Baclofen as a potential inhibitor of Insulin secretion

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Congenital hyperinsulinism in infancy (CHI) is a potentially lethal condition that usually appears in newborn and early childhood (Kapoor et al., 2009) and is characterised by oversecretion of insulin by β-cells in relation to the level of glycaemia (Dunne et al., 2004). In 1938 Laidlaw used the term "nesidioblastosis" to describe recurring episodes of hypoglycaemia associated with inappropriately high levels of serum insulin, C-peptide and proinsulin (Laidlaw, 1938). While this is still relevant to CHI, the term "nesidioblastosis" now describes neodifferentation of islets of Langerhans from pancreatic ductal epithelium which does not only occur in children. Consequently, since 1938, "nesidioblastosis" has been replaced by various synonyms such as persistent hyperinsulinaemic hypoglycaemia of infancy (PHHI), congenital hyperinsulinsm in infancy (CHI) and hyperinsulinism in infancy (HI) (Dunne et al., 2004).

The genetic basis underlying inappropriately high insulin secretion in CHI involves abnormalities in key genes whose products are involved in regulation of insulin secretion from β cells. The most common mutations in CHI are those in ABCC8 and KCNJ11 genes which encode SUR1 and Kir6.2 respectively; the two subunits that make up ATP- sensitive potassium channels KATP (James et al., 2009). These mutations cause defects in KATP channels directly and are referred to as channelopathies as opposed to mutations in HADH (encoding for 3-hydroxyacyl-coenzyme A), GLUD1 (glutamate dehydrogenase), GCK (glucokinase), HNF4A (hepatocyte nuclear activating factor 4A) and SLC16A1 (monocarboxylate transporter 1) which are referred to as metabolopathies, as their effects on KATP channels are indirect (Dunne et al., 2004, James et al., 2009).

Histological examination of CHI has led to classification of the condition in two major subtypes: (1) diffuse CHI which affects the entire pancreas and (2) focal which only affects a small region of the pancreas (Arnoux et al., 2010). Whether diffuse or focal CHI presents with the following biochemical features which occur as a result of inappropriately high insulin levels in the blood: (a) hypoglycaemia (<3mmol/L) due to (i) increased insulin-driven glucose uptake into insulin- sensitive tissues like skeletal muscle, adipose tissue and the liver and (ii) inhibition of glucose production by glycogenolysis and gluconeogenesis (b) low fatty acid release due to inhibition of lipolysis and (c) low ketone body levels due to inhibition of ketogenesis(Kapoor et al., 2009). Taking these into consideration, it is of no surprise that one of the main risks of CHI patients is brain injury, as the brain is not only deprived from glucose but also from ketone bodies which form an alternative fuel source (Hussain and Cosgrove, 2005).

Following diagnosis of CHI, priority is given to maintenance of glycaemia within normal ranges to avoid severe brain damage (Arnoux et al., 2010). This is primarily achieved by intravenous administration of concentrated glucose solution or intramuscular glucagon in case of emergency where veins are not easily accessible (Kapoor et al., 2009). Long-term treatments of CHI are assigned on the basis of medical responsiveness and type of CHI. The first line treatment is oral diazoxide, a drug that binds to the SUR1 subunit of KATP channels, keeping the latter open and thus preventing depolarisation and consequent events leading to insulin secretion. Despite diazoxide being the mainstay of medical therapy, patients with mutations in ABCC8 and KCNJ11 i.e. channelopathies do not usually respond to diazoxide (Arnoux et al., 2010). Other medical treatments available include octreotide (somatostatin analogue) nifedipine (calcium antagonist) and corticosteroids (Dunne et al., 2004, Arnoux et al., 2010). Focal forms of CHI can be treated by partial pancreatectomy, involving the surgical removal of the focal lesion, whereas medically-unresponsive diffuse CHI requires total pancreatectomy (Kapoor et al., 2009)

Development of new therapies for CHI lies upon understanding of the various pathways leading to insulin secretion from pancreatic β-cells. This tightly regulated process involves various protein components and requires the concerted action of nutrients like glucose and amino acids together with hormones and neurotransmitters which act on G-protein coupled receptors (GPCRs) (Lang, 1999). Insights into the mechanisms and integration of the various intracellular pathways leading to exocytosis of the different pools of insulin granules form the basis of present and potential treatments for CHI. In an attempt to find new drugs that could potentially decrease insulin secretion to an extent such that pancreatectomy could be avoided one could search for substances that act as inhibitory mediators of β-cell insulin secretion.

