At present, mutation of the first-line anti-tuberculosis drug resistance genes have been extensively studied at home and abroad, such as INH, RFP, EMB and SM. Studies showed that the mutation sites of the drug resistance genes of the above anti-tuberculosis drugs are quite different in different regions [21].This study showed that the mutation rates of isoniazid resistance gene katG, rifampicin resistance gene rpoB, ethambutol resistance gene embB, streptomycin resistance gene rpsL and RRs in multidrug resistance group were higher than those in full-sensitivity group, which consistent with the results of domestic or foreign research [22–25].However, there was no mutation in inhA gene of isoniazid-resistant group and full-sensitive group, which was inconsistent with the results of Asho Ali et al. The results showed that the mutation rate of inhA gene was about 8% [26]. This difference may relate with geographical differences. At the same time, the results showed that the mutation site of INH resistance gene katG was 315. The base pairs of INH resistance gene katG were converted from AGC to AAC and AGC to ACC, while the amino acids were converted from Ser to Asn and Thr, this result consistent with Mekonnen D et al [27]. Studies showed that the anti-tuberculosis mechanism of INH is mainly to inhibit the synthesis of cell wall of MTB and make it lose its proliferative ability, which leads to the apoptosis of MTB [28]. In this process, the KatG gene of MTB (catalase peroxidase encoding gene) plays a key role. The above-mentioned anti-tuberculosis effect of INH could be accomplished only by converting it into its biological active isonicotinic acid under the action of catalase-peroxidase. The codon mutation at the 315 locus of katG gene resulted in the conversion of amino acids from serine to asparagine (Asn) and threonine (Thr), caused a change in the protein structure encoded by the katG gene, leaded a decrease in catalase peroxidase activity, which ultimately prevented the conversion of isoniazid to its bioactive isoniazid, resulted in the loss of isoniazid resistance [29]. Because there was no inhA gene mutation in this study, there was no combined mutation of katG gene and inhA gene, which resulted in isoniazid resistance in MDR-MTB. This could provide some reference to future genetic diagnosis of MDR-TB in Urumqi. The highest mutation rate of RFP resistance gene rpoB was 531. The base of RFP was transformed from TCG to TTG, and the amino acid was transformed from Ser to Leu. This was consistent with the results of Nwofor AC and Hu Qiong et al [30]. In terms of mechanism, serine at the 531 site of rpoB gene was the most critical target for RFP to act on MTB. Once mutation occurs, it will directly lead RFP resistance in MTB. At the same time, the mutation at 526 site of rpoB gene also occurred in this study, suggesting that there may be a joint mutation at 526 site and 531site. The embB mutation of the EMB resistance gene was 206; the base group was mainly converted from ATG to ATA, and the amino acid was converted from Met to Ile. This was inconsistent with Khosravi AD et al.'s research, which showed that 306 site is a common mutation site of embB of the EMB resistance gene [31].It was suggested that the embB gene of multi-drug resistant MTB may have some regional specificity, which can be used as a reference site for gene screening and rapid diagnosis of drug-resistant TB in Urumqi. The results are consistent with those of Smittipat N, Rezaei F and others [32–33]. The results showed that the mutation sites of rpsL and RRs of SM resistance gene were 43 and 1401 ; the mutation bases of 43 sites were transformed from AGG to AAG, and the mutation base pairs of 1401 sites were transformed from A to G. The mutations of rpsL and RRs resulted in the encoding disorders of the glycosomal protein S12 and ribosomal 16SrRNA, respectively, leading to the resistance of MTB to SM, while the most common mutations of rpsL and RRs were 43 and 1401 [34].It was also found that there were joint mutations in the SM resistance genes rpsL and rrs. Mutation analysis of drug resistance genes of second-line drugs showed that only gyrA gene of OFX resistance mutated in four second-line anti-tuberculosis drugs, and the mutation sites were 74, 90, 91 and 94. This is consistent with the results of Wang Z, Sirous M and others [35–36].From the above analysis of drug resistance gene site we could see that there were not only the genetic and molecular biological characteristics of the mainstream MDR-TB strains at home and abroad, but also has a few special characteristics. These characteristics will provide a new idea for rapid detection of MDR-TB in Urumqi in the future. With the deepening of molecular biology research on multi-drug resistant MTB strains, it was also suggested that the specialized TB prevention and control institutions and specialized hospitals in Urumqi should make full use of modern MDR-TB diagnosis technology to make rapid and accurate diagnosis, so as to gain valuable time for follow-up treatment. According to the genotype distribution characteristics of MTB: In this study, 12-locus MIRU-VNTR genotyping method was used to genotype multidrug-resistant MTB in Urumqi. It was found that 19 strains of multidrug-resistant MTB could be divided into four genotypes, named Beijing type, T1, T2 and idiotype. Among them, the best potential genotype of multidrug-resistant MTB in Urumqi was Beijing type. This result consistent with Yuan L et al [37]. MTB with Beijing genotype has the characteristics of rapid transmission, strong virulence and wide epidemic range. According to the results of genotyping, the dominant strain of MDR-TB in Urumqi is Beijing, which give a new challenge to the prevention and control of MDR-TB in Urumqi. It also suggests that the disease prevention and control agencies should pay close attention to monitoring drug resistance (especially RNH and RFP) in the treatment and management of TB patients whose strain genotype