New insights into the role of connexins in pancreatic islet function and diabetes

Abstract

Multi-cellular systems require complex signaling mechanisms for proper tissue function, to mediate signaling between cells in close proximity and at distances. This holds true for the islets of Langerhans, which are multicellular micro-organs located in the pancreas responsible for glycemic control, through secretion of insulin and other hormones. Coupling of electrical and metabolic signaling between islet β-cells is required for proper insulin secretion and effective glycemic control. β-cell specific coupling is established through gap junctions composed of connexin36, which results in coordinated insulin release across the islet. Islet connexins have been implicated in both Type-1 and Type-2 diabetes; however a clear link remains to be determined. The goal of this review is to discuss recent discoveries regarding the role of connexins in regulating insulin secretion, the regulation of connexins within the islet, and recent studies which support a role for connexins in diabetes. Further studies which investigate the regulation of connexins in the islet and their role in diabetes may lead to novel diabetes therapies which regulate islet function and β-cell survival through modulation of gap junction coupling.

A. Introduction

The pancreatic islets of Langerhans are responsible for blood glucose homeostasis through the regulated secretion of insulin and other hormones. In β-cells within the islet, membrane depolarization and calcium influx drives the coordinated release of insulin. Gap junctions are intercellular channels, composed of two connexon hemi-channels each made up of six connexin subunits, which readily transfer ions, metabolites and other small molecules between cells. For a more in depth review on connexins in general and their role in the endocrine system, we refer readers to the review article by Bosco et al. [1]. In the islet, gap junctions composed of connexin36 (Cx36, also known as Gjd2) provide electrical and metabolic coupling between β-cells which regulates electrical activity and insulin secretion. Under high glucose, gap junctions facilitate the coordination of electrical activity across the islet which leads to synchronized release of insulin from individual β-cells. Under low glucose, gap junctions repress electrical activity which contributes to the inhibition in the release of insulin. While the role of Cx36 gap junctions is becoming increasingly established in the islet, a link between gap junction coupling and diabetes is also beginning to emerge. Recent studies in mouse models have shown a correlation between decreased gap junction coupling, disruptions to glucose homeostasis and altered islet function; similar to that observed in models of diabetes. This suggests that gap junction coupling may be one underlying factor, of many, which contributes to disease development. As such, this review will discuss the specific role of Cx36 gap junctions in islet function and insulin secretion, the known pathways for regulation of Cx36, and current evidence for the role of Cx36 in Type 1 and Type 2 diabetes. We also note that discussion of a number of roles Cx36 plays in the islet can also be found in reviews by Meda and colleagues [2, 3] [4] [5] or by Perez-Armendariz [6]. Current evidence suggests that regulation of insulin secretion from the islet can be modulated through gap junction coupling. Understanding the regulation of Cx36 gap junctions and the role they play in diabetes has the potential to uncover new therapies which restore more physiological blood glucose control and which could potentially deter the onset of disease.

B. The role of connexins in the islet

1. Connexins and gap junctions

In the human genome, 22 connexin species have been identified, while only 19 have been identified in the mouse [7, 8]. Connexin species are identified by the molecular weight of the protein, where for example Cx36 has a molecular weight of 36 kDa [9]. Most connexins share a very similar gene structure and therefore a similar protein structure, where all connexins are composed of 4 membrane spanning domains connected by 2 extracellular and 1 intracellular loop [8]. The N-terminal, 2 extracellular loops and 4 transmembrane regions are highly conserved between different connexins, while the intracellular loop and C-terminal are highly variable [10]. Both the N- and C-terminal are located in the cytoplasm, positioning them for post-translational modifications by intracellular kinases and phosphatases [11], and for interactions with connexin-associated proteins [12].

Connexin proteins oligomerize around a central hydrophilic space into a 6-protein hemi-channel called a connexon during trafficking from the ER to the Golgi complex [13]. Connexons may be homomeric, composed of only one type of connexin protein, or heteromeric, composed of different species of connexins [14]. Connexons are then trafficked to the plasma membrane via Golgi-derived vesicles mediated through microtubules or the actin cytoskeleton [15]. Connexons are inserted into a lipid raft in the membrane [13], where they can rapidly diffuse and aggregate at tight junctions or other non-junctional areas of the membrane. At tight junctions, adhesion molecules such as zonula occludens-1 on the surface of neighboring cells have decreased the gap between cells to ~2nm [16, 17]. At this point, connexon hemi-channels may remain uncoupled [18], or they may interact with a connexon from a neighboring cell via extracellular loops and form an intercellular channel. In many cell types uncoupled hemi-channels can play a role in cell signaling. However, no known role for uncoupled Cx36 hemi-channels has been established in the islet to date and recent studies suggest that β-cells do not have functional uncoupled hemi-channels [19].

