Role of pericytes in the retina

Diabetic retinopathy is a major severe ocular complication associated with the metabolic disorder of diabetes mellitus.¹ The lack of a detailed knowledge about the cellular and molecular mechanisms involved in diabetic retinopathy restricts the design of effective treatments. Understanding the roles of retinal cells during this process is of utmost importance, since gaining control of specific cell populations may allow us to arrest or even induce reversion of diabetic retinopathy.

Pericyte dropout or loss has been suggested to have great consequences on blood vessel remodeling, and possibly causes the first abnormalities of the diabetic eye which can be observed clinically in diabetic retinopathy.² Nevertheless, a concreate evidence to support this concept is not available. Surprisingly, in a recent article in Nature Communications, Park and colleagues demonstrated that pericytes are not essential in the adult stable retinal blood vessels; and their selective depletion did not lead to a phenotype similar to diabetic retinopathy.³ The authors used a transgenic mouse model which can be used to specifically ablate PDGFRβ-expressing pericytes (PDGFRβ-CreER/DTA mice). Several studies suggest that PDGFB released from vascular endothelial cells recruits PDGFRβ-expressing pericytes to facilitate vascular stabilization during blood vessel development.⁴ Nonetheless, whether this PDGFB/PDGFRβ signaling continues to be necessary for proper pericyte attachment to stable adult retinal vasculature was unknown. Park and colleagues used VE-Cadherin (Endothelial specific)-CreER/PDGFB floxed mice and intra-vitreal administration of PDGFRβ blocking antibody to show that PDGFB/PDGFRβ signaling is not required for the maintenance of the interaction between pericytes and endothelial cells, and for the integrity of the blood-retinal-barrier in adults.³

In contrast, Park and colleagues demonstrated using state-of-the-art techniques, including deletion of several genes from endothelial cells, that PDGFB/PDGFRβ signaling is indispensable in the formation and maturation of blood-retinal-barrier at the postnatal stage through active recruitment of pericytes onto the growing retinal vessels.³ Additionally, the authors revealed that pericytes are important in the adult retina as regulators, as they control the expression of several genes (FOXO1, Ang2, and VEGFR2) to protect retinal vessels against injuries and stresses.³

Here, we discuss the findings from this work, and evaluate recent advances in our understanding of pericytes roles in the retina.

Perspectives/future directions

The findings from this study are based on the expression of PDGFRβ in pericytes. Several other pericytic markers have been characterized, such as nerve/glial antigen 2 (NG2) proteoglycan (CSPG4),⁵ aminopeptidase N (CD13),⁶ alpha smooth muscle actin (αSMA),⁷ ATP-sensitive potassium-channel Kir6.1,⁸ Glutamate aspartate transporter GLAST,⁹ desmin,⁹ leptin receptor,¹⁰ Nestin,¹¹ and many others. ¹² However, there is no specific molecular marker that can be used to unmistakably identify pericytes. For instance, retinal microglia express NG2 and Nestin.¹³ Thus, it is possible that in the retina other cell populations are confused with pericytes. Additionally, pericytes that do not express NG2 proteoglycan were recently described.¹⁴ Currently, a state-of-the-art identification of pericytes in tissue preparations relies on a combination of anatomical localization (covering endothelial cells), morphology, and at least two pericytic molecular markers. Additionally, in adult mice, PDGFRβ expression is not restricted to pericytes. Several stromal cells, such as fibroblasts,¹⁵ and vascular smooth muscle cells^{16, 17} express this cell-surface tyrosine kinase receptor.¹⁸ PDGFB plays an important role in the proliferation and differentiation of aortic and venous vascular smooth muscle cells.^{16, 17} Also, not all cells in perivascular position are pericytes. In addition to pericytes, other cellular types have been described as perivascular: for instance, macrophages,¹⁹ microglia,²⁰ fibroblasts,²¹ adventitial cells,²² and vascular smooth muscle cells.²³ Moreover, in earlier stages, PDGFRβ is broadly expressed throughout the animal in multiple cellular lineages.²⁴ Thus, PDGFB/PDGFRβ signaling may occur between other cell populations as well, not necessarily pericytes. Altogether this highlights the possibility that some of the observations by Park et al.³ are due to a different, non-pericytic, cell population.

