The bradykinin system in stress and anxiety in humans and mice

Abstract

Pharmacological research in mice and human genetic analyses suggest that the kallikrein-kinin system (KKS) may regulate anxiety. We examined the role of the KKS in anxiety and stress in both species. In human genetic association analysis, variants in genes for the bradykinin precursor (KNG1) and the bradykinin receptors (BDKRB1 and BDKRB2) were associated with anxiety disorders (p < 0.05). In mice, however, neither acute nor chronic stress affected B1 receptor gene or protein expression, and B1 receptor antagonists had no effect on anxiety tests measuring approach-avoidance conflict. We thus focused on the B2 receptor and found that mice injected with the B2 antagonist WIN 64338 had lowered levels of a physiological anxiety measure, the stress-induced hyperthermia (SIH), vs controls. In the brown adipose tissue, a major thermoregulator, WIN 64338 increased expression of the mitochondrial regulator Pgc1a and the bradykinin precursor gene Kng2 was upregulated after cold stress. Our data suggests that the bradykinin system modulates a variety of stress responses through B2 receptor-mediated effects, but systemic antagonists of the B2 receptor were not anxiolytic in mice. Genetic variants in the bradykinin receptor genes may predispose to anxiety disorders in humans by affecting their function.

Figure 1

The effect of acute stress on mouse brain bradykinin receptor gene expression. mRNA levels of Bdkrb1, Bdkrb2 and Fos in cortex (Cx) and hippocampus (Hp) directly (2 h), 1 h (2 h + 1 h) or 5 h (2 h + 5 h) after 2 h restraint stress in C57BL/6NCrl mice. N = 8 in Cx Bdkrb1, N = 7–12 in Hp Bdkrb1 and Hp Bdkrb2, and N = 3 in Hp Fos. The mRNA expression level of the control group was defined as 1. Bars represent the mean ± SEM of the normalised mRNA expression level. *p < 0.05 compared to non-stressed control mice.

Figure 2

The effect of acute stress on bradykinin receptor protein expression in the hippocampus. We measured normalized protein expression levels of BDKRB1 and BDKRB2 using Western blot in hippocampus (a,b) and control stress-associated genes in hypothalamus (c–e) directly (2 h), 1 h (2 h + 1 h) or 5 h (2 h + 5 h) after 2 h restraint stress in C57BL/6NCrl mice. BRKRB1 HMW = bradykinin receptor B1 high molecular weight, BRKRB1 LMW = bradykinin receptor B1 low molecular weight, CRYAB = α-B-crystallin, FOS = FBJ osteosarcoma oncogene, HMGB1 = high mobility group B1. N = 3–5 in each experiment. Bars represent mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001compared to non-stressed control mice. For uncropped blots, see Supplemental Fig. 1.

Figure 3

Expression levels of KLK-related genes after chronic social defeat stress in mice. Agtr2, Bdkrb1, Klk1, Klk12, Klkb1, Kng1 and Kng2 were included in the analysis but their expression levels were below the detection level in both brain regions. F12, Bdkrb2 and Aplnr expression levels were below the detection level in the mPFC. LogCPM = normalised expression level in log Counts Per Million from RNA-sequencing. P-values (from moderated t-test for differential expression) are corrected for multiple testing with Benjamini-Hochberg method. Box plots show minimum, 1st quartile, median, 3rd quartile, and maximum expression value for each group. Data outliers (values > 1.5 interquartiles from the IQR range) are depicted by a dot. vHPC = ventral hippocampus, mPFC = medial prefrontal cortex, B6 = C57BL/6NCrl, D2 = DBA/2NCrl. N of mice: mPFC (B6: controls 6, resilient 6, susceptible 6; D2: controls 6, susceptible 8) and vHPC (B6: controls 6, resilient 8, susceptible 3; D2: controls 6, susceptible 5). *p < 0.05, **p < 0.01, ***p < 0.001 compared to the same-strain control group.

Figure 4

The effect of bradykinin receptor antagonists on anxiety-like behaviour. We injected DBA/2NCrl mice subcutaneously with B1 receptor non-peptide receptor antagonist ELN-441958 or B2 receptor non-peptide antagonists bradyzide or WIN 64338 and carried out behavioural testing 30 min later in the open field test (N = 8–12 per group) (a,b), in the novelty suppressed feeding test (N = 8–10 per group) (c) and in the elevated plus maze test (N = 7–9 per group) (d,e). Bars represent the mean ± SEM. *p < 0.05, **p < 0.001 compared to the vehicle (PBS/0.05% DMSO) group.

Figure 5

The effect of bradykinin B2 receptor antagonists on mouse stress-related behaviour. DBA/2NCrl mice were injected intraperitoneally with drug or vehicle, subjected to 30 min restraint stress and let recover for 1 h, followed by behavioural testing in the open field (a,b) or forced swim test. (c) N = 8–9 per group. Effect of WIN 64338 on body temperature in stress induced hypothermia (SIH) model. (d) Temperature difference ∆T (T₂-T₁) is shown. N = 13 per group. Bars represent the mean ± SEM. *p < 0.05 compared to the vehicle control group.

Figure 6

Stress-induced changes in the brown adipose tissue (BAT) gene expression levels. (a) Expression levels of Kng2 variants and Bdkrb2 transcripts in the interscapular BAT of cold-stressed and control C57BL/6NCrl mice (N = 11–13 per group). The y-axis shows the relative mRNA expression levels. Due to different amounts of input cDNA, expression levels of the different transcripts are not readily comparable. ***p < 0.001 compared to the room temperature (RT) control. (b) The effect of stress-induced hyperthermia (SIH) and WIN 64338 on BAT gene expression levels of DBA/2NCrl mice. N = 6–9 per group. For statistical significances, see the main text. Bars represent the mean ± SEM. *p < 0.05 calculated by Bonferroni post-hoc comparison.