OBJECTIVE—The counterregulatory response to insulin-induced hypoglycemia is mediated by the ventromedial hypothalamus (VMH), which contains specialized glucosensing neurons, many of which use glucokinase (GK) as the rate-limiting step in glucose's regulation of neuronal activity. Since conditions associated with increased VMH GK expression are associated with a blunted counterregulatory response, we tested the hypothesis that increasing VMH GK activity would similarly attenuate, while decreasing GK activity would enhance the counterregulatory response to insulin-induced hypoglycemia.

RESEARCH DESIGN AND METHODS—The counterregulatory response to insulin-induced hypoglycemia was evaluated in Sprague-Dawley rats after bilateral VMH injections of 1) a GK activator drug (compound A) to increase VMH GK activity, 2) low-dose alloxan (4 μg) to acutely inhibit GK activity, 3) high-dose alloxan (24 μg), or 4) an adenovirus expressing GK short hairpin RNA (shRNA) to chronically reduce GK expression and activity.

RESULTS—Compound A increased VMH GK activity sixfold in vitro and reduced the epinephrine, norepinephrine, and glucagon responses to insulin-induced hypoglycemia by 40–62% when injected into the VMH in vivo. On the other hand, acute and chronic reductions of VMH GK mRNA or activity had a lesser and more selective effect on increasing primarily the epinephrine response to insulin-induced hypoglycemia by 23–50%.

CONCLUSIONS—These studies suggest that VMH GK activity is an important regulator of the counterregulatory response to insulin-induced hypoglycemia and that a drug that specifically inhibited the rise in hypothalamic GK activity after insulin-induced hypoglycemia might improve the dampened counterregulatory response seen in tightly controlled diabetic subjects.

Hypoglycemia is a major complication of insulin therapy for diabetes, and the incidence of hypoglycemia has increased as clinicians have attempted to maintain tighter control of blood glucose levels in diabetic subjects (1). During hypoglycemia, declining glucose levels are detected by peripheral glucosensors and by specialized glucosensing neurons within select areas of the brain (2). A subset of these glucosensing neurons reside in the ventromedial hypothalamus (VMH), a brain area that includes the arcuate hypothalamic nucleus (ARC) and ventromedial hypothalamic nucleus (VMN). Many previous studies have confirmed the importance of the VMH in mediating the counterregulatory response to hypoglycemia (3,4) and suggest that the glucosensing neurons in this area might be critical regulators of these responses.

Glucokinase (GK) (hexokinase IV) is a likely candidate as a regulator of glucosensing in VMH neurons (510). In vitro, inhibiting GK activity with either RNA interference (RNAi) or with a variety of drugs reduces the sensitivity of VMN glucosensing neurons. On the other hand, pharmacological activation with Compound A increases their sensitivity to glucose (58,10). Also, pharmacological inhibition of GK activity with alloxan in the raphe pallidus/obscurus produces an avid feeding response, whereas prior application of higher toxic doses of alloxan completely abolishes this response (11). Importantly, situations in which VMH GK mRNA is increased are associated with a dampened counterregulatory response to hypoglycemia (7,1115). In addition, prior hypoglycemia increases the sensitivity of some glucosensing neurons in parallel with an increase in GK mRNA within the VMN and arcuate hypothalamic nuclei (15). This caused them to be less activated at the higher glucose levels at which control neurons first responded. This result suggested that such neurons from animals with prior insulin-induced hypoglycemia and elevated GK activity would not trigger as great a counterregulatory response when glucose levels fell to comparable levels as naive animals during subsequent bouts of insulin-induced hypoglycemia. Despite such evidence, no studies have directly assessed the effects of manipulating VMH GK activity on the counterregulatory response to hypoglycemia. The current studies were carried out to test the hypothesis that the increasing VMN GK activity in vivo with a GK activator drug would dampen, whereas reducing GK mRNA expression and/or activity with drugs or RNAi would enhance, the counterregulatory response to insulin-induced hypoglycemia. Proof of these postulates would suggest that either preventing the increase in hypothalamic GK activity after single bouts of insulin-induced hypoglycemia or specifically lowering GK activity before subsequent bouts would provide a potential therapeutic target for improving the dampened counterregulatory response seen in tightly controlled diabetic subjects (16).

Male Sprague-Dawley rats (Charles River Labs) were used with starting weights between 300 and 450 g. They were housed at 22–24°C on a 12:12 h light:dark cycle with lights off at 2000 and were provided with ad libitum Purina lab chow (#5001) and water. All procedures were approved by the East Orange VA Medical Center Institutional Animal Care and Use Committee.

Placement of hypothalamic cannulae and vascular catheters.

Experimental groups contained 6–10 rats each. Animals were anesthetized with chloropent (pentobarbital, chloral hydrate, magnesium sulfate) and given buprenorphine postoperatively. For VMH cannula placements, 26-gauge guide cannulae or injection cannulae were angled at 20 ° to the vertical aiming at the junction between the ARC and VMN (−2.9 mm bregma, ±3.7 mm midline and −8.5 mm dura). These injection sites are referred to as the “VMH,” since preliminary injections of 1.0 μl cresyl violet dye into this site spread to both the ARC and VMN. After 2–3 days, animals were again anesthetized, and jugular venous catheters were implanted and filled with heparinized polypropylene glycol by previously described methods (13). Animals were allowed 6–7 days with daily handling to recover their preoperative body weight. Terminally, all microdialysis probe and injection cannulae placements were verified histologically.

Drug effects on the counterregulatory response to insulin-induced hypoglycemia.

