Ingestive Classics
Smith and Glucoprivic Feeding

SMITH, G.P. and A.N. EPSTEIN.  Increased feeding in response to decreased glucose utilization in the rat and monkey.  American Journal of Physiology 217; 1083-1087, 1969.

SMITH, G.P., J. GIBBS, A.J. STROHMAYER, and P.E. STOKES.   Threshold doses of 2-deoxy-D-glucose for hyperglycemia and feeding in rats and monkeys.  American Journal of Physiology 222; 77-81, 1972.

Comments by Gerard P. Smith (November 20, 2013)

I first used 2-deoxy-D-glucose (2-DG) to investigate its stimulation of gastric secretion in rats with chronic gastric fistulas (Feinblatt et al. 1966).  During these experiments, I learned that 2-DG increased blood glucose and decreased intracellular glucose utilization in a number of tissues, especially in the brain (Brown, 1962).  This suggested that 2-DG was a useful tool to test Mayer’s hypothesis (Mayer, 1953) that decreased glucose utilization, particularly in the ventromedial hypothalamic nucleus, not the concentration of circulating glucose, was the critical stimulus for the initiation of eating. The best evidence for this hypothesis at that time was changes in the peripheral arterio-venous (AV) glucose differences (Van Itallie et al., 1952).  But peripheral AV glucose differences did not distinguish the uptake of glucose from its intracellular utilization.  Since 2-DG produced dose-related decreases in glucose utilization and increased circulating glucose, I thought that 2-DG could give a decisive answer to this question.

I began with rhesus monkeys (Macaca mulatta) because they were the primary research animal in my laboratory in John Brobeck’s Department of Physiology at the University of Pennsylvania.  When I got positive results in monkeys, I walked a block to Alan Epstein’s laboratory in the Department of Biology to ask him to collaborate with me on similar experiments with rats.   He was enthusiastic.   The first paper, “Increased Feeding in Response to Decreased Glucose Utilization in the Rat and Monkey”, describes the results (Smith and Epstein, 1969).
2-DG increased food intake and circulating glucose in rats and monkeys.  On the basis of our results in rats and monkeys, and those in mice (Likuski et al., 1967), we concluded that “an acute decrease of glucose utilization must now be considered one of the sufficient conditions for feeding in animals (p. 1086).”  Although sufficient, we thought it was not necessary because rats recovered from lateral hypothalamic lesions had normal controls of food intake despite failing to respond to acute decreased glucose utilization in the brain produced by insulin hypoglycemia (Epstein and Teitelbaum, 1967).  Subsequent experiments with 2-DG confirmed its lack of necessity for initiating spontaneous feeding (see Smith, 1976 for review).

In July, 1968, I moved to the Bourne Laboratory at Cornell Medical College and I took the 2-DG problem with me.  The problem was that although 2-DG stimulated food intake by decreasing glucose utilization, I did not know if this effect of 2-DG modelled the initiation of spontaneous eating during ad lib access to food or whether the effect modelled the initiation of eating as part of a homeostatic response to the metabolic emergency of cerebral glucoprivation.
I was also concerned about the relation between increased food intake and increased blood glucose produced by 2-DG.  Although both responses could be considered homeostatic (Richter, 1927) in that both could provide glucose to compete with 2-DG to counteract the decrease in glucose utilization, there was an important difference in their relationship to 2-DG—the hyperglycemic response was dose-related, but the food intake response was not.  Given that 2-DG produced a dose-related decrease in glucose utilization, this seemed important.

Although the dose-response functions of hyperglycemia and food intake were different, I reasoned that a comparison of the threshold doses could be used to investigate whether increased food intake produced by 2-DG contributed to spontaneous eating or whether it was part of the homeostatic response to the metabolic emergency of decreased glucose utilization.  My hypothesis was that if the feeding response to 2-DG were part of the control of spontaneous eating, the threshold dose for eating should be smaller than the threshold for hyperglycemia.  If the threshold dose for feeding were significantly larger than the threshold dose for hyperglycemia, this would be evidence that the feeding response was part of the homeostatic response to the metabolic emergency.

