Identification of immune associated potential molecular targets in proliferative diabetic retinopathy

Diabetic retinopathy (DR), resulting from chronic hyperglycemia, is one of the most common microvascular complications of diabetes and causes of blindness among adults aged between 20 and 74 in developed countries [1, 2]. Depending on the degree of related ischemic injury and microvascular lesions, DR can be divided into two phases, non-PDR (proliferative diabetic retinopathy) and PDR [3]. DR is recognized as a microvascular, inflammatory, and neurodegenerative complication of diabetes [4], which can be triggered by mitochondrial damage, endoplasmic reticulum stress, and oxidative stress, among others [5]. However, the pathogenesis of DR has not been fully elucidated. Inflammation and angiogenesis play key roles during DR pathogenesis [6]. Thus, the morphological and molecular alterations in DR caused by these two processes are receiving great attention, becoming the focus of extensive research [7, 8].

Chronic low-grade inflammation is present during both the early and advanced stages of DR, eventually resulting in retinal vasculopathy, which is characterized by increased retinal vascular permeability and neovascularization [9, 10]. Inflammation is undoubtedly implicated in the dysregulated pathological angiogenesis during early and advanced stages of DR [11]. Neovascularization and inflammation share several common mediators and signaling pathways [11, 12]. Inflammatory responses drive angiogenic processes through the production of pro-angiogenic cytokines and growth factors [11‐13].

Shao et al. identified a subset of differentially expressed genes (DEGs) from both active and inactive fibrovascular membranes (FVMs) with normal retinas, which were enriched for angiogenic factors [hypoxia inducible factor-1 subunit alpha (HIF-1α) and placental growth factor (PGF)] [14]. Pathological secretion of vascular endothelial growth factor A was shown to promote the expression of pro-angiogenic transcription factors and growth factors, which in turn induced retinal neovascularization [15]. In the present study, we first analyzed an RNA-seq dataset of NVMs from PDR patients from the GEO database. We identified genes associated with the immune system in order to elucidate the role of inflammatory processes during DR pathogenesis and identify novel diagnostic and therapeutic markers for DR.

In this study, we first separately analyzed immune-related gene expression in NVMs from PDR of both types in order to narrow down and identify potential genes implicated in PDR pathogenesis. The most significantly enriched pathways were mainly implicated in chemokine signaling and cytokine-cytokine receptor interactions. The current study provides better insight into the immune mechanisms underlying DR progression, with potential implications for diagnosis and treatment.

Chemokines are critical mediators of immune cell migration, with essential roles in immune surveillance, development, and inflammation. Chemokines exert their effects via transmembrane G protein-coupled receptors (GPCRs) present on a wide variety of cell types. Upon binding, conformational changes in trimeric G proteins trigger intracellular signaling pathways, promoting cellular movement and activation [17]. Based on the locations of conserved cysteine residues near the amino terminus, chemokines are divided into four subfamilies: C-C chemokine motif receptor (CCR), C-X-C chemokine motif receptor (CXCR), CX3CR, and XCR [18]. GPCRs regulate leukocyte trafficking and promote immune responses, mediating cell chemotaxis. GPCRs can be categorized into classical and chemokine subfamilies according to the ligand source. Classical GPCRs include formyl peptide receptors (FPR1, FPR2, and FPR3), platelet-activating factor receptor (PAFR), activated complement component 5 receptor (C5aR), and leukotriene B4 receptors (BLT1 and BLT2) [19].

In our study, 11 chemokines and their cognate receptors were identified as related to the pathogenesis of DR, including CCR1, CCR7, CCL5, CCL20, CXCL1, CXCL3, CXCL8, CXCL9, CXCL10, FPR1, and GNAI3.