Connexons which have formed intercellular channels aggregate in the membrane and form plaques, termed gap junctions [20]. Gap junctions provide electrical and metabolic coupling between cells [21]. Electrochemical gradients drive the rapid diffusion of ions, metabolites, nucleotides, small peptides and other small molecules [16, 22], where the selectivity and conductance of gap junctions to these molecules is largely determined by the species of connexin that comprises the hemi-channels [23]. Although gap junctions are made up of many intercellular channels, the individual connexin channels are only open about 10% of the time [24]. Gating of connexin channels can be controlled through environmental factors, such as electric current and voltage across the channel, cytosolic pH, or changes in intracellular free-calcium activity ([Ca²⁺]_i) [25, 26]; as well as through post-translational phosphorylation [11]. Connexin channels are also very short lived, with a half-life of ~4 hours. Gap junctions are internalized from the center of a plaque where the hemi-channel and connexin protein are degraded, while new hemi-channels are inserted at the periphery of the plaque [13, 14]. While gap junctions play a role in cell signaling through electrical and metabolic coupling, recent evidence suggests that connexin hemi-channels may also effect changes in gene expression and cell proliferation through interactions with other proteins or through insertion into non-junctional regions of the plasma membrane [27].

2. Connexin36 gap junctions

Insulin-secreting β-cells within the pancreatic islet are exclusively coupled by Cx36 gap junctions in mice, and strongly coupled by Cx36 gap junctions in humans [28-30]. To date, Cx36 has been found to be highly expressed in the brain [31], the adrenal medulla [32], the retina [33], in the murine carotid body, ileum, and colon [34], as well as the pancreatic islet [30]. While Cx36 is expressed by insulin producing β-cells in the islet, it has also been reported to be expressed in non β-cells [30]. Data in studies examining electrical activity within the islet suggest that coupling may occur between α-cells and β-cells based on synchronous oscillations being observed [35]. However, direct evidence of functional gap junctions has only been found between β-cells [36]. Also recent studies from our lab have demonstrated that no functional gap junctions exist between α- and β-cells [submitted]. Cx36 gap junction coupling has been found to be very heterogeneous in mouse islets [28, 36]. Cx36 gap junctions studied in β-cells are distinct from gap junctions composed of other connexin species in that they preferentially exchange cationic molecules [37], are minimally voltage sensitive with a half-activation voltage of only ±85mV [38], and have a small unitary conductance (~6pS) [36]. Compared to gap junctions formed by other species of connexins, Cx36 gap junction only have an open probability of ~0.8% [39].

C. Connexin36 regulation of insulin secretion

1. Role of connexin36 under high glucose

In β-cells, a series of metabolic and electrical events regulates insulin secretion in response to elevations in blood glucose. Glucose uptake and metabolism leads to a generation of ATP, closure of ATP-sensitive potassium channels (K_ATP), subsequent membrane depolarization, activation of voltage gated calcium channels and an elevation of [Ca²⁺]_i, as outlined in Figure 1. This [Ca²⁺]_i elevation triggers insulin granule exocytosis and release from individual β-cells. The dynamics of insulin release are tightly coupled with electrical activity and therefore calcium signaling across the islet. Glucose driven oscillations in electrical activity modulated by K_ATP channel gating leads to oscillations in [Ca²⁺]_i [40]. Glucose stimulated insulin secretion is biphasic, with a burst release in the 1^st phase and a pulsatile release in the 2^nd phase [41]. While the mechanism underlying biphasic release of insulin are under debate, one proposed mechanism is that different phases originate from two separate pools of insulin granules; one population which is docked to the plasma membrane and is readily available for insulin release, and one population of reserve granules which must be trafficked to the cell membrane for insulin release. The initial pulse of glucose stimulated [Ca²⁺]_i stimulates release of insulin from docked granules, resulting in the initial increase of insulin characterized as 1^st phase release. The subsequent oscillations in [Ca²⁺]_i trigger exocytosis of insulin granules in a pulsatile manner, characterized as the 2^nd phase insulin release [41]. Another theory is that the biphasic nature arises from two subpopulations of β-cells which exclusively respond to during either the first or second of insulin secretion [42]. However, the action of Cx36 gap junctions in coordinating β-cell [Ca²⁺]_i and insulin responses across the islet argues against this.

Figure 1.

