Exploratory development of PCR-fluorescent probes in rapid detection of mutations associated with extensively drug-resistant tuberculosis

This study aims to evaluate the clinical value of PCR-fluorescent probes for detecting the mutation gene associated with extensively drug-resistant tuberculosis (XDR-TB). The molecular species identification of 900 sputum specimens was performed using polymerase chain reaction (PCR)–fluorescent probe. The mutations of the drug resistance genes rpoB, katG, inhA, embB, rpsL, rrs, and gyrA were detected. The conventional drug susceptibility testing (DST) and PCR-directed sequencing (PCR-DS) were carried out as control. DST demonstrated that there were 501 strains of rifampicin resistance, 451 strains of isoniazid resistance, 293 strains of quinolone resistance, 425 strains of streptomycin resistance, 235 strains of ethambutol resistance, and 204 strains of amikacin resistance. Furthermore, 427 (47.44%) or 146 (16.22%) strains were MDR-TB or XDR-TB, respectively. The mutations of the rpoB, katG, inhA, embB, rpsL, rrs, and gyrA genes were detected in 751 of 900 TB patients by PCR-fluorescent probe method, and the rate of drug resistance was 751/900 (83.44%). No mutant genes were detected in the other 149 patients. Compared with DST, the mutant rates of rpoB, katG/inhA, rpsL, rrs, embB, and gyrA of six drugs were higher than 88%; five of six drugs were higher than 90% except for SM (88.11%). The MDR and XDR mutant gene types were found in 398 (42.22%) and 137 (15.22%) samples. PCR-DS was also employed and confirmed the PCR-fluorescent probe method with the accordance rate of 100%. The PCR-fluorescent probe method is rapid and straightforward in detecting XDR-TB genotypes and is worthy of being applied in hospitals.


Background
Tuberculosis (TB) is a chronic infectious disease caused by Mycobacterium tuberculosis. According to the Global Tuberculosis Report 2019, in 2018, there were 10 million new TB cases [1]. It was reported that 3.4% of new patients and 18% of retreat-patients were multidrug-resistant tuberculosis (MDR-TB) and rifampicin-resistant TB (RR-TB), among which extensively drug-resistant tuberculosis (XDR-TB) accounted for 6.2%, and 1.45 million cases died. Nearly 210,000 patients died due to MDR/RR-TB [1]. There were 866,000 new TB cases in China, 66,000 rifampicin resistance cases, and 38,800 deaths, ranking the second in the world for many years. However, only one-third of the patients had access to treatment, far from achieving the goal of eliminating TB. MDR-TB is defined as infected M. tuberculosis resistant to at least both isoniazid (INH) and rifampicin (RFP) in vitro. XDR-TB refers to the infected M. tuberculosis that is not only resistant to INH and RFP but also resistant to at least one kind of fluoroquinolones and second-line anti-TB drug injection (kanamycin, amikacin, capreomycin). Once healthy people inhaled resistant M. tuberculosis, which was exhaled into the air by tuberculosis patients through cough, sneeze, or spit, it will develop as resistant tuberculosis in a certain period of their lives [2].
With the slow growth of M. tuberculosis, diagnosis and drug resistance detection become a challenging problem in clinical treatment [3]. The delayed diagnosis resulted in improper treatment of TB patients and increasing the rate of drug resistance of M. tuberculosis which enormously influenced the treatment's effect. The emergence and transmission of MDR-TB and XDR-TB hindered TB infection control, thereby developing a stubborn disease. The MDR/XDR-TB is the critical reason for the high lethality of TB [4].
Generally, there are three types of drug resistance mechanisms in M. tuberculosis: (1) reducing cell membrane permeability and efflux pump mechanism; (2) generating catabolic and inactivated enzymes; and (3) altering drug target locus. Chromosome-mediated drug resistance is the primary basis of M. tuberculosis drug resistance [5]. At present, M. tuberculosis drug resistance mechanism research is mainly focused on the drug-targeted locus and the mutation-related genes. The current first-line drugs for TB treatment include isoniazid, streptomycin, rifampicin, pyrazinamide, and ethambutol. The second-line d rugs are consisting of fluoroquinolones (levofloxacin, moxifloxacin), ethionamide and protionamide, and injectable drugs such as aminoglycoside (kanamycin and amikacin) and polypeptide antibiotic (capreomycin). Currently, significant drug-resistant genes of M. tuberculosis have been analyzed and identified. The drugresistant mechanism of isoniazid, an anti-TB chemotherapy drug, is rather complicated, with 92% of the INH-resistant isolates associated with gene mutations from katG, inhA, and ahpC [6][7][8][9]. Rifampicin (RFP) targets a DNAdependent RNA polymerase subunit β (rpoB) in M. tuberculosis, with 95% of the RFP-resistant isolates associated with rpoB gene mutations. The detection of rifampin resistance is a paramount indication of the MDR-TB [10,11]. EmbB gene mutation interpreted the major molecular mechanism of the 50-60% of ethambutol (EMB) resistance [12,13]. Moreover, 80% of clinical streptomycin (SM)-resistant isolates of M. tuberculosis were detected mutations in the rpsL or rrs gene [14]. Fluoroquinolones (FQs) included moxifloxacin (Mfx), levofloxacin (Lfx), and others. Mutations of the gyrA gene are related to drug concentration and structure, causing medium and high drug resistance of FQs. In contrast, gyrB gene mutation might result in low drug resistance of FQs by altering drug accumulation in the cell [15,16]. Mutations associated with resistance to amikacin (AM) are located within rrs, which encodes the ribosomal 16S rRNA. The 60.5% of mutation was a single base substitution of 1401 (A→G). A minority of isolates were 1402 (C→T or A), 1484 (G→T), which occurred mainly on the high resistance strains [17]. The mutation of pncA accompanied by a decrease or loss of PZase activity is the main reason for pyrazinamide (PZA) resistance [18].
Currently, the routine clinical bacteriological laboratory examinations on TB are microscopic smear and culture methods. Conventional mycobacterium culture and drug sensitivity detection method, BactecMGIT960, has been used as the gold standard for drug resistance diagnosis in TB laboratories. Still, its detection cycle is too long to provide timely detection results for clinical practice [19]. GeneXpert MTB/ RIF assay (Cepheid, Sunnyvale, CA) is a new technology that can detect M. tuberculosis and rifampicin resistance. However, its biggest drawback is that it can only detect rifampicin resistance, not other first-line and second-line drugs [20][21][22]. In this study, we developed a mutation detection system of resistance with isoniazid (INH), rifampicin (RFP), streptomycin (SM), ethambutol (EMB), amikacin (AM), and fluoroquinolones (FQs) including moxifloxacin and/or levofloxacin by rapid detecting clinical sputum specimens with PCR fluorescence probe method, providing guidance to both establish the suitable MDR/XDR mutations detection system and carry out the practical and individualized treatment in the early stage.

