Role of Magnesium in the Regulation of Hepatic Glucose Homeostasis

The liver comprises of hepatocytes, biliary epithelial cells, stellate cells (or Ito cells), Kupffer cells, sinusoid endothelial cells, and pit cells [1,2]. Most of the clinically quantifiable liver functions such as metabolic processes and protein synthesis take place within the hepatocytes, while non-hepatocyte cells are responsible for other functions including inflammatory response (Kupffer cells), collagen deposition (Ito cells), and cell orientation [2-5]. Regulation of blood glucose is one of the main functions exerted by the liver. The organ contains a dynamic storage of glycogen that is rapidly dismissed into the circulation as glucose to maintain glycemia and support brain functions. Hence, hepatocytes are enzymatically specialized to switch rapidly between glycogenolysis and glycogenosynthesis based upon hormonal stimuli and metabolic conditions.


The liver and glucose metabolism
The liver comprises of hepatocytes, biliary epithelial cells, stellate cells (or Ito cells), Kupffer cells, sinusoid endothelial cells, and pit cells [1,2]. Most of the clinically quantifiable liver functions such as metabolic processes and protein synthesis take place within the hepatocytes, while non-hepatocyte cells are responsible for other functions including inflammatory response (Kupffer cells), collagen deposition (Ito cells), and cell orientation [2][3][4][5]. Regulation of blood glucose is one of the main functions exerted by the liver. The organ contains a dynamic storage of glycogen that is rapidly dismissed into the circulation as glucose to maintain glycemia and support brain functions. Hence, hepatocytes are enzymatically specialized to switch rapidly between glycogenolysis and glycogenosynthesis based upon hormonal stimuli and metabolic conditions. Glucose enters the hepatocytes through the low-affinity transporter GLUT2 (Km=15-20 mM, Table 1). At variance of GLUT1 and GLUT4 glucose transporter that possess a Km=1-5mM and are therefore constitutively active near their maximal rate under euglycemic conditions (i.e. between 60 to 100 mg/dl), GLUT2 is maximally activated following a meal. The high Km of GLUT2 (~15-20mM) correlates well with the high Km glucokinase responsible for the conversion of glucose to glucose 6-phosphate [6].
Glucose 6-phosphate (G6P) can be routed towards glycogenosynthesis, glycolysis, or oxidation by the cytoplasmic glucose 6-phosphate dehydrogenase, de facto entering the pentose shunt pathway, an alternative path that generates ribose 5-phosphate utilized in nucleic acid formation (cell cycle) or return as glucose 6-phosphate to be used once again as most convenient for the cell (Fig.1). Glucose 6-phosphate is also transported into the endoplasmic reticulum to undergo hydrolysis via glucose 6-phophatase, or oxidation via the hexose 6-phosphate dehydrogenase, the reticular variant of the glucose 6-phosphate dehydrogenase [7]. The modality whereby glucose 6-phosphate enters the hepatic E.R. lumen is still debated. The substrate-transport theory postulates that G6P enters the ER lumen via a specific transporter (T1) distinct from the glucose 6 phosphatase. In this model, T1 represents the rate-limiting factor for the G6Pase system [8]. The conformational flexibility substrate-transport theory proposes that the G6Pase enzyme possesses a hydrophilic region that spans the E.R. membrane and project into the cytoplasm. This region is specific for substrate binding and is distinct from the hydrolytic site. Upon binding to glucose 6-phosphate this cytoplasmic site of the protein undergoes a conformational change and delivers the substrate to the intra-luminal catalytic site. According to this model, the substrate binding site and a hydrolytic site of the G6Pase are two parts of the same protein, and the enzyme is not specific for a particular substrate [9]. Irrespective of the modality of entry, the hydrolysis of glucose 6-phosphate by the hydrolytic site of the glucose 6-phosphatase generates two byproducts, glucose and inorganic phosphate (Pi), which are released into the cytoplasm via two additional, specific transport mechanisms [8][9][10][11]. The glucose exported out of the ER is either earmarked for glucose output from the hepatocyte into the bloodstream or is converted anew to glucose 6-phosphate by the glucokinase thereby contributing to the glucose-glucose 6-phosphate futile cycle [12]. The inorganic phosphate (Pi) is either exported out of the ER lumen through its specific transporter, or forms a complex with the Ca 2+ ions that are actively transported into the ER lumen by the SERCA pumps [13]. Far from being static and irreversible, this Ca*Pi complex promotes an enlargement of the reticular Ca 2+ pool within the hepatocyte, and it can be dynamically reversed to Ca 2+ and Pi, with both moieties being mobilized out of the ER following IP 3 -induced Ca 2+release [13]. Thus, this enlargement of the reticular Ca 2+ pool is an integral part of the hepatic response to hormones such as vasopressin or norepinephrine that tap into the IP 3 -related Ca 2+response for metabolic and functional purposes [14].