Role of Cx36 in regulating electrical activity and insulin secretion under high glucose conditions. (A) Under high glucose, gap junctions synchronize membrane depolarization (V_m), [Ca²⁺]_i, and insulin secretion, by promoting a depolarizing current (I_Ca), including diffusion of Ca²⁺. (B) Coordinated [Ca²⁺]_i oscillations in cells of a wild-type islet with normal gap junction coupling (Cx36^+/+, upper panel) generating overall pulsatile [Ca²⁺]_i across the islet. In cells of an islet lacking gap junction coupling (Cx36^-/-, lower panel) there is a lack of any synchronization between heterogeneous and irregular [Ca²⁺]_i oscillations, leading to a lack of overall pulsatile [Ca²⁺]_i across the islet. Data in B reproduced from [22].

Under stimulatory levels of glucose, electrical coupling provided by Cx36 gap junctions allows for transfer of a depolarizing current, synchronization of K_ATP channel-regulated membrane depolarization, and the coordination [Ca²⁺]_i oscillations across the islet [28, 29, 43], as depicted in Figure 1. As such insulin secretion is released in a coordinated and oscillatory fashion. This has been demonstrated in isolated islets from mice with knockout of the Cx36 gene, Gjd2. In one study, islets from Cx36 deficient mice showed a loss of glucose-stimulated [Ca²⁺]_i oscillations and a loss of pulsatile insulin secretion [29]. Further studies in Cx36 deficient isolated islets have shown the loss of [Ca²⁺]_i oscillations arises from a loss of synchronization between β-cells, where individual β-cells still show irregular, heterogeneous oscillations that lack any synchronization [28]. This is similar to the irregular and heterogeneous oscillations observed in isolated β-cells, dissociated from the islet [44] and is consistent with Cx36 being the sole connexin that forms gap junctions to coordinate electrical activity between β-cells. Further, islets from Cx36 deficient mice show an initial first phase release with reduced amplitude occurring over a longer duration, which is consistent with the initial [Ca²⁺]_i elevation lacking coordination among β-cells. Cx36 gap junctions also likely coordinate [Ca²⁺]_i oscillations in human islets [45]. K_ATP channel-independent mechanisms of insulin release have been identified, often collectively termed as ‘amplifying’ pathways [46]. One such amplifying pathway involves cyclic adenosine monophosphate (cAMP) which regulates insulin granule trafficking to the plasma membrane. Oscillations in cAMP have been observed which contributes to pulsatile insulin release [47], and these are coordinated across β-cells within the islet [48], however it is unknown whether gap junctions or an alternative mechanism synchronizes these oscillations in cAMP, as well as the dynamics of other amplifying pathways.

Insulin is also released in vivo in a biphasic fashion, which is linked to the dynamics of insulin release from isolated islets. For example the pattern of insulin oscillations in a mouse correlates strongly with the pattern of calcium oscillations in islets isolated from that mouse [49]. In humans, it is well established that the first phase of insulin secretion is important for insulin action. The pulsatile nature of the second phase of insulin secretion in vivo has also been linked to enhanced insulin action [50], particularly in terms of enhancing hepatic insulin signaling [51]. In vivo studies in Cx36 deficient mice reveal that Cx36 is important for glucose homeostasis, as deletion of Cx36 leads to significant glucose intolerance [52]. Mice retain normal fasting glucose levels and are therefore not overtly diabetic, but have a ‘prediabetic’ phenotype. Importantly a deletion of Cx36 significantly impacts the dynamics of in vivo insulin secretion [52]. Specifically, a decrease in 1^st phase insulin secretion and decrease in pulse size (but not frequency) in the 2^nd phase of insulin secretion is observed, consistent with measurements in isolated islets. However mice retained normal plasma insulin levels and insulin sensitivity. The glucose intolerance that follows a loss of Cx36 therefore likely results from the disrupted insulin dynamics, which is consistent with the importance of 1^st phase insulin for insulin action as discussed above, as well as pulsatile insulin. Interestingly, such disruptions to insulin secretion dynamics are also observed in Type 2 diabetes and pre-diabetes [53-55], suggesting a potential role for Cx36 in islet dysfunction underlying diabetes [56]. Further discussion on this topic can be found in the section on Type 2 diabetes below.