Rats (n = 6–10 per group) were provided with ∼45 g food at lights off (2000). At 1400 on the following day, any remaining food was removed, intracranial injection cannulae attached to tubing were inserted into the guide cannulae, and venous catheters were attached to tubing for blood sampling. At 1500, either vehicle (1.0 μl saline, pH 3.0, or 1.0 μl 1% DMSO) or drug (4 μg alloxan in saline or 0.5 nmol Compound A in 1% DMSO, respectively) were injected bilaterally over 5 min simultaneously at 0.2 μl/min. Compound A (Merck Research Labs) is an analog of a previously characterized GK activator (17). After another 1 h (1600), 0.6 ml baseline blood was obtained, followed by infusion of 5 units/kg i.v. pork insulin (Lilly). Blood samples (0.6 ml) were drawn at 30, 60, 90, and 120 min thereafter for subsequent measurements of plasma glucose, norepinephrine, epinephrine, and glucagon. After each sampling, plasma volume was replaced with saline, and washed red cells from the same rat's previous sampling period were returned to maintain red cell volume as previously described (13).

At 24 h after the injection of insulin, those rats that had received low-dose VMH alloxan (4-μg) injections plus insulin-induced hypoglycemia were then injected with 24 μg alloxan in 1 μl saline, and those that had received saline were again injected with saline bilaterally through the VMH cannulae. Although low-dose alloxan produces a pharmacological inhibition of GK in vitro (8), the higher dose causes a permanent reduction in GK mRNA (11). After an additional 6–7 days, both groups were subjected to a second bout of insulin-induced hypoglycemia with blood collection over 120 min. Within 1–3 days, brains were removed for histological verification of injection sites, and micropunches of the VMH (ARC plus VMN) for real-time quantitative PCR were taken according to our previous description (8).

Adenoviral vector injections.

Adenoviruses expressing short hairpin RNA (shRNA) for either GK or a scrambled sequence were grown and quantitated according to the methods of Bain et al. (18). A total of 15 million infectious units of virus in 1.0 μl were injected bilaterally in both the ARC (−9.6 mm dura) and VMN (−9.3 mm dura) at 5.9 mm anterior to the intra-aural line and 0.3 mm lateral to the midline. After preliminary studies established that maximal (65–75%) selective reductions in GK mRNA occurred between 7–10 days after adenoviral injections, additional rats were tested for their counterregulatory response to insulin-induced hypoglycemia 7–10 days after comparable VMH adenoviral injections. They were killed for assessment of VMH mRNA expression at 14 days after viral injections.

Assays for GK enzyme activity.

Brains from naive rats (n = 8) were killed in the semi-fasted state (see above), and the VMH was dissected bilaterally, homogenized, pooled, and aliquoted into quadruplicate samples. These samples were assayed immediately for GK enzyme activity at concentrations of 0.5, 1, 5, 10, 20, and 50 mmol/l glucose. Additional samples were incubated in 0.5, 1.0, and 10 mmol/l glucose in the presence of the GK activator, Compound A (0.5 mmol/l in 0.3% DMSO) (8). No higher glucose concentrations were used with Compound A, since it was anticipated that GK activity would be maximal at 10 mmol/l glucose in its presence.

Freshly dissected VMH tissue from other naive rats (n = 6) was pooled, aliquoted, and incubated at 0.5 or 20 mmol/l glucose in the presence or absence of 4 nmol/l alloxan. We previously showed that this concentration of alloxan reversibly alters the responsiveness of dissociated VMN glucosensing neurons to glucose (7,8) and is 1,000-fold higher than the dose that half-maximally inhibits purified GK (19). A third set of rats were injected bilaterally in the VMH with adenovirus expressing GK shRNA or nonsense RNA (18) (n = 6 per group) and were killed 10 days later. VMH punches were removed for GK enzyme assay and were frozen at −80°C for up to 14 days for GK assays.

GK enzyme activity in VMH tissue was assessed using a fluorometric assay that monitors NADPH production, by minor modifications of the methods of Zelent et al. (20). Activity measured at 0.5 mmol/l glucose (saturating concentration for hexokinase I) was taken as representing hexokinase I activity and that at 20 mmol/l (near-saturating glucose concentration for GK) minus that at 0.5 mmol/l as representing GK activity. GK activity was expressed as relative units (Compound A studies) or as microunits per microgram protein, where authentic hexokinase I (Sigma) was used as an internal standard (GK shRNA and acute alloxan studies). Protein levels were measured using the BCA Protein Assay Kit (Thermo Scientific).

Assays of blood and brain tissues.

Plasma catecholamines were measured using high-performance liquid chromatography with electrochemical detection as previously described (13). Glucagon levels were measured by radioimmunoassay (Linco) and glucose by automated glucose analyzer (Analox). The brains of rats treated with high-dose (24 μg) alloxan and with adenoviral vectors had the VMH micropunched and assayed for real-time quantitative PCR by previously described methods (8,21).

Statistical analysis.

Serial blood determinations were analyzed by repeated measures one-way ANOVA with post hoc correction by Bonferroni's test. Area under the curve (AUC) was calculated using the trapezoidal method for the change in glucose, norepinephrine, epinephrine, and glucagon levels from baseline. For single measures, groups were compared by an unpaired t test.

Effects of increasing VMH GK activity on the counterregulatory response.

Initial in vitro studies showed that there was a concentration-dependent increase in calculated GK activity between 1 and 50 mmol/l glucose (Fig. 1), where GK activity was taken as 0 at 0.5 mmol/l glucose, the concentration at which all of the measured hexokinase activity was assumed to be due to hexokinase I. The level of calculated GK activities at 10 and 20 mmol/l glucose were 12 and 19% of total hexokinase activity, respectively (after subtracting presumptive hexokinase I activity at 0.5 mmol/l glucose). Thus, maximum calculated hexokinase I activity at 20 mmol/l glucose comprised a minimum of 81% of total VMH hexokinase activity. However, in the presence of the GK activator, Compound A, GK activity was reliably detected at 0.5 and 1 mmol/l glucose after subtracting for the hexokinase activity at 0.5 mmol/l glucose measured in the absence of the drug (Fig. 1). At 10 mmol/l glucose, calculated GK activity was sixfold greater in the presence of Compound A than in its absence. Finally, no GK activity was found in samples of frontal cerebral cortex (data not shown). This is in keeping with the absence of GK mRNA in this brain area, as assessed by both RT-PCR (9) and in situ hybridization (7).

