Personalized Proteomics in Proliferative Vitreoretinopathy Implicate Hematopoietic Cell Recruitment and mTOR as a Therapeutic Target

Abstract

Purpose: To profile vitreous cytokine expression of proliferative vitreoretinopathy (PVR) patients.

Design: Case-control study.

Methods: Liquid biopsies were collected from 2 groups: control subjects (n = 3) undergoing pars plana vitrectomy to remove an epiretinal membrane (ERM), and test subjects (n = 7) with varying degrees of PVR. A high-throughput cytokine screen measured expression of 200 cytokines. Cytokine expression patterns were prospectively validated in separate cohorts of control patients and those with PVR-A, PVR-B, and PVR-C (n = 10 for each group). Expression changes were evaluated by analysis of variance (significant P value < .05), hierarchical cluster algorithm, and pathway analysis, to identify candidate pathways for prospective studies.

Results: In PVR vitreous, 29 cytokines were upregulated compared to controls. Early PVR vitreous showed upregulation of T-cell markers, profibrotic cytokines, and cytokines downstream of mTOR activation (IL-2, IL-6, and IL-13), whereas in late PVR vitreous, cytokines driving monocyte responses and stem-cell recruitment (SDF-1) prevailed. Prospective validation confirmed the differential expression of specific cytokines from PVR-A to C.

Conclusions: Early PVR is characterized by activation of T cells and mTOR signaling, whereas advanced PVR is characterized by a chronic monocyte response. PVR might be treated by rational repositioning of existing drugs that target mTOR and IL-6. Our analysis demonstrates that successful therapeutic intervention will be highly dependent on the specific therapeutic target and the stage of PVR. This study provides insights into cytokines that will serve as biomarkers and therapeutic targets. These biomarkers will help design clinical trials that intervene at appropriate times.

Figure 1. Phenotypic Ascertainment

Clinical features of grade A proliferative vitreoretinopathy (PVR-A). Fundoscopic exam reveals retinal detachment (rd) and open arrow denotes pigment cells on retina (top left). Clinical features of PVR-B. Open arrow denote rolled edges of an open retinal break (middle left). Clinical features of PVR-C. Open arrow denote (bottom left). Hierarchal clustering of reveals 29 upregulated cytokines (p<0.05) in seven patients with PVR versus three control cases (right). Results are represented as a heatmap and display protein expression levels on a logarithmic scale. Orange indicates high expression while dark green/black indicates low or no expression.

Figure 2. Differential cytokine expression in early proliferative vitreoretinopathy (PVR) reveals mTOR signaling involvement

Hierarchal clustering of cytokines differentially expressed (p<0.05) in three patients with early-PVR versus three control cases. A total of 8 cytokines were downregulated and 50 were upregulated (left). Pathway analysis of differentially-expressed in early-PVR (top right). Pathways are organized by the log (p-value) obtained from the right-tailed Fisher Exact Test. The connecting line represents the ratio of enrichment, which is equal to the number of observed proteins divided by the number of expected proteins from each pathway that is represented. The cytokines downstream of mTOR in early-PVR were leukemia inhibitory factor (LIF), transforming growth factor beta 2 (TGFB2), tumor necrosis factor (TNF), interferon gamma (IFNg), interleukin-2 (IL-2), interleukin-13 (IL-13), platelet-derived growth factor beta (PDGF-BB), Fas, Endoglin, interleukin-13 receptor (IL-13R), E-cadherin, interleukin-13 receptor subunit alpha 1 (IL13RA1), colony stimulating factor 1 receptor (CSF1R), osteopontin (OPN), interleukin-13 receptor 2 (IL-13 R2), and macrophage chemotactic factor receptor (MCF-R). Protein interactome of cytokines in early-PVR colored by immune function (bottom left). The protein interactome was constructed by interrogating protein-protein interaction databases using network analysis software (STRING). Each protein is represented as a circle (node) while known interactions between proteins are represented by lines (edges). Proteins that do not fall in the central network of circles form fewer interactions than those in the center. Coloring individual proteins by immune function reveals that a majority (33%) of cytokines in early-PVR are involved in T-cell recruitment and differentiation.

Figure 3. Differential cytokine expression in advanced proliferative vitreoretinopathy (PVR)

Hierarchal clustering of cytokines differentially expressed (p<0.05) in four patients with advanced-PVR versus three control cases. A total of 14 cytokines were upregulated (top left). Hierarchal clustering of cytokines differentially expressed (p<0.05) in three patients with early-PVR versus four patients with advanced-PVR (bottom left). A total of 16 cytokines were downregulated and 7 were upregulated, highlighting proteins that are downregulated as PVR progresses to stage C (advanced). Pathway analysis of differentially-expressed cytokines in advanced-PVR (top right). Pathways are organized by the log (p-value) obtained from the right-tailed Fisher Exact Test. The connecting line represents the ratio of enrichment, which is equal to the number of observed genes divided by the number of expected genes from each pathway that is represented. Protein interactome of cytokines in advanced-PVR colored by immune function (bottom right). Coloring individual proteins by immune function reveals that all the cytokines in cytokines PVR are involved in monocyte recruitment and differentiation.

Figure 4. Stage-specific expression of validated cytokines

Comparison of cytokines in our validation study based on their picogram levels at different stages. Data were analyzed using 1-way ANOVA followed by Bonferroni’s multiple comparison test. Alpha was adjusted for multiplicity. Individual cytokine levels are grouped based on their cellular function: Monocyte recruitment (top row), T-cell recruitment (middle row), Wound healing (middle row), and Fibrosis (bottom row).

Figure 5. A proliferative vitreoretinopathy (PVR) pathway model for therapeutic testing

Downstream mTOR effectors in PVR categorized by disease stage (left). A constructed disease model highlighting the molecular phenotype with associated cytokines and correlation to clinical phenotype of PVR (right). Potential therapeutics which have already been approved for use in other diseases are represented based on upregulated cytokines and pathways.