2. Role of connexin36 under basal glucose

At low (basal level) glucose, Cx36 gap junctions coordinate K_ATP-driven hyperpolarization across β-cells within the islet [57], as depicted in Figure 2. It has long been known that isolated β-cells are highly heterogeneous in the regulation of insulin release at high and low glucose levels. As mentioned above, the electrical activity of individual β-cells or β-cells in islets lacking Cx36 has been found to be heterogeneous, with spontaneous bursts of [Ca²⁺]_i occurring at basal glucose [44, 58]. While the source of this heterogeneity is unknown, Cx36 gap junctions have been found to suppress this spontaneous [Ca²⁺]_i at basal glucose [29, 59]. It has been proposed that Cx36 channels mediate the transfer of hyperpolarizing currents between β-cells, and therefore allow less active cells to prevent membrane depolarization and suppress more active cells [57-59]. For example, studies in isolated islets discovered that inhibiting K_ATP channels in the majority of β-cells in an islet did not result in elevated [Ca²⁺]_i at low glucose. Those neighboring cells with normal K_ATP channel regulation, which are coupled through Cx36 gap junctions, served to suppress the electrical activity in excitable cells that had inhibited K_ATP channels [57]. Given the extensive heterogeneity present in islets, this mechanism involving Cx36 closely regulates insulin secretion under basal glucose levels by overcoming β-cell heterogeneity to limit the range of glucose levels that stimulate insulin secretion.

Figure 2.

Role of Cx36 in regulating electrical activity and insulin secretion under low glucose conditions. (A) Gap junctions mediate a suppression of electrically active cells with reduced K_ATP activity through a hyperpolarizing current (I_K), and thus suppress [Ca²⁺]_i and insulin secretion. (B) Fraction of cells exhibiting dynamic changes in [Ca²⁺]_i as a function of glucose concentration in wild-type islets with normal gap junction coupling (Cx36^+/+) and islets lacking gap junction coupling (Cx36^-/-), showing a right-shift and sharpening of the dose-response in the presence of Cx36. (C) [Ca²⁺]_i in cells of an islet at 5mM glucose showing largely silent behavior in wild-type islets with normal gap junction coupling (Cx36^+/+, upper panel), but significant presence of spontaneous [Ca²⁺]_i elevations in cells of an islet lacking gap junction coupling (Cx36^-/-, lower panel). Data in B,C reproduced from [51].

While some studies have reported a small increase in basal insulin release in Cx36 deficient isolated islets [29, 60], other studies have reported that no increase in basal insulin release is observed in Cx36 deficient islets, despite [Ca²⁺]_i levels being elevated [58]. In-vivo, fasting insulin and glucose is also unchanged in Cx36 deficient mice [52]. This contrasts with elevated [Ca²⁺]_i and insulin release observed in isolated β-cells. This apparent contradiction is thought to be due to other mechanisms of inter-cellular signaling, possibly including EphA-ephrin-A regulation of insulin granule trafficking and exocytosis, which acts to suppress insulin release at basal glucose levels even in the absence of gap junction coupling [58, 61].

Over-activity of K_ATP channels in β-cells underlies the majority of cases of neonatal diabetes [62]. Similar to the ability of Cx36 gap junctions to suppress electrical activity at basal glucose, Cx36 gap junctions exacerbate the clamping of β-cell electrical activity and insulin release at elevated glucose levels under conditions of K_ATP–overactivity [63, 64]. Reducing Cx36 gap junction coupling in mice expressing overactive K_ATP channels led to a partial recovery in glucose-stimulated [Ca²⁺]_i and insulin release and improved glucose control. This shows the importance in the proper balance of Cx36 gap junctions and K_ATP channels in regulating islet electrical activity. While the role of Cx36 in regulating electrical activity and insulin secretion at low and high glucose levels is increasingly well established; connections between the loss of coupling and development of disease remains to be fully determined.

3. Role of other connexins in the islet

Although Cx36 is well established to be expressed in islet β-cells to regulate insulin secretion, other connexins are also found within the islet. Previous studies have found that cells within the islet contain mRNA for Cx43 and Cx45 in mice and Cx30.3, Cx31, Cx31.1, Cx31.9, Cx37, Cx43 and Cx45 in human pancreas [7]. Some species, such as Cx43 and Cx32 are expressed in vasculature or exocrine tissue respectively, and are not directly expressed in β-cells. However heterotypic channels (composed of connexons of two different connexin species) could potentially form between exocrine tissue or vasculature and β-cells [65, 66]. For example, a recent study has found a correlation between Cx36 expression and changes in glucose stimulated islet blood cell flow rates [67]. This suggests a connection between β-cell electrical coupling and vascular regulation of blood flow; however the mechanism behind this regulation is still unclear

Recently, however, another connexin isoform has been found to be expressed in β-cells. Connexin 30.2 (Cx30.2) was found to be highly expressed in mouse β-cells and vascular endothelial cells and was spatially associated with Cx36. This suggests that these connexin species may form heteromeric connexon hemi-channels and gap junctions [68]. These findings could explain the heterogeneity of Cx36 gap junction coupling across the islet, as mouse Cx30.2 gap junctions have a lower unitary conductance than Cx36 channels in connexin expressing cell lines [69]. Gap junctions may also be formed between β-cells and vascular endothelial cells through heterotypic Cx30.2 and Cx36 connexon interactions. This may suggest a role for Cx30.2 in glycemic control as previous studies in rat retina have shown high glucose down-regulates Cx30.2 protein production. However further study is required to better understand the role of Cx30.2 in islet function and insulin secretion [70], as well as the role of other connexins in the islet.