Some papers report that γ-aminobutyric acid (GABA), the main inhibitory neurotransmitter in the mammalian central nervous system, is present in pancreatic β-cells and acts as both as a paracrine inhibitor of glucagon secretion from pancreatic α-cells through its action on ionotropic GABAA receptors and as an autocrine inhibitor of glucose stimulated insulin secretion (GSIS) from β-cells through metabotropic GABAB receptors; the latter still being under investigation (Braun et al., 2004). The role of calcineurin, a serine/threonine protein phosphatase, as a candidate to mediate GABA mediated regulation of insulin secretion has also been reported in some studies (Bernal-Mizrachi et al., 2010).

In this review, the triggering and amplifying pathways of insulin secretion will be described in an attempt to explain the pathophysiology, clinical presentation, diagnosis and current treatments of CHI. The structure and function of GABAB receptors and calcineurin will also be presented in relation to bacofen/GABAB -mediated action, as they are potential therapeutic targets for patients with channelopathy-associated CHI who are unresponsive to diazoxide.


Insulin is a peptide hormone secreted by pancreatic β-cells in order to maintain normal body glucose homeostasis. It is stored in large dense core vesicles (LDCVs), also known as secretory granules and is released from β-cells by exocytosis. The latter is a multistage process involving transport of granules, docking, priming and eventually fusion of the granules with the membrane (Lang, 1999).

It has been well established by now that there are at least two different pools of insulin-secretory granules, namely the readily releasable pool (RRP) and the reserve pool which account for the first (triggering) and second (amplifying) phases of insulin release respectively(Hou et al., 2009). According to (Ohara-Imaizumi et al., 2009), the types of fusing granules are determined by the pattern of rise of subplasma membrane [Ca2+]. The RRP of insulin granules are pre-docked at the cell surface membrane complexed with SNARE (soluble N-ethylmaleimide-sensitive factor attachment protein receptor) and Ca2+ -regulated proteins allowing for rapid fusion of primed insulin granules with the membrane (Hou et al., 2009). In mouse β-cells this initial rapid phase of insulin secretion involves only 50 out of 10000 insulin-containing granules clustered close to voltage gated Ca2+ channels (VGCC) ready to be released (Barg et al., 2002). The second and more prolonged phase of insulin secretion occurs after the reserve pool of insulin-containing granules has been recruited to the release sites (Hou et al., 2009). The signalling events and molecular mechanisms involved in biphasic insulin secretion have been intensively studied but several questions regarding the repertoire of regulatory proteins and fusion reactions required for insulin exocytosis still need to be answered.

Owing to disorders like CHI and Diabetes mellitus, glucose-stimulated insulin secretion (GSIS) has been the most well-studied pathway in β-cells, thus it will be the first pathway to explain. Insulin secretion however, has been additionally coupled to the action of other hormones and neurotransmitters acting on GPCRs (Lang, 1999). The effects of glucose are enhanced by some of these factors, examples of which include acetylcholine (Ach), glucagon-like peptide 1 (GLP-1), which act through GPCR in an attempt to optimise insulin secretion. It should be noted that the Ach and GLP-1 pathways which will be described in some detail below, only operate in the presence of glucose and seem to exert their effects on the amplifying pathway of insulin secretion thereby enhancing the secretory response (Winzell and Ahren, 2007). Somatostatin, adrenaline and galanin also exert their action via GPCRs but their effects are inhibitory and will not be described in this review (Renstrom et al., 1996).

1.1 Triggering and amplifying pathways of insulin secretion:

1.1.1 Triggering pathway

The principal regulator of the triggering pathway of insulin secretion is glucose metabolism and its effect on KATP channels, which were firstly identified and linked to GSIS in 1984 (Cook and Hales, 1984). The latter are heteroctameric complexes formed from 2 distinct families of proteins; four inwardly rectifying potassium channel subunits known as Kir 6.2 and four sulphonyluria receptor 1 subunits (SUR1) which are encoded by the KCNJ11 and ABCC8 genes respectively (Kapoor et al., 2009). Under normal conditions, KATP channels couple glucose metabolism to insulin secretion via changes in the membrane potential of β-cells (Ashcroft et al., 1984).

