CUMS induced depressive-like behavior
Mice underwent 3 weeks of CUMS exposure, then were weighed and tested for sickness or depression-like behavior before being euthanized to collect brain samples and serum. The CUMS-induced sickness was measured on the first day of week one by assessing body weight loss. (Fig. 2A) and one day after the final stress session by assessing the sucrose solution consumption (Fig. 2B). To investigate CUMS-induced depressive-like behaviors, all mice underwent behavioral tests including OFT, TST, and FST (Fig. 2C-2F). As expected, there was a significant difference in body weight change after CUMS (p < 0.0001). Similarly, the sucrose consumption of CUMS-exposed mice was significantly reduced (p < 0.0001), which reflected anhedonia (Fig. 2B). After CUMS exposure, mice spent less time on the center zone than controls during the OFT test (p < 0.05) (Fig. 2C), however, the distance traveled in the open field was comparable among the groups, which indicated that mice's motor function was not impaired (p > 0.05) (Fig. 2D). The mice exposed to CUMS exhibited increased time spent immobile in TST (p < 0.01) (Fig. 2D) and FST (p < 0.0001) (Fig. 2E) compared to the control mice. These results indicate that mice exposed to CUMS displayed significant depressive-like behaviors,
CUMS induces central and peripheral cytokine expression
To validate that our model was successful and verify the relationship between central and peripheral cytokines and depression, we measured the mRNA expression and protein level of pro-inflammatory cytokines, including IL-6, IL-1β and TNF-α. As shown in Fig. 3A-C, CUMS increased the expression of IL-6 (p < 0.05), IL-1β (p < 0.05) and TNF-α (p < 0.001) mRNA compared with the Control+PBS group. Meanwhile, our data showed that stress stimuli dramatically increased the levels of IL-6 (p < 0.01), IL-1β (p < 0.0001) and TNF-α (p < 0.01) (Fig. 3D-I) in the brain and serum.
CUMS increased the release of GLU in mouse serum and brain
GLU, is a major excitatory neurotransmitter, that plays a vital role in the central nervous system[20]. However, the changes in GLU levels in depressed mice serum and brain remain poorly understood. We observed central and peripheral GLU levels by HPLC and found that GLU was a significantly elevated in the serum of CUMS mice model (Fig. 5I) (p < 0.01). In contrast, GLU was elevated in brain tissues in the model group but did not differ from control mice (p > 0.05) (Fig. 5D).
CUMS activated IDO and mediated KP and its metabolism
Given that IDO is the first rate-limiting and inflammatory-inducing enzyme in the KP[37], we explored the molecular mechanisms underlying stress-induced depression. Accordingly, we measured two IDO isoforms, IDO1and IDO2, KP and its metabolites in the serum and brain tissue of mice harvested at the end of behavioral experiments by q-PCR, ELISA and HPLC. We found that brain (Fig. 4A-B, 4D-E) and serum (Fig. 4C, 4F) expression of IDO1 and IDO2 were significantly increased in CUMS induced-mice (p < 0.01), indicating that CUMS-exposure elevated IDO expression in mice.
In addition to IDO, other downstream enzymes in the KP—including TRP, KYN, KYNA and QA, were analyzed in the CUMS group. As expected, KYN/TRP was an indicator of IDO activity[38], we found the ratio of KYN to TRP was increased in the brain and serum (Fig. 5A, E) (p < 0.01), suggesting it has a significant value as a surrogate marker of IDO activation[32, 39]. Moreover, the level of KYN was increased (p < 0.05) (Fig. 5B) while QA exhibited no significant change (p > 0.05) (Fig. 5C) during HPLC. However, the KP exhibited significant changes in the peripheral region, the level of TRP was decreased (p < 0.001) (Fig. 5F) while QA was elevated (p < 0.0001) (Fig. 5G). Of note, there was a difference observed for KYNA/QA ratio (Fig. 5H) (p < 0.001), a measure of NMDA agonist/antagonist balance. These findings established that IDO could be activated by stress and then mediated the change in KP. To better verify these results, we also performed the correlation analysis between the level of QA and GLU, QA and cytokines respectively (see SF1 and ST1)
1-MT specifically inhibits IDO and depressive-like behavior
To directly target IDO for in vivo experiments, mice were injected with DL-1-MT (50mg/kg) or L-1-MT (15mg/kg) for 3 weeks. First, we verified the inhibitory effect on IDO1 and IDO2 by q-PCR, as shown in Fig. 4A and 4D, respectively, 1-MT specifically inhibited the mRNA expression and protein level of IDO1 (p < 0.001) and IDO2 (p < 0.01); and through HPLC, we found that L-1-MT attenuated the increased ratio of kynurenine to tryptophan in the brain (Fig. 5B) (p < 0.05) and peripheral blood (Fig. 5C, F) (p < 0.001), however, DL-1-MT only affected the peripheral ratio value and not the brain, demonstrating L-1-MT had more potent inhibitory effect. What’s more, the ratio of KYNA/QA could be reversed by two inhibitors, we also did not find a difference in KYN or QA levels in the brain in our inhibitors group, suggesting the KP was activated strongly in the peripheral blood and propelled kynurenine transport from the blood to the brain, which agreement with the Robert Danzer’s results[40].
Depressive-like behaviors were also measured post-1-MT injection. In contrast with mice treated with PBS, the 1-MT intervention did not affect the CUMS-induced reduction in body weight (Fig. 2A) (p > 0.05). However, injection with DL-1-MT (p < 0.0001) or L-1-MT (p < 0.001) significantly reversed the decrease in sucrose preference percentage induced by CUMS (Fig. 2B), increased the time spent in the center area during the open field test (Fig. 2D) (p < 0.01), and reduced the time spent during TST (p < 0.01) and FST (p < 0.001) compared with the CUMS group (Fig. 2E-F). Overall, these findings indicated that IDO inhibitors could ameliorate depressive-like behavior.
1-MT blocks peripheral CUMS-induced cytokine and GLU content
Then, to determine whether DL-1-MT or L-1-MT yielded an anti-inflammatory effect, IL-6, IL-1β and TNF-α mRNA expression (Fig. 3A-C) and protein levels (Fig. 3D-I) were measured in the brains and serum in mice exposed to CUMS. Our results found that 1-MT did not modify the levels and mRNA expression of cytokines; however, as shown in Fig. 3GH, L-1-MT inhibited the protein level of IL-6 (p < 0.01) and IL-1β (p < 0.05), suggesting L-1-MT had a definite anti-inflammatory effect.
We also assessed the function of inhibitors on GLU and found that DL-1-MT or L-1-MT significantly decreased GLU levels in the serum (Fig. 5I) (p < 0.001) than in brain tissues (Fig. 5D) (p > 0.05), showing the inhibitors had a more significant effect on peripheral GLU levels.