hypoglycemia: dual role of hypothalamic nitric oxide
ntensive insulin therapy is used to obtain tight glycemic control in both Type 1 and Type 2 diabetic patients and significantly reduces the onset and progression of diabetes-related complications. However, intensive insulin therapy also causes a clinically adverse effect: hypoglycemia. The brain is mostly responsible for the detection of hypoglycemia and the initiation of counter-regulatory responses that restore normal blood glucose levels. We have recently shown that a specific population of neurons produces the messenger gas nitric oxide in response to a decrease in glucose concentration.
The purpose of our work was to determine whether the production of nitric oxide plays a role in the initiation of the counter-regulatory response to hypoglycemia. Using different rodent models of acute hypoglycemia, we showed that physiologic nitric oxide production in a specific area of the brain (the ventro-medial hypothalamus) is necessary to fully initiate the counter-regulatory response. On the other hand, we also showed that excessive production of nitric oxide impairs this mechanism. Thus our data suggests that nitric oxide production during hypoglycemia plays a dual role in the counter-regulatory response.
Hypoglycemia can be the result of intensive insulin therapy. The brain is particularly vulnerable to such hypoglycemia since glucose is its preferred fuel. Powerful neuroendocrine and autonomic counterregulatory mechanisms that prevent and correct hypoglycemic conditions protect the brain from hypoglycemia. These corrective mechanisms, known as the counterregulatory response (CRR), involve the release of counterregulatory hormones (e.g. glucagon, epinephrine), which restore euglycemic status by stimulating hepatic glucose production and inhibiting peripheral glucose uptake. However, these mechanisms become impaired as a result of recurrent episodes of hypoglycemia. This impairment is a component of the life-threatening syndrome: hypoglycemia-associated autonomic failure (HAAF). During HAAF, the glycemic threshold for the CRR shifts to lower glucose levels. As a result, glucose levels are allowed to drop, without detection, to dangerously low or even lethal levels before the CRR is initiated. Thus, increasing susceptibility to recurrent hypoglycemia is a major limiting factor in the management of Types 1 and 2 diabetes mellitus. For this reason, it is critical to understand mechanisms of hypoglycemia detection in order to propose new molecular targets, which will avoid the development of HAAF in diabetic patients.
The ventromedial area of the hypothalamus, within the brain, is mainly responsible for the detection of hypoglycemia and the initiation of the CRR. The VMH contains specialized glucose sensitive neurons (GSNs) that alter their electrical activity in response to a change in extracellular glucose concentration. It has been proposed by our laboratory, as well as others, that the VMH GSNs enable the brain to sense hypoglycemia and initiate the CRR. However, the molecular mechanisms by which VMH GSNs detect hypoglycemia and signal for the initiation of the CRR are not fully understood. We have recently shown in vitro that a specific subtype of GSNs, the glucose-inhibited neurons which increase their activity as glucose decreases, produce nitric oxide (NO) in response to a decrease in glucose level. NO is a messenger within the brain and other tissues known to take part in many physiological mechanisms, such as the control of energetic homeostasis. Thus, we hypothesize that the VMH NO production plays a role in the initiation of the CRR in vivo.
To address this question, we used complementary approaches in vivo in rats or mice. First, using rats, we blocked NO production in the whole brain (intracerebroventricular injection, ICV) or directly in the VMH and monitored the recovery from hypoglycemia induced by an intravenous (IV) injection of insulin (1 U/kg). We showed that the inhibition of the nitric oxide synthase (NOS) by a non-selective NOS inhibitor (N-monomethyl-L-arginine, LNMMA) injected either in ICV or directly in the VMH significantly slows down the recovery from hypoglycemia (Fig 1). We also showed that the NO receptor, the soluble guanylate cyclase (sGC), is involved in this mechanism. The administration of a selective inhibitor of the sGC (1H-[1,2,4]Oxadiazolo[4,3-a]quinoxalin-1-one, ODQ) also slows down the recovery from hypoglycemia when injected in the VMH (Fig 1). These data support our hypothesis that NO is important for generation of a full CRR.
