UC combined with anxiety and depression affects patients' quality of life
A total of 137 patients with UC were included, including 90 patients with active stage and 47 patients with remission. Age, gender and weight were not associated with disease activity in UC patients (P > 0.05) (Additional file 2: Table S2).
We found the proportion of anxiety during the active stage was significantly higher than in remission (P < 0.05), with notable differences in anxiety severity across varying disease activity levels (P < 0.05). Similarly, the proportion of depression in the active stage exceeded that in remission (P < 0.05), and differences in depression severity were also significant among different disease activity levels (P < 0.05). Furthermore, the proportion of sleep disturbances was significantly higher in the active stage compared to remission (P < 0.05), although no differences were found across disease activity levels (P > 0.05). Additionally, the proportion of individuals experiencing poor quality of life was significantly greater in the active stage (P < 0.05) than in remission (Additional file 2: Table S3, Table S4 and Fig. 1A).
Moreover, the GAD-7, PHQ-9, and PSQI scores were higher in active UC than in remission (P < 0.05), and the IBD-Q scores were smaller in active UC (P < 0.05) than in remission (Additional file 2: Table S5). Linear regression analyses of GAD-7 and PHQ-9 with PSQI and IBD-Q scores revealed that anxiety (R2 = 0.379, P < 0.0001), depression (R2 = 0.475, P < 0.0001) and quality of life were linearly correlated, while anxiety (R2 = 0.072, P < 0.0001), depression (R2 = 0.145, P < 0.0001) and sleep disturbances were linearly correlated (Fig. 1B). These results suggested that UC combined with anxiety and depression affected the progression of disease and the quality of life.
Bidirectional two-sample Mendelian randomization revealed UC has causal relationship on anxiety
Previous investigations have unveiled a proximal interplay among anxiety, depression, and UC through observational methods. However, the precise direction of causation remains obscure. Herein, we expounded upon the causal associations between anxiety, depression, and UC employing a bidirectional two-sample Mendelian randomization approach. According to the selection criteria of IVs, the SNPs were used as IVs (Additional file 2: Table S6-S9). Our study disclosed that UC exhibited a notable linkage with anxiety (OR: 1.069, 95% CI: 1.023-1.117, P < 0.01), while its correlation with depression was statistically nonsignificant (OR: 1.047, 95% CI: 0.928-1.183, P = 0.454) in the forward regression. Notably, no inheritable predisposition toward anxiety manifests a relationship with UC upon reverse analysis (OR: 1.005, 95% CI: 0.839-1.206, P = 0.949). In contrast, our bidirectional Mendelian randomization assessments fail to corroborate a significant association between UC and depression ( Fig. 1C ). Subsequent scrutiny through tests for heterogeneity (P > 0.05) and horizontal polytropy (P > 0.05) on IVs attests to the absence of heterogeneity or horizontal polytropy. Mendelian randomization exclusion sensitivity analysis revealed that removing any specific SNP did not affect the results (Additional file 1: Fig. S1).
Magnoflorine has potential ability to treat UC and anxiety
Utilizing the GeneCards and Batman databases, we employed a targeted approach for pharmaceutical screening to address the therapeutic management of UC and anxiety disorders (Fig. 2A). Initially, a comprehensive search operation was conducted in the GeneCards database to identify genes associated with UC and anxiety, setting a stringent threshold (Scores > 5), utilizing the keywords “Ulcerative colitis” and “Anxiety”. This exploratory phase yielded a remarkable total of 728 genes linked to UC and 190 genes linked to anxiety, with an intriguing convergence of 46 genes overlapped between the two conditions, thereby implicating these specific genetic loci in the co-occurrence of UC and anxiety. Subsequent interrogation of the Batman database targeted TCM associated with the identified 46 overlap genes. Intriguingly, our analysis underscored Ziziphus jujuba as the TCM exhibiting the strongest association with these genetic targets (P = 7.41e-28). Delving into the pertinent literature pertaining to Ziziphus jujuba, we uncovered documented evidence regarding its potential efficacy in ameliorating symptoms of UC and anxiety17,18. While the precise mechanistic underpinnings remain unexplored, our investigative foray unveiled Magnoflorine as the principal bioactive constituent of Ziziphus jujuba, renowned for its diverse pharmacological attributes encompassing anti-inflammatory and antidepressant properties19,20. Motivated by these intriguing findings, we prioritize Magnoflorine as the focal point of our investigative pursuits to delineate its therapeutic impact on experimental colitis and anxiety-related behaviors through meticulously designed animal experimentation protocols.
