Metabolism: How removal of damaged cells impacts energy availability in the retina

At its core, metabolism is a set of life-sustaining chemical reactions that transforms chemical energy into energy for cellular processes. These chemical reactions allow organisms to grow, reproduce, and respond to their environments. Therefore, it may not be surprising that a critical tissue for responding to environments — the eye — is highly active metabolically. Photoreceptors in the retina rely on aerobic glycolysis to meet their energy demands to convert light into signals that stimulate physiological responses^1,2. These photoreceptor cells, however, are separated from glucose and other essential metabolites in the underlying layer of blood vessels by the retinal pigment epithelium (RPE), which forms the blood–retina barrier. Consequently, the RPE maintains the health of the sensitive photoreceptors by transporting molecules across this barrier and by removing damaged photoreceptor outer segments (POS). A new study published in Nature Metabolism led by the labs of Kodi Ravichandran and Jayakrishna Ambati reports a molecular link between these two functions of the RPE³. They discovered that RPE cells provide a phagocytosis-induced source of insulin in the retina, and this source of insulin is necessary for metabolic homeostasis in the retina during starvation (Figure 1). These influential findings add to our understanding of how cell death, removal and recycling modulate tissue metabolism.

Within the vertebrate eye, photoreceptor cells absorb photons through G-protein coupled receptors called opsins that activate a phototransduction cascade in the POS. High levels of light, however, induce photo-oxidative reactions within the photoreceptor cells. To deal with these cytotoxic reactions, photoreceptor cells shed their POS daily, and the POS are subsequently phagocytosed by the RPE (Figure 1B)⁴. The magnitude of this homeostatic process is tremendous; estimates suggest that every phagocyte of the RPE engulfs hundreds of thousands of POS discs over a human lifetime⁴. Thus, POS phagocytosis has been linked to metabolic changes in the retina. Previous insights revealed that engulfment of POS not only fuels the RPE through fatty acid metabolism⁵, but also regulates the transport of circulating glucose through the RPE to the retina⁶. The direct molecular pathways that connect POS phagocytosis and metabolic homeostasis in the retina, however, were poorly understood.

Keenly aware of the connection between RPE phagocytosis and metabolism, Iker Etchegaray et al.³ reanalyzed RNA-sequencing data from their previous study examining the effects of the loss of the phagocytic receptor Mer tyrosine kinase (MerTK) on retinal degeneration⁷. The gene Ins2 (one of two genes that encodes insulin) was significantly downregulated in the RPE of mice lacking MerTK, while other factors related to growth and metabolism were not altered. Furthermore, Ins2 is distinctly expressed in the RPE and not expressed in the photoreceptors or neurons of mouse and human retinas. Interestingly, Ins2 mRNA level fluctuated in the retina at different time points during the day, following a similar circadian pattern previously reported for POS phagocytosis. This observation led the investigators to determine whether POS phagocytosis directly induces Ins2 expression. Inhibition of two phagocytic receptors MerTK or CD36 significantly decreased Ins2 expression. On the other hand, enhanced MerTK activity significantly increased both POS phagocytosis and Ins2 expression in the RPE. Collectively, these studies demonstrated that POS phagocytosis induces Ins2 expression in the RPE (Figure 1C).

Given these interesting results, we are left to ponder the direct molecular links of phagocytic receptor activation, POS engulfment and Ins2 expression. Is Ins2 expression only dependent upon the process of POS engulfment or must specific phagocytic receptors be activated? The RNA-sequencing data that served as the impetus of these studies suggest that these transcriptional changes might be unique to MerTK, as transgenic expression of another phagocytic receptor does not rescue retinal degeneration⁷. But, as the authors note, the Mertk knockout mice used in these studies carry multiple coincidental changes within their genome, including functional loss of the Mertk paralog Tyro3⁸. The investigators therefore utilized a separate mouse model deficient for the phagocytic receptor Cd36 and observed decreased Ins2 expression, although the effects were less pronounced than in Mertk (and possibly also Tyro3) knockout mice. Further suggesting a direct role of MerTK, enhanced MerTK activity significantly increased Ins2 expression. Whether this is due to increased POS uptake or MerTK-dependent signaling remains unclear (Q1 in Figure 1). Outside of its role in POS phagocytosis, MerTK has recently been linked to immunosuppressive functions in retinal homeostasis⁹. Thus, does MerTK activation, aside from inducing POS phagocytosis, also induce Ins2 expression and could this be targeted therapeutically? Overall, Iker Etchegaray et al.³ provide a solid foundation for future studies to interrogate the molecular mechanisms by which phagocytic receptors modulate Ins2 expression in the RPE.

