Date: April 8th, 2010

Speaker's Name and Affiliation:

• Vanessa Routh,
• UMDNJ-New Jersey Medical School

Seminar Title: "Hypothalamic Glucose Sensing Neurons: From turning a blind eye to crying wolf"

Chair: Harry R. Kissileff, Ph.D.

Rapporteur: Kathleen L. Keller, Ph.D.

Attendees and their Affiliation:

Kathleen Keller	Columbia/Obesity Research Center
Harry Kissileff	Obesity Research Center
Tony Sclafani	Brooklyn College
Laurence Nolan	Wagner College
JA Grinker	University of Michigan
Karen Ackroff	Brooklyn College
Allan Geliebter	NYORC
Khalid Touzani	Brooklyn College
Steven Zukerman	Brooklyn College
John Glendinning	Barnard
Joe Vasselli	ORC St. Luke's
Majella Okeefe	Columbia / ORC
Charlesa Gibson	NYORC
Annemarie Olsen	NYORC
Clemence Blouet	Einstein

Summary: (Prepared by the Rapporteur)

The central nervous system, particularly the hypothalamus, plays an key role in integrating hormonal signals from the periphery that allow organisms to regulate food intake. Glucose is the obligate fuel for the brain, so it makes sense that glucose sensing neurons are present in this organ. Dr. Routh has been studying the characteristics of glucose sensing neurons in the hypothalamus. Her hypothesis is that glucose sensing neurons are a part of a safety network that protects against the effects of a severe glucose deficit. Further, glucose sensing neurons have evolved to protect organisms from effects due to famine and fasting. Moreover, she proposes that in the modern world, glucose sensing neurons have an important role in restoring euglycemia following iatrogenic insulin-induced hypoglycemia. However, under normal conditions of energy balance, it is important that glucose sensing neurons do not respond to small glucose decreases associated with meal-to-meal fluctuations in glucose. The prediction is that during metabolic syndrome (obesity, diabetes, etc), glucose sensing neurons are hypersensitive to small fluctuations in glucose levels. Inappropriate activation of energy sparing mechanisms under conditions of energy sufficiency leads to excess energy storage.

Dr. Routh¡¯s talk discussed the following: 1. What are the mechanisms underlying glucose sensing? 2. What are the effects of satiety hormones on glucose sensing? 3. What happens to glucose sensing neurons during an energy deficit? 4. What happens to glucose sensing neurons during diabetes and/or the metabolic syndrome? First, glucose sensing occurs by integration from two types of neurons, glucose excitatory (GE) neurons and glucose inhibited (GI) neurons. The GE neurons are found in the VL-VMN and the ¡°cell poor¡± region between the VMN and the ARC. Wang et al., (Diabetes 2004) found that GE neurons are downstream mediators of melanocortins. KATP channels are thought to be involved with glucose sensing by GE neurons. Approximately 40% of ARC NPY neurons are GI neurons, however the phenotype of VMN GI neurons is not known. Glucose sensing by GI neurons involves an interaction between the cellular fuel sensor, AMPK, and nitric oxide (NO). Decreased glucose activates AMPK which phosphorylates neuronal nitric oxide synthase (nNOS) leading to increased NO production. NO causes a further activation of AMPK which leads to closure of the cystic fibrosis transmembrane regulator chloride channel (CFTR). The effects of NO are mediated by its receptor, soluble guanylyl cyclase (sGC) and cyclic GMP (cGMP). The observation that NOS inhibition blocks the effects of glucose on GI neurons and that nNOS knockout mice lack GI neurons indicates that NO-sGC-cGMP signaling is crucial for glucose sensing by these neurons.

What effect do satiety hormones have on glucose sensing neurons? Murphy et al., 2009 (AJP) showed that leptin attenuates the response of GI neurons to a glucose deficit by AMPK inhibition. Insulin attenuates the response of GE neurons to decreased glucose levels via the phosphoinositol 3 kinase signaling pathway. Thus, the ability of glucose sensing neurons to sense glucose deficit is blunted in the presence of hormonal signals of satiety and positive energy balance.

In the next part of the talk, Dr. Routh discussed the glucose sensing during an energy deficit. During fasting, response of GI neurons to a glucose deficit is increased, compared to the response in fed animals. Changes that occur in GI neurons during fasting are mediate by AMPK. Fasting also enhances NPY release and the response to NPY-GI neurons to decreased glucose levels. Thus, in the fasted state, NPY neurons would be activated to a greater extent in the presence of low glucose, leading to activation of compensatory mechanisms. Furthermore, the ability of glucose sensing neurons to sense glucose deficit is attenuated under conditions where the ability of the brain to sense hypoglycemia and initiate the counterregulatory response is impaired. These data suggest a role for glucose sensing neurons in sensing severe energy deficit (e.g., fasting, hypoglycemia).

In contrast during type 2 diabetes mellitus (T2DM), the response of VLVMN GE neurons to decreased glucose is enhanced. This is due to the insulin resistance of GE neurons which occurs during T2DM since normalizing insulin sensitivity also normalizes the glucose sensitivity of GE neurons. Thus, in T2DM GE neurons are hypersensitive to glucose decreases which would normally have little or no effect.

