Abstract:

Objective: To analyze the phenotypic characteristics of linezolid resistance and the mutation profiles of resistance-associated genes in clinical Mycobacterium tuberculosis (MTB) isolates, and to investigate the correlation between phenotypic and genotypic resistance. Methods: A total of 757 MTB clinical isolates collected from Southern Xinjiang Uygur Autonomous Region between June 2017 and December 2020 were included. The minimum inhibitory concentration (MIC) of linezolid was determined using the microplate broth microdilution method. Whole-genome sequencing was performed to analyze mutations in three known resistance-related genes (rrl, rplC, and tsnR) as well as in genes interacting with rplC and/or tsnR (interaction score ≥0.70). Statistical analyses were applied to evaluate the MIC distribution, phenotypic and genotypic resistance profiles, and their correlations. Results: Among the 757 isolates, the Beijing genotype (lineage 2) was predominant (62.22%). The overall resistance rate to linezolid (MIC≥1 μg/ml) was 6.34% (48/757), with both MIC₅₀ and MIC₉₀ values of 0.5 μg/ml and MIC₉₉ of 2 μg/ml. Among multidrug-resistant (MDR) and pre-extensively drug-resistant (pre-XDR) strains, the linezolid resistance rates were 6.09% (12/197) and 17.57% (13/74) respectively. Their MIC₅₀ values were 0.25 μg/ml and 0.5 μg/ml, MIC₉₀ values were 0.5 μg/ml and 1 μg/ml, and MIC₉₉ values were 2 μg/ml and 4 μg/ml, respectively. Pre-XDR phenotype was significantly associated with linezolid resistance (χ²=17.407, P<0.001). Mutation frequencies in rrl, rplC, and tsnR genes were 8.85% (67/757), 0.53% (4/757), and 1.32% (10/757), respectively. Among the 48 linezolid-resistant isolates, 43.75% (21/48) harbored mutations in these genes. Specifically, mutations in rrl and tsnR were detected in 18 (37.50%) and 3 (6.25%) isolates, respectively, while no mutations were found in rplC. Among the susceptible isolates, 24 types of rrl mutations, 4 types of rplC mutations, and 5 types of tsnR mutations were identified. Using the broth microdilution method as the gold standard, the sensitivity and specificity of rrl mutations alone for predicting linezolid resistance were 37.50% (18/48) and 93.09% (660/709), respectively. Combining rrl and tsnR mutations increased the sensitivity to 43.75% (21/48). Among the 152 interacting genes, ten genes (Rv1540, dnaJ1, fbiC, fmu, fusA2, infC, nusG, purA, rpoC, and rpsF) were significantly associated with linezolid resistance (for nusG, Fisher’s exact test was used, P<0.05; for the other genes, the chi-square test was used, with χ² values were 5.941, 4.985, 5.027, 4.732, 3.944, 4.673, 19.814, 5.789, and 29.234 respectively, with all P<0.05). Conclusion: Pre-XDR strains have a significantly increased risk of linezolid resistance. Molecular detection methods based on rrl, rplC, or tsnR genes showed limited sensitivity for predicting phenotypic resistance, indicating the existence of unelucidated resistance mechanisms. It is recommended that in clinical practice, a combined diagnostic strategy incorporating phenotypic drug susceptibility testing and genomic analysis be adopted for high-risk patients such as those with pre-XDR, and research on novel linezolid resistance mechanisms should be accelerated to improve the accuracy of linezolid resistance identification and provide reliable evidence for rational clinical drug use.

Key words: Mycobacterium tuberculosis, Linezolid, Drug resistance, Mutation, Minimum inhibitory concentration