In resting, unstimulated β-cells KATP channels are open and together with Na+/K+ ATPase they establish a resting membrane potential of -65mV (Dunne et al., 2004) . The sequence of events leading to the triggering or KATP-dependent pathway of GSIS is the following, also shown schematically in Figure 1: Postprandial increase in glucose concentration in the blood causes an increase in glucose flux in β-cells via the glucose transporter GLUT2 (in rodents) and GLUT1/3 (in humans) in the β-cell membrane(De Vos et al., 1995). Glucose is then converted to glucose-6-phosphate by glucokinase. Further glucose metabolism both in the cytoplasm and mitochondria results in an increase in the intracellular ATP/ADP ratio leading to closure of KATP channels and depolarisation of the cell with subsequent opening of L-type VGCC and Ca2+ influx down the electrical gradient (Ashcroft et al., 1984, Cook et al., 1988). Intracellular Ca2+ oscillations trigger oscillations in insulin secretion (Henquin, 2000, Ravier et al., 1999). The rise in [Ca2+]i allows fusion of predocked insulin granules with the plasma membrane in a SNARE complex dependent process which results in insulin release into the bloodstream (Wang and Thurmond, 2009).

1.1.2 Amplifying pathway

(A) Glucose-mediated augmentation of insulin secretion

It has been shown that glucose can stimulate insulin secretion through KATP-independent pathways, also known as amplifying pathways some of which are Ca2+- dependent and some Ca2+ independent but do not involve further rise in [Ca2+]I (Bratanova-Tochkova et al., 2002). One of the mechanisms of the amplifying pathway involves elevated [Ca2+]i and serves to enhance the effect of Ca2+ on exocytosis (Hussain and Cosgrove, 2005). In the absence of elevated [Ca2+]i insulin secretion is stimulated via activation of protein kinases A and C ( PKA and PKC) which phosphorylate proteins involved in exocytosis of insulin granules in addition to mobilising intracellular Ca2+ stores (Yu et al., 2000). Studies on the pathways of insulin secretion have shown that all metabolised nutrients including amino acids and fatty acids are capable of activating the amplifying pathway which nevertheless requires metabolism of glucose to operate. Other second messengers besides cAMP, PKA and PKC have been proposed to be important in the amplifying pathway such as adenine and guanine nucleotides, calmodulin-dependent protein kinase II (CAMKII) (Straub et al., 2001) as well as phosphorylated and dephosphorylated forms of certain proteins in β-cells implicating the involvement of protein kinases and phophatases (Sato et al., 1998)

There seems to be a hierarchy between the triggering and amplifying pathways of GSIS since the amplifying pathway remains functionally silent in the absence of the triggering pathway preventing inappropriate insulin secretion in the presence of low glucose concentrations (Henquin, 2000).

(B) Acetylcholine-mediated augmentation of insulin secretion

In addition to glucose stimulation of insulin secretion in pancreatic β-cells, Ach, the major neurotransmitter of the peripheral parasympathetic nervous system has also been found to modulate insulin secretion due to the presence of muscarinic Ach receptors (mAchRs) expressed on the membrane of β-cells (Henquin and Nenquin, 1988). Ach is released from intrapancreatic nerve endings, which are under vagous control, during preabsorptive and absorptive phases of feeding. Studies on mutant mice selectively lacking the M3 mAchRs in β-cells as well as transgenic mice overexpressing the same receptor have demonstrated its importance in maintaining proper release of insulin and glucose homeostasis (Gautam et al., 2006).

Ach-mediated insulin secretion is dependent on glucose concentration. Stimulating the vagous nerve in vivo when blood glucose levels are low has minimal effect on insulin secretion. Increase in secretion is only observed under elevated blood glucose concentration. This phenomenon implicates that Ach most probably affects the amplifying, but not the triggering pathway of insulin secretion. It has also been observed that Ach is capable of increasing insulin secretion in response to nutrients besides glucose, e.g the amino acid leucine (Gautam et al., 2006).

Upon binding of Ach to M3 receptors, insulin secretion is increased via two mechanisms: (1) several transduction pathways are activated, the major pathway involving phospholipase C (PLC) which generates inositol triphosphate (IP3), diacylglycerol (DAG) and phospholipase A2 (PLA2) resulting in the production of arachidonic acid and lysophosphatidylcholine. DAG, arachidonic acid and lysophosphaditylcholine activate protein kinase C (PKC), thus increasing the efficiency of cytosolic Ca2+ on insulin granule exocytosis (Gilon and Henquin, 2001) whilst IP3 results in mobilisation of intracellular calcium stores (Lang, 1999). The latter effect results in myosin light chain kinase activation (possibly via Ca2+-calmodulin) which leads to phosphorylation of myosin light chains in microfilaments and subsequent insulin granule movement and exocytosis (Iida et al., 1997) (2) Surprisingly, Ach was found to further stimulate insulin secretion via depolarisation of the β-cell membrane by Na+ or other non-specific cationic mechanisms provided that the plasma membrane is already partially depolarised via GSIS. This additional depolarisation allows sustained elevation of cytosolic Ca2+ and thus further enhances GSIS (Gilon and Henquin, 2001).

