Sputum microbiota as a potential diagnostic marker for multidrug-resistant tuberculosis

Abstract

The prevalence of drug-resistant Mycobacterium tuberculosis (Mtb) strains makes disease control more complicated, which is the main cause of death in tuberculosis (TB) patients. Early detection and timely standard treatment are the key to current prevention and control of drug-resistant TB. In recent years, despite the continuous advancement in drug-resistant TB diagnostic technology, the needs for clinical rapid and accurate diagnosis are still not fully met. With the development of sequencing technology, the research of human microecology has been intensified. This study aims to use 16 rRNA sequencing technology to detect and analyze upper respiratory flora of TB patients with anti-TB drug sensitivity (DS, n = 55), monoresistance isoniazide (MR-INH, n = 33), monoresistance rifampin (MR-RFP, n = 12), multidrug resistance (MDR, n = 26) and polyresistance (PR, n = 39) in southern China. Potential microbial diagnostic markers for different types of TB drug resistance are searched by screening differential flora, which provides certain guiding significance for drug resistance diagnosis and clinical drug use of TB. The results showed that the pulmonary microenvironment of TB patients was more susceptible to infection by external pathogens, and the infection of different drug-resistant Mtb leads to changes in different flora. Importantly, seven novel microorganisms (Leptotrichia, Granulicatella, Campylobacter, Delfitia, Kingella, Chlamydophila, Bordetella) were identified by 16S rRNA sequencing as diagnostic markers for different drug resistance types of TB. Leptotrichia, Granulicatella, Campylobacter were potential diagnostic marker for TB patients with INH single-resistance. Delftia was a potential diagnostic marker for TB patients with RFP single drug-resistance. Kingella and Chlamydophila can be used as diagnostic markers for TB patients with PR. Bordetella can be used as a potential diagnostic marker for identification of TB patients with MDR.

Keywords: 16S rRNA sequencing; Tuberculosis; drug resistance; sputum microbiota.

Figure 1

Alpha diversity analysis of sputum flora diversity index. (A) Chao1 index analysis of the sputum flora diversity, DR group has higher sputum flora abundance than DS group, **P < 0.01. (B) ACE index analysis of sputum flora abundance, DR group has higher sputum flora diversity than DS group, **P<0.01.

Figure 2

PCoA primary coordinate analysis based on Unweighted UniFrac distance. PC1 15.1%, PC2 11.11%. (blue: DS group, red: DR group)

Figure 3

Comparison of the relative abundance of major bacteria in each group of samples. (A) The flora in human sputum is mainly composed of Proteobacteria, Firmicutes, Bacteroidetes, and Actinobacteria. (B) The human sputum sample mainly contains 14 bacterial genera, Ralstonia of the Ralstoniaceae family, Ralstonia, Ochrobactrum, Prevotella, Streptococcus and Delftia of the Burkholderiaceae family account for about 50% of all the bacterial genera.

Figure 4

LEfSe analysis of sputum flora differences between samples. (A) Differential display of flora in MR-INH and DS groups. (B) Differential display of flora in MR-RFP and DS groups. (C) Differential display of flora in MDR and DS groups. (D) Differential display of flora in PR and DS groups. The classification level tree displayed by cladogram describes the hierarchical relationship of all the flora from phylum to genus (successively ordered from the inner circle to the outer circle) in the sample community. The node size corresponds to the average relative abundance of the flora, red and green respectively indicate flora with high abundance, and the difference is significant. The letters make the names of the flora with significant differences between the groups.

Figure 5

ROC analysis which can be used to distinguish potential diagnostic flora of respiratory microecology in the drug-sensitive group and different drug-resistant groups. (A) ROC curve between MR-INH and DS. (B) ROC curve between MR-RFP and DS.(C) ROC curve between MDR and DS. (D) ROC curve between PR and DS.