Identification and Characterization of Human Fructose or Glucose Taste Variants with Hypogeusia for One Monosaccharide but Not for the Othera

Human psychophysical functions for sweetness are similar for sucrose and fructose, but different for glucose, and suggest different mechanisms for fructose and glucose. Drosophila behavioral and electrophysiological data are similar to the human data and indicate separate receptor cell mechanisms for the monosaccharides. Moreover, fructose 'nontasters' (NTs) and glucose NTs have been identified in two Drosophila species. Identification of human NTs would confirm separate mechanisms and could lead to identification of proteins in human sweet taste by molecular genetic techniques. To identify human NTs, we first obtained responses for sucrose, fructose and glucose from 20 subjects. They tasted seven concentrations of each sugar (2-128 mM), paired with water, and indicated the sweeter of each pair. Functions for recognition indices (RIs) (proportion of subjects recognizing the sugar as sweeter) were similar for sucrose and fructose and different for glucose; this result agrees with the previous studies and supports different mechanisms for the monosaccharides. At 128 mM, RIs for all three sugars were 1.0; this result is consistent with the monogeusia reported by Breslin et al. for concentrations higher than those tested here. Eleven rising-phase concentrations (10-35 mM fructose, 10-90 mM glucose) then were tested on 32 subjects. A statistically significant interaction indicated different regression slopes and supported different monosaccharide mechanisms. From these data, positive identification values (PIDs) (lowest concentration at which the sugar always was judged sweeter than the water) were determined for each subject. The fructose log(PID) and glucose log(PID) data were not well correlated; thus separate mechanisms were supported further. Next, NT traits were defined by log(PID)s > or = 2 SD above the mean for one sugar, while the PID for the other remained within 1 SD of the population mean log(PID). Ninety-two subjects were screened to identify 12 glucose NTs and four fructose NTs. Two glucose NTs and three average subjects were tested in six additional sessions. The NTs showed an experience-induced change: there was a statistically significant reduction of glucose PIDs, but not of fructose PIDs. No change occurred in PIDs of the average subjects for either sugar.