基于新疆维吾尔自治区南疆地区169株结核分枝杆菌探讨基因缺失对贝达喹啉的耐药机制

Mechanism of gene deletion induced resistance to bedaquiline based on 169 strains of Mycobacterium tuberculosis in southern Xinjiang

目的挖掘与结核分枝杆菌(MTB)对贝达喹啉耐药相关的基因缺失片段。方法2017—2020年期间从新疆维吾尔自治区南疆地区收集807株MTB分离株,对菌株进行全基因组测序,采用TB-Profiler 4.1.2检测菌株的家系和预测异烟肼、利福平、氟喹诺酮类等7种药物的敏感性;选择44株准广泛耐药(Pre-XDR)菌株和125株随机菌株,采用Delly 0.9.1软件检测基因缺失情况,采用C++脚本程序检测Rv0678、atpE和pepQ基因突变情况,采用微量肉汤稀释微孔板法检测菌株对贝达喹啉的最小抑菌浓度(MIC)及耐药情况。采用χ^(2)检验和秩和检验等方法分析基因缺失与贝达喹啉耐药的相关性及不同家系菌株对贝达喹啉耐药性及敏感程度的差异。结果44株Pre-XDR株中共鉴定出23株贝达喹啉耐药株(MIC≥0.25μg/mL)和49种基因缺失类型,发生缺失菌株数>5株的缺失类型共有8种,其中2729618 bp~2729833 bp(Rv2434c)缺失组贝达喹啉耐药率和MIC水平低于未缺失组(χ^(2)=6.145,P=0.013和Z=−2.224,P=0.026);2235169 bp~2235233 bp(ctpG)、1710766 bp~1711557 bp(Rv1519-Rv1520)和3952924 bp~3954223 bp(echA19-Rv3517)多基因缺失组贝达喹啉耐药率和MIC水平均高于未缺失组(χ^(2)=4.322,P=0.038和Z=−2.526,P=0.012)。但是在125株验证用菌株(含贝达喹啉耐药株23株)中均未发现差异具有统计学意义(χ^(2)=1.660,P=0.198和Z=−1.152,P=0.249)。将44株菌株和125株菌株合并分析,仅发现2235169 bp~2235233 bp(ctpG)、1710766 bp~1711557 bp(Rv1519-Rv1520)和3952924 bp~3954223 bp(echA19-Rv3517)多基因缺失组贝达喹啉耐药率高于未缺失组(χ^(2)=6.396,P=0.011),该多基因缺失仅发生于L3家系菌株,同时两两比较发现L3家系菌株的贝达喹啉耐药率显著高于L4家系(P<0.017)。结论2235169 bp~2235233 bp(ctpG)、1710766 bp~1711557 bp(Rv1519-Rv1520)和3952924 bp~3954223 bp(echA19-Rv3517)多基因缺失为南疆地区L3家系菌株特有的基因缺失类型(与该地其他家系相比),其与贝达喹啉耐药性的关系不稳健,L3家系菌株可能比L4家系对贝达喹啉耐药的易感性更高。由于样本量较小,需扩大样本量做进一步验证,以上结果为了解贝达喹啉的耐药机制及控制相关耐药菌株的传播提供了线索。

Objective To understand gene deletion associated bedaquiline resistance in Mycobacterium tuberculosis.Methods During 2017−2020,a total of 807 strains of M.tuberculosis were collected from the southern area of Xinjiang Uygur autonomous region for whole genome sequencing,then TB-Profiler was used to detect the strain lineages and predict the strains’susceptibilities to 7 antibiotics,including isoniazid,rifampicin,fluoroquinolone,and so on.In the 807 isolates,44 preextensively drug resistant(Pre-XDR)strains were selected and 125 isolates were randomly selected for the tests of gene deletion by using Delly 0.9.1 software,the mutations in Rv0678,atpE and pepQ by using C++script program,and the minimum inhibitory concentrations(MICs)of bedaquiline to the strains by using the micro-broth dilution microplate method.The association between gene deletion and bedaquiline resistance and the difference in bedaquiline resistance rates and MIC levels among the strains in different lineages were analyzed byχ^(2) test and rank-sum test.Results A total of 49 gene deletion types,including 8 types involving>5 strains,and 23 bedaquiline resistant strains were identified in 44 Pre-XDR strains.In the 44 Pre-XDR isolates,the bedaquiline resistance rate and MIC level of the deletion group of 2729618 bp–2729833 bp(Rv2434c)were lower than those of the non-deletion group(χ^(2)=6.145,P=0.013 and Z=−2.224,P=0.026);multi-gene deletion group of 2235169 bp–2235233 bp(ctpG),1710766 bp–1711557 bp(Rv1519–Rv1520)and 3952924 bp–3954223 bp(echA19–Rv3517)showed higher bedaquiline resistance rate and MIC level compared with non-deletion group(χ^(2)=4.322,P=0.038 and Z=−2.526,P=0.012).However,no significant difference was found among the 125 strains(including 23 bedaquiline resistant strains)used for validation(χ^(2)=1.660,P=0.198 and Z=−1.152,P=0.249).Combined analysis on the 44 strains and 125 strains revealed that only the multi-gene deletion group of 2235169 bp–2235233 bp(ctpG),1710766 bp–1711557 bp(Rv1519–Rv1520)and 3952924 bp–3954223 bp(echA19–Rv3517)had higher bedaquiline resistance rate compared with non-deletion group(χ^(2)=6.396,P=0.011),and the multi-gene deletion occurred only in the lineage 3 strains,meanwhile,one-by-one comparisons revealed that the resistance rate to bedaquiline was significantly higher in the lineage 3 strains than in the lineage 4 strains(P<0.017).Conclusion Multi-gene deletions of 2235169 bp–2235233 bp(ctpG),1710766 bp–1711557 bp(Rv1519–Rv1520)and 3952924 bp–3954223 bp(echA19-Rv3517)were unique in the lineage 3 strains in southern Xinjiang(compared with other lineages in this area),and its relationship with bedaquiline resistance was not stable.The lineage 3 strains might be more inclined to resistant to bedaquiline compared with the lineage 4 strains.The sample size of the study was small and further validation in larger samples is needed.These results provide clues for the understanding of the mechanism of bedaquiline resistance and the control of the spread of bedaquiline resistant strains.