Serial feature positive and feature negative discrimination learning in a taste avoidance preparation: implications for interoceptive control of behavior

Background: Psychoactive drugs produce interoceptive stimuli that can guide appropriate behaviors by initiating or inhibiting responding.

Objective: The current study investigated whether an interoceptive morphine state produces similar patterns of serial feature positive (FP) and feature negative (FN) discrimination learning under comparable conditions in a taste avoidance design.

Methods: Male Sprague-Dawley rats were trained under 10 cycles of FP or FN discrimination. In the FP task, morphine (10 mg/kg, IP) signaled that a saccharin solution was followed by LiCl (1.2 mEq, IP), while the vehicle (saline) signaled that the LiCl was withheld. In the FN task, the contingency was reversed.

Results: The FP-trained rats acquired the discrimination after three training cycles, consuming significantly less saccharin on morphine, than on vehicle, sessions ( P < 0.05). The FN-trained rats acquired the discrimination after six training cycles, consuming more on morphine than on vehicle sessions ( P < 0.05). However, FN-trained rats never recovered saccharin consumption to baseline levels and 40% of the rats continued to avoid saccharin (consuming 0 ml) on morphine sessions. Control rats that never received LiCl consumed high levels of saccharin on morphine and vehicle sessions, indicating that morphine did not produce unconditioned suppression of saccharin consumption.

Conclusion: The difficulty to acquire FN discrimination might reflect the limitations of learning about safety contingencies in the taste avoidance design. The rapidity of FP learning when a drug state signals an aversive contingency may have implications for the general role of interoceptive stimuli in the control of behavior.