Magnoflorine alleviated DSS‑induced colitis
To evaluate the effect of magnoflorine on colitis, we induced a mouse model of colitis with 3% DSS orally for 7 days, and observed the phenotype of mice after administration of magnoflorine (10 mg/Kg/d) by gavage. The experimental procedure is shown in Fig. 2B. We determined the optimal concentration of magnoflorine in our previous pre-experiments.. In contrast to DSS, magnoflorine significantly alleviated DSS-induced colitis, and the relevant evidence was that magnoflorine reversed weight loss, decreased DAI scores (Fig. 2C), and alleviated colon length shortening (Fig. 2D). Pathological analyses further indicated that magnoflorine reduced colonic inflammatory cell infiltration, decreased crypt destruction, and lowered pathological scores (Fig. 2F). To further investigated the effect of magnoflorine on colonic inflammation, we examined the levels of pro-inflammatory factors in colonic tissues. As shown in Fig. 2E, compared with the DSS group, the mRNA levels of TNF-α, IL-1β, and IL-6 were reduced in the Mag+DSS group.Taken together, these results demonstrated that magnoflorine can alleviate colitis symptoms and colonic damage in mice with colitis
The colonic epithelial barrier plays a pivotal role in the pathogenesis of IBD, and tight junctions are critical components of the colonic epithelial barrier. In our study, we observed reduced expressions in tight junctions including ZO-1 and Claudin 3 in colitis mice through western blotting, while magnoflorine treatment effectively reversed this decline (Fig. 3A). Additionally, we employed immunofluorescence staining to investigate the expression of ZO-1. Our data exhibited a decline in ZO-1 expression in colitis mice (Fig. 3B), which was attenuated by magnoflorine treatment. The mucins secreted by goblet cells are crucial for maintaining intestinal barrier integrity and preventing the infiltration of intestinal microbiota. In this study, we employed the PAS staining to assess the numbers of goblet cells in colonic tissues. As shown in Fig. 3C, our findings revealed that magnoflorine effectively counteracted the reduction in goblet cell numbers observed in colitis. Furthermore, we employed immunohistochemical staining to examine the expression of MUC2, a key mucin protein, in the colonic tissues of colitis-induced mice. The results demonstrated a significant decrease in MUC2 expression, which was mitigated by the administration of magnoflorine. Collectively, these findings suggest that magnoflorine exerts its protective effects against DSS-induced colitis by preserving goblet cell function and maintaining the integrity of tight junctions.
Magnoflorine alleviated anxiety-like behaviors in mice with colitis
Patients with colitis often experience comorbid neurological disorders, such as anxiety and depression. In our study, we aimed to assess anxiety and depressive behaviors in mice subjected to colitis induction. To evaluate anxiety-like behaviors, we conducted the OFT and observed a reduction in exploration time and distance in the central area among colitis-induced mice. Additionally, in the EPT, colitis mice exhibited decreased exploration time and distance in the open arms, as well as a decrease in the number of entries into the open arms. These findings collectively indicated the presence of anxiety-like behaviors in colitis mice. However, treatment with magnoflorine reversed these results (Fig. 4A, Additional file1: Fig. S2), as evidenced by increased exploration time and distance in the center area of the OFT, as well as increased exploration time in open arms, time in close arms, and the number of entries into the open arms in the EPT. Thus, our results demonstrated that magnoflorine ameliorated anxiety-like behaviors in colitis mice.
Furthermore, we investigated depressive-like behaviors in mice using the FST and the TST (Additional file1: Fig. S2). Surprisingly, acute colitis induced by DSS did not elicit depressive-like behaviors in the mice. Moreover, magnoflorine did not exert any impact on these behaviors.