Other than the pancreas, the eye has the highest expression of Ins2 compared to other tissues, implying that Ins2 may have a unique role in eye homeostasis. Ins2 encodes insulin, a peptide hormone that increases glucose uptake by translocation of glucose transporters to the plasma membrane via insulin receptor signaling (Figure 1D,E). It is unclear how insulin regulates the metabolically active ecosystem of the retina, a tissue where Krebs first described the release of large amounts of the glycolysis by-product lactate, even in the presence of oxygen (aerobic glycolysis)². Photoreceptors rely heavily on glucose to perform aerobic glycolysis, yet these photoreceptors are physically separated from circulating glucose in the blood by the RPE. Therefore, the RPE has evolved to utilize lactate for energy metabolism and to allow for glucose to be transported across the barrier to the photoreceptors¹⁰. Yet, even though the retina relies heavily on glucose, previous studies reported that the retina was insensitive to changes in systemic insulin levels¹¹. Retinal homeostasis, however, was severely affected after inhibition of insulin receptor signaling¹². These somewhat conflicting conclusions imply that the retina contains a local source of insulin regulated separately from systemic insulin.

Iker Etchegaray et al.³ ultimately discovered that the RPE is the local producer of insulin in the retina. The ramifications of this finding are clearly apparent during starvation, when the concentration of insulin in the blood significantly decreased while the concentration of insulin in the retina remained unchanged. However, depletion of Ins2 caused the retina to become sensitive to starvation, as observed by decreased insulin receptor activation and decreased glucose transport (Figure 1D,E). Physiologically, loss of Ins2 resulted in defects in visual processing and the response to light under starved conditions and may also contribute to the progression of retinitis pigmentosa. Consequently, the RPE may serve as an ‘emergency’ source of insulin to sustain glucose uptake in photoreceptor cells, maintaining aerobic glycolysis during times of starvation. Supporting this hypothesis of a compensatory source of phagocytosis-derived insulin is the finding that starvation increased the amount of internalized POS and Ins2 expression in RPE cells. But how does the RPE detect and respond to systemic changes in metabolite availability during starvation (Q2 in Figure 1)? Do photoreceptors increase the shedding of their outer segments or does the RPE increase their phagocytic machinery? Understanding the molecular mechanisms by which the RPE senses and responds to changes in systemic metabolic changes is an important next step to understand how the retina is affected by metabolic disorders such as diabetes and obesity.

In sum, these impactful findings add another piece to the puzzle towards our understanding of how cell death, removal and recycling are fundamental processes controlling tissue metabolism and homeostasis. The process of cell death actively releases metabolites as ‘goodbye’ signals that alter neighboring cells’ metabolism and function¹³. In the heart, cardiomyocytes release their dysfunctional mitochondria, which are phagocytosed by resident macrophages via MerTK to maintain homeostasis¹⁴. Additionally, phagocytic engulfment of dying cells supplies abundant lipids, proteins, and carbohydrates that are metabolized to support further phagocyte metabolism and function. Phagocytosis induces the expression of various solute carriers to promote glucose uptake and to release lactate into the extracellular space¹⁵. Additionally, the bolus of lipids from a dying cell may undergo fatty acid oxidation to fuel the electron transport chain to polarize phagocytes towards a pro-reparative phenotype¹⁶. The RPE may even be unique among phagocytes in that RPE cells utilize fatty acids from digested POS for ketogenesis⁵. Adding to these recent findings, Iker Etchegaray et al.³ provide valuable insight into the impact of phagocytosis of damaged cells on tissue metabolism in the unique context of the vertebrate eye.