Skip to main content
NCBI home page
Journal of Clinical Microbiology logoLink to Journal of Clinical Microbiology
. 2023 Dec 19;62(1):e01183-23. doi: 10.1128/jcm.01183-23

Single-tube protocol for culture-independent spoligotyping of Mycobacterium tuberculosis based on MeltArray

Zihan Xia 1,#, Biyi Su 2,#, Chunxia Tu 1,#, Siyang Sun 1, Yaoju Tan 2, Ye Xu 1,✉, Qingge Li 1,✉
Editor: Melissa B Miller3
PMCID: PMC10793361  PMID: 38112521

ABSTRACT

Spacer oligonucleotide typing (spoligotyping), the first-line genotyping assay for Mycobacterium tuberculosis (MTB), plays a fundamental role in the investigation of its epidemiology and evolution. However, the traditional spoligotyping protocol was established by using the reverse dot blot hybridization technique, which is characterized by a complex procedure, long turnaround time, and subjectivity in result interpretation, thus hindering its widespread use in low- and middle-income countries (LMICs) with a high tuberculosis burden. In this study, we established a single-tube spoligotyping assay using MeltArray, a highly multiplex polymerase chain reaction (PCR) approach that runs on a real-time PCR thermocycler. The MeltArray protocol included an internal positive control, gyrB, to indicate the abundance of MTB via the quantification cycle (Cq) and 43 spacers to identify the spoligotype via melting curve analysis. The entire protocol was completed in a single step within 2.5 hours. The lowest detectable copy number for the tested strains was 20 copies/reaction. In a blind evaluation of 318 MTB isolates, the MeltArray assay yielded 98.1% (312/318) concordance with the traditional approach and 100.0% with Sanger sequencing. Further evaluation of 151 liquid culture-matched sputum samples showed that all qualified (Cq <35) sputum samples (80.8%, 122/151) yielded results consistent with those of the liquid culture samples, including 97.5% (119/122) single spoligotypes and 2.5% (3/122) mixed spoligotypes. We conclude that MeltArray-based spoligotyping could be used immediately in LMICs, given its easy access, improved throughput, and potential applicability to clinical samples.

IMPORTANCE

Spacer oligonucleotide typing (spoligotyping), the first-line genotyping assay for Mycobacterium tuberculosis (MTB), plays a fundamental role in the investigation of its epidemiology and evolution. In this study, we established a single-tube spoligotyping assay using MeltArray, a highly multiplex polymerase chain reaction (PCR) approach that runs on a real-time PCR thermocycler. The MeltArray protocol included an internal positive control, gyrB, to indicate the abundance of MTB via the quantification cycle and 43 spacers to identify the spoligotype via melting curve analysis. The entire protocol was completed in a single step within 2.5 hours. The lowest detectable copy number for the tested strains was 20 copies/reaction and thus sufficient for analyzing both culture and sputum samples. We conclude that MeltArray-based spoligotyping could be used immediately in low- and middle-income countries with a high tuberculosis burden, given its easy access, improved throughput, and potential applicability to clinical samples.

KEYWORDS: Mycobacterium tuberculosis, spoligotyping, MeltArray, sputum

INTRODUCTION

The discovery of a region within the Mycobacterium bovis BCG strain, characterized by the presence of short repeats, interspaced by unique highly polymorphic sequences, has fostered the invention of spacer oligonucleotide typing called spoligotyping (1–3). As a rapid polymerase chain reaction (PCR)-based method, spoligotyping is one of the most widely adopted methods for genotyping Mycobacterium tuberculosis (MTB) complex. One distinct feature of spoligotyping is that the presented results are digital and, thus, readily shared among laboratories, thereby enabling the creation of large international databases. Additional strengths of this method include low cost, adequate level of overall strain differentiation, high-throughput capacity, and ability to provide species-level identification within the MTB complex (1, 4, 5). Despite its inability to accurately differentiate within large strain families such as the Beijing family, spoligotyping is widely recommended as the first-line genotyping tool in countries with a high tuberculosis (TB) burden (6–11). Spoligotyping was standardized as a reverse dot blot hybridization technique 20 years ago. The difficulty in obtaining specialized membranes and instrumentation, complex multistep manual post-PCR manipulations, and the limited ability in the use of sputum samples have collectively hindered its further expansion in low- and middle-income counties (LMICs) with a high TB burden.

