Cognitive impairment (CI) is very common in patients with Parkinson’s Disease (PD) and progressively develops on a spectrum from mild cognitive impairment (PD-MCI) to full dementia (PDD). Identification of PD patients at risk of developing cognitive decline, therefore, is unmet need in the clinic to manage the disease. Previous studies reported that oral microbiota of PD patients was altered even at early stages and poor oral hygiene is associated with dementia. However, data from single modalities are often unable to explain complex chronic diseases in the brain and cannot reliably predict the risk of disease progression. Here, we performed integrative metaproteogenomic characterization of salivary microbiota and tested the hypothesis that biological molecules of saliva and saliva microbiota dynamically shift in association with the progression of cognitive decline and harbor discriminatory key signatures across the spectrum of CI in PD. We recruited a cohort of 115 participants in a multi-center study and employed multi-omics factor analysis (MOFA) to integrate amplicon sequencing and metaproteomic analysis to identify signature taxa and proteins in saliva. Our baseline analyses revealed contrasting interplay between the genus Neisseria and Lactobacillus and Ligilactobacillus genera across the spectrum of CI. The group specific signature profiles enabled us to identify bacterial genera and protein groups associated with CI stages in PD. Our study describes compositional dynamics of saliva across the spectrum of CI in PD and paves the way for developing non-invasive biomarker strategies to predict the risk of CI progression in PD.

认知障碍 (CI) 在帕金森病 (PD) 患者中非常常见，并逐渐发展为从轻度认知障碍 (PD-MCI) 到完全痴呆 (PDD)。因此，识别有认知能力下降风险的帕金森病患者是临床治疗该疾病的未满足的需求。先前的研究报告称，帕金森病患者的口腔微生物群甚至在早期就发生了改变，口腔卫生不良与痴呆有关。然而，单一模式的数据通常无法解释大脑中复杂的慢性疾病，也无法可靠地预测疾病进展的风险。在这里，我们对唾液微生物群进行了综合的宏蛋白质组学表征，并测试了这样的假设：唾液和唾液微生物群的生物分子随着认知能力下降的进展而动态变化，并在 PD 的 CI 范围内具有歧视性的关键特征。我们在一项多中心研究中招募了 115 名参与者，并采用多组学因子分析 (MOFA) 来整合扩增子测序和宏蛋白质组分析，以识别唾液中的特征分类单元和蛋白质。我们的基线分析揭示了整个 CI 谱系中奈瑟氏球菌属与乳酸杆菌属和 Ligilactobacillus 属之间的相互作用。群体特异性特征谱使我们能够识别与 PD CI 阶段相关的细菌属和蛋白质群体。我们的研究描述了 PD 中 CI 范围内唾液的成分动态，并为开发非侵入性生物标志物策略来预测 PD 中 CI 进展风险铺平了道路。