UC Davis

Insidious uveitis is a form of equine recurrent uveitis (ERU), an inflammatory eye disease, that disproportionally affects the Appaloosa horse breed. Previous research indicates that these horses are eight times more likely to be affected by ERU and four times more likely to become blind as a direct result of the disease compared to other breeds studied. Previous research also suggests that an insertion in TRPM1 (LP), which leads to the breed’s characteristic leopard complex spotting pattern (LP), is a risk factor for ERU. TRPM1 plays a known role within normal eye function as an ion channel that is necessary for transmission of the light detection signals between rod photoreceptor cells and ON-bipolar nerve cells. Specifically, TRPM1 is needed for proper vision in low light (scotopic) conditions, and lack of the functional protein in homozygous LP horses, therefore, causes congenital stationary night blindness (CSNB). Several investigations in the Appaloosa breed also indicate that LP is associated with ERU disease status; however, LP genotype is not sufficient to explain the distribution of insidious uveitis among Appaloosa horses. It is, therefore, suspected that additional genetic loci may contribute to disease risk.Investigations of the cellular mechanisms of uveitis and genetic investigations in humans suggest that regulatory elements may play a large role in ocular immune privilege and intraocular inflammation. While extensive efforts have focused on annotating genes since the sequencing of the equine genome, recent efforts have focused on functionally annotating non-coding elements in the horse genome in a tissue-specific manner as part of the Functional Annotation of Animal Genomes (FAANG) initiative. The aim of the equine portion of the FAANG project is to generate publicly available data to improve upon the existing annotation of the equine genome and thereby advance research of putative functional variants outside of coding regions for investigations of equine diseases and traits of interest. Toward this aim, chromatin immunoprecipitation sequencing (ChIP-Seq) was chosen to evaluate four chemical modifications (H3K4me1, H3K4me3, H3K27ac, and H3K27me3) of histones proteins within eight prioritized tissues and four additional “adopted” tissues. The four histone modifications are associated with key genomic regulatory elements, such as promoters (H3K4me3) and enhancers (H3K4me1), and transcriptional activity, such as active (H3K27ac) and inactive (H3K27me3) genomic elements. Through these efforts, thousands of novel putative regulatory elements from two Thoroughbred mares were identified within the equine genome, and these histone mark peaks represent a region of increased interest outside of gene boundaries for variant identification in research on ERU and other equine diseases. Prior to any variant investigations for insidious uveitis, a prospective observational study was performed in the Appaloosa breed to better understand the major characteristics of ERU. In that investigation, we identified increasing age as a significant risk factor for disease (OR = 1.15, 95% CI [1.06 - 1.24], P = 0.001) and recapitulated the association with the LP locus when comparing LP/LP and lp/lp horses (OR = 19, 95% CI [2.8 - +INF], P = 0.009). Although uveitis cases were more likely to share common ancestors compared to the control horses as determined by coancestry coefficients (P = 0.02), affected horses did not have higher levels of inbreeding (P = 0.8), and a pedigree analysis did not identify a clear mode of inheritance from the data available. Heritability was estimated to understand the role of genetic variance in the distribution of insidious uveitis among Appaloosa horses. Using a combined and imputed dataset of 142 Appaloosas (59 cases and 83 controls) genotyped on either the Illumina Equine SNP70 BeadChip (n = 46) or the Axiom Equine 670K Genotyping Array (n = 96), SNP-based narrow-sense heritability of insidious uveitis was estimated to be 1.09 (SE = 0.16, P = 1.56x10-4). The h2 estimate accounted for sex and age as covariates and incorporated LP genotype as a fixed effect in the model, which contributed 0.20 of the overall heritability of ERU. The high estimate indicates that additional genetic loci outside of LP may be contributing to ERU in the dataset, warranting further investigation of genetic risk factors. Using genotyping array data (Axiom Equine Genotyping Array) from 96 Appaloosas (36 cases and 60 controls), a genome-wide association study (GWAS) with LP genotype, sex, age, and relatedness as covariates in a mixed linear model (MLM) identified a significantly associated region on the