Further investigation is required to fully elucidate the functional implications of the reticular hexose 6-phosphate dehydrogenase. This enzyme also utilizes the glucose 6-phosphate transported into the E.R., oxidizing it to 6-phosphogluconolactone [7]. Essentially, this enzyme performs the first two steps of the pentose shunt pathway within the E.R. [7.] and is responsible for maintaining a reduced pyridine nucleotide pool (NADPH/H + ) within the E.R. to be utilized in various reticular functions including E.R. stress regulation [15]. Presently, it is unknown whether the expression and activity of the hexose 6-phosphate dehydrogenase (H6PD) change as a result of hormonal stimuli, metabolic conditions, or liver pathologies.
The liver plays a critical role in maintaining blood glucose levels within the normal range during the fed-fast cycle. During early fasting, hepatic glycogenolysis and glucose output from the organ maintains glycemia within a suitable range for brain function and metabolism. As the amount of glycogen stored within the liver (i.e. ~10% of the organ weight) is not sufficient to maintain glycemia over an extended period of time or prolonged fasting, gluconeogenesis becomes essential to synthesize glucose from amino acids, lactate and pyruvate dismissed into the circulation by skeletal muscles through glycogenolysis and glycolysis, and from glycerol dismissed by adipose tissue through lipolysis.
The complex metabolic scenario of fed to fast cycling is maintained through the antagonistic roles of insulin on one side, and glucagon, catecholamine and glucocorticoids on the other side. All these hormones modulate liver metabolism through the glucose to glucose 6-phosphate futile cycle [12], with insulin inhibiting the glucose 6-phosphatase activity and expression, and the pro-glycemic hormones increasing them.
During fed and postprandial states, elevation in blood glucose level promptly increases insulin secretion from pancreatic β-cells, which in turn, decreases glucagon release from pancreatic α-cells. The combined effect of these hormonal changes decreases hepatic glucose output and production by suppressing gluconeogenesis and glycogenolysis while increasing glucose storage within skeletal muscles via glycogenosynthesis and adipocytes via lipogenesis. In addition, insulin promotes glucose utilization in peripheral tissues through activation of glycolysis [16].

Cellular magnesium distribution
Our body absorbs minerals through food and drinks consumed daily. However, industrial food processing techniques limit to a varying extent the dietary content and intake of minerals and vitamins, making necessary the utilization of supplements. This is indeed the case of the macro mineral magnesium. Overall, Mg 2+ is the fourth most abundant cation in vertebrates and the second most abundant cation within cells after potassium. In humans, total body magnesium (Mg 2+ ) is found mostly in the bones (60-65% of total content), soft tissues and cells in general [17]. Only 1% of total body magnesium is found in the extracellular fluid, thus making serum magnesium level a poor indicator of total magnesium content and availability in the body. Of the 1% total body Mg 2+ present in the extracellular fluid, about sixty percent (60%) is free, the reminder (~33%) being bound to proteins, citrate, bicarbonate, ATP 1 and phosphate (≤7%) [18].
Whole body Mg 2+ homeostasis changes overtime. At an early stage, most Mg 2+ in the bones can readily exchange with serum, representing an optimal store to compensate for occasional dietary deficiency. As age progresses, however, the proportion of readily exchangeable Mg 2+ in the bones decreases significantly due to a change in bone crystal size with age [19]. In individuals consuming Mg 2+ enriched diet, a positive association between bone mineral density and Mg 2+ content within the erythrocytes has been reported [20].