is Beijing in the future, and adjust the treatment plan in time to avoid the deterioration of the disease and the multi-drug resistance of MTB. This study also found that the multidrug resistance rate of the idiotype was second just behind Beijing type. This phenomenon may be caused by the large floating population and frequent population flow in our city, leaded a wide range of genotypes of MTB strains, and the idiotype produced by genotype-related hybridization of each strain. This idiotype retains the predominant resistance characteristics of the original genotypes, which make it easier for the idiotype to produce multi-drug resistance. In terms of clustering, MIRU-VNTR technique at 12 locus was used to cluster 19 strains of MTB isolated from multidrug-resistant patients. The results showed that 2 of 19 MDR-TB patients could be clustered into one cluster. It was inferred that the homology of MDR-TB in Urumqi is 5.263%, which is slightly lower than similar studies [38]. This may be related to the implementation of the new Trinity TB management model in our city since 2012. The Trinity TB management model refers to the establishment of a clear division of labor and coordinated service system for TB prevention and control agencies, designated medical institutions and grass-roots medical and health institutions. The construction of this prevention and control system greatly improved the efficiency of TB from prevention, diagnosis, treatment to management, and could discover the source of infection in time, control it at the same time, effectively blocking the spread of tuberculosis. Under this model, the incidence of MDR-TB and the probability of homologous transmission of MDR-TB could be reduced [39]. As an effective modern TB management mode, the Trinity management mode should be implemented by the disease prevention and control agencies unremittingly. On this basis, combined with Urumqi TB epidemic situation, regional characteristics and other aspects of improvement, in order to adapt the actual needs of TB prevention and control work, greatly meet and ensure the actual interests of TB patients. Further analysis of the minimum spanning tree under complex paths showed that there were 3 optimal transmission paths for 19 MDR-TB patients. At the same time, combined with the results of genotyping, it was found that there might be a relationship between infection and infection between patients 107 and patients 195. By reviewing the questionnaires of patients No. 107 and No. 195, easy to find that they did not live in the same district and county, but the occupations of patients No. 107 and No. 195 were worker and farmers, which suggested that there may be employment relationship between them. The field epidemiological case study found that there was no social relationship between them, such as relatives, friends and colleagues, but there was employment relationship between them. At the same time, we know that No. 195 earlier infection TB at December 10, 2013, and the confirmation time was September 10, 2014. First time of No. 107 infection TB was September 2, 2014, and the confirmation time was September 19, 2014. Meanwhile, patients 107 worked in villages where patients No. 195 lived from January 2012 to September 2014, and patients No. 195 employed patients 107 during 2014. They had close contact with each other, mainly in farmland and the residence of patients No. 195. Two people recalled that the contact time between the two people in the residence of patient No. 195 was longer than that in the farmland, but the specific contact time could not be confirmed. We believe that patients 107 are more likely to be infected in the residence of patients 195.The preliminary success of the Etiology Survey provides relevant experience for the future investigation of TB outbreaks, especially MDR-TB outbreaks in our city. According to the latest reports, breakthroughs have been made in the application of molecular epidemiological methods such as gene fingerprinting, gene homology and cluster analysis in source tracking and source identification, and they have been gradually applied in practice, especially in source identification and chain tracking of major infectious disease outbreaks [40–41]. Moreover, this method had high specificity, high efficiency and high speed, which had won valuable time for rapid eradication of the epidemic situation [42]. Therefore, molecular epidemiological techniques should be further developed in the investigation of TB outbreaks in order to achieve scientific and rapid prevention and control.
At present, the study on molecular epidemic characteristics of multidrug-resistant MTB in Urumqi city level is still blank. This was the first report on the molecular biology of MDR-TB in Urumqi. At the same time, this study provided relevant experience for the future prevention and control of MDR-TB and outbreak investigation (mainly referring to the tracing of infectious sources) in Urumqi, and also provided reference for the research to other related infectious diseases. In addition, the results of this study will have great scientific value and social and economic effects on the prevention and control of MDR-TB in Urumqi. The study was influenced by time, long treatment cycle and low incidence of MDR-TB. In fact, only 319 eligible patients were surveyed in limited research time. Only 19 cases of MDR-TB were found. Although the sample size can basically meet the needs of analysis, the sample size was still insufficient. In the future, the investigation time can be prolonged appropriately, more MDR-TB cases can be collected, and more accurate data can be obtained. Secondly, although the results of this study were the same as those of the mainstream research at home and abroad, they also have their own characteristics. Because the research in this field in Urumqi is still blank, it could not provide evidence for this study. Therefore, further studies are needed to provide evidence for these specific results.