Magnoflorine ameliorated neuroinflammation, reduced microglia activation and maintained blood-brain barrier
Anxious behavior in colitis mice has been reported to be associated with the blood-brain barrier21. Thus, we investigated on the blood-brain barrier, specifically focusing on ZO-1 and PV1(Fig. 4B). To demonstrate the impact, we employed immunofluorescence staining and our findings revealed a reduction in the expression of ZO-1 and PV1 within the choroid plexus of the lateral ventricle in mice with colitis. Intriguingly, the administration of magnoflorine successfully counteracted this decrease, and these results suggested that magnoflorine could maintain the blood-brain barrier.
Moreover, we assessed the activation level of microglia in the CA1 region of the hippocampus using immunohistochemical staining to label the expression of IBA1. As shown in Fig. 4C, our findings revealed that microglia in the CA1 region of the hippocampus of colitis mice exhibited significant activation, and treatment with magnoflorine reversed this activation, as evidenced by the reduction in endpoints, and an increase of protrusion length , indicating its potential to modulate microglial activation.
Given the association between neuroinflammation and anxiety behavior, we sought to investigate the levels of pro-inflammatory factors in the hippocampal region of each experimental group. Our results indicated that colitis mice exhibited elevated expression of mRNA levels of TNF-α, IL-1β, and IL-6 (Fig. 4D). However, treatment with magnoflorine significantly reduced the expression of mRNA levels of pro-inflammatory factors. Collectively, these results suggest that magnoflorine may exert its protective effects against anxiety behavior by modulating neuroinflammation and microglial activation in the hippocampal region.
Magnoflorine regulated gut microbiota and promoted the enrichment of secondary bile acids‑producing bacteria
Next, we further explored the impact of magnoflorine on the gut microbiota composition of each group via 16S rRNA gene sequencing. Magnoflorine increased the Alpha diversity of gut microbiota as shown by shannon index (Additional file1: Fig. S3A). Principal Coordinate Analysis (PCoA) revealed a distinct segregation among all experimental groups (Additional file1: Fig. S3B). Simultaneously, we observed a consistent fecal bacterial composition among all groups of mice at both the phylum and genus levels. But in mice with colitis, the administration of magnoflorine resulted in a reduction of the Firmicutes phylum, an increase in the Verrucomicrobia phylum, and a decreased the Firmicutes/Bacteroidetes ratio (Fig. 5A). As shown in Fig. 5B and C, the application of LEfSe analysis revealed an enrichment of specific bacterial taxa within the Mag group at the family level, including Odoribacteraceae, Clostridiales_Incertae_Sedis_XIll, Erysipelotichaceae, Streptomycetaceae, Eubadteriaceae, Pdrphyromohadaceae, and Sutterellaceae, additionally, an enrichment of Ihubacter, Odoribacter. Longibaculum, Amedibacillus, Streptomyces, Anaerofustis, Parabacteroides and Parasutterella were observed at the genus level (LDA score > 2.0, P < 0.05). After DSS treatment, Bifidobacteriaceae, Oxalobacteraceae and SpirocHaetaceae were enriched at the family level in Mag+DSS group, and Bifidobacterium, Ruminococcus, Butyricicoccus, Lachnospira and Rectinema were enriched at the genus level in Mag+DSS group(LDA score > 2.0, P < 0.05). Bifidobacterium, Ruminococcus, Oxalobacteraceae and Butyricicoccus play a role in the regulation of bile acid metabolism22–25. In addition, we utilized the STAMP software to conduct an in-depth analysis of the dissimilarities between the experimental groups. Consistently, our findings revealed an enrichment of Odoribacteraceae in the Mag+DSS group (Additional file1: Fig. S3C), which is known to possess a secondary bile acid metabolism capability26. Subsequently, we employed the PICRUSt technique to predict the functional attributes of the groups, followed by a differential analysis using the STAMP software. Intriguingly, our results indicated a higher expression of the baiH gene in the Mag+DSS group compared to the DSS group (Additional file1: Fig. S3D). Notably, this gene is associated with secondary bile acid metabolism.