At the cellular level, Mg 2+ is highly compartmentalized within nucleus, endoplasmic or sarcoplasmic reticulum, mitochondria, and cytoplasm [18], the only notable exception being the erythrocytes, in which Mg 2+ is merely cytoplasmic [21]. In the majority of mammalian cells examined, including the hepatocytes, total cellular Mg 2+ concentrations range from 15 to 20mM as measured by various techniques including electron X ray microprobe analysis (EXPMA), fluorescent dyes, and scanning fluorescence x-ray microscopy [21]. Total Mg 2+ concentrations between 15 and 20mM have also been measured within the nucleus, the mitochondria, and the rough endoplasmic reticulum of various cell types by EPXMA [21]. In the cytoplasm, Mg 2+ is present as a complex with ATP (~4-5mM=Mg*ATP) and other phosphonucleotides [22]. Consequently, the free Mg 2+ concentration ([Mg 2+ ] i ) within the cytoplasm and the mitochondrial matrix ranges between 0.5 and 1.2 mM [21,23], i.e. slightly below or at the concentration present in the extracellular environment. These measurements suggest that the majority of mammalian cells are near zero trans conditions as far it concerns the cellular distribution of Mg 2+ .
Despite the large amount of Mg 2+ present within the majority of mammalian cells, limited information is available about the physiological role of Mg 2+ for specific cell function. In liver cells, Mg 2+ controls ATP production by the mitochondria and its utilization by various ATPases including the Na + /K + -ATPase [21] and the reticular Ca 2+ -ATPase [21]. As a result, 90% of cytoplasmic ATP is in the form of a complex with Mg 2+ [24]. Moreover, in hepatocytes Mg 2+ is a cofactor for many enzymes involved in energy metabolism, including glycolysis and Krebs cycle [25]. The list of Mg 2+ -regulated glycolytic enzymes includes hexokinase, phosphofructokinase, aldolase, phosphoglycerate kinase and pyruvate kinase [26]. The regulation of specific enzymes or channels by Mg 2+ is not restricted to the cytoplasm occurring also in the cellular organelles in which Mg 2+ is compartmentalized. In liver mitochondria, changes in matrix Mg 2+ content regulate the activity of succinate and glutamate dehydrogenases but not αketoglutarate dehydrogenases [23]. In addition, Mg 2+ regulates the opening of the inner mitochondrial anion channel (IMAC), the permeability transition pore (PTP), K ATP -channels, and possibly the H + /K + exchanger, thus regulating the organelle volume [23]. It is still unresolved as to whether Mg 2+ is required for the adenine nucleotide translocase to operate [23]. At the level of the hepatic rough endoplasmic reticulum (R.E.R.), Mg 2+ regulates Ca 2+ uptake via the Ca-ATPase, and its release through the IP 3 receptor [27], as well as the rate of protein synthesis and dismissal into the cytoplasm via the translocon [28]. Experimental evidence suggests that Mg 2+ inversely regulates the rate of glucose 6-phosphate entry into the E.R. lumen, thus providing higher level of substrate to the glucose 6-phosphatase (G6Pase), and the hexose 6-phosphate dehydrogenase (H6PD) under conditions in which cellular Mg 2+ levels are reduced [29]. In the nucleus, changes in Mg 2+ content have been associated with inhibition of specific endonucleases and chromatin folding [21]. Less known is the function of Mg 2+ within the Golgi lumen. The recent localization of a Mg 2+ transporter in the Golgi cisternae, however, suggests a possible role of the cation in regulating protein glycosylation [23]. As for endosomal and lysosomal vesicles, nothing is known about the Mg 2+ concentration within these vesicles and its role in modulating their physiological processes.
Despite its large total concentration within the cell, Mg 2+ is not a static cation. Major Mg 2+ fluxes have been detected across the cell membrane of the hepatocyte and other mammalian cells.
Various hormones and pharmacological agents modulate total and free Mg 2+ concentrations within the hepatocyte, supporting the hypothesis that many of the metabolic changes elicited by these agents are attained by changing the concentration of Mg 2+ within the cells and/or within specific cellular compartments, which then results in the up-or down-regulation in the activity of Mg-sensitive enzymes.

Cellular magnesium transport mechanisms
The current understanding of Mg 2+ transport across the hepatocyte cell membrane indicates that Mg 2+ exits the liver cell via a Na + /Mg 2+ exchanger [30,31], which functionally depends on the physiological concentration of extracellular Na + [30,32] and the cellular level of cAMP [33], which activates the exchanger through phosphorylation [34]. Under conditions in which limited inward Na + gradient is present across the cell membrane, or Na + transport is inhibited by agents like amiloride or imipramine, cellular Mg 2+ is extruded via a Na + -independent mechanism that utilize different cations or anions in counter-transport for or co-transport with Mg 2+ , respectively [32].