In order to confirm that the VMH NO production is necessary to initiate the CRR, we next performed hyperinsulinemic/hypoglycemic clamp, which is the gold standard technique for measuring the CRR. This technique consists of inducing and maintaining a hypoglycemia around 50 mg/dl for one to two hours. Thus, a high concentration of insulin (20 mU/kg/min) is simultaneously perfused by IV with sufficient glucose to maintain the glycemia around 50 mg/dl. The rate of glucose added to maintain the glycemia (glucose infusion rate, GIR) predicts the strength of the CRR. That is, the concentration of glucose required to maintain glycemia (GIR) is inversely proportional to the strength of the CRR. Our data show that the selective inhibition of the neuronal NOS isoform (nNOS) in the VMH increases the GIR necessary to maintain a hypoglycemia around 50 mg/dl and decreases the plasma levels of the CRR hormone epinephrine. These data strongly support our hypothesis that VMH NO production is necessary to fully initiate the CRR. Moreover, they suggest that the nNOS isoform is involved in this pathway.
To verify that the nNOS is specifically the NOS isoform involved, we performed hyperinsulinemic/hypoglycemic clamps in wild type (WT), eNOS and nNOS knock-out mice. Our results show that the GIR necessary to maintain a hypoglycemia around 50 mg/dl is the same in WT and eNOS KO mice. In contrast, the GIR is significantly higher in nNOS KO mice, confirming that the neuronal NOS isoform is involved in the initiation of the CRR. Collectively, our data show that during hypoglycemia, VMH NO production through the nNOS isoform is necessary to fully initiate the CRR. We also show that activation of the sGC by NO is necessary in this pathway.
In other tissues such as muscle, the “yin-yang” theory of NO signaling has been proposed. This theory suggests that NO has opposite effects depending on the concentration being produced. Here, we wanted to determine whether high NO production could have a negative effect on the CRR. NO is the byproduct of conversion of L-arginine to L-citrulline by the NOS. Increasing the concentration of L-arginine has been shown to triple NO production in brain slices exposed to oxygen-glucose deprivation. Therefore, L-arginine (5 mM) was infused ICV or in the VMH. Our experiments show that L-arginine slows down the recovery from hypoglycemia (Fig 2A). In order to confirm that L-arginine worked through an increase of NO production, rats were perfused ICV with both L-arginine and LNMMA, the non-selective NOS inhibitor. As shown in Fig 2A, LNMMA reverses the effect of L-arginine on the recovery from hypoglycemia. Taken together, these data suggest that excessive NO production during hypoglycemia impairs the CRR.
The last step was to determine how a high NO concentration can inhibit the CRR. S-nitrosylation is a non-enzymatic reaction that takes place under elevated concentration of NO. S-nitrosylation is an addition of a NO moiety to a free thiol of a cystein of proteins. It has been shown that different proteins involved in glucose sensitivity can be S-nitrosylated. Moreover, in our department, Dr. Annie Beuve showed that S-nitrosylation of the sGC desensitizes this enzyme. Thus, we next checked the level of S-nitrosylation of the sGC in the ventral hypothalamus in rats infused ICV with vehicle or L-arginine during hypoglycemia. Our data show that L-arginine increases the S-nitrosylation of the sGC during hypoglycemia (Fig 2B). This is consistent with an impairment of the CRR since we have previously shown that the sGC is involved.
Altogether, our work shows that NO produced in the VMH during hypoglycemia plays a crucial and dual role in the control of the CRR. We show here that a physiological equilibrium in NO production is necessary to correctly regulate the CRR to hypoglycemia. Our future work is to determine whether an excessive NO production, which could be seen during recurrent hypoglycemia, could lead to the impaired CRR that occurs during the HAAF syndrome. This work will bring up new pharmaceutical targets, which would help reduce the major limiting factor in treating diabetes with intensive insulin therapy.
Xavier Fioramonti earned a PhD in molecular, cellular and integrative physiopathology from the University Paul Sabatier in Toulouse, France, in 2005. He joined the laboratory of Dr. Routh in the Pharmacology and Physiology Department, UMDNJ-New Jersey Medical School, in April 2006 as a postdoctoral fellow. He is currently funded by the Juvenile Diabetes Research Foundation to perform this work.