Expression of the Niacin Receptor GPR109A in Retina: More than Meets the Eye?

GPR109A was discovered recently as the G-protein coupled receptor for niacin (nicotinic acid), a drug used widely in the treatment of hyperlipidemia. Upon its initial discovery, expression of the receptor was thought to be restricted primarily to adipocytes and immune cells (monocytes/macrophages), a pattern of localization consistent with the known actions of niacin – anti-lipolytic and anti-atherogenic. Of late however, several new reports have arisen detailing expression of the receptor in other cell and tissue types. Interestingly, with the exception of dermal Langerhans cells, the cells responsible for skin flushing, an unwanted side effect of high-dose niacin therapy, the function of the receptor in the additional cell types described is largely anti-inflammatory in nature. The receptor might also have a role in cancer; silencing of the receptor has been reported in colon and breast cancers, and forced expression of the receptor in tumor cells induces apoptosis, thereby suggesting a tumor-suppressive role for the receptor. This supports strongly not only the critical importance of GPR109A expression and activity under normal, basal conditions, but also the strength in impact that therapies capable of augmenting or optimizing its expression and activation may have in thwarting the development and progression of inflammation and cancer. Given the key causative role of inflammation in diabetic retinopathy, and the critical lack of viable strategies for intervening early in this pathology, new therapies, particularly those targeting inflammation, are sorely needed. Herein, we describe preclinical and clinical studies documenting the expression of GPR109A, the pleiotropic effects elicited in response to its activation and the underlying mechanisms to explain these actions. This information we discuss in the context of its relevance to diabetic retina, ultimately providing insight into strategy for future targeting of the receptor and development of new therapies for prevention and treatment of retinopathy in diabetes. *Corresponding author: Pamela M Martin, Georgia Regents University, 1120 15th St., CN-1160, Augusta, GA 30912 USA, Tel: 706-721-4220; Fax: 706-7216608; E-mail: pmmartin@gru.edu Received April 04, 2013; Accepted April 27, 2013; Published April 29, 2013 Citation: Martin PM, Gambhir D, Promsote W, Ganapathy V, Moore-Hill D (2013) Expression of the Niacin Receptor GPR109A in Retina: More than Meets the Eye? Clin Exp Pharmacol S3:003. doi:10.4172/2161-1459.S3-003 Copyright: © 2013 Martin PM, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.


Introduction
Diabetic retinopathy (DR), one of the most common and feared complications of diabetes, is the leading cause of irreversible vision loss and blindness among adults of working-age in the U.S. and other industrialized countries [1,2]. The incidence of DR is high; nearly all type 1 patients and greater than 60 percent of patients with type 2 diabetes develop DR within the first 1-2 decades of their disease. In fact, at the time of diagnosis, signs of retinopathy are already detectable in many type 2 diabetic patients. Hyperglycemia is a primary factor in DR pathogenesis and while correcting it through maintenance of tight glycemic control (a task that is difficult in many patients) helps, it does not mitigate entirely DR development and progression [1,3]. There are at present treatments for DR (e.g., laser photocoagulation, vitrectomy, intravitreal anti-VEGF therapy); however in addition to being associated with adverse effects, these therapies are applicable only at the late stages of the disease when the signs and symptoms of proliferative microvascular disease present. But unfortunately, at these late stages, DR has already reached a relatively advanced state. Hence, there is a critical need for new, viable strategies for DR prevention and treatment.
Inflammation is implicated as a key causative factor both in the development and progression of DR [3][4][5][6]. Indeed, diabetes itself is recognized as a chronic, low-grade inflammatory disease. Congruent with this is a burgeoning literature suggesting strongly that therapies that reduce inflammation in retina may block early cellular and biochemical alterations in this tissue long before they become clinically evident. Thus, strategies to intervene within this "early" time frame have a high likelihood of effectively preventing or slowing progression of DR to advanced stage retinal disease and thereby, preserving vision and quality of life in a large number of persons. We identified recently in retina a new target, the G-protein coupled receptor GPR109A, that when activated therapeutically in diabetes could potentially fit this bill. In this short review, we discuss the clinical and experimental evidence that forms the basis of this rationale.

