肺腺癌（LUAD）的发展和抗肿瘤治疗疗效中，微生物组发挥着作用。肿瘤中的异常糖酵解可能促进乳酸产生，改变肿瘤微环境，影响微生物组、癌细胞和免疫细胞。我们旨在构建肿瘤内微生物组评分以预测LUAD患者的预后，并全面调查糖酵解和乳酸特征与LUAD免疫细胞浸润的关联。肺癌基因组图谱（TCGA-LUAD）微生物组数据从cBioPortal下载并进行分析，以研究其与总生存率的关联，创建一个预后评分模型。基因集富集分析（GSEA）用于查找每组涉及的主要机制。然后，我们调查了基于19个与肿瘤微环境（TME）表型和免疫疗法结果相关的基因的LUAD患者的糖酵解和乳酸模式。我们开发了一个糖酵解-乳酸风险评分和特征签名，以准确预测TME表型、预后和免疫疗法反应。通过单变量Cox回归分析，鉴定出了具有预后价值的38个属的丰度，并从TCGA-LUAD-微生物组数据集中开发了一个肺驻留微生物评分（LMS）。高LMS组中富集了糖酵解特征途径，并生成了三种不同的糖酵解-乳酸模式。Cluster1中的患者显示了不利的预后，并且可能对免疫疗法不敏感。构建了糖酵解-乳酸评分以准确预测预后，并在外部队列中进行了验证。开发了基因签名，此签名在单细胞水平上尤其在肿瘤组织中的上皮细胞中升高。最后，我们发现糖酵解-乳酸签名水平与组织学亚型的恶性程度一致。我们的研究证明了一个18微生物组预后评分和一个19基因的糖酵解-乳酸签名，用于预测LUAD患者的预后。我们的LMS、糖酵解-乳酸评分和糖酵解-乳酸签名在LUAD患者的精准治疗中具有潜在作用。版权所有©2023邓、陈、罗、阙和陈。

Microbiome plays roles in lung adenocarcinoma (LUAD) development and anti-tumor treatment efficacy. Aberrant glycolysis in tumor might promote lactate production that alter tumor microenvironment, affecting microbiome, cancer cells and immune cells. We aimed to construct intratumor microbiome score to predict prognosis of LUAD patients and thoroughly investigate glycolysis and lactate signature's association with LUAD immune cell infiltration.The Cancer Genome Atlas-LUAD (TCGA-LUAD) microbiome data was downloaded from cBioPortal and analyzed to examine its association with overall survival to create a prognostic scoring model. Gene Set Enrichment Analysis (GSEA) was used to find each group's major mechanisms involved. Our study then investigated the glycolysis and lactate pattern in LUAD patients based on 19 genes, which were correlated with the tumor microenvironment (TME) phenotypes and immunotherapy outcomes. We developed a glycolysis-lactate risk score and signature to accurately predict TME phenotypes, prognosis, and response to immunotherapy.Using the univariate Cox regression analysis, the abundance of 38 genera were identified with prognostic values and a lung-resident microbial score (LMS) was then developed from the TCGA-LUAD-microbiome dataset. Glycolysis hallmark pathway was significantly enriched in high-LMS group and three distinct glycolysis-lactate patterns were generated. Patients in Cluster1 exhibited unfavorable outcomes and might be insensitive to immunotherapy. Glycolysis-lactate score was constructed for predicting prognosis with high accuracy and validated in external cohorts. Gene signature was developed and this signature was elevated in epithelial cells especially in tumor mass on single-cell level. Finally, we found that the glycolysis-lactate signature levels were consistent with the malignancy of histological subtypes.Our study demonstrated that an 18-microbe prognostic score and a 19-gene glycolysis-lactate signature for predicting prognosis of LUAD patients. Our LMS, glycolysis-lactate score and glycolysis-lactate signature have potential roles in precision therapy of LUAD patients.Copyright © 2023 Deng, Chen, Luo, Que and Chen.