Chronic gut inflammation impairs contextual control of fear

Abstract

Chronic inflammatory diseases are highly comorbid with anxiety in humans. The extent to which chronic inflammation is responsible for this relationship remains to be determined. We therefore tested the hypothesis that prolonged, but not brief, gut inflammation is sufficient to evoke anxiety-related behaviours in mice. We used the discriminative fear to context conditioning paradigm to assess fear generalization, which is a prominent feature of anxiety disorders. Gut inflammation was induced by exposure to dextran sodium sulfate (DSS) in the drinking water, a well-established rodent model of ulcerative colitis evoking prolonged inflammation. Neither acute (1 × 5 day cycle) nor chronic (3 × 5 day cycles) exposure to DSS affected fear responses when tested shortly after conditioning. Mice in all groups generated more fear responses (freezing) in a chamber previously paired with mild shock, as compared to a chamber with no pairing. This suggests DSS exposure had no effect on acquisition or expression of conditioned fear. Acute and control animals showed this same contextual control of freezing when tested 9 days later. In contrast, at this remote time point, the chronically treated animals exhibited increased freezing in the unpaired chamber such that freezing was equivalent in both contexts. These animals, however, showed intact preference for the unpaired chamber when allowed to freely move between chambers. These data suggest that some mnemonic process engaged after training, such as memory consolidation, is affected by past chronic inflammation so as to generalize negative associations and engage fearful responding in inappropriate contexts, despite intact knowledge that the chambers have different affective associations sufficient for place preference.

Figure 1

Timeline of experimental procedure. One cycle of DSS consists of 2.75–3% wt/v DSS in the drinking water for 5 days (orange line), followed by regular drinking water. The severity of disease activity peaks 6–8 days post exposure to DSS within a cycle (indicated by red line), at which point some clinical signs of disease begin to resolve (in blue). Mice assigned to the chronic DSS treatment receive two cycles of DSS prior to pre-exposure (day 26) and training (days 27–34) in the fear to context paradigm. On day 35, chronic DSS mice receive their third cycle of DSS, and acute DSS mice receive their first cycle of DSS. Controls receive regular drinking water throughout the experiment. Fear to context testing and place preference testing occur during recent (days 40–42) and remote intervals (days 49–51). N = 12 per treatment group.

Figure 2

Schematic representation of the discriminative fear to context apparatus during training and testing. Two chambers differ in shape (rectangle, triangle), colour (white, black stripes), and odor (cinnamoaldehyde, gerenyl-formate), respectively. The scented gauze is placed at the top of the chambers, and held in place by a Plexiglas lid. (A) Pre-exposure: animals freely explore both chambers, connected by a Plexiglas hallway, for 10 min. (B) Training: paired and unpaired context training occur on alternate days (in separate rooms), for 5 min each day over 8 days. (C) Fear testing; freezing behaviour was recorded in the paired and unpaired context on separate days in the safe room (room 1). (D) Preference test; dwell time in each context was recorded (testing conducted in room 1).

Figure 3

Indicators of gastrointestinal dysfunction and inflammation. (A) Body weight and (B) Disease activity index, and (C) fecal marker of GI inflammation (lipocalin-2 levels on day 51) among treatment groups. Data represents the mean ± SEM, and are the aggregate from fully counterbalanced cohorts run in two separate experiments (n = 11–12 per group). Asterisks ‘*’ indicates p < 0.05 versus control, and hash ‘#’ indicates p < 0.05 versus acute DSS, as determined by two-way RM ANOVA with Tukey’s post-test (A), mixed model analysis with Tukey’s post-test (B), and Brown-Forsythe ANOVA (C), with Dunnett’s post-test.

Figure 4

Freezing behaviour and chamber preferences at the recent time interval 6–7 days after training. (A) Total time spent in chambers when allowed to freely explore the apparatus, prior to fear training. (B) Total time freezing in paired and unpaired contexts during discriminative fear to context testing at the recent test interval. (C) Discrimination index of freezing behaviour, computed as the relative fraction of time freezing in the paired context versus the unpaired context. (D) Dwell time in paired and unpaired chambers. *p < 0.05; **p < 0.01; ***p < 0.001. ****p < 0.0001 computd by mixed effects repeated measures ANOVA and Sidak post-test (A,B,D), or one-way ANOVA and Tukey’s post-test. Data are mean ± SEM (n = 11–12).

Figure 5

Freezing behaviour and chamber preferences at the remote testing interval (17–18 days after training). (A) Total time freezing in paired and unpaired contexts during discriminative fear to context testing at remote time point. (B) Discrimination index of freezing behaviour; Time spent freezing in unpaired-paired context, divided by the total freezing time at remote time point. (C) Dwell time in paired and unpaired chambers Mixed effects repeated measures ANOVA with Sidak post-test; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001 computed by mixed effects repeated measures ANOVA and Sidak post-test (A,C), or one-way ANOVA and Tukey’s post-test (B). Data are mean ± SEM, n = 9–12.