The experimental results in this second paper were clear—the threshold dose of 2-DG for feeding was about 4 times larger than the threshold dose for hyperglycemia in rats and monkeys (Smith et al., 1972).  Thus, under these conditions, the decrease in glucose utilization required for initiating feeding was much larger than the decrease required for eliciting hyperglycemia.  Our strong inference from this was that if decreased glucose utilization were involved in the initiation of spontaneous eating, hyperglycemia should be present before each meal.  But this had not been observed “in any mammal at the present time” (p. 80).  Although subsequent experiments by others reported an overlap in the threshold doses of 2-DG for feeding and hyperglycemia in rats (see Smith, 1976 for review), no one has found conditions in which the threshold dose of 2-DG for feeding was significantly smaller than the threshold for hyperglycemia--the result that would support the possibility that decreased glucose utilization stimulated spontaneous feeding during ad lib access to food.  Our rejection of the glucostatic hypothesis for the control of short-term food intake is consistent with the latest research (see Commentary by Levin, 2013).

Although not discussed in the 1972 paper, hyperglycemia and increased food intake in response to decreased glucose utilization have specific homeostatic functions in monkeys (Smith and Root, 1969).  Food intake did not change the magnitude of the hyperglycemic response to 300 mg/kg of 2-DG, but it did change the distribution of glucose because food intake increased circulating insulin, presumably as a result of enteric stimuli, and this enhanced glucose uptake and utilization in peripheral tissues.  In contrast, peripheral glucose utilization was inhibited during hyperglycemia when no food intake was permitted after 2-DG because adrenomedullary epinephrine inhibited the insulin response to hyperglycemia (Smith et al., 1973).   Thus, the hyperglycemic response conserves circulating glucose for the brain, while ingesting food during the hyperglycemia permits the distribution of glucose to peripheral tissues as well.


Brown, J.  Effects of 2-deoxy-D-glucose on carbohydrate metabolism: review of the literature and studies in the rat.  Metabolism 11: 1098-1112, 1962.
Epstein, A.N. and P. Teitelbaum.  Specific loss of the hypoglycemic control of feeding in recovered lateral rats.  Am.J.Physiol. 213: 1159-1167, 1967.
Feinblatt, J., T.Gelfand, and G.P.Smith.  Gastric acid response to 2-deoxy-D-glucose in chronic fistula rats.  Proc. Soc. Exptl. Biol. Med.  123: 241-242, 1966.
Levin,B.  Comments on Ingestive Classic: MAYER, J. Glucostatic mechanism of regulation of food intake. New England Journal of Medicine 249 : 13-16, 1953.   doi 1 November 2013.
Likuski, H.J., A.F. Debons, and R.J. Cloutier.  Inhibition of gold thioglucose-induced hypothalamic obesity by glucose analogues. Am.J.Physiol. 212: 669-676, 1967.
Mayer, J. Glucostatic mechanism of regulation of food intake. N.Engl.J.Med., 249: 13-16, 1953.
Richter, C.P.  Animal behavior and internal drives.  Quart. Rev. Biol. 2: 307-343, 1927.
Smith, G.P.  Humoral hypotheses for the control of food intake.  In: G.Bray (Ed.), Obesity in Perspective, Vol.2, Pt.2.  Bethesda: National Institute of Health, 1976, pp. 19-29.
Smith, G.P. and A.N. Epstein.  Increased feeding in response to decreased glucose utilization in the rat and monkey.  Am. J. Physiol.  217: 1083-1087, 1969.
Smith, G.P. and A.W. Root.  Effect of feeding on hormonal responses to 2-deoxy-D-glucose in conscious monkeys.  Endocrinology 85: 963-966, 1969.
Smith, G.P., J. Gibbs, A.J. Strohmayer, and P.E. Stokes.   Threshold doses of 2-deoxy-D-glucose for hyperglycemia and feeding in rats and monkeys.  Am. J. Physiology 222:77-81, 1972.
Smith, G.P., J. Gibbs, A.J. Strohmayer, A.W. Root, and P.E. Stokes.  Effect of 2 deoxy-D-glucose on insulin response to glucose in intact and adrenalectomized monkeys.  Endocrinology 92: 750-754, 1973.
Van Itallie T.B., R. Beaudoin, and J. Mayer.  Arteriovenous glucose differences, metabolic hypoglycemia and food intake in man.  J. Clin. Nutr.  1: 208-217, 1952.