X chromosome after multiple testing correction (ECA X: 14.5 Mb, P = 2.11x10-8), and the association was still significantly associated with disease phenotype in a sex-stratified analysis (P = 1.35x10-8). Furthermore, a Cochran-Mantel-Haenszel test for genetic interaction indicated that LP is epistatic to the locus on ECA X (P = 1.72x10-6). To interrogate the X locus further, whole genome sequencing (WGS) data from 18 Appaloosas (nine cases and nine controls) were evaluated for potential causal single-nucleotide variants (SNVs). SNVs within 150 Kb of the locus of interest were prioritized if they were identified by both variant callers (Freebayes and BCFtools) and had genotype frequency differences of 30% or greater in cases compared to controls. To investigate the association further, 130 prioritized variants from the region of interest were genotyped in a larger cohort of 157 horses (70 cases and 87 controls). One variant was identified within a coding region, and four variants were found within histone mark peaks based on the publicly available FAANG data. However, no SNVs were perfectly concordant with phenotype in the region of interest. Although the original GWAS SNPs remained significant in this larger cohort (P = 4.06x10-5), none of the 102 SNVs that passed quality filtering were significantly associated with disease. Based on these results, it is suspected that the top three markers may be tagging a structural variant or another uncharacterized mutation, and additional work is needed to further characterize the association on ECA X for putative causal variants. To compliment the clinical investigation of insidious uveitis in Appaloosas, a prospective observational study was performed in the Knabstrupper breed, and we discovered that the characteristics of ERU are similar between the Appaloosa and the Knabstrupper. In particular, signs of ocular discomfort and inflammation, such as aqueous flare, conjunctival hyperemia, hypotony, and miosis, were more common in uveitis affected Appaloosas and Knabstruppers compared to controls (P < 0.001), while two “historic” indicators of ERU were not significantly more common (bullet-hole lesions, P > 0.2 and butterfly lesions, P > 0.3). Furthermore, LP/LP Knabstruppers were at significantly higher odds of developing ERU compared to true solid horses (lp/lp) (OR = 7.64, 95% CI [0.8 - +INF], P = 0.04), and horses in the age range of 11-20 years old had significantly higher odds of ERU compared to the reference age group (< 5 years old) (11-15 years OR = 8.19, 95% CI [1.2 – 105.6], P = 0.02 and 16-20 years OR = 13.36, 95% CI [1.4 – 213.4], P = 0.009). These results are consistent with findings from the Appaloosa breed. Additionally, modes of inheritance were investigated in a pedigree analysis, yet it was not possible to conclusively exclude any simple models of inheritance. The data suggest that a simple sex-linked or maternal effect mechanism does not explain the inheritance pattern of ERU in the Knabstrupper; however, investigations of additional families with phenotyping across multiple generations is necessary. Although inbreeding was not significantly different between the cases and the control horses (P > 0.9), Knabstruppers with ERU were more likely to share common relatives compared to the controls (P = 0.01). Given the similarities between the disease phenotype identified across breeds with LP, a GWAS for insidious uveitis was performed on a combined dataset of 250 horses (111 cases and 139 controls) from the Appaloosa, Pony of the Americas (POA), and Knabstrupper breeds. The only region to reach genome-wide significance contained the LP locus on ECA 1 (P = 6.58x10-9), and a haplotype analysis identified a 76 Kb haplotype, termed “Haplotype A,” in all affected horses. However, Haplotype A was also present in many unaffected horses and was not more concordant with ERU compared to the LP locus (PHapA = 9.04x10-21 versus PLP = 1.08x10-27). While it is suspected that the LP allele itself may play a role in ERU, further refinement and functional assessments of the associated region are needed to determine the underlying role of that locus in insidious disease. Taken together, all these data support that there is a strong genetic underpinning to insidious uveitis, and the LP locus is a major risk marker for insidious uveitis across LP breeds. However, the causative variant in the region containing LP is suspected to have incomplete penetrance for ERU, and identification of other breed-specific risk factors, such as the associated region on ECA X, will help to better explain the distribution of insidious uveitis within the Appaloosa, Knabstrupper, and POA breeds.