Sugar preference: beyond the tongue

In studies, rodents generally prefer sugars over amino acids/proteins when given the choice, and this seems to reflect their ability to bind to the sweet T1R2/T1R3 receptor on the tongue.  Postingestion, however, nutrients obviously differentially effect hormones involved in fuel mobilization (e.g. insulin and glucagon), so mechanisms should exist to influence appetite for specific macronutrients.  This has been demonstrated with restricted essential amino acids, for example; there is a compensation and greater intake.  So consumption isn’t all about taste, but there is still a limited sense of relative contributions of these postingestive mechanisms.

Ren et al. set out to find if these mechanisms can reinforce sugar intake independent of taste.  They knocked out TRPM5 in mice, which is necessary for bitter, sweet, and most L-amino acid taste signaling.  If the mice developed a preference for food dispensers with glucose over those with an isoenergenic L-amino acid (serine, which is also sweet, or arginine, which is not), it suggests that taste is not the only way for sugar preference to form.

In C57BL/6 mice, 10 minute 2-bottle feed preference tests were performed to test if the knockout mice displayed a preference for one of the macronutrients (the short nature reduces postingestive effects).  Wildtype mice clearly preferred the glucose, but knockout mice consumed them equally (see figure B below).   Conditioning experiments were performed to test if knockout mice could develop a preference for serine or glucose, which involved associating a side of the cage with a macronutrient for 30 minutes per day for 6 days, followed by a 10 minute 2-bottle feed, both with water (they chose the bottle more often with the side they associated with glucose even though they couldn’t taste it).  Though knockout mice didn’t lick the glucose bottle as many times as wildtype, the pattern and preference ratio does not differ (see figure C and D), suggesting that taste is not necessary to develop a preference for glucose.  An extinction test found that in these same mice, water substitution for glucose decreased the preference for that sipper.

A longer (21 hour) nutrient availability test was also performed, and shows similar results (figure A and B below), while energy expenditure via indirect calorimetry was measured.  During glucose consumption, higher respiratory quotients (and heat production) (figures C, D, and E) reflected a higher use of glucose as a fuel substrate.

Blood glucose, and liver glycogen were also measured.  Unsurprisingly, blood glucose and glycogen was higher in both knockouts and wildtypes after glucose ingestion versus serine (figures A and B).  Scatter plots show an association between number of licks of glucose and respiratory quotient (glucose being used as fuel, figure C) but not for serine (figure D); one for licks of glucose and blood glucose but not for serine (figures E and F); and one for licks of glucose and glycogen but not for serine (figures G and H).  Interestingly, the association was strongest for the respiratory quotient and glucose intake, suggesting that glucose utilization rather than an increase in blood glucose may play a larger role in reinforcing its intake.

Additional tests reasonably eliminated the possibility of other sensory properties (like odor or texture) influencing intake.

Finally, to examine which brain systems might be involved, various microdialysis experiments were performed and dopamine measured.  Extracellular dopamine in the striatum was measured during intragastric (bypasses mouth).  Similar results for glucose preference were shown (figure A and B),  Dopamine % changes in the ventral but not dorsal striatum was increased after glucose infusion (figures C, D, E, and F) but not serine.

In order to eliminate an influence from gut hormones like ghrelin on intakes, a jugular (intravenous) infusion experiment was performed.  As expected, glucose was preferred (figure A).  When an antimetabolic glucose analog (2-DG, which inhibits glucose metabolism) is infused, dopamine decreases in the dorsal striatum which is reversed with a glucose infusion (figure B).  With an intraperitoneal injection of 2-DG, glucose intake was significantly higher versus a saline injection, suggesting an increased reward value of glucose dependent on dopamine concentration (figure C). In other words, glucose utilization (oxidation) seems to influence striatal dopamine concentrations which influences glucose intake.

Importance?

There are some important considerations for these results.  First, as the authors discuss, noncaloric artificially sweetened beverages do not necessarily curb sugar intake in all research, suggesting that simply satisfying your sweet tooth does not uninforce taste independent mechanisms that drive sugar intake.  The fact that most long-term weight reductions fail on low calorie diets may suggest altered reward systems because of these mechanisms.