As for Mg 2+ entry, hepatocytes appear to utilize predominantly the TRPM7 channel [35]. Protein kinase C (PKC) appears to regulate this channel directly via phosphorylation of its C terminus, or indirectly by removing RACK1-inhibition [36].
Several other Mg 2+ entry mechanisms have been observed to be present in liver cells [32] but it is still unclear to which extent these mechanism cooperate with TRPM7 in mediating Mg 2+ entry and in regulating hepatic Mg 2+ homeostasis.

Regulation of magnesium transport
The specific modality of operation and regulation of the various Mg 2+ transport mechanisms have been extensively addressed in recent review articles [32,[37][38][39], and we refer to those reviews for further information. For the purpose of this chapter, we will only mention that in liver cells both Mg 2+ entry and extrusion are under hormonal control. Hormones like catecholamine and glucagon, which increase cAMP level within the hepatocyte, all promote Mg 2+ extrusion by phosphorylating the Na + /Mg 2+ exchanger mentioned above [40]. Activation of α 1 -adrenoceptors by catecholamine also induces Mg 2+ extrusion. Stimulation of this class of adrenergic receptors activates PLCγ, which in turn hydrolyzes phosphatidyl-inositol bisphosphate (PIP2) to diacylglycerol and inositol 1,4,5-trisphosphate (IP 3 ). The IP 3 -induced Ca 2+ release from the endoplasmic reticulum and the subsequent capacitative Ca 2+ entry through the hepatocyte plasma membrane promote Mg 2+ extrusion from the hepatocyte, most likely by displacing Mg 2+ for Ca 2+ from organelle and cytosolic binding sites [41]. This is consistent with the observation that distinct Na + and Ca 2+ -dependent Mg 2+ extrusion mechanisms operate in the basolateral and the apical portion of the hepatocyte cell membrane, respectively [34,42]. The differential activation of the Na + -dependent (β-adrenergic receptors and glucagon receptor) and the Ca 2+ -dependent Mg 2+ extrusion mechanism (α 1 -adrenergic receptors) points to the ability of the hepatocytes to activate Mg 2+ extrusion by different modalities and circumvent possible inhibitory mechanisms. It has to be noted, in fact, that insulin pre-treatment abolishes the Mg 2+ extrusion mediated by cAMP but not that mediated via α 1 -adrenoceptor activation [43]. Conversely, hormones or agents that maximize Ca 2+ release from the ER elicit a time-dependent inhibition of α 1 -adrenergic receptor mediated Mg 2+ extrusion that leaves unaffected the extrusion occurring via β-adrenergic receptors stimulation and cellular cAMP elevation [44]. In this contest, it has to be noted that cytoplasmic free [Mg 2+ ]i modulates adenylyl cyclase activation in a variety of cell types including hepatocytes [45]. Under resting conditions, cytoplasmic [Mg 2+ ]i is insufficient to activate the adenylyl cyclase maximally. Following β-adrenoceptor or glucagon receptor stimulation the cytoplasmic Mg 2+ pool increases markedly but transiently via the release of Mg 2+ from other cellular pool (namely mitochondria and endoplasmic reticulum) promoting adenylate cyclase activity and cAMP synthesis [45]. Elevation of cytoplasmic [Mg 2+ ]i also inhibits IP 3 -induced Ca 2+ -release [27] most likely via a direct modulatory effect of Mg 2+ on the IP 3 receptor subunits.
Cellular Mg 2+ accumulation is also under hormonal regulation. Among the hormones involved in the process there are insulin and vasopressin. These hormones either counteract cAMP production by acting at the level of the β-adrenergic receptor (inhibition) or the cytoplasmic phosphodiesterase that converts cAMP to AMP (stimulation), and/or activate PKC signaling, which acts as cAMP alter ego. Due to its ubiquitous presence and abundance, the TRPM7 channel is the mechanism most likely responsible for Mg 2+ accumulation in the hepatocyte [46]. It is presently unclear whether PKC activates the channel by binding RACK1 and removing this protein from a specific site near the C terminus of the channel through which RACK1 inhibits TRPM7 conductance [36], or whether phosphorylation of the channel C-terminus is also required for full activation [46]. In the case of insulin, a direct modulatory effect on the putative Na + /Mg 2+ exchanger has also been observed [47].