(C) Glucagon like peptide (GLP-1) mediated augmentation of insulin secretion

GLP-1 is a gut-derived hormone (incretin), released from the gastrointestinal (GI) tract and more specifically from the L-cells of the small intestine (Hou et al., 2009) in response to oral but not intravenous stimulation of insulin secretion by nutrients. It acts to increase insulin secretion by stimulation of insulin exocytosis from pancreatic β-cells as well as to increase insulin gene expression, β-cell growth and differentiation (Kjems et al., 2003).

GLP-1 acts on pancreatic β-cells by binding on its recptor GLP-1R, which in turns activates adenylyl cyclase (AC), thereby increasing cAMP concentration. PKA (Hou et al., 2009) and Epac (Kang et al., 2008) are activated by increased cAMP resulting in stimulation of insulin secretion (Hou et al., 2009). Studies performed using low but physiological concentrations of GLP-1 have shown that the peptide can also stimulate insulin secretion independently of cAMP and PKA, and this is most probably the main pathway operating in vivo (Shigeto et al., 2008).

1.2 Insulin granule exocytosis

Images of insulin-containing granule fusion with the plasma membrane following stimulation by glucose and other modulators described above were first published in 1957 (Lacy and Davies, 1957). It was not until 1993 however that the involvement of SNARE proteins as a minimal requirement for vesicle exocytosis was determined (Hou et al., 2009, Wang and Thurmond, 2009). SNARE proteins were initially thought to be specific to neuron cells. However, vesicle SNARES (v-SNARES) and cognate-target SNARES on the plasma membrane (t-SNARES) were also identified in β-cells using immunostaining (Lang, 1999). Insulin release was shown to involve syntaxin and SNAP25 (t-SNARES) and VAMP2 (v-SNARE). Additionally, the involvement of SNARE accessory proteins in vesicle exocytosis was also established as the core SNARE complex could not account for rapid vesicle exocytosis by itself (Hou et al., 2009).

A Examples of these accessory proteins include the Ca2+/phospholipid binding proteins synaptotagmin III, IV, VII, VIII and syncollin on insulin granules which mediate Ca2+-dependent secretion. Ca2+ binding to synaptotagmin results in blocking of the repressive effect of complexin on SNARE and allows fusion of granules with the membrane. Furthermore, L-type VGCC are physically associated with syntaxin, synaptotagmin and SNAP25 while KATP channels were found to interact with syntaxin, implicating a role for the two channels in release of the RRP of insulin granules (Hou et al., 2009).

BC:\Users\admin\Desktop\diagram 1.jpg

Figure : (A) The triggering and amplifying pathways of insulin secretion. Triggering (purple arrows): Glucose enters the pancreatic β-cell via the GLUT transporter and following glycolysis it gets converted to pyruvate which then enters the mitochondria. The latter produce ATP from pyruvate, thus increasing the intracellular ATP/ADP ratio, leading to closure of KATP channels leading to depolarisation, Ca2+ influx and triggering of insulin exocytosis. Amplifying (pink arrows) : The process is enhanced by other hormones and neurotransmitters such as Acetylcholine (Ach) and glucagon-like peptide 1 (GLP-1) which exert their effects via G-protein coupled receptors on the membrane. These in turn stimulate insulin exocytosis via signal transduction pathways involving second messengers. (B) Insuline granule exocytosis. SNARE proteins involved in secretion of insulin granules include v-SNARES present on the insulin granile membrane and t-SNARES present on the plasma membrane of β-cells. Regulatory proteins of SNARES are also involved in the process. DAG ( diacylglycerol), PIP2 (phosphatidylinositol 4,5-bisphosphate), IP3 (inositol triphosphate), mAchR (muscarinic acetylcholine receptor), GLP1R (glucagon like peptide-1 receptor), AC (adenylate cyclase), PKA (protein kinase A), PKC (protein kinase C), Protein-P (protein phosphorylation) C:\Users\admin\Desktop\diagram 3.jpg



The main feature of CHI which requires urgent treatment to prevent severe brain damage and general distress is hypoglycaemia, which in children is defined by plasma glucose levels below 3mmol/L (Arnoux et al., 2010). Hypoglycaemia in CHI occurs as a result of unregulated insulin secretion from pancreatic β-cells. The process of insulin secretion is usually regulated such that blood glucose levels are maintained within the normal range, 3-5mmol/L (James et al., 2009).

