Mendelian Randomization Shows a Causal Effect of Low Vitamin D on Non-infectious Uveitis and Scleritis Risk.

PURPOSE: To investigate a causal relationship between Vitamin D levels and non-infectious uveitis and scleritis using Mendelian randomization (MR) techniques. DESIGN: Two-sample Mendelian randomization case-control study. METHODS: The study setting was a biobank of an academic, integrated health care system. The patient population comprised 375 case patients with a non-infectious uveitis and/or scleritis diagnosis and no diagnosis of infectious, trauma-related, or drug-induced uveitis/scleritis. In addition, there were 4167 controls with no uveitis or scleritis diagnosis. Causal effect estimates of low 25-hydroxy Vitamin D (25OHD) on uveitis/scleritis risk were calculated. RESULTS: We found an association of genetically decreased 25OHD with uveitis/scleritis risk (odds ratio [OR] = 2.16, 95% CI = 1.01-4.64, P = .049, per SD decrease in log25OHD). In a first sensitivity MR analysis excluding the genetic variants that are unlikely to have a role in biologically active 25OHD, effect estimates were consistent with those from the primary analysis (OR = 2.38, 95% CI =1.06-5.36, P = 0.035, per SD of log25OHD). Furthermore, in a second sensitivity analysis using only the 6 variants within the CYP2R1 locus (which encodes 25OHD hydroxylase, the liver enzyme responsible for converting Vitamin D to 25OHD), genetically decreased 25OHD was strongly associated with increased uveitis/scleritis risk (OR = 6.42, 95% CI = 3.19-12.89, P = 1.7 × 10-7, per SD of log25OHD). CONCLUSIONS: Our findings suggest a causal relationship between low Vitamin D levels and higher risk of non-infectious uveitis and scleritis. Vitamin D supplementation may be a low-cost, low-risk intervention to mitigate non-infectious uveitis and scleritis risk, and should be explored in a prospective trial.

Association of Low Vitamin D Levels with Noninfectious Uveitis and Scleritis.

Purpose: To determine whether an association between Vitamin D and noninfectious ocular inflammation exists. Methods: Retrospective case-control study with 765 patients (333 uveitis cases, 103 scleritis cases, 329 controls). Logistic regression models examined the relationship between hypovitaminosis D and ocular inflammation. Results: The odds of having uveitis were 1.92 times higher for patients with hypovitaminosis D compared to patients with normal Vitamin D levels in the multivariate analysis [odds ratio (OR) = 1.92, 95% Confidence Interval (CI) = 1.36-2.72, p = 2.32 × 10-4]. A secondary analysis demonstrated that the odds of developing uveitis or scleritis were 5% lower and 4% lower, respectively, for every unit increase in Vitamin D level (uveitis: OR = 0.95, 95% CI = 0.94-0.97, p = 9.87 × 10-6; scleritis: OR = 0.96, 95% CI = 0.93-0.99, p = 0.009). Conclusion: Hypovitaminosis D was associated with increased risk of ocular inflammation in this retrospective study.

Phage Immunoprecipitation Sequencing of Autoantigens in Autoimmune Retinopathy.

PURPOSE: To identify autoantigens in autoimmune retinopathy patients by phage immunoprecipitation sequencing (PhIP-Seq), a new technique for autoantigen discovery. METHODS: PhIP-Seq was used to sequence putative autoantibodies in plasma from 11 patients with autoimmune retinopathy and eight controls. We compared the autoantibodies' molecular weights with those of proteins detected by Western blot. RESULTS: Several autoantigens were found in cases and not detected in the controls. Autoantigens RTN3, PRPF6, TRPC6, and B3GNT8, four proteins expressed in the retina, were detected in plasma as autoantibodies from one patient each and no controls. Only one patient had an autoantibody, B3GNT8 (43.4 kDa), within a similar weight range as that detected by antiretinal antibody Western blot (42 kDa). Autoantibody POLR3A, which has a well-characterized role in scleroderma, was detected in two cases and no controls. CONCLUSION: PhIP-Seq detected autoantigens that are expressed in the retina as well as scleroderma-related autoantigens in autoimmune retinopathy patients.

Whole exome sequencing identification of novel candidate genes in patients with proliferative diabetic retinopathy.

Rare or novel gene variants in patients with proliferative diabetic retinopathy may contribute to disease development. We performed whole exome sequencing (WES) on patients at the phenotypic extremes of diabetic retinal complications: 57 patients diagnosed with proliferative diabetic retinopathy (PDR) as cases and 13 patients with no diabetic retinopathy despite at least 10years of type 2 diabetes as controls. Thirty-one out of the 57 cases and all 13 controls were from the African American Proliferative Diabetic Retinopathy Study (AA). The rest of the cases were of mixed ethnicities (ME). WES identified 721 candidate genes with rare or novel non-synonymous variants found in at least one case with PDR and not present in any controls. After filtering for genes with null alleles in greater than two cases, 28 candidate genes were identified in our ME cases and 16 genes were identified in our AA cases. Our analysis showed rare and novel variants within these genes that could contribute to the development of PDR, including rare non-synonymous variants in FAM132A, SLC5A9, ZNF600, and TMEM217. We also found previously unidentified variants in VEGFB and APOB. We found that VEGFB, VPS13B, PHF21A, NAT1, ZNF600, PKHD1L1 expression was reduced in human retinal endothelial cells (HRECs) cultured under high glucose conditions. In an exome sequence analysis of patients with PDR, we identified variants in genes that could contribute to pathogenesis. Six of these genes were further validated and found to have reduced expression in HRECs under high glucose conditions, suggestive of an important role in the development of PDR.