CSF3, COL18A1, CXCR2, CCR1, FGF23, CXCL11, and IL13 were previously reported as related to PDR pathogenesis based on a Laplacian heat diffusion algorithm [20]. Therapeutic strategies targeting MIP1γ to inhibit CCR1-related signaling in retinal endothelial cells might have potential against DR progression [21]. CCR7 significantly enhanced neovascularization and the non-perfusion area in oxygen-induced retinopathy [22]. CCL5 could be measured in the blood, vitreous body, retina, aqueous humor, and tears of patients with DR [23]. C-C-chemokine receptor 6 is the only receptor interacting with CCL20. Treatment with CCL20-neutralizing antibodies or PG inhibits CCL20 expression, alleviating retinal degeneration and inflammation [24]. CCL20, CXCL2, and other core genes have been described as playing key roles in DR pathogenesis [25]. Activation of the P2X7R-NLRP3 pathway significantly increased the production of TNF-α, CXCL-1, CSF-1, IL-6, IL-1β, IL-18, and other pro-inflammatory cytokines in retinal microglia [26]. CXCL3, VEGF, CXCL5, and other inflammatory mediators were increased in DR and retinopathy of prematurity [27]. CXCL8 (also known as IL-8) is known to be elevated in the vitreous humor of patients with DR [28, 29]. Autocrine CXCL9 and CXCL10 signaling in retinal endothelial cells were enhanced in DR [30]. Recent research suggests that the level of CXCL10/IP-10 in normal vitreous humor was significantly higher than that in serum. Further, CXCL10/IP-10 levels in the vitreous humor of proliferative vitroretinopathy and PDR patients were much higher than that in patients with rhegmatogenous retinal detachment. CXCL10/IP-10 was also significantly upregulated in patients with active PDR compared to those with inactive disease [31]. CXCL10 derived from platelet-rich plasma exosomes can cause retinal endothelial injury, which was considerably alleviated by antagonizing CXCL10 with a neutralizing antibody [32]. UPARAR is a peptide inhibitor of the uPAR system, which could reverse the upregulation of uPAR, FPR1, and FPR2, thus slowing DR development [33]. The GαI1/3 protein plays a key role in vascular endothelial growth factor-induced endocytosis, signal transduction, and angiogenesis. High GαI1/3 protein expression was previously reported in the proliferative retinal tissue of PDR patients [34].