In most CHI patients loss of regulation of insulin secretion resulting in impaired glucose homeostasis occurs due to of loss - of function mutations in ABCC8 and KCNJ11 genes, known as channelopathies, which as previously mentioned code for the two subunits of KATP channels. These mutations have a variety of effects on the channels including defects in channel biogenesis and turnover, channel trafficking and channel regulation (James et al., 2009) which ultimately lead to spontaneous membrane depolarisation, unregulated opening of VGCC and uncontrolled insulin release (Cosgrove et al., 2002). The majority of KATP channel mutations are recessive, but dominantly inherited mutations have also been reported (James et al., 2009). In addition to mutations in the two genes coding for the two subunits of KATP channels (which are also referred to as channelopathies), mutations in other genes such as glucokinase (GCK), glutamate dehydrogenase (GDH) which will be described below, as well as hydroxyacyl coenzyme A dehydrogenase (HADH), solute carrier family 16 member 1 (SLC16A1) and hepatocyte nuclear factor 4 alpha (HNF4A) mutations have been described and are known as metabolopathies (Dunne et al., 2004, James et al., 2009)

Mutations in GCK and GDH are dominant activating in nature. Glucokinase is a glycolytic enzyme which catalyses the phosphorylation of glucose to glucose-6-phosphate i.e. the first step in the process. Activating mutations in glucokinase are usually clustered in the allosteric activator site of the enzyme cause increased affinity of the enzyme for glucose (thus decreased Km) and lower threshold for glucose stimulated insulin secretion due to increased rate of glycolysis (Kapoor et al., 2009). Glutamate dehydrogenase is a mitochondrial enzyme that catalyses the oxidative deamination of glutamate to α-ketoglutarate and ammonia. α-ketoglutarate then enters the tricarboxylic acid cycle leading to an increase in ATP/ADP ratio, opening of KATP channels and cell depolarisation stimulating insulin exocytosis. Activating mutations in the enzyme occur as a result of reduced allosteric enzyme inhibition by leucine leading to increased glutamate oxidation in the presence of leucine and thus increased insulin secretion (James et al., 2009).

CHI most commonly presents in newborn with hypoglycaemia being quite severe but it can also appear in infancy and early childhood. Hypoglycaemia usually occurs following oral feeding and normoglycaemia can only be maintained by administration of concentrated dextrose solutions (Arnoux et al., 2010). Hypoglycaemic symptoms include poor feeding, lethargy and irritability in the newborn while more severe cases can present apnoea, seizures and even coma. A typical feature of hyperinsulinaemic newborns is macrosomia; nevertheless non-macrosomic newborns could also have CHI (Kapoor et al., 2009). Hepatomegaly and hypertrophic cardiomyopathy may also be observed (Arnoux et al., 2010).

2.2 Diagnosis and current management of CHI

Left untreated CHI can lead to death and early recognition of newborn with CHI is important for preventing hypoglycaemic brain injury (Kapoor et al., 2009). One tool used to assess unregulated insulin secretion is calculation of the intravenous glucose infusion rate necessary to maintain normoglycaemia. A glucose infusion rate greater than 8mg/kg/min compared to 4-6mg/kg/min which is normal is indicative of HH (Kapoor et al., 2009). The criteria used to determine the presence of CHI are (a) high plasma insulin levels during fasting and post-prandial hypoglycaemia (b) positive response to glucagon administration (c) absence of ketone bodies in urine and plasma (d) low serum fatty acids and (e) low levels of serum insulin growth factor binding protein (IGFBP-1) (Kapoor et al., 2009, Arnoux et al., 2010).

Once diagnosis of CHI is made, priority is given to maintenance of normal blood glucose levels by oral, enteric or intravenous administration of glucose. If this is insufficient, glucose administration can be combined with glucagon administration to allow mobilisation of hepatic glucose (Cosgrove, 2007) and thus achieve the required 3.5mmol/L plasma glucose concentration in neonates (Arnoux et al., 2010).

The first specific treatment for HI is oral diazoxide; yet though most neonatal cases of CHI are diazoxide unresponsive (Arnoux et al., 2010). In addition to diazoxide being only suitable in metabolopathy-associated CHI but not channelopathy-associated CHI, the drug has adverse side effects including fluid retention, hyperuricaemia, thrombocytopenia, hypotension, facial changes and hypertrichosis thus the compound is poorly tolerated. Attempts have been made to develop diazoxide analogues, an example being BPDZ 154 in order to avoid the side effects of the drug (Cosgrove et al., 2002). Somatostatin analogues such as octreotide and sandostatin are next in line before considering surgical intervention when the patient is diazoxide-unresponsive. Nifedipine, a VGCC blocker is another available medical option and could possibly be of therapeutic value in addition to corticosteroids which increase gluconeogenesis (Cosgrove, 2007).