To delve deeper into the intricate association between microbial colonies and their corresponding traits, we employed the WGCNA approach. This method enabled us to cluster and establish a network of Operational Taxonomic Units (OTUs). Furthermore, we performed power calculations with a value of 7 to enhance the accuracy and reliability of our analyses (Additional file1: Fig. S3E). Finally, we successfully constructed sixteen OTU coexpression modules through our analysis. As shown in Fig. 5D, among these modules, the blue module (eigengene value = 0.77, P = 0.001) exhibited the strongest association with EPT(frequency of visits to open arms), while the red module (eigengene value = 0.62, P = 0.02) demonstrated the highest correlation with Mag+DSS. These two modules were subsequently chosen for further investigation and analysis. At the family taxonomic level, the Red module encompasses a total of 7 microbial colonies, while the Blue module consists of 47 colonies. Notably, 4 colonies are overlapped between these two modules (Additional file1: Fig. S3F). These overlapping colonies are identified as unclassified_Bacteroidales, Lachnospiraceae, Ruminococcaceae, and Muribaculaceae.
Magnoflorine alleviated colitis and anxiety-like behavior depending on microbiota
In light of the previous findings demonstrating the potential of magnoflorine to enhance the intestinal microbiota in mice, we sought to further investigate the role of the intestinal microbiota in this context. To achieve this, we employed a three-step experimental approach. Firstly, we administered an ABx to the mice as a pretreatment, followed by the administration of magnoflorine and induction of colitis using DSS(Fig. 6A). Our results showed no significant differences in body weight, DAI scores, colon length, and pathological damage between the ABx-D and ABx-MD groups (Fig. 6B, C and Additional file1: Fig. S4A and B ). Furthermore, behavioral tests were conducted to assess the impact of the interventions (Fig. 6D). Notably, no significant differences were observed in the time in center area of OFT and the time in the open arms of EPT between the ABx-D and ABx-MD groups, interestingly, there were significant difference between the ABx-C and ABx-D groups. These findings provide valuable insights into the potential influence of the intestinal microbiota on the observed effects of magnoflorine in the context of colitis.
To elucidate the potential involvement of the microbiota in the observed effects, we conducted FMT experiments (Fig. 6E). Remarkably, our findings demonstrated that FMT-Mag effectively ameliorated the weight loss, DAI elevation, colon shortening, and histological damage observed in the colitis mice (Fig. 6F, G and Additional file1: Fig. S4C and D). Moreover, behavioral assessments revealed notable improvements (Fig. 6H), as evidenced by an increased the time in the center area of OFT and the time in the open arms of EPT.
The previous findings indicated that FMT-Mag exhibited potential in ameliorating anxiety-like behaviors in colitis mice. However, the underlying mechanism by which the microbiota exerted its effects remained unclear, as the bacteria were unlikely to directly cross the blood-brain barrier. Therefore, we hypothesized that the beneficial effects might be mediated by the metabolites produced by the microbial colony. To investigate this, we conducted the SFF experiment (Fig. 6I). Remarkably, our results demonstrated that SFF-Mag exhibited the same effects as FMT-Mag, SFF-Mag effectively ameliorated the weight loss, DAI elevation, colon shortening, and histological damage observed in the colitis mice (Fig. 6J, K and Additional file1: Fig. S4E and F), and improved colitis-induced anxiety behiviours (Fig. 6L). That suggested that the metabolites derived from the magnoflorine-treated microbiota were responsible for the observed improvements in anxiety behavior.
These findings provide valuable insights into the potential mechanisms underlying the beneficial effects of microbiota regulated by magnoflorine and highlight the importance of microbial metabolites in mediating gut-brain axis communication.
The key bacteria altered by magnoflorine is associated with genes
We conducted RNA-seq analysis on colon tissue samples obtained from Mag+DSS group and DSS group. The results illustrated significant differences in the transcriptomes between two groups (Fig. 7A). Subsequently, we conducted a comprehensive analysis of the differentially expressed genes to identify enriched Gene Ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways (Additional file1: Fig. S5A and B). As shown in Fig. 7B, we identified specific KEGG pathways that were significantly associated with the differentially expressed genes(P < 0.05), including neuroactive ligand-receptor interaction, PI3K-Akt signaling pathway and MAPK signaling pathway. These results provide valuable insights into the molecular mechanisms underlying these biological processes and pathways, which may have implications for next study.