Both Mg 2+ extrusion and Mg 2+ accumulation are quantitatively and timely limited processes [48,49], implying the movement of Mg 2+ from and to specific cellular compartments. The cytoplasm is but one of the cellular compartments involved in Mg 2+ transport out of the cell or into the cells [21,22], other compartments being the mitochondria and the endoplasmic reticulum [21]. This notion is supported by the observation that the co-stimulation of hepatic β 2 -and α 1 -adrenergic receptors by the mix agonist epinephrine results in a Mg 2+ extrusion that is quantitatively similar to the sum of the Mg 2+ amounts mobilized by the stimulation of each adrenoceptor class by specific agonists [40]. However, the mechanisms involved in Mg 2+ transport in-and-out of these compartments have not been fully elucidated. It is known that mitochondria accumulate Mg 2+ through Mrs2, a Mg 2+ -specific channel, the absence of which affects complex I expression and activity [50]. Less certain is whether Mg 2+ extrusion from the mitochondria occurs via the adenine nucleotide translocase [21]. As for the cytoplasm, this compartment acts as a temporary step-in between the extracellular compartment and the cellular organelles both in the extrusion and in the accumulation of Mg 2+ due to the high concentration of ATP that buffers Mg 2+ with a very high Kd (~75μM) and the presence of other phosphonucleotides and binding proteins [22]. The role of ATP is further supported by the observation that pathological conditions that decrease cellular ATP content through dysmetabolic processes (namely diabetes and alcoholic liver disease) ultimately cause Mg 2+ loss from the cell [51][52][53].

Insulin signaling in the liver
Insulin signaling is mediated by a complex and highly integrated signaling network that controls several processes including whole body glucose homeostasis. The liver is the first organ 'seen' by insulin following its release from the β-cells into the portal vein, and is responsible for the clearance of 50% of the released insulin at the first pass. Stimulation of the insulin receptor in liver cells is a key event to regulate hepatic glucose homeostasis. In addition, insulin acts indirectly on hepatic glucose homeostasis in that insulin released from β-cells inhibits glucagon release from pancreatic α-cells thus limiting the drive on hepatic gluconeogenesis. The impairment of both these processes observed in insulin resistance is linked to major health problems including type 2-diabetes.
Insulin initiates its signaling cascade by interacting with the insulin receptor on the cell surface.
Binding of insulin to the extracellular α-subunits of the insulin receptor results in a conformation change that translates to the intracellular β-subunits of the receptor. The consequent activation of the kinase domain in the β-subunits of the receptor results in the autophosphorylation of specific tyrosine residues in the intracellular β-subunits. The phosphorylated insulin receptor now recruits the insulin receptor susbstrate (IRS), which upon phosphorylation on tyrosine residues acts as a docking unit for numerous cellular proteins including the phosphatidyl inositol 3-kinase (PI3K) [54]. Recruitment of these proteins to the IRS results in their activation. Activation of PI3K results in the phosphorylation of PIP2 to PIP3 and in the subsequent activation of protein kinase B (PKB or Akt), which then phosphorylates Forkhead box protein O1 (FoxO1), preventing its translocation to the nucleus. In its un-phosphorylated state FoxO1 localizes in the nucleus, binds to the insulin response element sequence of gluconeogenesis-related genes, chiefly glucose 6 phosphatase and PEPCK, and increases their transcription rate, indirectly increasing the rate of hepatic glucose production. In its phosphorylated state, FoxO1 is unable to translocate to the nucleus and to activate the gluconeogenesis-related genes. Inhibition of FoxO1 could then improve hepatic metabolism in cases of insulin resistance and metabolic syndrome [55].

Role of Mg 2+ on insulin receptor activation and signaling
The human insulin receptor homodimer is heavily glycosylated and contains a total of 19 predicted N-linked glycosylation sites in each monomer. The presence of sialic acid residues on molecules and cells is critical to their biological function and the presence of sialic acid residues on glycoproteins is partly responsible for the binding and transport of molecules, masking of the surface charge, aggregation and shape of cells [56]. Most recently, neuraminidase-1(Neu-1) an enzyme responsible for hydrolyzing sialic acid (neuraminic acid), has been associated with the positive regulation of insulin signaling [57]. Neu-1 is transported to the cell surface and gets involved in the regulation of cell signaling. Insulin binding to its receptor rapidly induces interaction of the glycan chains of the receptor with Neu-1 which hydrolyzes sialic acid residues in the glycan chains of the receptor consequently inducing activation of the insulin receptor. Impaired insulin-induced phosphorylation of Akt, thus identifies Neu1 as a novel component of the signaling pathways of energy metabolism and glucose uptake. Insulin binding to the insulin receptor has been shown to induce the interaction of the receptor with a pool of Neu-1 near the cell surface [57]. Also, insulin signaling is partially impaired in tissues of Neu-1-deficient mice [3], and desialylation of the insulin receptor by Neu1 promote the receptor activation [57]. While CaCl 2 has no significant effect on human liver neuraminidase activity, 10mM MnCl 2 or MgCl 2 shows a mild stimulatory effect (112% and 125% over control activity, respectively) [56].