Patients who are resistant to all medical therapies need to be screened and assessed for focal and diffuse forms of CHI. The focal form of CHI which affects only a small region of the pancreas is associated with Ch11p15 gene defects while the diffuse form which affects the entire pancreas is associated with Ch11p15, 7p15-p13, 10q23.3 and 4q22-q26 gene abnormalities (Dunne et al., 2004). The diffuse form can be detected by genetic analysis which determines whether the patient is homozygous or compound heterozygous for ABCC8 and KCNJ11 mutations (Kapoor et al., 2009). 18F-Fluoro-L-DOPA PET scan can also be used to distinguish the focal from the diffuse form (Arnoux et al., 2010). Another suggested test for distinguishing the two major CHI forms is assessing the response of the patient to tolbutamide (a KATP channel blocker) (Straub et al., 2001) but this is not recommended as the PET scan is more effective in both diagnosing and localising the focal lesion in focal CHI (Arnoux et al., 2010). The focal form of CHI only requires limited pancreatectomy cases of medical unresponsiveness and the surgery is carried out either with the open approach or using laparoscopic techniques. Near-total pancreatectomy is only performed in diffuse cases of CHI where all medical intervention has failed but it is often associated with postoperative diabetes mellitus and exocrine insufficiency (Kapoor et al., 2009).

Since manipulation of KATP channel activity is only an option in metabolopathy-associated CHI, further attempts need to be made in order to find alternatives-to-surgery for channelopathy-associated patients whose only option is surgical intervention. A possible approach is manipulation of GABAB receptor regulation of insulin secretion through the use of baclofen, a GABAB receptor specific agonist (Braun et al., 2004).

3. γ- Aminobutyric acid (GABAΒ) receptors and calcineurin structure and function in relation to their role in pancreatic β-cells

3.1 GABA: synthesis uptake and release

Apart from being the main inhibitory neurotransmitter in the mammalian central nervous system (CNS), GABA and its synthesising and degrading enzymes glutamic acid decarboxylase (GAD) and GABA transaminase respectively, are present at high levels in pancreatic insulin-producing β-cells (Braun et al., 2004). In addition to the presence of intracellular GABA, the neurotransmitter might also be secreted in the pancreas from GABAergic neurons that are associated and can penetrate the islet mantle (Bonaventura et al., 2008). Within the β-cell, GABA is mostly localised in synaptic-like microvesicles (SLMVs) where it accumulates by active transport by the vesicular GABA/glutamate transporter, VGAT (Franklin and Wollheim, 2004). GABA active transport is dependent on a proton gradient generated by vesicular H+-ATPase (Bonaventura et al., 2008).

Intracellular GABA is synthesised in the cytosol from glutamate decarboxylation, a reaction catalysed by GAD and enters the "GABA shunt" in the mitochondria (Franklin and Wollheim, 2004) after its transamination with 2-oxoglutarate to form succinate semialdehyde (SSA). SSA can then be oxidised to succinic acid and enter the tricarboxylic acid cycle to be used as an energy source(Pizzaro-Delgado Javier, 2010). It has also been reported that GABA may have an exocrine role in pancreatic β-cells (Park and Park, 2000). Additionally, GABA metabolism might exert an effect on insulin secretion (Pizzaro-Delgado Javier, 2010). Studies have shown that the neurotransmitter is released from the β-cells in response to glucose (Bonaventura et al., 2008) by Ca2+-dependent exocytosis (Braun et al., 2004) similar but not identical to GSIS (Franklin and Wollheim, 2004). In some studies glucose was found to inhibit GABA release thus the results are controversial(Wang et al., 2006)

3.2 GABA receptors

GABA is known to act on three types of receptors, namely, GABAA, GABAB and GABAC. The former and latter are ionotropic in nature and they belong to the family of ligand-gated ion channels; more specifically they are Cl- gated ion channels (Braun et al., 2004). GABAA receptors are expressed on pancreatic α-cells in humans and are pentameric ion channels. Each of the five subunits consists of four transmembrane domains (TMDs); the second subunit of each TMD combines with the rest to form the wall of the channel pore. The receptor is selectively blocked by the antagonist biculline (Chebib and Johnston, 1999). Some recent studies have reported its role in insulin secretion in INS-1 cells but the results are still controversial (Dong et al., 2006). GABAA expression in the pancreatic islets is still not fully elucidated either. GABAA subunits have been found in all of α,β and δ cells of rat pancreatic islets whereas in guinea pig immunostaining revealed GABAA in α and δ but not β cells (Franklin and Wollheim, 2004). The presence of GABAA receptors on human glucagon-secreting α cells implicated their involvement in inhibition of glucagon release by GABA secretion from nearby β-cells (Wendt et al., 2004, Annette et al., 1991).