When injected bilaterally into the VMH in vivo, 0.5 nmol Compound A had no effect on baseline glucose, epinephrine, norepinephrine, or glucagon levels compared with vehicle controls (data not shown). Also, Compound A produced no overall difference in glucose AUC (Table 1) or overall glucose levels during the entire 120 min of testing as assessed using a one-way ANOVA for repeated measures. Although this lack of difference precludes valid post hoc analysis of individual time points, the glucose nadir during insulin-induced hypoglycemia was 30% higher than vehicle controls at the 90-min time point in rats injected with Compound A (Fig. 2). While glucose values were marginally (and not significantly) higher at 30 min (22%) and 60 min (4%) in rats treated with Compound A, the corresponding epinephrine values were reduced by 47 and 57%, the norepinephrine values by 46 and 48%, and glucagon values by 34 and 38%, respectively. Also, the AUC over 120 min for epinephrine, norepinephrine, and glucagon levels were 50, 62, and 50% lower in rats injected with Compound A (Table 1) and the epinephrine, norepinephrine, and glucagon maxima were 60% (P = 0.01), 30% (P = 0.03), and 31% (P = 0.04) lower than in vehicle-injected rats over the 120 min after insulin injections (Fig. 2). Thus, although there might have been a somewhat reduced stimulation of the counterregulatory response due to slightly higher glucose levels, the time point by time point and overall reductions for epinephrine, norepinephrine, and glucagon were much greater than the marginal increases in glucose. Whereas this issue might have been clarified by clamp studies, the overall results strongly suggest that increasing VMH GK activity with infusions of Compound A reduces the counterregulatory responses to insulin-induced hypoglycemia to a much greater degree than they increase plasma glucose levels.

Effects of reducing VMH GK mRNA and “activity” with alloxan.

In vitro incubation of pooled VMH tissue samples with 4 nmol/l alloxan produced a 19% reduction in measurable GK activity (control: 0.95; alloxan: 0.77 μU/μg protein). We next injected rats bilaterally in the VMH with 4 μg alloxan per side. This is four times the dose that stimulates a robust food intake response when injected into the raphe pallidus/obscurus in the medulla (11). These acute alloxan injections had no effect on baseline levels of glucose, epinephrine, norepinephrine, or glucagon (data not shown). However, the acute injections increased peak epinephrine levels by 27% (P = 0.05; Fig. 3) and epinephrine AUC values by 28% (Table 2) compared with saline-injected controls during insulin-induced hypoglycemia. On the other hand, there was no effect of these acute alloxan injections on the decline in glucose or rise in norepinephrine or glucagon levels during insulin-induced hypoglycemia.

At 6–7 days after bilateral VMH injections of 24 μg alloxan, there were no differences in baseline levels of glucose, epinephrine, norepinephrine, or glucagon (data not shown). However, peak epinephrine levels were increased by 28% (P = 0.04) and AUC by 44% compared with controls over the 120 min of insulin-induced hypoglycemia (Fig. 4, Table 2). As with the pharmacological dose of alloxan, there were no effects on glucose, norepinephrine, or glucagon levels. Rats were killed at 7–10 days after the initial injections of 24 μg alloxan, and GK mRNA levels were 74% of saline-injected levels (Table 2). This was presumed to be due to the toxic effect of alloxan on glucosensing neurons at high dosages (11).

Reduction of VMH GK mRNA expression with GK shRNA.

Initial studies demonstrated a progressive reduction of VMH GK mRNA in the GK shRNA-injected rats, which reached a nadir at 25–35% between 7 and 10 days and rose to 73% of that in scrambled mRNA rat levels at 14 days after injection (Fig. 5). These effects were selective for GK mRNA, since neither hexokinase I nor cyclophilin mRNA expression was affected (Fig. 5). However, although injections of adenovirus expressing GK shRNA into the VMH reduced GK mRNA expression by 75% after 10 days, similar injections had no effect on measurable GK activity in VMH tissues taken from additional rats (n = 6 per group) injected with this virus (2.12 + 0.31 μU/mg protein) compared with that expressing scrambled RNA (2.21 + 0.52 μU/mg protein). When tested at baseline, prior GK shRNA injections had no effect on baseline levels of glucose, epinephrine, norepinephrine, or glucagon at 10–14 days (data not shown).

Despite our inability to demonstrate a decrease in GK activity in the face of a 75% decrease in VMH GK mRNA, rats injected with GK shRNA had peak epinephrine levels that were increased by 40% (P = 0.01; Fig. 6), AUC levels that were increased by 25%, and peak glucagon levels that were increased by 26% (P = 0.05) compared with values in rats injected with scrambled RNA rats (Table 3). On the other hand, there was no effect on glucose, norepinephrine levels, AUC, or glucagon AUC values between the groups during insulin-induced hypoglycemia.

The current studies represent the first evidence that direct manipulation of VMH GK mRNA expression or activity can alter the counterregulatory response to insulin-induced hypoglycemia. Based on our previous studies demonstrating that situations in which increased VMH GK mRNA were associated with a reduced responsiveness of VMH glucosensing neurons (15) and a dampened counterregulatory response to comparable levels of glucose during insulin-induced hypoglycemia compared with controls (7,12,15), we postulated that increasing VMH GK activity pharmacologically would also blunt these responses. We found that the GK activator drug, Compound A, produced a marked activation of GK activity in vitro and led to a substantial blunting of the epinephrine, norepinephrine, and glucagon responses to insulin-induced hypoglycemia when injected into the VMH in vivo. On the other hand, reducing VMH GK mRNA or activity by three different methods had a more modest and relatively selective effect on the counterregulatory response. Pharmacological inhibition of GK with low-dose alloxan (7,10,19) led to a 19% reduction in measurable GK activity in vitro, but acute VMH injection of a comparably low-dose alloxan produced only a modest and selective increase in the epinephrine response to insulin-induced hypoglycemia in vivo. When VMH GK mRNA was chronically reduced by 26% after high-dose alloxan injections, the response to insulin-induced hypoglycemia was the same as that seen after acute alloxan injections. Reducing VMH GK mRNA levels further to 65–75% of control levels by injecting adenovirus expressing GK shRNA into the VMH produced a comparable increase in the epinephrine peak and AUC to both alloxan studies. In addition, there was a modest increase in the peak, but not overall, AUC glucagon response to insulin-induced hypoglycemia. For unexplained reasons, these injections failed to reduce GK activity in vitro, despite the large decreases in GK mRNA they caused.