CCR4, CXCR6, C3AR1, LPAR1, C5AR1, and P2RY14 have been implicated in a number of eye diseases, but not in PDR. Upregulated CCR4 expression was previously observed in keratoconjunctivitis, glaucoma, and uveitis [35‐37], in addition to its downregulation in dry eyes [38]. The IFN-γ and IL-17 expression of CD4⁺ T cells was significantly increased in patients with age-related macular degeneration. IFN-γ-expressing Th1 cells and IL-17-expressing Th17 cells could be selectively enriched based on the expression of surface CCR3⁺, CCR4⁺, CCR6⁺ [39]. CXCR6 was upregulated in the primary culture of orbital fibroblasts from patients with Graves’ orbitopathy, following treatment with pro-inflammatory cytokines IL-1β and TNF-α [40]. Furthermore, CXCR6 expression was highly confined to memory Th1 cells, which can be categorized into activated memory Th1 and Tc1 cells secreting IFN-γ [41]. A previous study reported that CD4⁺ and CXCR6⁺ cells were decreased in T1D patients [42]. A possible reason for the reduction in CD4⁺ cells expressing CXCR6, CXCR3, and CCR5 could be the selective recruitment of Th1 cells into the pancreas [43]. In human T cells, intracellular C5AR1 signaling induces ROS production through the mitochondria. ROS in turn trigger assembly of the NLRP3 (NACHT, leucine-rich repeat and pyrin domain-containing protein 3) inflammasome. Inflammasome formation initiates caspase-1-dependent IL-1β secretion, which promotes IFN-γ production and Th1 differentiation in an autocrine manner. Secreted C5a/C5a-Desarg interacts with surface-expressed C5AR2, which negatively controls NLRP3 activation through a currently undetermined mechanism [44]. Further, there is growing evidence that microglia-mediated inflammatory responses are associated with deleterious effects implicated in DR [7]. In fact, increased hypertrophic amoeba-like microglia were observed in the outer retina and subretinal space of human DR patients [45]. Overactivated amoeba-like microglia lead to a dysregulation of the complement system by upregulating the expression of activators C3, CFB, C1q, and C5AR1, while downregulating that of complement inhibitors CFH, CFI, CD46, and CD93 [46]. Subsequently, microglial overactivation establishes a pro-inflammatory environment conducive to further invasion of retinal microglia and exogenous monocyte infiltration [47]. The accumulation of subretinal microglia derived paracrine factors can trigger NLRP3 inflammasome activation in the retinal pigment epithelium [48]. Studies previously showed that complement may modulate the production of inflammatory factors and angiogenic factors via C5AR on Müller cells, which are implicated in DR pathogenesis [49]. C3AR1 is considered an injury-induced neuroinflammatory factor, whose interaction with IL-10 signaling and other immune-related pathways might be a major regulator of microglial activity and neuroinflammatory function [50]. In DBA/2 J mice, significant damage occurred with in the optic nerve head (ONH) prior to in other regions of the optic nerve [51]. At the same time point, the expression of C3AR1 in the ONH increased, with no increase observed in the retina [52]. In healthy brains, cell types other than the microglia exhibited low or no expression of C3AR1 [53]. The involvement of C3AR1/C5AR1 signaling in angiogenesis was reported on day 5 in ocular and retinal angiogenesis neonatal mouse models [54]. In our study, the sampled area included the area around the optic nipple. Previous works have shown that six G-coupled protein receptors (LPAR1–6) can be activated by lysophosphatidic acid (LPA). LPA and its receptors play vital roles in the central nervous system, cancer, and macular edema [55]. A link between LPA and retinopathy was previously demonstrated, as LPA1 and LPA2 expression were significantly increased in retinal ganglion cells after retinal ischemia in adult rats, leading to LPA1-mediated retinal ganglion cell death in preterm infants. In contrast, LPAR1–3 expression of retinal pigment epithelial cells promoted retinal healing. It was suggested that LPA exerts either a neuroprotective or neurodegenerative effect on the retina by binding to different LPA receptors on different cell types [56]. ATX, AGK, and LPA1 receptors are expressed in vascular endothelial cells and stromal cells within PDR epiretinal membranes. LPA-producing enzymes or LPA were shown to play a key role in the development of PDR and PVR [57]. The expression of P2Ry14 (purinergic receptor P2Y, G-protein coupled, 14) in the trabecular meshwork is higher than that of other purine receptors, suggesting that the protein product reduces IOP in monkeys, an observation that has not been further confirmed [58]. P2Ry14 was down regulated during both oxygen-induced pathologic neovascularization and physiological angiogenesis of the retina [59]. Purine receptor P2Y14 is highly expressed in collecting tube insertion cells and mediates renal aseptic inflammation [60].

He et al. previously suggested that differential ALDH2/SIRT1 expression might be responsible for the differences in DR severity between chronic inflammation-related T1 diabetes mellitus and T2 diabetes mellitus. Retinal IL-1 and IL-6 production in the T1 diabetes mellitus group was significantly increased compared to that in the T2 diabetes mellitus group [61]. In our study, we found no significantly enriched gene sets between type 1 and type 2 PDR groups. Li et al. previously reported that the prevalence of DR in diabetes patients was affected by diabetes duration, diabetic nephropathy occurrence, and regular DR screening. Diabetes type indirectly affected DR occurrence through its influence on diabetes duration and diabetic nephropathy occurrence [62].

In the present study, we found that six immune-related genes, namely, CCR4, CXCR6, C3AR1, LPAR1, C5AR1, and P2RY14, were upregulated based on bioinformatics analysis. CCR4, C5AR1, CXCR6, C3AR1, and LPAR1 expression was further confirmed via qRT-PCR, which was not the case for P2RY14, necessitating further study. Moreover, the upregulation of LPAR1 was found to have both neuroprotective and PDR promoting properties. Therefore, whether LPAR1 acts as a protective or pathogenic factor in PDR remains unclear.

In summary, our findings provide insight into the molecular pathogenesis of DR, which may be of value for disease diagnosis and treatment. The roles of P2RY14 and LPAR1 in the pathogenesis of DR require further study. Our data provide a new idea for the diagnosis and treatment of DR in the future.

In our study, Nevertheless, the current study had some limitations, one being the limited number of neovascularization samples from patients with type 1 DR and the normal retina samples. We analyzed hub gene expression only via qRT-PCR, and it should be further validated via western blotting. Meanwhile, the alterations of immune-elated pathways might be explored in the future.