Furthermore, we employed the Spearman correlation analysis to examine the associations between the differentially abundant microbiota and the differentially expressed genes (Fig. 7C). Specifically, we observed a positive correlation between Gabrg2 and all the differentially abundant microbiota, except for Akkermansia. Conversely, Twist1 displayed a negative correlation with the remaining components of the differentially abundant microbiota. These results shed light on the complex dynamics between host genes and the intestinal microbiota, emphasizing the potential significance of these interactions in shaping host physiology and health.
Magnoflorine altered the bile acid metabolism in brain
Utilizing ADMETlab 2.027, we computed a LogP value of 0.395 and a BBB Penetration of 0.147 for magnoflorine, suggesting its limited ability to traverse the blood-brain barrier. Additionally, a comprehensive literature review revealed consistently low levels of magnoflorine and its metabolites within brain tissue28,29, further corroborating its restricted brain accessibility. Given the previous findings highlighting the potential involvement of the intestinal microbiota and its predicted functions, particularly in bile acid metabolism, we hypothesized that magnoflorine may exert its effects on anxiety behavior in colitis mice by modulating bile acid metabolism. To test this hypothesis, we employed targeted metabolomics to assess the levels of bile acids in the brain tissues of colitis mice. Our results revealed the detection of a total of 27 bile acids, we identified a significant increase in HDCA and 7-ketolithocholic acid (fold change >2, P < 0.05) in the Mag+DSS group (Fig. 8A). However, HDCA showed the most significant difference in abundance between groups, so we chose HDCA for the follow-up study. These findings provide novel insights into the potential mechanisms underlying the effects of magnoflorine on anxiety behavior in colitis mice, highlighting the potential role of bile acid metabolism in mediating gut-brain axis communication.
HDCA alleviated anxiety-like behavior and neuroinflammation via TLR4/Myd88 pathway
To evaluate the effect of HDCA on colitis-induced anxiety, we induced a mouse model of colitis with 3% DSS orally for 7 days, and observed the phenotype of mice after administration of HDCA (500 mg/Kg/d) by gavage (Fig. 8B). HDCA could alleviate colitis (Additional file1: Fig. S6A and B) and colitis-induced anxiety (Fig. 8C), as evidence by increasing exploration time in the center area of the OFT, as well as increasing exploration time in open arms of EPT. HDCA significantly reduced the expression of mRNA levels of pro-inflammatory factors as shown in Fig. 8D. These suggested that HDCA had a role in alleviating colitis-induced anxiety
Previous studies and the literature have reported the role of microglia-mediated neuroinflammation involved in the development of anxiety30, we conducted an experiment using BV2 cells. We exposed these cells to LPS and simultaneously treated them with HDCA. Our results revealed that LPS administration led to a significant upregulation of TNF-α, IL-1β, and IL-6 mRNA levels and active BV2 cells (Additional file1: Fig. S6C and D). However, the presence of HDCA inhibited this increase in a dose-dependent manner. This suggested that HDCA could alleviate microglia-mediated neuroinflammation.
To elucidate the underlying mechanism by which HDCA ameliorated microglia mediated neuroinflammation, we conducted a thorough investigation of relevant genes in microglia using the Genecard database and performed KEGG pathway analysis (Fig. 8E). Our analysis revealed a significant enrichment of relevant genes in the Toll-like receptor signaling pathway (P < 0.05). Subsequently, we conducted molecular docking experiments and observed that the ligand-binding domain (LBD) of TLR4/MD2 complex to HDCA was consistent with that to LPS (Fig. 8F). Based on these findings, we hypothesized that HDCA may competitively inhibit the binding of LPS to the TLR4/MD2 complex, thereby suppressing the downstream transmission of TLR4 signaling. To validate this hypothesis, as shown in Fig. 8G, we induced an inflammatory response in BV2 cells using LPS and observed a significant increase in the protein levels of Myd88 and the downstream effector molecule IL-6. The same phenomenon has been observed in animal experiment (Fig. 8H), indicating HDCA’s potential role in alleviating neuroinflammation. So, treatment with HDCA inhibited the increase in these proteins, providing further evidence for our hypothesis.