Most importantly, the GK activator results support our hypothesis that the increase in VMH GK mRNA expression that follows a prior bout of hypoglycemia is likely to be associated with increased GK activity, which contributes to the blunting of the counterregulatory response during subsequent bouts (7,15). Further support for this hypothesis was provided by our previous studies in which a reversible increase in VMH GK mRNA after intraventricular injections of alloxan was paralleled by a time-linked reversible blunting of the counterregulatory response (12). On the other hand, the much less robust increase in the counterregulatory response when VMH GK mRNA or activity were reduced may be because brain glucose levels fall at the lowest end of the GK activity curve during hypoglycemia (0.3–0.5 mmol/l) (19,20,22). Thus, experimentally reducing VMH GK activity to levels below its already very low (and almost immeasurable) physiological levels would be less likely to affect the counterregulatory response than would increasing VMH GK activity. Clinically, this suggests that a potential intervention that could inhibit GK mRNA or activity during a single bout of insulin-induced hypoglycemia would have only a limited effect on enhancing the counterregulatory response. However, since prior insulin-induced hypoglycemia raises GK mRNA levels, such an intervention might provide a means of preventing the downregulation of the counterregulatory response that occurs after recurrent bouts of insulin-induced hypoglycemia (16).

It is worth noting that our inability to reliably measure decreased GK activity, even in the face of 75% reductions in GK mRNA over several days, is likely because the threshold sensitivity of virtually all GK assays is just below the normally low amounts of GK found in the VMH by either mRNA or protein expression (7,9,2326). It is unlikely that the failure to document a decrease in GK activity in the face of large reductions in GK mRNA was due to prolonged stability of GK protein. We previously demonstrated that similar reductions of GK mRNA using RNA interference in vitro virtually abolished VMH neuronal glucosensing within 72 h (8), suggesting that GK protein activity could be completely eliminated in no more than 72 h. In both the high-dose alloxan and shRNA studies, we tested animals well after this time frame. The problem is more likely because GK activity is not measured directly by available assays. Instead, it is estimated as the difference between hexokinase activity at 0.5 mmol/l glucose, at which hexokinase I, the predominant brain hexokinase (24,26), should be fully saturated, and 20 mmol/l glucose, at which GK should be nearly saturated (19,20). In fact, it is questionable if previous or current estimates suggesting that GK activity comprises upwards of 20% of hypothalamic hexokinase activity are correct (24,26). GK mRNA levels are extremely low in whole VMH tissue. It is present only in glucosensing neurons in the VMH, and these neurons make up no more than 20–30% of all neurons in the VMH (7,9,10). Additionally, the level of VMH GK protein is below the detection threshold for immunocytochemical techniques (25) (personal observations, B.E.L., A.A.D.-M.). Thus, available assays are likely to overestimate true VMH GK activity.

The fact that manipulations of VMH GK did not more markedly affect the counterregulatory response to insulin-induced hypoglycemia is likely because GK is only one of several potential mechanisms that regulate the counterregulatory response to hypoglycemia. First, not all glucosensing neurons in the VMH express GK (10). Other regulators of neuronal glucosensing such as AMP-activated protein kinase have been identified in glucosensing neurons (27), and manipulation of its activity in the VMH alters the counterregulatory response to hypoglycemia (28). Given its exquisite sensitivity to low-energy states, it is likely that this enzyme, which is downstream of GK, may take on a very important role in altering neuronal activity during hypoglycemia (29). Second, a number of neurons that are not necessarily glucosensing in type may be influenced by changes in local neurotransmitters such as γ-aminobutyric acid (GABA) (30,31), glutamate (32), and norepinephrine (33) or neuropeptides such as corticotrophin-releasing hormone and urocortin (34), which are altered during insulin-induced hypoglycemia. Finally, although glucose-excited neurons clearly use the ATP-sensitive K+ channel to mediate their glucosensing (7,10,35,36), these channels are present on nonglucosensing neurons (10). This may cause such neurons to become inactivated at the low ambient glucose levels present in the brain during hypoglycemia, even though they have no intrinsic glucosensing properties (37). Thus, both nonglucosensing and glucosensing neurons in the VMH that do not use GK may be called into play to produce the full counterregulatory response to insulin-induced hypoglycemia.

In the current and previous studies, we used bolus injections of insulin to produce hypoglycemia in an attempt to mimic the responses seen in diabetic subjects who inadvertently become hypoglycemic after injecting insulin. In our hands, this provides a reproducible and self-limited period of hypoglycemia and associated increases in plasma epinephrine, norepinephrine, and glucagon levels (13,14). A major disadvantage of this method is that it obviates the measurement of changes in peripheral glucose utilization, as can be assessed using a hyperinsulinemic-hypoglycemic clamp (32,34,38,39). However, this should not be a problem for evaluating the results of our current studies, since glucose levels generally fell to the same levels, regardless of whether there were changes in epinephrine, norepinephrine, and glucagon responses. Even where nadir glucose levels were higher after VMH injections of Compound A, this elevation occurred only at the 90-min time point during insulin-induced hypoglycemia, whereas the maximal reductions in epinephrine, norepinephrine, and glucagon occurred during the first 60 min of the hypoglycemic response. Thus, it is likely that the differences in counterregulatory response to both increasing and decreasing putative GK activity in the VMH were a response to those manipulations and not to differences in the levels of glucose attained.