D. Regulation of connexins in the islet

1. Gene expression and transcription

While most connexins share a common gene structure, the spatial and temporal expression of connexin genes varies greatly by cell type [8, 16]. Selective regulation of Gjd2 expression, the gene that codes for Cx36, has been observed with maturation of mouse islets. Young and adult β-cells were found to express higher levels of Gjd2 than fetal and newborn β-cells, while no changes in other connexin species, such as Cx43, were observed [71, 72]. This role of Cx36 gene expression in early development has not been studied in depth, however recent studies have linked the activation of Cx36 gene expression with the increased production of transcription factors which determine β-cell differentiation [72]. This suggests that Cx36 may be a marker of β-cell differentiation. In mature β-cells, basal expression of Gjd2 is transcriptionally regulated by the neuron-restrictive silencer element (NRSE) located in the promoter region of the Cx36 gene, where RE-1 silencing transcription factor (REST) can bind this region and inhibit expression of Gjd2 [73].

The molecular regulation of Gjd2 expression and transcription has also not been well studied in β-cells. One study has shown transcriptional control of Gjd2 with estrogen and progesterone in neuronal cells; however evidence for similar regulation in β-cells has not been established [74]. Another study has found increased expression of Cx36 mRNA with glucocorticoids in β-cells of insulin resistant mice, however the mechanisms of this regulation are not clear [75]. While few signaling pathways underlying connexin gene expression have been identified across various cell types, gene transcription has been shown to be regulated by a variety of molecular treatments. For example, Cx43 transcription is elevated by treatment with cAMP elevating agents [76], glucocorticoids [77], estrogen [78], parathyroid hormone, and prostaglandin [79]. Transcriptional regulation of Cx43 may also be transactivated by the Wnt-1 signaling pathways, however further study is required to determine which signaling elements are involved [80]. Although little is known about Gjd2 gene or transcriptional regulation, future studies utilizing known regulators of other connexins may elucidate these mechanisms in β-cells. Regulation of Cx36 gene expression is outlined in Figure 3.

Figure 3.

Potential mechanisms of regulation for Cx36 gap junctions in the islet. 1. Regulation of gene expression; 2. Regulation through Phosphorylation; 3. Regulation through connexon trafficking; 4. Environmental Factors; and 5. Pharmacological Regulators. Mechanisms of regulation which have been verified for Cx36 in the islet or in MIN6 cells are indicated in italics. Each regulator or potential regulator of Cx36 has an arrow indicating if the mechanism of action acts to increase or decrease Cx36 function.

2. Gap Junction Gating through Connexin Phosphorylation

The majority of connexins have been found to be regulated by phosphorylation [11]. Phosphorylation occurs on the cytoplasmic C-terminal of the connexin protein, on either serine/threonine or tyrosine residues [16]. While no direct evidence has been found to support phosphorylation of Cx36 in the islet, studies have shown that Cx36 contains serine/threonine and tyrosine residues in different locations on the C-terminal, whose sites in other connexins are readily phosphorylated. This highly suggests phosphorylation of C-terminal residues as mechanism of Cx36 gap junction regulation [16]. Phosphorylation of connexin proteins is thought to control connexon trafficking, assembly into gap junctions, and most notably channel gating. Different protein kinases are thought to phosphorylate different residues on the C-terminal, differentially affecting channel conductance via these mechanisms [81]. In mouse retina, Cx36 can either be phosphorylated by protein kinase A (PKA) which decreases Cx36 gap junction coupling or PKA can activate protein phosphatase 2A which dephosphorylates Cx36 and increases gap junction coupling [82, 83]. In neuronal cells Cx36 can be phosphorylated by calcium/calmodulin-dependent protein kinase (CaMKII) which increases Cx36 gap junction coupling [84]. Currently no other kinase which phosphorylate Cx36 have been identified; however protein kinase C (PKC) and mitogen activated protein kinase (MAPK) have been shown to phosphorylate several other connexin species and have been implicated in the regulation of Cx36 gap junctions (Figure 3) [85]. Future studies in the islet are required to determine if these mechanisms of phosphorylation and channel gating occur via similar mechanisms in islet β-cells.