History of the Niacin Receptor, GPR109A
In 2003, GPR109A, also known as hydroxycarboxylic acid receptor 2 (HCA2) and formerly as HM74A in humans and PUMA-G (Protein Upregulated in MAcrophages by Interferon-γ) in rodents, was identified as the high-affinity receptor for niacin [7][8][9][10]. A second receptor with a high degree of homology to HM74A but a much lower affinity for niacin, HM74 (GPR109B), was also reported; however, its exact physiologic role remains unclear. Upon its initial discovery, GPR109A expression was thought to be limited primarily to adipocytes, the cell type in which the anti-lipolytic effects of niacin are most warranted. Niacin has been used widely to treat hyperlipidemia and remains today, five decades following its initial introduction, as one of the most effective agents for lowering low-density lipoprotein (LDL; "bad" cholesterol) and increasing High-Density Lipoprotein (HDL; "good" cholesterol). In contrast, newer therapies such as HMG-CoA reductase inhibitors (statins) and fibric acid derivatives are effective only in reducing the LDL cholesterol [11,12]. Congruent with its improvement of lipid/lipoprotein characteristics, niacin is touted also for its ability to reduce vascular inflammation and thrombosis and therefore lower significantly the risk of atherosclerosis, stroke and other cardiovascular events, properties that we now know to be a direct consequence of its interaction with and subsequent activation of the G i -protein coupled receptor GPR109A [13]. This entire highlighted sentence should be revised to read "The efficacy of niacin in slowing the progression of atherosclerosis and reducing cardiovascular events and mortality in patients with hyperlipidemia and the involvement of GPR109A in these actions is supported by a wealth of clinical and experimental evidence [14], though some recent studies have challenged these views.

Journal of Clinical and Experimental Pharmacology
Along with the identification of GPR109A as the receptor for niacin, came additional studies devoted to characterizing in detail the expression and function of the receptor in various cell and tissue types. Out of these studies came the identification of β-hydroxybutyrate and butyrate as endogenous receptor agonists [15,16], and importantly, the revelation that the benefits of activating the receptor therapeutically extend far beyond the alteration of lipid profiles [17,18]. In immune cells, cells central also to the development and progression of atherosclerotic lesions/cardiovascular disease, GPR109A activation was found to elicit robust anti-inflammatory responses by way of suppressed pro-inflammatory cytokine and chemokine expression and signaling [19]. The same anti-inflammatory effect was described also in adipocytes [20]. Additional studies involving mouse and human endothelial cells demonstrated the potentiation of responses that were both antioxidant and anti-inflammatory in nature in association with GPR109A activation [11,21,22]. Investigations into the role of the receptor under normal physiologic conditions and the consequences of its activation by endogenous ligands (e.g., β-hydroxybutyrate, butyrate), which happen to be intermediates of normal metabolism, led to the discovery of a role for the receptor also in nutrient sensing and energy regulation, and in tumor suppression [17,23,24].

GPR109A Expression in Retina and Rationale for Therapeutic Targeting in Diabetes
As detailed above, the major cellular properties attributed to GPR109A-dependent signaling to date can be described in sum as: lipid-modulatory, anti-inflammatory, anti-oxidative, and energy regulating. The underlying molecular mechanisms to explain these actions include regulation of the activity of lipolytic enzymes (i.e., hormone-sensitive lipase in adipocytes) and cholesterol efflux transport proteins [7][8][9][10]15,25]; suppression of Nuclear Factor Kappa-B (NF-kB) signaling [16]; Peroxisome Proliferator-Activated Receptor-Gamma (PPAR-γ) activation [26]; and regulation of monocarboxylate transporter expression and functional activity [24]. Alone, each of the aforementioned responses is highly desirable in diabetic retina; however, the discovery of a single receptor with the potential to influence them collectively is extremely significant with respect to developing new, targeted therapies for DR prevention and treatment.