In conclusion, GK is an important mediator of neuronal glucosensing in the VMH in vitro. Here we demonstrate that in vivo manipulation of GK mRNA expression and enzyme activity in the rat VMH alters the counterregulatory response to insulin-induced hypoglycemia, as measured by plasma glucose, epinephrine, norepinephrine, and glucagon levels. The most profound effect was a blunting of these counterregulatory responses when VMH GK activity was increased, as presumably occurs after prior bouts of insulin-induced hypoglycemia (7). The low levels of GK expression in the normal brain (7,2326) and the low affinity of GK in relationship to the levels of brain glucose, particularly during insulin-induced hypoglycemia (22), probably explain the smaller effect of lowering than raising VMH GK activity. In addition, as opposed to increased VMH GK (7,12), there are no known physiological or pathological states in which decreased VMH GK activity or mRNA expression have been documented. Whereas GK is a critical mediator of neuronal glucosensing (5,7,8,10), there are clearly other mediators of function in glucosensing and nonglucosensing neurons that are affected by the low levels of brain glucose that occur during insulin-induced hypoglycemia. Thus, it may be that, because of its low affinity for glucose compared with hexokinase I, GK functions best as a neuronal glucosensing regulator when brain glucose levels are within the normal physiological range. However, during severe hypoglycemia, GK may play a less prominent role in stimulating the counterregulatory response than enzymes such as AMP-activated protein kinase, which are highly sensitive to low levels of ATP. According to this scenario, only when its expression is increased after a prior bout of hypoglycemia would GK play a significant role in dampening the counterregulatory response by increasing the flux through the glycolytic pathway in glucosensing neurons. Our current challenge is to define further the cellular machinery by which neurons sense and respond to low levels of glucose and the neuronal networks involved in the counterregulatory response to hypoglycemia. Our current and previous findings (15) suggest that a pharmacological intervention that could provide either primary prevention or secondary lowering of the elevated hypothalamic GK levels that occur after even a single bout of insulin-induced hypoglycemia would enhance the dampened counterregulatory response to recurrent bouts of hypoglycemia seen in diabetic subjects with tight blood glucose control.

FIG. 1.

In vitro assay of VMH GK activity in the presence or absence of the GK activator, Compound A. Pooled samples of VMH tissue were incubated with 0.5, 1, 5, 10, 20, or 50 mmol/l glucose. GK activity was defined as the difference between hexokinase I activity at 0.5 mmol/l glucose and total activity at each of the increasing glucose concentrations >0.5 mmol/l glucose. Additional samples were incubated with 0.5, 1, and 10 mmol/l glucose in the presence of Compound A (10 mmol/l). Activity is expressed as arbitrary units of NADPH formed (20).

FIG. 1.

In vitro assay of VMH GK activity in the presence or absence of the GK activator, Compound A. Pooled samples of VMH tissue were incubated with 0.5, 1, 5, 10, 20, or 50 mmol/l glucose. GK activity was defined as the difference between hexokinase I activity at 0.5 mmol/l glucose and total activity at each of the increasing glucose concentrations >0.5 mmol/l glucose. Additional samples were incubated with 0.5, 1, and 10 mmol/l glucose in the presence of Compound A (10 mmol/l). Activity is expressed as arbitrary units of NADPH formed (20).

FIG. 2.

Effect of bilateral VMH injections of 0.5 nmol Compound A on plasma glucose, epinephrine, norepinephrine, and glucagon levels during insulin-induced hypoglycemia. Data are means ± SE. *P < 0.05 or less scrambled vs. GK shRNA by post hoc t test after repeated measures ANOVA demonstrated a significant inter-group difference.

FIG. 2.

Effect of bilateral VMH injections of 0.5 nmol Compound A on plasma glucose, epinephrine, norepinephrine, and glucagon levels during insulin-induced hypoglycemia. Data are means ± SE. *P < 0.05 or less scrambled vs. GK shRNA by post hoc t test after repeated measures ANOVA demonstrated a significant inter-group difference.

FIG. 3.

Effects of bilateral VMH injections of 4 μg alloxan on the response to insulin-induced hypoglycemia. Semi-fasted rats (n = 6–10 per group) were injected bilaterally in the VMH with 1 μl saline or saline containing alloxan and then injected with 5 units/kg insulin i.v. *P = 0.05 or less for values at a given time period by post hoc t test after repeated measures ANOVA demonstrated significant inter-group differences. Data are means ± SE.

FIG. 3.

Effects of bilateral VMH injections of 4 μg alloxan on the response to insulin-induced hypoglycemia. Semi-fasted rats (n = 6–10 per group) were injected bilaterally in the VMH with 1 μl saline or saline containing alloxan and then injected with 5 units/kg insulin i.v. *P = 0.05 or less for values at a given time period by post hoc t test after repeated measures ANOVA demonstrated significant inter-group differences. Data are means ± SE.

FIG. 4.

Effect of injecting a “cytotoxic” dose of alloxan (24 μg) bilaterally into the VMH on the counterregulatory response to insulin-induced hypoglycemia 6–7 days later. *P = 0.05 or less by post hoc t test for values at a given time period after repeated measures ANOVA demonstrated significant inter-group differences. Data are means ± SE.

FIG. 4.

Effect of injecting a “cytotoxic” dose of alloxan (24 μg) bilaterally into the VMH on the counterregulatory response to insulin-induced hypoglycemia 6–7 days later. *P = 0.05 or less by post hoc t test for values at a given time period after repeated measures ANOVA demonstrated significant inter-group differences. Data are means ± SE.

FIG. 5.

Time course of changes in VMH cyclophilin, GK, and hexokinase I (HKI) mRNA after injecting adenoviruses expressing either scrambled RNA or shRNA for GK bilaterally into the ARC and VMN. Values for GK and HKI are normalized to internal standards for each generated from pooled samples and expressed as a function of cyclophilin in the same sample. *P = 0.05 or less when the ratio of GK to cyclophilin mRNA levels were compared with levels in uninjected controls (assigned as 1.0 at day 0). Data are means ± SE.

FIG. 5.