3. Trafficking

While the cycle of connexon trafficking has been fairly well established, it has not been studied significantly in β-cells. The life cycle of a connexon involves trafficking from the trans-Golgi network (TGN) to the plasma membrane, insertion into the membrane and assembly into gap junctions, internalization, and trafficking to the lysosome or proteasome for degradation [86]. Trafficking of connexons to the plasma membrane occurs via one of two mechanisms: 1) microtubule-dependent vesicle trafficking or 2) trafficking directly from the TGN (Figure 3) [15]. While connexon trafficking via microtubules is well established, more recent studies also suggest a role for microtubules in directing the location of gap junction formation [15]. The secondary pathway for connexon trafficking involves the direct transport of connexon containing vesicles to the plasma membrane. Recent evidence suggests that consortin, a TGN cargo receptor present in transport vesicles and on the plasma membrane, is responsible for targeting connexons to the plasma membrane and facilitating this trafficking mechanism [87]. As these mechanisms appear to be conserved across several species of connexins, the regulation of Cx36 trafficking and recycling may follow a similar pathway in β-cells.

The regulation of these connexon trafficking pathways is not well understood, even for connexin species and cell types which have a defined trafficking mechanism. However a few regulatory pathways have been identified. In cardiac cells, studies have shown that increases in intracellular cAMP and the resulting activation Epac2 lead to increased Cx43 trafficking [88]. Epac2 is active in β-cells and important for insulin secretion [89], and therefore may represent a potential mechanism for the regulation of gap junction trafficking, assembly and coupling. In cardiomyocytes, EphB signaling down-regulates gap junction coupling [90]. It has been suggested that EphB regulation of gap junction coupling is due to changes in trafficking of connexons to and from the membrane [91]. However, additional studies are required to distinguish between effects on connexon trafficking and gap junction assembly as well as to verify this mechanism in β-cells. Factors which effect TGN vesicle formation, microtubule formation, and vesicle trafficking, may therefore be attractive targets for the regulation of connexins and gap junction coupling.

4. Environmental Factors

Environmental factors have been shown to play a role in the regulation of Gjd2 expression and transcription in β-cells. Increases in Cx36 gap junction coupling have been shown under short periods (~ 1 hour) of high glucose [92]. In contrast, prolonged periods (~ 24 hours) of high glucose have been shown to decrease Gjd2 expression. This suggests that under chronic hyperglycemia, such as is seen in Type 2 diabetes, decreases in Cx36 gene expression may be observed [93]. Similarly, conditions of lipotoxicity, caused by chronic exposure to free fatty acids, also decreased Cx36 gene expression [45]. Recent studies have shown that free fatty acids and hyperglycemia upregulate the transcriptional regulator inducible early repressor 1 (ICER-1), and decrease Gjd2 gene expression in mouse islets [93-95]. As both hyperglycemia and hyperlipidemia are present in Type 2 diabetes, this supports a role for Cx36 gap junction coupling in the disruption of insulin secretion associated with Type 2 diabetes. Interestingly, some studies have shown a correlation between Gjd2 gene and insulin gene expression; however further studies are required to elucidate this link and to identify other molecular regulators of Cx36 gene expression [30, 72].

While gap junctions have been shown to be regulated through gene transcription, environmental factors may regulate gap junction coupling in other ways. For example, as discussed above, EphA-ephrin-A communication between islet β-cells regulates insulin secretion [61]. However, studies in mice have shown that Cx36 gap junction coupling is required for EphA-ephrin-A regulation of insulin secretion [61]. Similarly, EphB-ephrinB signaling is required for gap junction coupling in mesenchymal stem cells [91]. While this may represent another mechanism of Cx36 regulation, the mechanistic link between Eph-ephrin signaling and the regulation of Cx36 in the islet is still not clear. Environmental regulation of Cx36 is outlined in Figure 3.

5. Pharmacological modulators of connexins

Pharmacological modulators of gap junction coupling are generally non-specific and have poorly understood mechanisms of action. Therefore, great care is required for interpreting results with putative gap junction blockers. For example, in neuronal cells, 18-α-glycyrrhetinic acid has been found to be a potent blocker of Cx36 gap junction coupling [96]. This blocker is often used in islets, however non-specific effects have been observed [97]. In neuroblastoma cells, Mefloquine, an anti-malarial drug, was found to be a potent blocker of Cx36 and Cx50 gap junction coupling [98]. However, the effect of Mefloquine on Cx36 gap junctions in β-cells has not been reported. Molecules which increase gap junction coupling have also not been studied in depth.