Additional experimental evidence indicates that Mg 2+ is required for the activated insulin receptor to phosphorylate IRS [54].

Magnesium and hepatic glucose metabolism
In liver cells, adrenergic stimulation of α 1 -and β-adrenergic receptors, and glucagon receptors elicit a Mg 2+ extrusion that is associated with activation of glycolysis and glucose output on functional and temporal bases [40]. Although the nature of this association requires further clarification, it is fairly evident that conditions that limit the amplitude of Mg 2+ extrusion decrease the amount of glucose outputted from liver cells, and vice versa [40]. This association is further supported by several pieces of observation. Overnight starvation, which depletes the liver of its glycogen content, decreases total hepatic Mg 2+ content by 10-15% [58] rendering liver cells unresponsive to any subsequent adrenergic stimulation [58]. Both type-1 and type-2 diabetes present with a marked decrease in hepatic Mg 2+ content [59], and treatment with the anti-diabetic drug metformin, which operates predominantly on liver metabolism, increases intra-hepatic Mg 2+ content [60]. The loss of hepatic Mg 2+ observed under diabetic conditions depends on the enhanced phosphorylation of the Na + .Mg 2+ exchanger [61,62], and can be attenuated to a significant extent by the presence of glycogen, amylopectin, or glucose within liver plasma membrane vesicle [62].
The functional association between Mg 2+ and glucose is also observed for Mg 2+ accumulation. Insulin, one of the hormones involved in Mg 2+ accumulation, is also responsible for glucose accumulation and conversion to glycogen [58]. Following insulin administration, Mg 2+ accumulation is directly proportional to the amount of glucose present in the system [63]. Conversely, decreasing Mg 2+ content in the extracellular system decreased the accumulation of glucose within the cells [40,63]. In part, the limited accumulation of glucose into insulinstimulated cells in the presence of low extracellular Mg 2+ concentration can be explained with the reduced activation of the insulin receptor occurring in these cells as Mg 2+ is essential for the proper autophosphorylation of the insulin receptor and the subsequent recruitment of the insulin receptor substrate to the activated receptor [54]. All together, these pieces of evidence and observation support an essential role of Mg 2+ in glucose regulation and pose for the cation as an important player in the onset and development of insulin resistance and diabetes in human patients.

Magnesium and enzyme activation in glucose metabolism
The physiological role of magnesium is principally related to enzyme activity. All enzymes utilizing ATP require Mg for substrate formation. Intracellular free magnesium also acts as an allosteric activator of enzyme action including critical enzyme systems such as adenylate cyclase, phosphofructokinase, phospholipase C, and Na + /K + -ATPase [64]. Magnesium is an enzyme substrate (ATPMg, GTPMg) to enzymes such as ATPase (Na + , K + ATPase, Ca 2+ ATPase), cyclases (adenylate cyclase, guanylate cyclase), and the kinases (hexokinase, protein kinase) [64]. Recently, our laboratory has provided evidence that Mg 2+ also modulates the amount of glucose 6-phosphate being routed into the endoplasmic reticulum (E.R) to be hydrolyzed to glucose plus Pi by the glucose 6-phosphatase, or to be converted to 6-phosphogluconolactone by the hexose 6-phosphate dehydrogenase, the reticular version of the G6PD. Moreover, our laboratory has provided significant evidence that both glucose and Mg 2+ homeostasis are altered under pathological conditions such as diabetes [61] and alcoholic liver disease [65].
