Vitreous metabolic perturbations during bacterial and fungal endophthalmitis

Project Summary Endophthalmitis is a devastating complication caused by a wide range of microorganisms which can result in irreversible vision loss if not diagnosed and treated promptly. The current diagnostic approach heavily relies on microbial culture to determine the etiology of endophthalmitis. However, only ~40% of suspected endophthalmitis cases are culture-positive, leaving many potentially infectious culture-negative cases. Therefore, the discovery of potential biomarkers is urgently needed for both the diagnosis of endophthalmitis and in guiding appropriate antimicrobial therapy. Here, we propose to use metabolomics, an emerging high-throughput technology that can identify, quantify, and characterize hundreds to thousands of low molecular weight biochemicals (metabolites), using targeted or untargeted analytical approaches. Since metabolism is directly or indirectly linked to every aspect of cell function, metabolomics is believed to reflect the phenotype of what is occurring in the body. Recently, we performed untargeted metabolomics in an experimental model of endophthalmitis and discovered distinct metabolic profiles of uninfected and bacterial-infected mouse eyes. Moreover, metabolomics analysis led to discovering a therapeutic target, i.e., itaconate/Irg-1 signaling, to ameliorate experimental bacterial endophthalmitis. Our pilot vitreous metabolomics analysis showed distinct metabolic profiles in patients with uveitis, endophthalmitis, and other inflammatory retinal diseases. Based on preliminary data and our prior publications, we hypothesize that “intraocular infections induce biochemical changes in the vitreous leading to distinct metabolic signatures with translational potential for therapeutic targeting and/or diagnosis of infectious endophthalmitis”. This will be tested by performing untargeted metabolomics in vitreous of culture-positive bacterial and fungal endophthalmitis (Aim-1). We will further devise a metabolic profile to aid in predicting endophthalmitis in culture-negative chohort. This will be accomplished this by developing an artificial intelligence (AI) model that can predict endophthalmitis' etiology based on vitreous metabolic profiles (Aim-2). The experiment will also involve validation using an independent cohort of culture-positive vitreous. The proposed study will be the first to utilize metabolomics profiling to identify the metabolic sigantures in the vitreous of endophthalmitis patients. The resulting profiles and/or biomarkers will be readily translatable for future clinical studies, ultimately guiding precision medicine for diagnosing and treating ocular infections.