Based on the role of Cx36 in synchronizing [Ca²⁺]_i oscillations, a novel screen of pharmacological regulators of synchronization recently identified several Cx36 blockers and activators [99]. These compounds were confirmed to modulate Cx36 translation, the size of gap junction plaques, and gap junction coupling through dye transfer measurements. Among many compounds, Quinine, Mebeverine, and Gedunin decreased gap junction coupling; and Zaprinast, Norcantharidin, and Glibenclamide increased gap junction coupling. In contrast to the dramatic effects these drugs had on gap junction coupling; only Norcantharidin and Zaprinast had an effect on Cx36 levels. This suggests that Glibenclamide, Quinine, Mebeverine, and Gedunin may regulate gap junction coupling through gating or trafficking, rather than expression/translation. Interestingly, these compounds all affected the observed size of the gap junction plaques [99]. Further investigations into the regulation of plaque formation may reveal their specific mechanisms of action. These compounds were screened in MIN6 cells, therefore future studies will be required to determine the efficacy of these drugs in regulating gap junctions in the islet. However this work has provided a useful platform for several promising candidates for gap junction modulation.

Studies in other cell types have also investigated the use of connexin peptide mimetics: short sections of the connexin protein which have been chemically synthesized for the modulation of gap junctions [100]. The advantage of using peptide mimetics is that they provide superior specificity towards connexin channels compared to chemical compounds. The exact mechanism by which these proteins inhibit gap junction coupling is not clear; however some have suggested that they bind to connexins and prevent proper docking in the plasma membrane, inhibiting channel formation [100]. Studies in epithelial cells have shown a reversible inhibition of [Ca²⁺]_i signaling using Cx43 peptide mimetics, indicating a decrease in gap junction coupling [101]. Peptide mimetics have also been used as a therapeutic treatment, where a Cx43 peptide mimetic increased Cx43 gene expression in cardiomyocytes preventing cardiac arrhythmia [102]. While this approach has not been tested in β-cells, this technique shows promise for specific modulation of Cx36 gap junctions in the islet. All pharmacological regulators of Cx36 are outlined in Figure 3.

E. Potential roles for islet connexins in diabetes

1. Type 1 Diabetes

Type 1 diabetes is characterized by the progressive destruction of insulin producing β-cells by infiltrating immune cells [103]. During the progression of the disease, immune cells which have infiltrated the pancreas produce large amounts of pro-inflammatory cytokines, including tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β), and interferon-γ (IFN-γ). These cytokines cause oxidative stress in the islet and can lead to β-cell apoptosis [104]. Previous studies in isolated islets have shown that pro-inflammatory cytokines inhibit insulin secretion and lead to cell death [105]. Chronic exposure to high levels of these cytokines is thought to play a role in the eventual destruction of insulin producing β-cells, leading to loss of insulin and hyperglycemia. Currently, insulin can be used to manage the consequences of Type 1 diabetes, but therapies are not available to prevent further destruction of β-cells in the islet, although some immunomodulatory therapies are in trials [106].

Recent studies have suggested a role for Cx36 gap junctions in modulating cytokine-induced apoptosis [107]. These studies have shown that a lack of cell coupling makes β-cells more susceptible to cell death under pro-inflammatory cytokines, but the overexpression of Cx36 gap junctions protects islets from ER and oxidative stress, and apoptosis induced by pro-inflammatory cytokines. Evidence suggests that this is due to Cx36 gap junctions regulating [Ca²⁺]_i, where overexpression of Cx36 inhibits formation of radical oxygen species and prevents depletion of ER Ca²⁺ stores [95]. However further studies are required to fully understand this mechanism. In mouse models of Type 1 diabetes, ER stress has been shown to precede the onset of diabetes and is associated with reduced insulin secretion [108]. Recent studies have attributed cytokine-induced ER stress to reduced uptake of Ca²⁺ into intracellular stores, via the sarco-endoplasmic reticulum Ca²⁺ ATPase (SERCA) pump [109]. As Cx36 gap junctions regulate [Ca²⁺]_i and have been shown to affect Ca²⁺ uptake in to ER stores, this suggests a possible role for the loss of gap junction coupling following pro-inflammatory cytokine induced ER stress in islet dysfunction and β-cell death during the progression of Type1 diabetes.

While a correlation between Cx36 gap junctions and Type 1 diabetes has not been implicated in humans, further study may reveal similar decreases in cell coupling in tissue from donors with recent onset Type1 diabetes. Therapies which increase Cx36 gap junction coupling may be able to provide protection against ER stress and β-cell death and potentially delay the onset of disease. Studies in human islets are required to verify the role of Cx36 in cytokine induced islet dysfunction, however current evidence supports investigating this further.