This also isn’t the only study on postingestive mechanisms; others have studied other carbohydrates and proteins (including L-glutamate), with different designs.

Other studies have established that dopamine in accumbens reflects the hedonic value of sugar even if  it is not absorbed, or independent of taste.  This study confirms other findings associating dopamine levels in the dorsal striatum with feeding behavior in animals and humans, but adds that dopamine responses can be directly stimulated by the GI tract, though the specific pathways remain to be elucidated.  The role of dopamine in feeding behavior is still being studied, but the of pleasure and primary reward has shifted toward more complex behavioral influences.  Recall the study on dopamine and estimating future pleasure that I discussed last week.  Though not directly related, it seems possible that in humans these can tie together somewhat.  In our environment, however, many other variables enter in that effect food intake, such as social communication, experience, more dramatic taste hedonics from the result of food science, variety, and much more, which must be considered here.

Clearly there are a lot of other possibilities yet to discover, and this is a very fascinating area.

Reference

Ren, X., Ferreira, J., Zhou, L., Shammah-Lagnado, S., Yeckel, C., & de Araujo, I. (2010). Nutrient Selection in the Absence of Taste Receptor Signaling Journal of Neuroscience, 30 (23), 8012-8023 DOI: 10.1523/JNEUROSCI.5749-09.2010

  • http://blogspot.varigenome.com Larry Parnell

    A couple of things to add to this excellent overview of the paper by Ren, et al.:- Genetic variation in taste receptor genes is just beginning to be explored. The results of these variants might be something akin to variants in cytochrome P450 genes with their differential, allele-specific metabolism rates of different drugs. See the recent paper by Shigemura, et al. (2009) for approaches with respect to umami taste perception. Such should be applicable to other tastes as well.- One of the hallmarks of the rat genome paper from a few years back was the differences in numbers of genes encoding olfactory receptors in rat, mouse and human. That paper does not discuss taste receptors, partly because little was known at time with respect to linking gene to a specific taste. Nonetheless, there are differences in taste perception among mammals and so one must be a bit cautious, as always when interpreting data from a model organism, that the mouse results can or cannot be extended to human. Obviously, tests on humans need to be done. The umami paper cited above states that the same genes encode the umami taste receptors in both human and mouse and that can strengthen the confidence in extending the glucagon results discussed here to H. sapiens.———————————————-ReferenceShigemura N, Shirosaki S, Ohkuri T, Sanematsu K, Islam AA, Ogiwara Y, Kawai M, Yoshida R, Ninomiya Y. (2009) Variation in umami perception and in candidate genes for the umami receptor in mice and humans. Am J Clin Nutr. 90:764S-769S.

  • http://blogspot.varigenome.com Larry Parnell

    A couple of things to add to this excellent overview of the paper by Ren, et al.:- Genetic variation in taste receptor genes is just beginning to be explored. The results of these variants might be something akin to variants in cytochrome P450 genes with their differential, allele-specific metabolism rates of different drugs. See the recent paper by Shigemura, et al. (2009) for approaches with respect to umami taste perception. Such should be applicable to other tastes as well.- One of the hallmarks of the rat genome paper from a few years back was the differences in numbers of genes encoding olfactory receptors in rat, mouse and human. That paper does not discuss taste receptors, partly because little was known at time with respect to linking gene to a specific taste. Nonetheless, there are differences in taste perception among mammals and so one must be a bit cautious, as always when interpreting data from a model organism, that the mouse results can or cannot be extended to human. Obviously, tests on humans need to be done. The umami paper cited above states that the same genes encode the umami taste receptors in both human and mouse and that can strengthen the confidence in extending the glucagon results discussed here to H. sapiens.———————————————-ReferenceShigemura N, Shirosaki S, Ohkuri T, Sanematsu K, Islam AA, Ogiwara Y, Kawai M, Yoshida R, Ninomiya Y. (2009) Variation in umami perception and in candidate genes for the umami receptor in mice and humans. Am J Clin Nutr. 90:764S-769S.