Time course of changes in VMH cyclophilin, GK, and hexokinase I (HKI) mRNA after injecting adenoviruses expressing either scrambled RNA or shRNA for GK bilaterally into the ARC and VMN. Values for GK and HKI are normalized to internal standards for each generated from pooled samples and expressed as a function of cyclophilin in the same sample. *P = 0.05 or less when the ratio of GK to cyclophilin mRNA levels were compared with levels in uninjected controls (assigned as 1.0 at day 0). Data are means ± SE.

FIG. 6.

Effects of bilateral VMH injection of adenovirus expressing GK shRNA compared with scrambled RNA on the response to hypoglycemia. *P < 0.05 or less scrambled vs. GK shRNA by post hoc t test after repeated measures ANOVA demonstrated a significant inter-group difference. Data are means ± SE.

FIG. 6.

Effects of bilateral VMH injection of adenovirus expressing GK shRNA compared with scrambled RNA on the response to hypoglycemia. *P < 0.05 or less scrambled vs. GK shRNA by post hoc t test after repeated measures ANOVA demonstrated a significant inter-group difference. Data are means ± SE.

TABLE 1

Effects of VMH GK activation on the counterregulatory response to hypoglycemia

DMSOGKA
Glucose AUC −8,145 ± 720 −6,765 ± 630 
Epinephrine AUC 127,440 ± 11,103 63,720 ± 5,870* 
Norepinephrine AUC 15,315 ± 1,120 5,820 ± 421* 
Glucagon AUC 9,979 ± 825 4,999 ± 329* 
DMSOGKA
Glucose AUC −8,145 ± 720 −6,765 ± 630 
Epinephrine AUC 127,440 ± 11,103 63,720 ± 5,870* 
Norepinephrine AUC 15,315 ± 1,120 5,820 ± 421* 
Glucagon AUC 9,979 ± 825 4,999 ± 329* 

Data are means ± SE. Nonfasted rats (n = 6–10 per group) were injected bilaterally in the VMH with 1.0 μl of 1% DMSO or 1% DMSO containing 0.5 nmol Compound A (GKA) and then 1 h later with 5 U/kg i.v. insulin. Plasma glucose, epinephrine, norepinephrine, and glucagon AUC levels were measured over 120 min after insulin administration.

*

P = 0.05 or less comparing DMSO vehicle to DMSO + Compound A.

TABLE 2

Effects of reducing VMH GK mRNA or “activity” with alloxan

SalineAlloxan
4 μg Alloxan + IIH   
    Glucose AUC (mg · dl−1 · 120 min−1−7,740 ± 630 −8,820 ± 742 
    Epinephrine AUC (pg · ml−1 · 120 min−1127,507 ± 3,453 162,843 ± 3,020* 
    Norepinephrine AUC (pg · ml−1 · 120 min−127,735 ± 2,534 25,815 ± 2,322 
    Glucagon AUC (pg · ml−1 · 120 min−114,745 ± 1,305 17,385 ± 1,653 
6–7d after 24 μg Alloxan + IIH   
    Glucose AUC (mg · dl−1 · 120 min−1−8,745 ± 753 −8,985 ± 777 
    Epinephrine AUC (pg · ml−1 · 120 min−1127,693 ± 3589 183,489 ± 3816* 
    Norepinephrine AUC (pg · ml−1 · 120 min−117,716 ± 1,623 19,470 ± 1,901 
    Glucagon AUC (pg · ml−1 · 120 min−114,940 ± 1,322 15,600 ± 1,483 
Terminal VMH GK/cyc 1.71 ± 0.11 1.26 ± 0.18* 
Terminal VMH HKI/cyc 1.88 ± 0.23 1.92 ± 0.14 
SalineAlloxan
4 μg Alloxan + IIH   
    Glucose AUC (mg · dl−1 · 120 min−1−7,740 ± 630 −8,820 ± 742 
    Epinephrine AUC (pg · ml−1 · 120 min−1127,507 ± 3,453 162,843 ± 3,020* 
    Norepinephrine AUC (pg · ml−1 · 120 min−127,735 ± 2,534 25,815 ± 2,322 
    Glucagon AUC (pg · ml−1 · 120 min−114,745 ± 1,305 17,385 ± 1,653 
6–7d after 24 μg Alloxan + IIH   
    Glucose AUC (mg · dl−1 · 120 min−1−8,745 ± 753 −8,985 ± 777 
    Epinephrine AUC (pg · ml−1 · 120 min−1127,693 ± 3589 183,489 ± 3816* 
    Norepinephrine AUC (pg · ml−1 · 120 min−117,716 ± 1,623 19,470 ± 1,901 
    Glucagon AUC (pg · ml−1 · 120 min−114,940 ± 1,322 15,600 ± 1,483 
Terminal VMH GK/cyc 1.71 ± 0.11 1.26 ± 0.18* 
Terminal VMH HKI/cyc 1.88 ± 0.23 1.92 ± 0.14 

Rats (n = 6–10 per group) were injected bilaterally in the VMH with 1 μl saline or saline containing 4 μg alloxan and then 1 h later, after baseline blood drawing, with 5 units/kg i.v. insulin, with blood drawing over 120 min for plasma glucose, epinephrine, norepinephrine, and glucagon levels. The next day, they were injected with saline or alloxan (24 μg) and then were retested 6–7 days later during insulin-induced hypoglycemia (IIH). Rats were killed after 2–5 days, and VMH, GK, and hexokinase I (HKI) mRNA expression was assessed relative to cyclophilin (cyc). Data are means ± SE AUC.

*

P < 0.05 or less saline vs. alloxan,

P = 0.05 or less 4 vs. 24 μg alloxan.

