM no. 614895)    Table 1), indicated by arrows in the adjacent schematic, followed by gel-electrophoresis confirming that CFW and Em/J mice were heterozygous and homozygous, respectively, for the p.P761L substitution. c) PCR-Sanger sequencing of exon-1 from Abhd12 confirming that Em/J mice were homozygous for a p.L30_A32delinsS substitution that was not present in CFW mice. encodes a postsynaptic density protein (PSD95), Drosophila disc large tumor suppressor (DlgA), and zonula occludens-1 protein (ZO-1) or PDZ domain protein known as periaxin that plays a key role in the maturation, packing, and hexagonal membrane organization of mouse lens fiber cells (Maddala et al. 2011). The Prx substitution (p.R167C) identified in Em/J mice was predicted in silico to have a borderline neutral/damaging effect on protein function. Further, mice lacking Prx did not develop cataract by 6-9 months of age and soon after they succumbed to severe coordination abnormalities and neuropathic pain necessitating euthanasia (Gillespie et al. 2000;Maddala et al. 2011). However, given its role in lens cell biology and its tentative association with congenital cataract in humans, we cannot exclude Prx as a candidate gene for Em. Mutations in ADAMTS10 (MIM no. 608990) on human chromosome 19p13.2 have been associated with Weill-Marchesani syndrome 1 (WMS1, MIM no. 277600)-a rare autosomal recessive connective tissue disorder characterized by short stature, brachydactyly, joint stiffness, occasional heart defects, and eye anomalies, including severe myopia (94%), microspherophakia (small, spherical lens, 84%), glaucoma (80%), ectopia lentis (lens displacement, 73%), and cataract (23%) (Faivre et al. 2003;Dagoneau et al. 2004). Adamts10 encodes a secreted metalloproteinase from the ADAMTS [a disintegrin-like and metallopeptidase (reprolysis type) with thrombospondin type 1 motif, 10] superfamily of proteins with diverse roles in embryonic development and human disease (Dubail and Apte 2015;Mead and Apte 2018). However, the Adamts10 missense substitution (p.P761L) identified in Em/J mice was predicted in silico to exert a benign effect on protein function. Whereas mice lacking or mutant for Adamts10 have not been reported to develop WMS1-like lens defects including microspherophakia, ectopia lentis, and cataract (Mularczyk et al. 2018;Wang et al. 2019), homozygous Em/J mice developed cataract with full penetrance-suggesting that Adamts10 is a marginal candidate gene for Em.
Mutations in ABHD12 (MIM no. 613599) on human chromosome 20p11.21 have been associated with a rare, autosomal recessive syndrome characterized by (demyelinating) polyneuropathy, hearing loss, (cerebellar) ataxia, retinitis pigmentosa, and (early-onset) cataract (PHARC-MIM no. 612674) (Fiskerstrand et al. 2010;Nishiguchi et al. 2014;Nguyen et al. 2021). PHARC syndrome is a neurodegenerative disease with variable onset, severity, and progression of neurological, auditory, and ophthalmological symptoms that may confuse diagnosis with other "deaf-blind" phenotypes including Refsum disease (MIM 266500) (Fiskerstrand et al. 2009(Fiskerstrand et al. , 2010Eisenberger et al. 2012;Yoshimura et al. 2015;Thimm et al. 2020). We note that Refsum disease is caused by mutations in the gene encoding phytanoyl-CoA hydroxylase (PHYH; MIM 602026), the mouse counterpart of which (Phyh) was excluded here (Supplementary  Table 3). Typically, polyneuropathy, hearing loss, and ataxia present in the first to third decades of life; however, cases of PHARC syndrome without manifest ataxia and hearing loss have been documented (Nguyen et al. 2021). Retinitis pigmentosa (RP) and cataracts usually present in the second or third decades; however, the ophthalmic spectrum of PHARC syndrome includes cases with congenital cataracts and cases of "non-syndromic" RP with posterior polar cataract in the absence of other symptoms (Nishiguchi et al. 2014;Nguyen et al. 2021). While we did not observe obvious gait defects and did not test for hearing loss in Em/J mice, we note that studies of an Abhd12-related retinal phenotype are confounded since Em/J mice are also homozygous for the Pde6b rd1 retinal degeneration mutation. In both mice and zebrafish, loss of ABHD12 function caused a PHARC-like syndrome including neurological, auditory, and retinal defects (Blankman et al. 2013;Tingaud-Sequeira et al. 2017). Whereas cataract was not reported in Abhd12-null mice from 5-18 months of age, inhibition of lens clarification during eye development was observed in abhd12 knock-down