Collection of sputum specimens
We collected 900 cases of morning sputum specimens from 900 TB patients (214 cases were initially treated, 686 cases were retreated) in the TB department of the 8th Medical Center of PLA General Hospital and Heilongjiang Chest Hospital from January to December 2018. The study protocol was approved by the Research Ethics Committee of the 8th Medical Center of Chinese PLA General Hospital. The informed consent was obtained from all adult participants and parents of the participant under 16 years old. All the 900 sputum samples were positive for acid-fast staining with Ziehl-Neelsen method [23].

Instruments and reagents
Mycobacterium nucleic acid detection reagent, fluorescence detection reagent of M. tuberculosis nucleic acid amplification, and real-time fluorescence quantitative PCR instrument (ABI7700) were provided by Capital Bio Corporation, Tsinghua University, Beijing. The flow chart of phenotypic drug susceptibility testing (DST), PCR-fluorescent probe method, and PCR-directed sequencing (PCR-DS) is shown in Fig. 1.

Phenotypic drug susceptibility testing
The rapid culture and drug sensitivity tests of the BACTEC MGIT 960 System (BD Diagnostic, USA) were conducted following the protocol described in the "TB Laboratory Standardization Operation and Network Establishment" [24].
Primers and probes for fluorescent probe method and reverse primers can be found in Table 1. Probes of these genes were designed based on the resistance-related high-frequency mutation sites. The 5′ ends of these probes were labelled with different fluorescent reporter groups according to the type of drug resistance. The 3′ ends were sequentially labelled with a non-fluorescent quencher (NFQ) and minor groove binder (MGB). The probe primers are listed in Table 2.
Detection of multiple mutation sites of drug resistance genes in M. tuberculosis by fluorescent probe method The RT-PCR amplification template was genomic DNA containing mutation sites of drug resistance genes or H37Rv wildstrain DNA. The mutant and wild-type genomes were sequenced. The PCR system consists of 5 sub-reaction systems, and the 5 sub-reaction systems are the same except for the different templates, primers, and probes (Table 1 and Table 2). The final concentration of each primer and probe in the PCR system in each of the above systems is 0.1 μM, and the concentration of mutant genomic DNA is 100 copies/μl. The PCR system is briefly described as follows:   The five sub-reaction systems were all performed on the ABI7500 fluorescent quantitative PCR instrument. The PCR program is as follows: 50°C 2 min, 95°C 10 min, 95°C 15 s, 63°C 1 min, and 40 cycles. Among them, 63°C 1 min is the fluorescence collection step. PCR negative control was 1× TE buffer.

