Effect of Tryptophan Depletion on Conditioned Threat Memory Expression: Role of Intolerance of Uncertainty

Abstract

Background: Responding emotionally to danger is critical for survival. Normal functioning also requires flexible alteration of emotional responses when a threat becomes safe. Aberrant threat and safety learning occur in many psychiatric disorders, including posttraumatic stress disorder, obsessive-compulsive disorder, and schizophrenia, in which emotional responses can persist pathologically. While there is evidence that threat and safety learning can be modulated by the serotonin systems, there have been few studies in humans. We addressed a critical clinically relevant question: How does lowering serotonin affect memory retention of conditioned threat and safety memory?

Methods: Forty-seven healthy participants underwent conditioning to two stimuli predictive of threat on day 1. One stimulus but not the other was subsequently presented in an extinction session. Emotional responding was assessed by the skin conductance response. On day 2, we employed acute dietary tryptophan depletion to lower serotonin temporarily, in a double-blind, placebo-controlled, randomized between-groups design. We then tested for the retention of conditioned threat and extinction memory. We also measured self-reported intolerance of uncertainty, known to modulate threat memory expression.

Results: The expression of emotional memory was attenuated in participants who had undergone tryptophan depletion. Individuals who were more intolerant of uncertainty showed even greater attenuation of emotion following depletion.

Conclusions: These results support the view that serotonin is involved in predicting aversive outcomes and refine our understanding of the role of serotonin in the persistence of emotional responsivity, with implications for individual differences in vulnerability to psychopathology.

Keywords: Emotion; Fear conditioning; Memory; Serotonin; Threat conditioning; Tryptophan depletion.

Figure 1

Task schematic. Each row represents a different phase of the experiment. Lightning bolts represent shock. The extinguished conditioned stimulus (CS+E) is the CS+ that was presented during the extinction phase. The not-extinguished CS+ (CS+N) is the CS+ that was not presented during the extinction phase. The CS− was never paired with shock. ATD, acute tryptophan depletion.

Figure 2

Robust tryptophan depletion was achieved, verified via plasma samples. More negative values indicate greater depletion of tryptophan. Error bars indicate ±1 SE. ΔTRP:LNAA, change in the ratio between tryptophan and all large neutral amino acids from before to after depletion.

Figure 3

(A) Skin conductance responses (SCRs) in the initial threat conditioning phase (acquisition) on day 1, conducted before serotonergic challenge. There were no differences between the future placebo and future acute tryptophan depletion (ATD) groups, and both groups showed significant threat conditioning to both paired conditioned stimuli (CS+s) compared with the CS−, as predicted. This equivalent baseline conditioning on day 1 enabled testing the effects of ATD on its retention on day 2. Brackets denote follow-up t tests contrasting stimuli within group, after observing a main effect of stimulus. ∗∗∗p < .001. Error bars indicate ±1 SE. (B) SCRs in the extinction phase on day 1. Smaller brackets refer to the beginning and end of extinction, and the larger bracket denotes a mild reduction in SCRs in late compared with early extinction. ∗p < .05. Error bars indicate ±1 SE. Raw data (following transformation) (see Supplement) are displayed. CS+E, extinguished CS+; CS+N, not extinguished CS+; n.s., not significant.

Figure 4

Tryptophan depletion reduced skin conductance response (SCR) expression. SCRs are displayed from the (A) spontaneous recovery, (B) reinstatement, and (C) reacquisition phases. Large brackets denote a main effect of stimulus. ∗∗p < .01; ∗p < .05. Error bars indicate ±1 SE. Raw data (following transformation) (see Supplement) are displayed, not adjusted values after controlling for intolerance of uncertainty, sex, or strength of initial conditioning on day 1. CS, conditioned stimulus; CS+E, extinguished CS+; CS+N, not extinguished CS+.

Figure 5

Skin conductance responses (SCRs) on day 2 plotted against the extent of depletion during the (A) spontaneous recovery, (B) reinstatement, and (C) reacquisition phases. Lower x-axis values indicate greater depletion, indexed by the change in the ratio of tryptophan to large neutral amino acids (ΔTRP:LNAA) assessed via plasma samples. Raw data (following transformation) (see Supplement) are displayed, not adjusted values after controlling for intolerance of uncertainty, sex, or strength of initial conditioning on day 1. The CS− is denoted by purple circles, the extinguished CS+ (CS+E) is denoted by blue triangles, and the not-extinguished CS+ (CS+N) is denoted by green squares. Significant relationships between depletion and SCR were seen in the spontaneous recovery phase for the CS+E and CS+N but not for the CS−, in reinstatement for the CS+E and CS−, and in reacquisition for the CS+E and CS−. CS, conditioned stimulus.

Figure 6

Skin conductance responses (SCRs) during spontaneous recovery (day 2) plotted against self-report on the Intolerance of Uncertainty Scale (IUS), shown separately for (A) placebo and (B) depletion. Raw data (following transformation) (see Supplement) are displayed. The CS− is denoted by purple circles, the extinguished CS+ (CS+E) is denoted by blue triangles, and the not-extinguished CS+ (CS+N) is denoted by green squares. CS, conditioned stimulus.