zebrafish larvae (Tingaud-Sequeira et al. 2017). Since Abhd12 encodes a serine hydrolase abhydrolase domain-containing protein 12 (or lysophosphatidylserine lipase ABHD12) that catalyzes the hydrolysis of 2-arachidonoyl glycerol (2-AG) to arachidonic acid and glycerol, PHARC syndrome has been classified as both an inborn error of endocannabinoid metabolism and an inborn error of phospholipid metabolism involving elevated levels of lysophosphatidylserine lipids and the endocannabinoid 2-AG (Fiskerstrand et al. 2010;Blankman et al. 2013;Lamari et al. 2013;Wortmann et al. 2015;Leishman et al. 2019). Unlike the variants in Prx and Adamts10 above, the Abhd12 variant (p.L30_A32delinsS) found in Em/J mice was predicted in silico to be strongly damaging at the protein function level. We note that genetic mutations in several other lipid metabolism enzymes are known to elicit cataracts in humans, including lanosterol synthase (LSS, MIM no. 600909; autosomal recessive congenital cataract, CTRCT44, MIM no. 616509) (Zhao et al. 2015;Zhao et al. 2021) and acylglycerol kinase (AGK, MIM no. 610345; Sengers syndrome, MIM no. 212350; and a nonsyndromic form of autosomal recessive congenital cataract, CTRCT38, MIM no. 614691) (Aldahmesh et al. 2012;Mayr et al. 2012). Further, loss of the lipid kinase phosphatidylinositol-4phosphate 3-kinase catalytic subunit type 2α (PI3K-C2α) leads to early senescence and cataract development in humans, mice, and zebrafish (Gulluni et al. 2021). Taken overall, the cross-species association with human cataract, predicted enzymatic dysfunction, and relatively strong expression in the mouse lens (Fig. 4) supports Abhd12 as a credible candidate gene for Em and suggests that available Em/J mice may serve as an animal model for the cataract associated with human PHARC syndrome.
There are certain caveats concerning phenotype inheritance that impact this candidate gene study of Em/J mice. First, we were unable to map the Em locus to a chromosome possibly due to delayed onset and/or reduced penetrance of the reported autosomal dominant cataract inheritance (Kuck et al. 1981). We note that two sub-strains of Em mice have been reported, one with "early-cataract" onset at 5-6 months of age and the other with late-cataract onset at 8-9 months (Kuck 1990). While the 5-to 6-month-cataract sub-strain was reportedly bred to genetic homogeneity, it remains unclear that the early-onset and late-onset cataract represent homozygous and heterozygous Em phenotypes, respectively (Kuck 1990). Second, since mutations in PRX, ADAMTS10, and ABHD12 are all associated with autosomal recessive syndromes in humans, we cannot rule out the possibility that Em exhibits autosomal recessive inheritance in mice. The Em phenotype arose spontaneously in an individual male (from an inbred CFW colony) that "exhibited bilateral cataracts at 11 months of age" (Kuck et al. 1981). Two male siblings of the affected male later developed cataract at 17 and 18 months; however one died. Progeny of the two male founders were entered into a large breeding program (∼1,000 mice) that necessitated mating of suspected cataract (i.e. pre-symptomatic) parental mice since cataract-onset occurred beyond prime breeding age. Such breeding over several years generated the Em sub-strain with "a continually improving yield of cataracts" and "a high probability of cataract formation" (Kuck et al. 1981) that may have included carriers of an autosomal recessive cataract phenotype. Finally, in the absence of firm genetic linkage data for the Em phenotype, we were limited to analyzing known genes for cataract/lens disorders and therefore cannot exclude the possibility that a currently unidentified gene for cataract is involved.
In conclusion, whole-exome sequencing and variant analysis has prioritized three candidate genes for Em that are associated with lens cell biology and human cataracts. Although we cannot formally exclude Prx and Adamts10, our data suggest that Abhd12 is a promising candidate gene for Em. Further studies, including gene-editing (e.g. CRISPR-Cas9) to engineer knock-in mouse models and/or induced pluripotent stem cell (iPSC) in vitro models of lens development and cataract, will be required to confirm or exclude the causative role of all three candidate genes in generating the Em/J lens phenotype. Ultimately, if Prx, Adamts10, and Abhd12 are excluded as causative genes for Em further variant analysis of the exome data presented here may contribute to the identification of a novel gene for cataract.