In conclusion, these data lead to the hypothesis that the role of glucose sensing neurons is to detect and respond to severe energy deficit (e.g., fasting, insulin-induced hypoglycemia) in order to maintain glucose availability for the brain. However, when glucose sensing neurons ¡°turn a blind eye¡± to glucose deficit the ability of the brain to restore euglycemia is impaired. Thus, impaired detection of glucose deficit may contribute to the impaired counterregulatory to hypoglycemia which occurs following repeated hypoglycemic insults.

On the other hand, under conditions of energy sufficiency (e.g., presence of leptin and insulin) the ability of glucose sensing neurons to sense glucose deficit is masked.When glucose sensing neurons show an enhanced response to glucose deficit during energy sufficiency (eg., ¡°crying wolf¡±) it may contribute to the development of obesity and diabetes

Question & Answers:

Q. What is the concentration of glucose in the VMH?
A. The euglycemic range for interstitial glucose is probably between 1 and 2 mM according to measurements using glucose oxidase electrodes or zero net flux microdialysis.

Q. Is there a tight correlation between brain and plasma glucose?
A. Yes, brain glucose is about 30% of plasma glucose over the physiological range

Q. What is the transport mechanism for glucose in the brain?
A. Facilitated transport. Most neurons use the GLUT 3 transport protein. Some neurons also have GLUT 4 and/or GLUT 2.

Q. When you said ¡°plasma glucose,¡± do you mean brain plasma glucose?
A. No, I was referring to plasma glucose as the periphery.

Q. If you measured extracellular glucose in the liver, how different would that be than the plasma?
A. In the portal vein, it would be similar.

Q. But what about liver tissue?
A. Not sure, it might be higher.

Q. So GE neurons are generally excited all the time, but if you lower the glucose, they will shut-off?
A. Yes, that is correct.

Q. Is this just 1 neuron type?
A. Yes.

Q. How are you delivering the glucose?
A. We are perfusing it.

Q. Are these rat models that you are using?
A. Yes, these are rats. We have done this in rats and mice. They are almost identical in response.

Q. Could you explain what you mean by the phenotype of a neuron?
A. We are looking at whether they are POMC or NPY, etc.

Q. Are the Glucose Inhibitory (GI) neurons in the same regions as the Glucose Excitatory (GE) neurons?
A. No. GE neurons are located in the VL-VMN whereas GI neurons are located throughout the VMN

Q. Are there more GI than GE neurons?
A. The GE neurons are concentrated in such a small area of the VMH whereas GI neurons are more diffuse. Thus, when you consider the whole VMH, you have many more GI neurons than GE neurons.

Q. Before you said that the GI neurons were more important?
A. We have a lot more data from GI neurons and they are more globally distributed, but I¡¯m not sure they are more important. We just know more about them.

Q. What about ghrelin neurons? Are they in the VMN?
A. I am not sure. I would think that ghrelin would modulate the NPY neurons, but I¡¯m not sure. The experiments have not been done.

Q. So the Nitric Oxide is in response to AMPK activation?
A. Yes, that is correct.

Q. Are you saying it¡¯s an intracellular signal?
A. Yes.

Q. I¡¯m not clear on what this membrane potential sensitive dye is?
A. It is a dye that can diffuse into the cell membrane and increases fluorescence once the cell depolarizes.

Q. Would that explain the fact that leptin inhibits food intake (referring to data that leptin attenuates the response of GI neurons to a glucose deficit)?
A. It is consistent with inhibition of food intake, though it is probably more linked to metabolic regulation.

Q. Are you actually going to model the two kinds of diabetes, or are you just changing the glucose environments?
A. Right now we are just changing the glucose environments.

Q. How long does this impaired response to hypoglycemia last? With every episode of hypoglycemia, are diabetics getting worse with glucose control?
A. Two weeks. Yes, that is correct. Glucose control gets worse with each episode of hypoglycemia.

Q. Why does N-acetyl cysteine help with acetaminophen overdose?
A. I¡¯m not sure. I think it prevents future liver damage when patient presents with an overdose. It protects the liver from additional damage.

Q. You don¡¯t think short-term glucose release is involved in food intake regulation? Why? What about Campfield¡¯s data?
A. In Barry Levin¡¯s data, they found that there was no correlation between brain glucose level and spontaneous food intake.

Q. What type of glucose sensing would you expect to see in bulimics? They are normal weight, but eat a lot during a binge.
A. I don¡¯t know much about bulimia nervosa, but in anorexic patients, I would predict that glucose sensing would be impaired.

Q. What do you think about the POMC knock-out animals? Do they sense glucose?
A. POMC neurons which lack AMPK show impaired glucose sensing.

Q. What other parts of the brain do glucose sensing neurons exist?
A. We haven¡¯t looked at other parts of the brain yet. However, glucose sensing neurons also exist in the amygdala and brain stem.

Q. Are there estrogen receptors in the VMH?
A. Yes, they are there and have an important role in reproductive health. In males, this area of the brain is larger than in females. It¡¯s possible that pesticides which are phytoestrogens or hormones in food have a greater effect in males because of this.

Q. Are there leptin receptors in the VMH?
A. Yes.