Criteria for determining the results of the fluorescence probe method
The cycle threshold (Ct) value of the positive control should be lower than 40, and the Ct value of the negative control should more than 40. If any of the control is false, all the samples' results in one experiment are defined as invalid and need to be redetected. In the sub-reaction system 1, if the FAM or VIC amplification curve appears and the Ct value is less than 40, it indicates a point mutation in the rpoB gene or katG gene. In the sub-reaction system 2, if the FAM or VIC amplification curve appears and the Ct value is less than 40, it indicates that there is a point mutation in the rpoB gene or inhA gene. In the sub-reaction system 3, if the FAM or VIC amplification curve appears and the Ct value is less than 40, it means that there is a point mutation in the gyrA gene or embB gene. If the FAM amplification curve appears in the subreaction system 4 and the Ct value is less than 40, it indicates a point mutation in the rpsL gene. If the FAM amplification curve appears in the sub-reaction system 5, and the if the Ct value is less than 40, it indicates a point mutation in the rrs gene.

PCR-directed sequencing
A total of 20 μl PCR products will be sent to the Invitrogen (Shanghai) Trading Co., Ltd. to verify sequencing further.

Evaluation method
With the BACTEC MGIT 960 System (BD, USA) drug sensitivity results as the standard, the sensitivity and specificity of the PCR fluorescence probe method and the detection coincidence rate of the two methods were evaluated. PCR-DS was used to verify the accuracy of the PCR fluorescence probe method and compare the consistency rate of drug resistance detection between the PCR fluorescence probe method and PCR-DS.

Molecular species identification
Nine hundred specimens of acid-fast staining-positive sputum were identified by PCR fluorescence probe method following a previous study [26]. The results indicated that all the 900 specimens belong to M. tuberculosis complexes.

Phenotypic drug susceptibility testing
Nine hundred clinical culture isolates were analyzed by conventional drug susceptibility testing (
The mutations of the rpoB, katG, inhA, rpsL, rrs, embB, and gyrA genes were detected in 751 of 900 specimens by PCR-fluorescent probe method. No mutant genes were detected in the other 149 samples as wild-types. The rpoB, katG, or inhA mutant types were found in 398 specimens (44.22%), MDR gene mutant types, and the related cases were MDR-TB. The rpoB, katG/inhA, gyrA, rpsL, and rrs mutant types were simultaneously found in 137 specimens (15.22%), and the related cases were XDR-TB.

Comparison of the coincidence rate
According to Table 4, we can conclude that the lowest positive coincidence rate of the PCR-fluorescent probe method comparing with DST for the detection of amikacin was 75%, and that of the other drugs were higher than 75%, which showed that these two methods have good consistency in positive coincidence rate. At a negative coincidence rate, all six drugs were larger than 97%, indicating that those methods have a good consistency. All six drugs were larger than 88% in total coincidence rate, five of which were larger than 90%. Above all, the results revealed that both the PCR-fluorescent probe method and DST have good consistency in drug resistance's total coincidence rate. As Table 5 showed, both the positive coincidence rate and the negative coincidence rate of the PCR-fluorescent probe method compared with PCR-DS were 100%. The total coincidence rate was 100%. Additionally, the statistical results illustrated that all the seven drug-resistant genes' detection results by PCR-fluorescent probe method have good consistency with PCR-DS.