On the other hand, GABAB receptors are metabotropic GPCRs and are coupled to Gi/Go proteins (Bonaventura et al., 2008). They were firstly identified in 1979 and cloned in the 1990s (Padgett and Slesinger, 2010). The function of GABAB depends on heterodimerisation of GABAB1 and GABAB2 subunits (Bonaventura et al., 2008) which were found to be coexpressed and, surprisingly for heptahelical receptors, they were only functional upon linking of their C-terminal tails (Brice NL, 2002). GABAB1 remains bound to the ER where it is synthesised due the C-terminal ER retention signal RXRR unless it heterodimerises with GABAB2 so that the signal is masked and the receptor can be translocated to the membrane (Bowery et al., 2002).

3.3 GABA effect on hormone secretion and the effects of baclofen

3.3.1 Proposed GABA mode of action through baclofen activation of GABAB receptors

GABA released from pancreatic β-cells was proposed to have an autocrine as well as a paracrine function in the regulation of insulin, glucagon and somatostatin secretion in vitro but contradictory results were obtained in different studies (Bonaventura et al., 2008). This may well be attributed on differences in the experimental design or the animal species used in each experiment, thus yielding dissimilar results (Gu et al., 1993). An attempt will nevertheless be made to present the various findings associated with the role of GABA in the regulation of hormone, mostly insulin, secretion which is usually studied using the GABAB receptor specific agonist baclofen (Braun et al., 2004).

The action of GABABRs was primarily studied presynaptically and postsynaptically in neurons. Presynaptic GABAB activation mediated suppression of neurotransmitter release due to inhibition of VGCC, while postsynaptic GABAB mediated inhibition of adenylyl cyclase via Gai and activation of Kir channels via Gβγ ((Brice NL, 2002, Gladkevich et al., 2006). Since these effectors are also present in pancreatic β-cells it is sensible to assume that reported baclofen-mediated inhibition of insulin exocytosis might proceed via the same pathways(Padgett and Slesinger, 2010). In support of this hypothesis, baclofen inhibits insulin secretion in the presence of tolbutamide thus implicating its effect on KATP channels. However the KATP channel blocker is known to have a direct effect on the secretory machinery thus questioning whether baclofen exerts any of its effects through KATP channel opening (Braun et al., 2004).

The involvement of Gi/Go proteins in baclofen mediated inhibition of insulin secretion is supported by the finding that treatment of β-cells with pertussin toxin prevents inhibition of insulin exocytosis (Braun et al., 2004). The inhibition of GSIS by baclofen seems to occur in a dose dependent manner where increasing concentration of baclofen results in increasing inhibition capacity (Franklin and Wollheim, 2004). Inhibition also seems to be dependent on glucose concentration. Baclofen was found to inhibit GSIS in the presence of 16.7mM and 25 mM glucose in two different studies (Brice NL, 2002, Gu et al., 1993)

3.3.2 Calcineurin as a GABAB effector upon baclofen activation of GABAB receptors

Another candidate for GABAB effectors in pancreatic β cells is the Ca2+-and calmodulin-dependent serine/threonine phosphatase protein, calcineurin (also known as protein phosphatase 2B, PP2B) which was firstly isolated and identified in 1979 (Rusnak and Mertz, 2000). It is highly expressed in the brain but its presence in many other mammalian tissues including the pancreas has also been reported (Sans and Williams, 2004). However, the major discovery regarding calcineurin which also comprises the target of immunosuppressive therapies is the inhibition of the protein's activity by Cyclosporin A (CsA) and FK506 which bind to regulatory proteins of the enzyme, Cyclophilin A and FKBP-respectively (Bernal-Mizrachi et al., 2010). Consequently these two drugs are used in various studies to elucidate the role of calcineurin in signal transduction pathways (Rusnak and Mertz, 2000).

Calcineurin is expressed in almost all cell types including insulin secreting β-cells where it is involved in gene regulation, exocytic secretion and cell survival and death (Ranta et al., 2008). It is a heterodimeric protein that consists of a catalytic subunit, Calcineurin A and a regulatory subunit, Calcineurin B which contains four Ca2+ binding sites. Calcineurin can be phosphorylated by protein kinase C (PKC), casein kinase I and casein kinase II in vitro. In mammals calcineurin is localised in cytoplasmic and microsomal fractions of cells as wells as in synaptosomes. It is also associated with the cytoskeleton where several of its substrates are also localised, such as tau, microtubule-associated protein 2, tubulin, dystrophin and dynamein. The enzyme is inactive in low [Ca2+]i primarily due to its inability to bind calmodulin which is also activated by Ca2+ ions and active in high [Ca2+]i (Rusnak and Mertz, 2000).