TABLE 3

Effects of reducing VMH GK mRNA expression with GK shRNA

Scrambled RNAGK shRNA
Glucose AUC −7,695 ± 710 −7,095 ± 630 
Epinephrine AUC (pg · ml−1 · 120 min−1164,850 ± 9,000 206,580 ± 8,920* 
Norepinephrine AUC (pg · ml−1 · 120 min−114,700 ± 1,395 11,265 ± 1,1195 
Glucagon AUC (pg · ml−1 · 120 min−120,880 ± 1,936 22,995 ± 2,066 
Scrambled RNAGK shRNA
Glucose AUC −7,695 ± 710 −7,095 ± 630 
Epinephrine AUC (pg · ml−1 · 120 min−1164,850 ± 9,000 206,580 ± 8,920* 
Norepinephrine AUC (pg · ml−1 · 120 min−114,700 ± 1,395 11,265 ± 1,1195 
Glucagon AUC (pg · ml−1 · 120 min−120,880 ± 1,936 22,995 ± 2,066 

Data are means ± SE. Rats (n = 6–10 per groups) were injected in the VMH with scrambled RNA- or GK shRNA-expressing adenoviruses and assessed 7–10 days later for their responses to insulin-induced hypoglycemia. Abbreviations are the same as in Table 1.

*

P = 0.05 or less comparing scrambled to GK shRNA rats.

Published ahead of print at http://diabetes.diabetesjournals.org on 21 February 2008. DOI: 10.2337/db07-1755.

B.E.L., J.E., and B.B.Z. hold stock in Merck.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

This work was funded by the Research Service of the Veterans Administration, the National Institute of Diabetes and Digestive and Kidney Diseases (DK-53181), and Merck Research Labs.

We thank Odeal Gordon, Laura Petrie, Charlie Salter, and Antoinette Moralishvilli for their expert technical assistance.