2. Type 2 Diabetes

Central to the development of Type 2 diabetes is chronic hyperglycemia and hyperlipidemia [110] as well as insulin resistance. Factors derived from lifestyle or genetics play a role in disease development, however inflammation resulting from high levels of glucose and circulating free fatty acids has been shown to be a major factor in disease development and progression [110, 111]. Linked with these overt symptoms is a progressive decline in islet function [112], however the order of progression of islet dysfunction and insulin resistance is still under debate. The gene coding for Cx36 is located on the 14q region of chromosome 15, a susceptibility locus for Type 2 diabetes, suggesting a possible connection between Cx36 gap junction coupling and the development of Type 2 diabetes [31, 113].

Although no direct for the involvement of gap junction coupling in Type 2 diabetes has been found in humans, several animal studies have indicated a potential link. In mice, prolonged hyperglycemia led to decreases in β-cell coupling, Gjd2 gene expression and Cx36 protein in isolated islets [93]. Similarly, mice fed a high fat diet, which had high levels of free fatty acids and were characterized as pre-diabetic with increased fasting plasma insulin and glucose, had decreased Cx36 gap junction coupling [56]. The islets in these studies function similarly to the Cx36 deficient mouse model, where decreased glucose tolerance was observed. Decreases in Cx36 gap junction coupling under hyperglycemic and hyperlipidemic conditions will lead to a loss of synchronization of electrical activity and [Ca²⁺]_i oscillations. This will in turn lead to decreased 1^st phase and disrupted pulsatile second phase insulin secretion and therefore reduced insulin action. Such disruptions to insulin dynamics are in fact observed in humans with Type 2 diabetes and prediabetes, as discussed in section C [53-55]. Studies in vitro indicate that hyperglycemia may act via cAMP/PKA regulated activation of ICER-1/ICER-1γ, which has been shown to bind to the Cx36 gene repressor and down-regulate Gjd2 expression [94]. The role of radical oxygen species has also been implicated in mediated electrical uncoupling; however the mechanisms behind this are unclear.

Further study is needed to determine if there are disruptions to Cx36 gap junction function in islets of Type 2 diabetes in humans. However, a recent study suggested that a disruption to [Ca²⁺]_i, consistent with a disruption in Cx36 gap junction coupling, occurs in humans with higher BMI suggesting this occurs in pre-diabetes [45]. Studies which modulate Cx36 gap junction coupling in the presence of hyperglycemia or hyperlipidemia may also help to identify Cx36 as a novel therapeutic target. Given the implications of Cx36 disruption in pre-diabetes and its importance in regulating insulin release and glucose homeostasis, therapies which increase Cx36 gap junction coupling may even delay the onset of disease.

F. Conclusions

The role of Cx36 gap junction coupling in physiological insulin release and islet function has become increasingly well-established over the last decade. Recent studies indicate that electrical coupling is tightly linked to insulin release dynamics and regulation, and that disruptions to this coupling can lead to a pre-diabetes-like state. Better understanding of the role gap junction coupling plays in islet function will further the understanding of islet biology. Computer modeling studies of the islet may also help to predict how changes in gap junction coupling may impact islet function [44, 114]. Of note, relatively little is understood regarding the regulation of gap junction coupling in the islet. Therefore studies are required to understand Cx36 regulation including the role of Cx36 interacting proteins, such as adhesion molecules and the actin cytoskeleton, as well as Cx36 gating and trafficking by factors such as cAMP, PKA, glucose among many others. Given that gap junction regulation, including Cx36 gap junctions, is better defined in other tissue systems, established pathways for gap junction regulation exist, whose potential role in the islet can be tested. Understanding the regulation of Cx36 gap junction coupling in the islet is of particular importance given the more recent evidence linking changes in Cx36 with islet dysfunction and cell death in type1 diabetes, type2 diabetes, and islet function in neonatal diabetes. This includes a novel role for Cx36 gap junction in the cytokine-mediated islet dysfunction linked to the progression of Type 1 diabetes, and down-regulation of Cx36 under conditions similar to pre-Type 2 and Type 2 diabetes. Further studies for the role of Cx36 in islet dysfunction during the progression of diabetes are needed. However, these findings suggest that Cx36 gap junction coupling may be a viable target for future diabetes therapies. The poor gap junction inhibitors available do present some barriers towards this, however understanding Cx36 gap junction regulation should yield alternative signaling pathways that can be exploited to develop Cx36 as a therapeutic target. Finally, limited studies of gap junction coupling have been made in human islets; therefore studies which validate these results in human islets are also vitally needed to determine the possibilities of Cx36 gap junction coupling as a therapeutic target for human diabetes.