The involvement of inflammation and oxidative stress in DR pathogenesis is undeniable. While there are conflicting opinions as to the existence of a concrete link between atherosclerosis/cardiovascular disease, the pathology for which therapeutic intervention via GPR109A activation is most noted, and DR incidence and severity [27], clinical and experimental evidences support overwhelmingly a definitive correlation between the two [28,29]. Hence, enthusiasm in this regard is in no way diminished. Indeed, altered lipid metabolism and lipoprotein abnormalities are not only well established features of diabetes but more specifically, of diabetic retina. Hypertriglyceridemia is common in type 1 diabetes and even more so in type 2 diabetes, with type 2 patients commonly presenting with reduced HDL-choleserol levels in addition to hypertriglyceridemia [30]. In retina specifically, the size and density of hard exudates in diabetic patients is thought to be reflective directly of the severity of alteration in retinal lipid profiles [31,32]. Whether patients under niacin therapy suffer less by diabetic retinopathy has not been evaluated specifically. However, fibrates and statins, well-known anti-hyperlipidemic agents, have been shown to reduce significantly the risk of advanced DR development and progression both in human patients and experimental models of disease [31][32][33][34][35]. There are an increasing number of new reports citing the benefits of niacin-GPR109A signaling in the treatment of pathologies other than hyperlipidemia (e.g., multiple sclerosis, septic shock, lung inflammation and fibrosis, renal failure, ischemia-induced neovascularization and peripheral vascular dysfunction in diabetes) [36][37][38][39][40]. This coupled with the fact that niacin, the prototypic GPR109A agonist, elicits additional, unique actions atop its lipid-lowering effects via its interaction with GPR109A raises the question as to whether similar or added benefit might be attained upon therapeutic targeting of the receptor in DR. This is a question well worthy of investigation. However, before progress can be made toward addressing this issue, the expression of GPR109A in retinal cell types and intact retina must be characterized fully and its functional significance therein elucidated.
Along these lines, in 2009, we reported for the first time ever, expression of GPR109A in retina [41]. Evaluation of cross sections of mouse retina and transformed human and rodent retinal cell lines revealed the localization of the receptor in Retinal Pigmented Epithelium (RPE) and more specifically, to the basolateral membrane of this cellular layer. This localization is ideally suited for interaction of the receptor with ligands, endogenous or pharmacologic, presenting from the choroidal circulation. Our subsequent inquiry into the functional significance of RPE-specific GPR109A expression led to the discovery of a potent role for the receptor upon its activation by exogenous niacin or β-hydroxybutyrate in the mediation of antiinflammatory signaling not only in cultured RPE cells [42], but also in intact mouse retina as evidenced by the suppressed expression (mRNA and protein) of pro-inflammatory molecules such as Interleukin-1β (IL-1β), Monocyte Chemotactic Protein-1 (MCP-1), Intercellular Adhesion Molecule-1 (ICAM-1) and, the reduction of other notable hallmarks of retinal inflammation such as increased leukostasis (unpublished results). Further investigation into the effects of diabetes on receptor expression led to the revelation that not only does retinal expression of GPR109A persist in diabetes (both type 1 and type 2 diabetes -determined by analysis of receptor expression in postmortem human retina and in rodent models of diabetes) but, that it appears to be upregulated in this condition [42]. While this confirms the presence and availability of the receptor for targeting in diabetic retina, the functional significance of its upregulation remains to be determined. We postulate that it represents a mechanism by which the tissue tries to stop the progression of the damage caused by the disease, one precluded by the absence of an endogenous agonist present at levels sufficient to augment maximally the beneficial effects of GPR109A signaling. In healthy individuals, β-hydroxybutyrate, a principal ketone and physiologic GPR109A agonist, is present in the circulation, but at a relatively low concentration, increasing only slightly during periods of fasting or prolonged exercise. Ketone bodies serve as alternate sources of energy to maintain crucial organs such as the brain, retina, heart, kidney cortex and skeletal muscle when carbohydrates are either in short supply or as is the case in diabetes, cannot be used effectively [43]. Hence, in well-controlled diabetes, significant increases in circulating β-hydroxybutyrate do not occur. Though seemingly protective in moderation, when elevated excessively, as in uncontrolled diabetes, the effects of β-hydroxybutyrate are thought to be largely detrimental (e.g., ketoacidosis). This suggests that despite the presence of both the receptor and the agonist in such conditions, activation of the receptor may not occur at the optimal level and hence the beneficial effects of receptor signaling are lost. It is also plausible, particularly at such high concentrations, that additional, GPR109A-independent mechanisms come in to play. Thus, if receptor activation can be augmented optimally through the use of pharmacological agonists, robust benefits may be reaped in diabetic retina. Support for this line of thought and therefore the therapeutic targeting of GPR109A in diabetic retina can be inferred from the recent report by Poplawski et al. [44] demonstrating pronounced and effective reversal of parameters associated with progression and increased severity of diabetic nephropathy, another major complication of diabetes, through employment of a ketogenic diet, a diet in which β-hydroxybutyrate is the predominant ketone. Additional support can be gathered from a study by Huang et al. [40] in which the authors report niacin-induced improvement of blood flow and peripheral vascular function in a diabetic rodent model of ischemia-reperfusion through mechanisms involving the modification of nitric oxide bioavailability and reduction of oxidative stress, and from clinical and experimental studies of neurodegenerative brain diseases touting the cellular protective effects of β-hydroxybutyrate when employed therapeutically [45].