In insulin secreting cells, calcineurin seems to function in two opposing ways. On one hand it inhibits insulin secretion via stimulation of inhibitory pathways mediated by Gi proteins and regulation of endocytic recycling of insulin secretory vesicles. On the other hand it has an indirect effect on insulin gene expression stimulation which is achieved by calcineurin-dependent dephosphorylation and activation of the transcription factor NFAT (Ranta et al., 2008, Heit et al., 2006).

A proposed function for calcineurin in relation to β-cell insulin exocytosis is the role of the enzyme in dephosphorylating kinesin heavy chain (KHC) on insulin granules and thereby inhibiting transport of the latter to the cell membrane to be released (Donelan et al., 2002). β-granules are transported to the membrane along microtubules driven by kinesin and other ATP-dependent motors. Kinesin consists of both heavy and light chains which have potential sites for phosphorylation and have been found phosphorylated in vivo. In a single study, it has been proposed that KHC are phosphorylated by casein kinase-2 and dephosphorylated by calcineurin thus regulating exocytosis of insulin. It should be of note that suppression of kinesin expression by oligotargetting causes a decrease in GSIS (Donelan et al., 2002).

These results are however contradicted by studies made using baclofen and cyclosporin A. The data obtained by Braun et al 2004 suggest that baclofen inhibits Ca2+-dependent exocytosis by activating calcineurin. Taking into account the findings on the action of calcineurin on KHC (Donelan et al., 2002) it seems that there's still doubt about the state of calcineurin (i.e. active or inactive) required to inhibit exocytosis. The contradictory results may simply reflect the two opposing roles of calcineurin in regulating insulin secretion or it may be the case that other mechanisms need to be identified in order to explain the function of calcineurin in this process.

C:\Users\admin\Desktop\diagram 2.jpg

Figure : Possible modes of action of the GABAB specific agonist baclofen. Upon binding of baclofen to GABAB receptor Gi/Go proteins are activated. These can inhibit adenylate cyclase thus inhibiting closure of KATP channels via PKA mediated pathways leading to inhibition of insulin exocytosis. It has been also suggested that Gi/Go proteins inhibit calcineurin thus prevent it from dephosphorylating KHC leading to inhibition of exocytosis. Calcineurin is also known to affect transcription of genes associated with insulin release via NFAT-mediated signaling. AC (adenylate cyclase), PKA (protein kinase A), KHC (Kinesin Heavy chain), GABABR (GABAB receptor).


CHI is a clinically heterogeneous condition with respect to the age of onset (although it most commonly presents in the neonatal and infancy period), severity, morphology and responsiveness to medical treatment (Dunne et al., 2004, Straub et al., 2001) and it refers to the oversecretion of insulin by pancreatic β cells in relation to the level of glycaemia (Kapoor et al., 2009). Metabolopathy-associated CHI can be treated with current available drugs, whereas channelopathy-associated CHI tends to be medically-unresponsive and requires near total pancreatectomy with the risk of postoperative exocrine insufficiency and diabetes mellitus (Dunne et al., 2004). Current and future treatments for CHI are based on manipulation of insulin secretion pathways, mainly the amplifying pathway, thus gaining insides into the latter will hopefully allow for more substantial and effective treatment of the disease. Although insulin secretion pathways have been intensively studied by several researchers all around the world, owing to their relevance in both CHI and diabetes mellitus, several questions regarding the exact mechanisms involved still remain unclear. The amplifying pathway is a complex integration of several transduction pathways involving a variety of second messengers, each of which exerts an effect on more than one protein in the cell thus making it difficult to manipulate the process and get effective results.

Since channelopathy-associated CHI is unresponsive to diazoxide, one needs to look into other ways of inhibiting insulin secretion such as GABA-mediated inhibition of insulin granule exocytosis which involves Gi and calcineurin as putative second messengers.

In this project we will perform an experiment on healthy mouse tissue in an attempt to assess insulin secretion in the presence and absence of the GABAB specific agonist baclofen. The latter was found to inhibit insulin secretion in previous studies but the results are still controversial so this step needs to be repeated. Insulin secretion will be stimulated using different concentrations of glucose, Ach, GLP-1 and possibly KCl (chemical depolarisation of β-cell). If it is established that the agonist can actually inhibit insulin secretion baclofen can be tested on tissue from CHI patients to confirm that similar results are obtained. This will provide insights into treating diazoxide-unresponsive CHI patients.