Alterations in the gut microbiota of AIDS patients with pneumocystis pneumonia and correlations with the lung microbiota

BACKGROUND: Due to the inability to be cultured in vitro, the biological characteristics and pathogenicity of Pneumocystis jirovecii remain unclear. Intestinal microflora disorder is related to the occurrence and development of various pulmonary diseases. This work explores the pathogenesis of pneumocystis pneumonia (PCP) in acquired immune deficiency syndrome (AIDS) patients from a microbiome perspective, to provide better strategies for the diagnosis, treatment, and prevention of PCP. METHODS: Subjects were divided into three groups: human immunodeficiency virus (HIV)-infected patients combined with PCP, HIV-infected patients without PCP, and HIV-negative. Stool and bronchoalveolar lavage fluid (BALF) samples were collected, total DNA was extracted, and 16S rRNA high-throughput sequencing was performed using an Illumina MiSeq platform. PICRUSt and BugBase were used to predict microflora functions, and correlation analysis of intestinal and lung bacterial flora was conducted. RESULTS: Compared with the HIV- group, prevotella and another 21 genera in the intestinal microbiome were statistically different in the HIV+ group; 25 genera including Escherichia-Shigella from HIV+PCP+ group were statistically different from HIV+PCP- group. The abundance of Genera such as Porphyromonas was positively or negatively correlated with CD16/CD56+ (μL). Compared with the HIV- group, identification efficiency based on area under the curve (AUC) >0.7 for the HIV+ group identified seven genera in the gut microbiota, including Enterococcus (total AUC = 0.9519). Compared with the HIV+PCP- group, there were no bacteria with AUC >0.7 in the lung or intestine of the HIV+PCP+ group. The number of shared bacteria between BALF and fecal samples was eight species in the HIV- group, 109 species in PCP- patients, and 228 species in PCP+ patients, according to Venn diagram analysis. Changes in various clinical indicators and blood parameters were also closely related to the increase or decrease in the abundance of intestinal and pulmonary bacteria, respectively. CONCLUSIONS: HIV infection and PCP significantly altered the species composition of lung and intestinal microbiomes, HIV infection also significantly affected intestinal microbiome gene functions, and PCP exacerbated the changes. The classification model can be used to distinguish HIV+ from HIV- patients, but the efficiency of bacterial classification was poor between PCP+ and PCP- groups. The microbiomes in the lung and gut were correlated to some extent, providing evidence for the existence of a lung-gut axis, revealing a potential therapeutic target in patients with HIV and PCP.