Many of the enzymes of glycolytic pathway that utilizes glucose have a requirement for Mg 2+ [26] and utilize MgATP 2-as a cofactor [66]. The Km values for Mg 2+ in the glycolytic enzymes of the human erythrocyte are between 1 and 2.3 mM for hexokinase, 0.025 mM for phosphofructokinase (PFK), 0.3 mM for phosphoglycerate kinase (PGK), and 1 mM for pyruvate kinase [26]. Magnesium ions (Mg 2+ ) and MgATP 2-regulate the most important glycolytic enzymes, namely hexokinase, phosphofructokinase, aldolase, phosphoglycerate kinase, and pyruvate kinase [66]. Glucokinase (Hexokinase IV or D), an enzyme expressed predominantly in liver and pancreatic β-cells of vertebrates, shows marked deviations from Michaelis-Menten kinetics when the glucose concentration is varied at a constant MgATP 2-concentration, but shows no deviations from Michaelis-Menten kinetics with respect to MgATP 2- [26,67]. Compared to the other hexokinase isoenzymes, this isoform has a low affinity for glucose (Table  1). Maximum binding of glucokinase and its regulatory protein to the hepatocyte matrix occurs at low [glucose] (<5mM) in a Mg 2+ -dependent manner ( Table 2, [68]). The regulatory protein binds to the hepatocyte matrix with ionic characteristics similar to those of glucokinase but, unlike glucokinase, it does not translocate from the binding site. Since the binding of glucokinase to its regulatory protein is associated with a decrease in the affinity of the enzyme for glucose, the bound enzyme in the presence of Mg 2+ represents an inactive state and the translocated enzyme a more active state [69].

Magnesium and gluconeogenesis
Gluconeogenesis is the process of glucose synthesis from non-carbohydrate precursors. Phosphoenolpyruvate Carboxy kinase (PEPCK), fructose1,6-bisphosphatase (F1,6BP), pyruvate carboxylase and glucose-6-phosphatase (G6Pase),catalyze irreversible reactions in the pathway and have lower activities compared to the other enzymes in the pathway and are thus considered rate limiting for glucose synthesis. Experiments by McNeill et al, [70] suggest that magnesium deficiency alters PEPCK by affecting secretion of pancreatic hormones. Of these four enzymes, Mg 2+ is required by three, that is, pyruvate carboxylase, PEPCK, and F1, 6BP reactions. Though hormones such as insulin, glucagon, glucocorticoids and epinephrine influence the key enzyme activities of gluconeogenic enzymes, Mg 2+ plays a role in the secretion of all these hormones [70]. Thus in Mg 2+ deficiency, enzyme activities may change, as a result of altered circulating levels of one or more hormones. In this study like in earlier studies, an increase in PEPCK activity was observed in magnesium deficient rats making Mg 2+ deficiency a possible contributing factor to the maintenance of low insulin levels an increased PEPCK in diabetes [70].

Conclusion
In the last two decades, our understanding of the importance of Mg 2+ ions for numerous cell and body functions has increased significantly. The initial experimental evidence has been corroborated to a significant extent in clinical conditions such as diabetes, alcoholism, and dysendocrinopathies. n the case of liver cells, we have moved from the initial observation that Mg 2+ is abundantly represented within the hepatocyte as a whole to the notion that the cation's homeostasis is controlled by hormones, which promotes the movement of Mg 2+ in-and-out of the cell membrane to support and regulate specific liver metabolic functions. The observation that Mg 2+ is highly compartmentalized within cellular compartments and organelles support the notion that the cation plays a key role in regulating enzymes, channel activities, and metabolic processes within each of these organelles. Figure 2 recapitulates the relevance of Mg 2+ for the regulation of glucose homeostasis and bioenergetics within the hepatocyte. In the cytoplasm, Mg 2+ regulates glucokinase and glycolytic enzymes but also ATP utilization. In the mtiochondria, Mg 2+ regulates mitochondrial dehydrogenases and pyruvate dehydrogenase by promoting the activity of the pyruvate dehydrogenase phosphatase, responsible for dephosphorylating the enzyme to its active conformation. In the endoplasmic reticulum (ER), Mg 2+ regulates protein synthesis and the entry of glucose 6 phosphate (G6P), the limiting step for the utilization of this substrate by the glucose 6-phosphatase (G6Pase) and the hexose 6phosphate dehydrogenase (H6PD), the reticular variant of the cytosolic glucose 6 phosphate dehydrogenase (G6PD). The oxidation of G6P by the H6PD generates high levels of NADPH within the ER lumen to be used for other metabolic processes within the organelle and in the rest of the cell, including fatty acid synthesis and cholesterol synthesis. Far from being complete, the picture is a dynamic scenario in need to further clarification and study in the years to come.