Stroke is a leading cause of mortality and disability worldwide, expected to result in 61 million disability-adjusted life-years in 2020. Rapid diagnostics is the core of stroke management for early prevention and medical treatment. Serum metabolic fingerprints (SMFs) reflect underlying disease progression, predictive of patient phenotypes. Deep learning (DL) encoding SMFs with clinical indexes outperforms single biomarkers, while posing challenges with poor prediction to interpret by feature selection. Herein, rapid computer-aided diagnosis of stroke is performed using SMF based multi-modal recognition by DL, to combine adaptive machine learning with a novel feature selection approach. SMFs are extracted by nano-assisted laser desorption/ionization mass spectrometry (LDI MS), consuming 100 nL of serum in seconds. A multi-modal recognition is constructed by integrating SMFs and clinical indexes with an enhanced area under curve (AUC) up to 0.845 for stroke screening, compared to single-modal diagnosis by only SMFs or clinical indexes. The prediction of DL is addressed by selecting 20 key metabolite features with differential regulation through a saliency map approach, shedding light on the molecular mechanisms in stroke. The approach highlights the emerging role of DL in precision medicine and suggests an expanding utility for computational analysis of SMFs in stroke screening.

Although cadmium (Cd)-induced hepatoxicity is well established, pronounced knowledge gaps remain existed regarding the inherent cellular signaling that dictates Cd toxicity. Specifically, the molecular basis for determining the equilibrium between prosurvival and proapoptotic signaling remains poorly understood. Thus, it is recently revealed that long non-coding RNA (lncRNA) MT1DP, a pseudogene in the metallothionein (MT) family, promoted Cd-induced cell death through activating the RhoC-CCN1/2-AKT pathway and modulating MT1H induction. Here, first the dependency of MT1DP induction on MTF1, an important transcriptional factor in driving the mRNA expression of MT1 members is defined. Additionally, a bridge molecule between MT1DP and nuclear factor erythroid 2-related factor 2 (Nrf2) is established: miR-365. Mechanistically, MT1DP induction under Cd stress decreases the nuclear factor erythroid 2-related factor 2 (Nrf2) level to evoke oxidative stress through the elevation of miR-365, which acted to repress the Nrf2 level via direct binding to its 3'UTR. In contrast to the competing endogenous RNA (ceRNA) mechanism, a new mechanism is proposed: MT1DP elevated the miR-365 level though stabilizing its RNA via direct binding. Collectively, the combined data demonstrate a crucial role of MT1DP in reducing the Nrf2-mediated protection of cells, and this is dependent on the interplay with miR-365. Hence, the study further expands the knowledge of inducible endogenous lncRNA in modulating oxidative stress.

Arsenic (As) is a widespread metalloid contaminant, and its internal exposure is demonstrated to cause serious detrimental health problems. Albeit considerable studies are performed to interrogate the molecular mechanisms responsible for As-induced toxicities, the exact mechanisms are not fully understood yet, especially at the epigenetic regulation level. In the present study, it is identified that long non-coding RNA (lncRNA) urothelial cancer associated 1 (UCA1) alleviates As-induced G2/M phase arrest in human liver cells. Intensive mechanistic investigations illustrate that UCA1 interacts with enhancer of zeste homolog 2 (EZH2) and accelerates the latter's protein turnover rate under normal and As-exposure conditions. The phosphorylation of EZH2 at the Thr-487 site by cyclin dependent kinase 1 (CDK1) is responsible for As-induced EZH2 protein degradation, and UCA1 enhances this process through increasing the interaction between CDK1 and EZH2. As a consequence, the cell cycle regulator nuclear factor of activated T cells 2 (NFATc2), a downstream target of EZH2, is upregulated to resist As-blocked cell cycle progress and cytotoxicity. In conclusion, the findings decipher a novel prosurvival signaling pathway underlying As toxicity from the perspective of epigenetic regulation: UCA1 facilitates the ubiquitination of EZH2 to upregulate NFATc2 and further antagonizes As-induced cell cycle arrest.

Respiratory tract microbiome is closely related to respiratory tract infections, while characterization of oropharyngeal microbiome in recovered coronavirus disease 2019 (COVID-19) patients is not studied. Herein, oropharyngeal swabs are collected from confirmed cases (CCs) with COVID-19 (73 subjects), suspected cases (SCs) (36), confirmed cases who recovered (21), suspected cases who recovered (36), and healthy controls (Hs) (140) and then completed MiSeq sequencing. Oropharyngeal microbial α-diversity is markedly reduced in CCs versus Hs. Opportunistic pathogens are increased, while butyrate-producing genera are decreased in CCs versus Hs. The classifier based on eight optimal microbial markers is constructed through a random forest model and reached great diagnostic efficacy in both discovery and validation cohorts. Notably, the classifier successfully diagnosed SCs with positive IgG antibody as CCs and is demonstrated from the perspective of the microbiome. Importantly, several genera with significant differences gradually increase and decrease along with recovery from COVID-19. Forty-four oropharyngeal operational taxonomy units (OTUs) are closely correlated with 11 clinical indicators of SARS-CoV-2 infection and Hs based on Spearman correlation analysis. Together, this research is the first to characterize oropharyngeal microbiota in recovered COVID-19 cases and suspected cases, to successfully construct and validate the diagnostic model for COVID-19 and to depict the correlations between microbial OTUs and clinical indicators.