Inflammation and reactive oxygen species production are normal physiologic processes. Inflammation is essential to cellular defense, repair and turnover and reactive oxygen species are produced normally as a consequence or byproduct of many normal biologic processes. Therefore, cells, including those of the retina, are naturally equipped to deal with these factors, as they must on a regular basis, in order to maintain tissue homeostasis. In pathologic conditions such as diabetes and DR however, the increased production pro-inflammatory and pro-oxidant factors coupled with the decrease in the cellular defense machinery, influences negatively the delicate balance between proinflammatory/pro-oxidant and anti-inflammatory/antioxidant factors; this has been documented and described on numerous occasions [4,5,46,47]. Therefore, could it be that when elevated in moderation, β-hydroxybutyrate activates GPR109A expressed by neurons, RPE, and other cell types, thereby conferring protection against inflammation, oxidative stress and other types of cellular insult but, in extreme excess (i.e., ketoacidosis) it tips the scale in the opposite direction, favoring increased inflammation, oxidative stress and cellular damage? The likelihood of an affirmative answer to this question is strengthened by the fact that niacin, when taken at lower doses (< 1 g daily) has proven to be quite effective at improving lipid profiles and reducing the risk of atherosclerosis and cardiovascular disease development and progression in a large number of patients; however, when given at very high doses (3.0 -4.5 g daily), use of the compound, while still effective in the aforementioned parameters, is associated with three major side effects [11]. The first is flushing, an effect that we now know to be due to the unwanted effects of this compound/GPR109A activation in dermal Langerhans and keratinocytes [48]. The second and third, respectively, are a reversible form of cystoid maculopathy (occurs only in 0.67% of patients), and modest hyperglycemia [11,49]. These undesirable effects have fueled the search for lipid-lowering therapies in lieu of niacin. However, of the alternative therapies discovered and employed, the fact remains that none have proven to be as effective as niacin at raising HDL levels and improving additional parameters consistent with reduced risk of advanced cardiovascular disease; for this primary reason, the clinical use of niacin has continually been revisited. Indeed, the recent introduction of immediate-release and extended-release forms of the niacin brings the compound back to the forefront as the incidence and severity of the adverse effects associated with its use has been reduced significantly even when the agent is employed at very high doses while retaining the benefits attributed to the original compound [11]. Of specific interest and relevance to our present topic, the potential employment of GPR109A agonists for therapeutic management of DR, is the fact that the incidence of ocular side effects (niacin maculopathy), which was already relatively low, appears to have diminished even further as evidenced by the few, rare reports that can be found in the published literature over the past decade. The same is true with respect to niacin-induced deterioration of glycemic control in diabetic patients, which is now reported as being only minor with no evidence of increased incidence of new onset diabetes [11,50]. Furthermore, the avoidance or lessening of the flushing response is important, as it may impact positively patient compliance rates. The discovery by Walters et al. [51] that niacin-induced signaling through beta-arrestin 1 is responsible for flushing but not for the anti-lipolytic effects of the compound suggests that relevant compounds that activate GPR109A in a beta-arrestin-independent fashion could be developed. Hence, targeting GPR109A for therapeutic management of DR, whether by the use of niacin, other newly identified pharmacologic ligands like monomethylfumarate [49,52] or alternately, intake of a ketogenic diet to increase the endogenous agonist β-hydroxybutyrate, remains as a feasible and seemingly viable option that should be explored further. The appeal for such is heightened by the fact that it might be accomplished successfully via the re-purposing of existing compounds or strategies that have already garnered FDA-approval for treatment of other indications and therefore much is known regarding the toxicological and pharmacokinetic properties in humans.