Curiously, Park and colleagues showed that after ablation of pericytes in the adult retina using PDGFRβ-CreER/DTA mice, no changes in vascular leakage were detected in the retinas and brain, while profound leakage was found in several other peripheral organs.³ This may be due to the strong attachment between endothelial cells in the central nervous system in comparison with other parts of the body.²⁵ These results also imply that the classical role of pericytes in vascular stability in adult blood vessels may be restricted to specific organs, but not others.

The population of pericytes has been uncovered to not be homogeneous whitin several organs. Interestingly, the functions of different pericyte subpopulations may vary. Specific pericyte subsets with distinct roles in myofiber regeneration,²⁶ blood vessel formation,²⁷ fat accumulation,²⁸ and fibrous tissue deposition²⁹ in the skeletal muscle have been described.¹¹ In the bone marrow and skin, NG2⁺ pericytes are distinct from NG2^- pericytes in their role for hematopoietic stem cells maintenance,³⁰ and in their interaction with innate immune system cells,¹⁴ respectively. In several other organs, only a subpopulation of pericytes gives rise to myofibroblasts in organ-fibrosis.³¹ Whether PDGFRβ⁺ pericytes are heterogeneous in the retina remains unknown. As mentioned above, it is possible that PDGFRβ^- pericytes may reside in the retina, and what is their function remains completely unknown.

In the mouse models that were analyzed in this study, the influence of PDGFB/PDGFRβ signaling from other tissues in the retina was neglected. Is it possible that the observations in the retinas of those transgenic mice come from an indirect effect from a phenomenon that occurred in another tissue? Future studies, using elegant techniques, including sophisticated whole-eye transplant experiments from transgenic mice into normal hosts,³² may address this issue. Thus, future studies should allow specific genetic interventions into the retina without affecting other tissues.

Interestingly, in a recent study, it was revealed that pericytes play a key role in the regulation of vascular tone and blood flow in the hypoxic spinal cord after injury.³³ Strikingly, blocking pericytes-specific enzymes diminished hypoxia, and improved motor function and locomotion of the spinal cord-injured animals.³³ As diabetic retinopathy can develop from ischemia-induced retinal hypoxia;³⁴ hypothetically, blocking pericytes would ameliorate this ischemic condition if retinal pericyte have a similar role to medullar pericytes. Therefore, future studies should focus on investigating whether targeting retinal pericytes in a mouse model of diabetic retinopathy will alleviate hypoxia.

Cell-based therapies have been proposed as an option for both preventing neurovascular damage, and promoting regeneration of damaged cells in the retina in diabetic retinopathy.³⁵ However, it is naive to think that injecting exogenous cells can work better than activating endogenous stem cells to work properly. The problem is that we do not know yet how exactly endogenous retinal cells work and how to regulate their functions. Understanding the interplay between various cellular components of the retinal microenvironment, including endogenous stem cells, will be important to develop methods to regulate and activate endogenous stem cells in the retina. Pericytes role as stem cells contributing to formation of other cell types has been reported in numerous publications in several tissues. For instance, pericytes have neural progenitors’ activity in response to ischenmia in vivo in the central nervous system.³⁶ However, whether endogenous retinal pericytes have the plasticity to form other cellular populations remains unknown. If yes, why their efficiency is not enough to revert the disease state in diabetic retinopathy?

In conclusion, the study by Park and colleagues reveals important roles of pericytes during development and in the adult retina microenvironment. However, our understanding of the importance of pericytes in the retina remains limited, and the complexity of the retinal cellular microenvironment should be elucidated in future studies. The best is to come. A big challenge for the future will be to translate mice research into humans. Improving the availability of human tissue samples will be essential to reach this goal (Figure 1).

Figure 1

Retinal vasculature in diabetic retinopathy. Pericytes are present around blood vessels in the normal retina. Pericyte dropout is one of the major hallmarks of diabetic retinopathy. Park and colleagues now suggest that pericytes are essential in the formation and maturation of blood-retinal-barrier at the postnatal stage through active recruitment of pericytes onto the growing retinal vessels.³ Nevertheless, pericytes are not indispensable in the adult stable retinal blood vessels; and their selective depletion did not lead to a phenotype similar to diabetic retinopathy. Future studies may reveal other role of retinal pericytes in much greater detail.