Data_Sheet_2_Gut Microbial Dysbiosis and Plasma Metabolic Profile in Individuals With Vitiligo.xlsx

Autoimmune diseases are increasingly linked to aberrant gut microbiome and relevant metabolites. However, the association between vitiligo and the gut microbiome remains to be elucidated. Thus, we conducted a case-control study through 16S rRNA sequencing and serum untargeted-metabolomic profiling based on 30 vitiligo patients and 30 matched healthy controls. In vitiligo patients, the microbial composition was distinct from that of healthy controls according to the analysis on α- and β-diversity (P < 0.05), with a characteristic decreased Bacteroidetes: Firmicutes ratio. Meanwhile, the levels of 23 serum metabolites (including taurochenodeoxycholate and L-NG-monomethyl-arginine) in the vitiligo patients were different from those in the healthy individuals and showed significant correlations with some microbial markers. We found that Corynebacterium 1, Ruminococcus 2, Jeotgalibaca and Psychrobacter were correlated significantly with disease duration and serum IL-1β level in vitiligo patients. And Psychrobacter was identified as the most predictive features for vitiligo by machine learning analysis (“importance” = 0.0236). Finally, combining multi-omics data and joint prediction models with accuracies up to 0.929 were established with dominant contribution of Corynebacterium 1 and Psychrobacter. Our findings replenished the previously unknown relationship between gut dysbiosis and vitiligo circulating metabolome and enrolled the gut-skin axis into the understanding of vitiligo pathogenesis.

自身免疫性疾病与异常肠道菌群（gut microbiome）及其相关代谢产物的关联日益受到学界关注。然而，白癜风（vitiligo）与肠道菌群之间的关联仍有待阐明。为此，本研究针对30例白癜风患者与30例匹配的健康对照者，通过16S rRNA测序（16S rRNA sequencing）与血清非靶向代谢组学分析开展了一项病例对照研究。α多样性与β多样性分析显示，白癜风患者的菌群组成与健康对照者存在显著差异（P < 0.05），且呈现出拟杆菌门（Bacteroidetes）与厚壁菌门（Firmicutes）比值显著降低的特征。与此同时，白癜风患者体内的23种血清代谢物（包括牛磺鹅去氧胆酸（taurochenodeoxycholate）与L-NG-单甲基精氨酸（L-NG-monomethyl-arginine））的水平与健康个体存在显著差异，并与部分菌群标志物存在显著相关性。本研究发现，棒杆菌属1（Corynebacterium 1）、瘤胃球菌属2（Ruminococcus 2）、杰氏杆菌属（Jeotgalibaca）以及嗜冷杆菌属（Psychrobacter）与白癜风患者的病程及血清IL-1β水平存在显著关联。经机器学习分析，嗜冷杆菌属（Psychrobacter）被确定为白癜风最具预测价值的特征，其重要性值为0.0236。最终，本研究整合多组学（multi-omics）数据构建了联合预测模型，该模型的预测准确率高达0.929，且棒杆菌属1（Corynebacterium 1）与嗜冷杆菌属（Psychrobacter）在其中发挥了主导作用。本研究的发现补充了此前未知的肠道菌群失调（gut dysbiosis）与白癜风循环代谢组之间的关联，并将肠-皮肤轴（gut-skin axis）纳入了白癜风发病机制的认知范畴。