The demonstration of GPR109A expression in RPE and its function there in modulating inflammation is a novel and important finding. RPE facilitates numerous functions that are critically essential to retinal health and visual function normally and in diabetes including the regulation of immunity and inflammation in the outer retina via the expression and secretion of a plethora of cytokines and chemokines, maintenance of outer-blood retinal barrier integrity, phagocytosis of shed photoreceptor discs, cholesterol transport/lipid homeostasis, light absorption, amino acid and nutrient transport, etc. [53,54]. However, it is clear that in DR other retinal cell types are largely affected [55]. There are a number of studies suggesting strongly that processes similar to those that promote and/or potentiate atherosclerogenesis occur also in diabetic retina [56,57]. A role for inflammation and oxidative stress in DR has been established. Atop this are clinical studies demonstrating retinal dysfunction and cellular damage in association with lipid accumulation in this tissue [31,32,58] and experimental studies demonstrating the critical causative role of dyslipidemia in endothelial cell dysfunction and vascular inflammation in retina [59]. Could GPR109A in retina be expressed by retinal cell types in addition to RPE, and upon activation therein contribute to the propagation of multiple, unique responses that are, like the demonstrated actions associated with GPR109A expressed in other organs, overwhelmingly beneficial (i.e., improved retinal lipid profiles, reduced inflammation and oxidative stress, protected and preserved retinal cell viability and vascular integrity/function)? Along these lines, our group has initiated additional studies. Using primary rat microglial cells isolated following  [60], a generous gift from the laboratory of Dr. Gregory Liou (Georgia Regents University, Augusta, GA), and our rabbit polyclonal GPR109A antibody [41], we evaluated GPR109A expression by Western blotting techniques per our published method [61] (Figure 1). Protein extracted from primary RPE cells that were isolated from normal mouse retinas as described previously [41] was included as a positive control. Microglia are in essence the resident macrophage cells of the retina; their critical involvement in potentiating inflammation and therefore the development and progression of diabetic retinopathy is well documented [62][63][64]. Our studies have shown that microglia express GPR109A robustly. Though novel, the finding of GPR109A expression in these cells did not come as a surprise given the original description of the receptor protein as PUMA-G (protein-upregulated in macrophages by interferon-γ) and the well established role of these cells in the mediation of inflammatory processes in retina. However, additional studies to ascertain fully the functional significance of GPR109A expression and activation in this retinal cell type normally and in diabetes are needed. Similar studies are also needed with retinal endothelial cells; our preliminary work in this regard leads us to believe that the consequences of GPR109A activation in these cells will be, as described for peripheral vascular endothelial cells, largely anti-inflammatory.

Conclusions
Based on existing clinical and experimental evidence and our own published and preliminary findings, we speculate that GPR109A expression in retina is essential to the normal regulation of multiple parameters, both lipid-dependent and lipid-independent (e.g., inflammation, the metabolism and handling of lipids, nutrient sensing and energy regulation). Hence, therapies to target this receptor in diabetes may be effective in preventing or slowing the development and progression of DR ( Figure 2). Additionally, there is the promise that such strategies could be of benefit also in the prevention and treatment of other debilitating retinal diseases in which inflammation, oxidative stress and/or altered lipid regulation are majorly involved such as for example, age-related macular degeneration.

RPE Microglia
GPR109A β-Actin  Hyperglycemia and dyslipidemia are major features of diabetes and are known to potentiate factors that contribute to the increased oxidative stress and inflammation characteristic of the disease. Of those factors, clinical and experimental studies show upregulated Nuclear Factor-kappa B (NF-kB) and reduced Peroxisome Proliferator-Activated Receptor-Gamma (PPAR-γ) activity to be crucial. Studies conducted using non-retinal cells and tissues show GPR109A to be effective at reversing and/or inhibiting these parameters upon its activation by specific agonists. Additionally, GPR109A may elicit other direct effects congruent with reduced inflammation and oxidative stress. In retina, GPR109A is expressed in cells critical to the regulation of tissue immunity and inflammation (e.g., Retinal Pigment Epithelial (RPE) and microglia) and similarly suppresses inflammation and oxidative stress when activated. Thus, therapies to augment this phenomenon in diabetic retina may prevent or delay the development and progression of retinopathy.