Discussion
The diagnosis and treatment of drug-resistant TB, especially MDR/XDR-TB, are critical and challenging factors in preventing and controlling TB. Currently, conventional culture-based techniques have long turnaround times. We cannot offer timely and effective treatment programs for TB patients, especially those combined with HIV patients [27]. In 2010, WHO endorsed the Gene Xpert MTB/RIF assay [28] that used real-time PCR to identify M. tuberculosis complex DNA and the mutations associated with RFP resistance directly from sputum specimens. Nevertheless, Gene Xpert MTB/  RIF assay cannot detect mutations related to other anti-TB drugs such as INH, EMB, SM, AM, and FQs. With the indepth research on the molecular mechanism of drug resistance, the establishment of a simple, fast, and accurate method for the detection of drug-resistant mutations becomes more and more significant to improve the cure rate, reduce the occurrence of drug resistance, and decrease the risk of recurrence and death rate. PCR fluorescence probe technology uses double PCR technology and the Taqman probe technique to detect M. tuberculosis and drug resistance by monitoring different fluorescent channels' fluorescence signal. This technique has strong specificity and sensitivity, and is easy to operate. Our previous study [26] showed that PCR fluorescence probe technology is an essential clinical value in the rapid diagnosis of TB in sputum specimens. However, few studies on the exploratory development of PCR-fluorescent probes in the rapid detection of mutations are associated with XDR-TB.
The conventional DST method takes 3 to 4 weeks. This study is based on the PCR-fluorescent probe method, which has a low cost, simple operation, and only 1.5 h to detect the nucleic acid in specimens. Besides, this method can significantly shorten the detection cycle comparing with phenotypic DST. In this study, we established and evaluated the detection system of multidrug resistance and extensively drug resistance in M. tuberculosis, reflecting that (1) the drug resistance of TB was detected by fluorescence PCR detection technology covered six anti-TB drugs with seven drug resistance genes. Additionally, a multidrug-resistant and extensively drugresistant mutation detection system was established, and the results showed that the system had high specificity and sensitivity values; (2) The clinical diagnostic performance of PCR detection system was evaluated by testing 900 clinical specimens of M. tuberculosis, and the results were compared with the absolute concentration method of DST. In total coincidence rate, all of the six drugs were larger than 88%, five of which were larger than 90%.
In this study, compared with phenotypic DST, the coincidence rates of rpoB (RFP), katG/inhA (INH), embB (EMB), gyrA (Lfx), rpsL (SM), and rrs (AM) detected by fluorescent probe method were 95.89%, 91%, 92.11%, 90.89%, 88.11%, and 93.22%, respectively. The coincidence rate of two methods of RFP, INH, EMB, FQs, and AM resistance testing is higher than 90%, only SM coincidence rate was 88.11%. The possible reason may be that rpoB, katG/inhA, embB, gyrA, and rrs gene mutations of M. tuberculosis were the main resistance mechanism of RFP, INH, EMB, FQs, and AM. In addition to rpsL gene mutation (50-78%), rrs gene mutation (20-30%) is also the main molecular mechanism of drug-resistant to SM [29,30]. However, this study had not detected the rrs gene mutation locus of drug resistance to SM of M. tuberculosis, which may be one reason for the low consistency of SM drug resistance detected by PCR-fluorescent probe method compared with phenotypic DST.
In this study, phenotypic DST was used as the standard, the detection rate of MDR-TB by phenotypic DST was 47.44% (427/900), and the detection rate of rpoB and katG/inhA mutation was 42.22% (398/900) by PCR fluorescence probe method. The sensitivity, specificity, and coincidence rate of the PCR-fluorescent probe method for detecting MDR-TB were 90.16%, 97.25%, and 93.89%, respectively. The detection rate of XDR-TB by phenotypic DST was 16.22% (146/900). The detection rate of rpoB, katG/inhA, rpsL, rrs, embB, and gyrA by PCR fluorescence probe method was 15.22% (137/900), the sensitivity and specificity of PCR-fluorescent probe method for detection of XDR-TB were 87.67%, 98.81%, and the coincidence rate was 97.00%.
It is well-known that drug-resistant gene mutations are an effective form of TB resistance. As for the PCR fluorescence probe method, the positive detection rate was lower than DST because drug-resistant gene mutation was just a form of drug resistance. Reducing cell membrane vulnerability and efflux pumps and inactivated enzyme changes were also the causes of TB resistance. On the other hand, only a few loci of 7 common drug-resistant genes of 6 drugs were detected, while other drug-resistant genes or other drug-resistant loci were not developed in this study, such as ndh, efpA, kasA, iniABC operon (for INH resistance) [31], rpoC (for RFP resistance) [32], embA, embC, ubiA (for EMB resistance) [33], and gyrB (for FQs resistance) [34]. These may be reasons for the lower detection rate of the PCR-fluorescent probe method than phenotypic DST.

Conclusion
In summary, with the advantages of simplicity, accuracy, specificity, and high throughput, the PCR-fluorescent probe method gives impetus to the widespread clinical application of molecular diagnostic technology. Our results revealed the association between mutation types, drug-resistant type and dosage, clinical treatment, and prognosis by distinguishing and investigating different mutation types of anti-TB drugs. These findings provided a novel perspective on anti-TB drug development to achieve TB prevention and control truly. Herein, we established a comprehensive detecting system of XDR-TB, including both first-line and second-line anti-TB drugs. Additionally, developed genetic tests will inevitably produce more rapid results for drug-resistant isolates, which will lead to faster identification of MDR and XDR strains, more tailored treatment regimens, and a reduction in the transmission of TB.