Automated High Throughput Generation of Concentration‐Response Functions for Bitter Taste in Human Subjects

Bitter taste in humans is thought to be mediated by a group of approximately 30 GPCRs expressed in type II taste cells. The molecular interactions between bitter tastant and cognate receptor should fundamentally be equivalent to those observed for any other ligand‐receptor pair. Bitter taste responses, then, in principle, should be amenable to the methods of pharmacology for quantitative analysis of receptor‐mediated processes. Applying the principles of pharmacology to the study of human taste in vivo has been impeded by the predominance of low‐throughput methodologies that rely upon large numbers of subjects subjectively evaluating a few samples at a time. The ability to accurately measure taste responses across multiple trials is a necessity for an efficient concentration‐response analysis. In this regard, bitter tastants present a further challenge due to their tendency to impart a temporally prolonged taste stimulus that can carry impact across trials. We have developed a rapid throughput technology and method for human taste measurement, called the TaStation™, with the capacity for establishing robust concentration‐response functions for bitter taste within single 45‐minute test sessions. Samples of tastant solutions (control standards of sucrose, NaCl, citric acid, quinine and water— representative stimuli for the basic tastes or sweet, salty, sour, and bitter, and neutral, respectively—and novel bitter tastant “test articles”) are distributed in a 96‐well plate, which is placed on an x‐y motion table. At the start of a trial the plate is moved to align a randomly selected well directly beneath an automated pipette mounted on a Z‐axis gantry. The pipette is lowered into the well, withdraws 200 ul of solution, and then is presented to a subject seated before a touch‐sensitive display. When prompted by a command that appears on the display, the subject removes the pipette from the gantry and self‐administers the 200 ul of taste stimulus. After tasting, the subject then is prompted to respond by touching the display. Underlying a visual field on the display is a Cartesian grid in which specific sets of coordinates have been designated to be associated with each of the control standards. Subjects have previously learned in training sessions the locations of these targets through trial‐and‐error. Correct responses on control standard trials are rewarded by the appearance of virtual poker chips carrying point values, and incorrect responses are penalized by reductions in the point tally. On trials in which test articles (novel bitter tastants) are presented, responses made anywhere on the display are rewarded. Using this approach we rapidly established robust concentrations‐response functions for 4 bitter tastants—denatonium, quinine, salicin, and caffeine—in a cohort of 5 adult subjects. Data were plotted as distance of each subject's responses from the x‐y coordinates assigned to the quinine standard and averaged across all 5 subjects. A nonlinear regression applied to the resulting data points indicated a vast range of potencies among these bitter tastants, with EC50 values of 78 nM, 56 uM, 640 uM, 26 mM, for denatonium, quinine, salicin, and caffeine respectivley. No loss of performance accuracy was detected across the 96 trials that might have been expected from cross‐trial carry over effects of lingering bitter stimulus. Our results demonstrate the effectiveness of conducting a concentration‐response analysis for bitter taste using the TaStation™. Support or Funding Information This research was supported entirely by internal funds at Opertech Bio