1.
Diabetes Control and Complications Trial Research Group: The effect of intensive treatment of diabetes on the development and progression of long-term complications of insulin-dependent diabetes mellitus.
N Engl J Med
329
:
977
–986,
1993
2.
Levin BE, Routh VH, Kang L, Sanders NM, Dunn-Meynell AA: Neuronal glucosensing: What do we know after 50 years?
Diabetes
53
:
2521
–2528,
2004
3.
Borg WP, Sherwin RS, During MJ, Borg MA, Shulman GI: Local ventromedial hypothalamic glucopenia triggers counterregulatory hormone release.
Diabetes
44
:
180
–184,
1995
4.
Borg MA, Sherwin RS, Borg WP, Tamborlane WV, Shulman GI: Local ventromedial hypothalamus glucose perfusion blocks counterregulation during systemic hypoglycemia in awake rats.
J Clin Invest
99
:
361
–365,
1997
5.
Yang X, Kow L-M, Funabashi T, Mobbs CV: Hypothalamic glucose sensor: similarities to and differences from pancreatic β-cell mechanisms.
Diabetes
48
:
1763
–1772,
1999
6.
Wang R, Liu X, Hentges ST, Dunn-Meynell AA, Levin BE, Wang W, Routh VH: The regulation of glucose-excited neurons in the hypothalamic arcuate nucleus by glucose and feeding-relevant peptides.
Diabetes
53
:
1959
–1965,
2004
7.
Dunn-Meynell AA, Routh VH, Kang L, Gaspers L, Levin BE: Glucokinase is the likely mediator of glucosensing in both glucose excited and glucose inhibited central neurons.
Diabetes
51
:
2056
–2065,
2002
8.
Kang L, Dunn-Meynell AA, Routh VH, Gaspers LD, Nagata Y, Nishimura T, Eikis J, Zhang BB, Levin BE: Glucokinase is a critical regulator of ventromedial hypothalamic neuronal glucosensing.
Diabetes
55
:
412
–420,
2006
9.
Lynch RM, Tompkins LS, Brooks HL, Dunn-Meynell AA, Levin BE: Localization of glucokinase gene expression in the rat brain.
Diabetes
49
:
693
–700,
2000
10.
Kang L, Routh VH, Kuzhikandathil EV, Gaspers L, Levin BE: Physiological and molecular characteristics of rat hypothalamic ventromedial nucleus glucosensing neurons.
Diabetes
53
:
549
–559,
2004
11.
Levin BE, Kang L, Sanders NM, Dunn-Meynell AA: Role of neuronal glucosensing in the regulation of energy homeostasis.
Diabetes
55
(Suppl. 2):
S122
–S130,
2006
12.
Sanders NM, Dunn-Meynell AA, Levin BE: Third ventricular alloxan reversibly impairs glucose counterregulatory responses.
Diabetes
53
:
1230
–1236,
2004
13.
Tkacs NC, Dunn-Meynell AA, Levin BE: Presumed apoptosis and reduced arcuate nucleus neuropeptide Y and pro-opiomelanocortin mRNA in non-coma hypoglycemia.
Diabetes
49
:
820
–826,
2000
14.
Tkacs NC, Levin BE: Obesity-prone rats have pre-existing defects in their counterregulatory response to insulin-induced hypoglycemia.
Am J Physiol
287
:
R1110
–R1115,
2004
15.
Kang L, Sanders NM, Dunn-Meynell AA, Gaspers LD, Routh VH, Thomas AP, Levin BE: Prior hypoglycemia enhances glucose responsiveness in some ventromedial hypothalamic glucosensing neurons
Am J Physiol Regul Intergr Comp Physiol
294
:
R784
–R792,
2008
16.
Amiel SA, Tamborlane WV, Simonson DC, Sherwin RS: Defective glucose counterregulation after strict glycemic control of insulin-dependent diabetes mellitus.
N Engl J Med
316
:
1376
–1383,
1987
17.
Kamata K, Mitsuya M, Nishimura T, Eiki J, Nagata Y: Structural basis for allosteric regulation of the monomeric allosteric enzyme human glucokinase.
Structure (Camb
) 
12
:
429
–438,
2004
18.
Bain JR, Takeuchi K, Schisler JC, Newgard CB, Becker TC: An adenovirus vector for efficient RNAi-mediated suppression of target genes in insulinoma cells and pancreatic islets of Langerhans.
Diabetes
53
:
2190
–2194,
2004
19.
Meglasson MD, Burch PT, Berner DK, Najafi H, Matschinsky FM: Identification of glucokinase as an alloxan-sensitive glucose sensor of the pancreatic β-cell.
Diabetes
35
:
1163
–1173,
1986
20.
Zelent D, Golson ML, Koeberlein B, Quintens R, van Lommel L, Buettger C, Weik-Collins H, Taub R, Grimsby J, Schuit F, Kaestner KH, Matschinsky FM: A glucose sensor role for glucokinase in anterior pituitary cells.
Diabetes
55
:
1923
–1929,
2006
21.
Levin BE, Dunn-Meynell AA, Banks WA: Obesity-prone rats have normal blood-brain barrier transport but defective central leptin signaling prior to obesity onset.
Am J Physiol
286
:
R143
–R150,
2004
22.
de Vries MG, Arseneau LM, Lawson ME, Beverly JL: Extracellular glucose in rat ventromedial hypothalamus during acute and recurrent hypoglycemia.
Diabetes
52
:
2767
–2773,
2003
23.
Jetton TL, Liang Y, Pettepher CC, Zimmerman EC, Cox FG, Horvath K, Matschinsky FM, Magnuson MA: Analysis of upstream glucokinase promoter activity in transgenic mice and identification of glucokinase in rare neuroendocrine cells in the brain and gut.
J Biol Chem
269
:
3641
–3654,
1994
24.
Roncero I, Alvarez E, Vazques P, Blazquez E: Functional glucokinase isoforms are expressed in rat brain.
J Neurochem
74
:
1848
–1857,
2000
25.
Maekawa F, Toyoda Y, Torii N, Miwa I, Thompson RC, Foster DL, Tsukahara S, Tsukamura H, Maeda K: Localization of glucokinase-like immunoreactivity in the rat lower brain stem: for possible location of brain glucose-sensing mechanisms.
Endocrinol
141
:
375
–384,
2000
26.
Sutherland VL, McReynolds M, Tompkins LS, Brooks HL, Lynch RM: Developmental expression of glucokinase in rat hypothalamus.
Brain Res Dev Brain Res
154
:
255
–258,
2005
27.
Claret M, Smith MA, Batterham RL, Selman C, Choudhury AI, Fryer LGD, Clements M, Al-Qassab H, Heffron H, Xu AW, Speakman JR, Barsh GS, Viollet B, Vaulont S, Ashford MLJ, Carling D, Withers DJ: AMPK is essential for energy homeostasis regulation and glucose sensing by POMC and AgRP neurons.
J Clin Invest
117
:
2325
–2336,
2007
28.
McCrimmon RJ, Fan X, Ding Y, Zhu W, Jacob RJ, Sherwin RS: Potential role for AMP-activated protein kinase in hypoglycemia sensing in the ventromedial hypothalamus.
Diabetes
53
:
1953
–1958,
2004
29.
Hayashi T, Hirshman MF, Fujii N, Habinowski SA, Witters LA, Goodyear LJ: Metabolic stress and altered glucose transport: activation of AMP-activated protein kinase as a unifying coupling mechanism.
Diabetes
49
:
527
–531,
2000
30.
Beverly JL, Beverly MF, Meguid MM: Alterations in extracellular GABA in the ventral hypothalamus of rats in response to acute glucoprivation.
Am J Physiol
269
:
R1174
–R1178,
1995
31.
Chan O, Lawson M, Zhu W, Beverly JL, Sherwin RS: ATP-sensitive potassium channels regulate the release of GABA in the ventromedial hypothalamus during hypoglycemia.
Diabetes
56
:
1120
–1126,
2007
32.
Tong Q, Ye C, McCrimmon RJ, Dhillon H, Choi B, Kramer MD, Yu J, Yang Z, Christiansen LM, Lee CE, Choi CS, Zigman JM, Shulman GI, Sherwin RS, Elmquist JK, Lowell BB: Synaptic glutamate release by ventromedial hypothalamic neurons is part of the neurocircuitry that prevents hypoglycemia.
Cell Metab
5
:
383
–393,
2007
33.
Beverly JL, de Vries MG, Beverly MF, Arseneau LM: Norepinephrine mediates glucoprivic-induced increase in GABA in the ventromedial hypothalamus of rats.
Am J Physiol Regul Integr Comp Physiol
279
:
R990
–R996,
2000
34.
McCrimmon RJ, Song Z, Cheng H, McNay EC, Weikart-Yeckel C, Fan X, Routh VH, Sherwin RS: Corticotrophin-releasing factor receptors within the ventromedial hypothalamus regulate hypoglycemia-induced hormonal counterregulation.
J Clin Invest
116
:
1723
–1730,
2006
35.
Routh VH, McArdle JJ, Levin BE: Phosphorylation modulates the activity of the ATP-sensitive K+ channel in the ventromedial hypothalamic nucleus.
Brain Res
778
:
107
–119,
1997
36.
Ashford MLJ, Boden PR, Treherne JM: Glucose-induced excitation of hypothalamic neurones is mediated by ATP-sensitive K+ channels.
Pflugers Arch
415
:
479
–483,
1990
37.
Mobbs CV, Kow LM, Yang XJ: Brain glucose-sensing mechanisms: ubiquitous silencing by aglycemia vs. hypothalamic neuroendocrine responses.
Am J Physiol Endocrinol Metab
281
:
E649
–E654,
2001
38.
McCrimmon RJ, Fan X, Cheng H, McNay E, Chan O, Shaw M, Ding Y, Zhu W, Sherwin RS: Activation of AMP-activated protein kinase within the ventromedial hypothalamus amplifies counterregulatory hormone responses in rats with defective counterregulation.
Diabetes
55
:
1755
–1760,
2006
39.
Chan O, Zhu W, Ding Y, McCrimmon RJ, Sherwin RS: Blockade of GABA(A) receptors in the ventromedial hypothalamus further stimulates glucagon and sympathoadrenal but not the hypothalamo-pituitary-adrenal response to hypoglycemia.
Diabetes
55
:
1080
–1087,
2006