Real-time PCR can detect amplification products without opening the reaction tube and therefore essentially eliminates carry-over contamination, saves time and labor, simplifies automation, and enables target quantification. The addition of a melting analysis step after PCR enables the detection of multiple targets based on their distinct melting temperature (Tm) values (12), which is used as a second dimension for target identification, in addition to the fluorescent dye used in real-time PCR. We previously described a real-time PCR-based spoligotyping protocol termed McSpoligotyping, based on multicolor melting curve analysis (13). This one-step method can be completed within 3 hours starting with purified DNA. A multicenter study showed that McSpoligotyping could be performed easily and reliably using standard microbiology techniques. However, McSpoligotyping requires three reactions to test one sample, yielding a lower throughput than traditional protocols. In addition, single-nucleotide polymorphisms, small indels, or IS6110 insertions present in spacers can cause Tm shift of the melting peak, leading to incorrect identification of spacers and incorrect typing. We later designed a ligation-based alternative that allowed two-reaction melting curve analysis for spoligotyping (14). The direct repeat (DR)-free design avoided false negatives caused by IS6110 insertion in the DR region. However, this strategy requires an additional ligation step before PCR, adding to the overall complexity and increasing the turnaround time. A common drawback of the aforementioned two real-time PCR-based methods is their non-quantitative endpoint detection format, which has prevented their use in clinical samples without knowledge of MTB abundance; thus, both are confined to isolated strains.

More recently, we proposed a highly multiplex PCR approach, termed MeltArray, which can detect 10-fold more targets than current real-time PCR assays in a single reaction. In addition, MeltArray provides both qualitative and quantitative results via successive real-time PCR detection and melting curve analyses (15). Here, we describe the development of a MeltArray-based spoligotyping assay that permits single-tube, one-step spoligotyping of MTB for the first time. Furthermore, we extended its application to culture-independent spoligotyping by coupling quantification with MTB spoligotyping. We systematically evaluated the analytical and clinical performance, particularly for sputum samples.

MATERIALS AND METHODS

MeltArray spoligotyping

MeltArray spoligotyping was performed using the SLAN-96 thermocycler (Zeesan Biotech, Xiamen, China). The reaction was performed in a 25 µL solution containing 1 × PCR buffer (Zeesan), 11.0 mM MgCl2, 0.5 mM dNTPs, 2.0 U of Taq 01 DNA polymerase (Zeesan), and 5 µL (~5,000 copies) of genomic DNA (gDNA) template. Each reaction contained four target-specific primers, 43 spacer-targeted mediator probes, one gyrB-targeted molecular beacon probe, and six pairs of universal molecular beacon reporters (sequences and concentrations are listed in Table S1). The PCR program was performed as follows: 95°C for 5 min, 45 cycles of 95°C for 20 s and 57°C for 35 s, 35°C for 30 min, and 45°C for 2 min, followed by a temperature increase from 45°C to 95°C at 0.16°C/s. The fluorescence intensity was measured in six detection channels (Atto 425, FAM, HEX, ROX, Cy5, and Quasar 705) at the denaturation stage of each of the 45 thermal cycles and at each step of the continuous temperature increase during the melting curve analysis. The results were analyzed using MeltPro Manager for the automatic export of a 43-digit binary code for the spoligotyping readout. The SITVIT 2 database (http://www.pasteurguadeloupe.fr:8081/SITVIT2/index.jsp) was used to analyze the spoligotyping data to determine the spoligotype and phylogenetic lineage (16).

Analytical evaluation

The lowest detectable copy number of MeltArray spoligotyping was estimated using four representative MTB strains: H37Rv (37 spacers), BCG (35 spacers), Beijing lineage spoligotype international type (SIT) 1 (9 spacers), and Manu lineage SIT 523 (43 spacers). Serial dilutions of gDNA from 50, 000, 5, 000, 500, 50, 20, 10, 5, and one copies/reaction were detected. The lowest concentration that provided correct typing results for all four strains was determined and used to derive the cut-off quantification cycle (Cq) for MeltArray spoligotyping. To study the reproducibility of the Tm values, H37Rv, BCG, and 50 MTB isolates with different SITs (5,000 copies/reaction) were detected, and the standard deviation (SD) of the Tm value for each spacer was calculated. The assay was repeated for 3 consecutive days. To evaluate the specificity, gDNA templates from 23 non-tuberculous mycobacterial species and 16 non-mycobacterial species (Table S2) at 5,000 copies/reaction were tested.

All the gDNAs were extracted according to a previously described protocol (17) and stored at –80°C as reference materials in our laboratory. The original gDNA concentrations were determined using an ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA).

Double-blind analysis of 318 archived gDNA

A subset of 318 gDNA samples purified from isolates from cultured MTB clinical samples was retrieved from our laboratory-archived storage. These gDNA samples, which were typed using a reverse dot blot hybridization protocol, were selected to cover the broadest possible spoligotypes in different lineages. Double-blind analysis was performed: DNA samples were renumbered by one supervisor (Y.X.), who was solely in charge of the data collection and statistical analysis. The coded DNA samples were analyzed by three students (Z.X., C.T., and S.S.). The results were reported to a third individual (Q.L.), who checked the data and calculated the concordance rate. The minimum spanning tree of the isolates was constructed based on the information of traditional spoligotyping using PHYLOViZ 2.0 (18). Detailed usage of PHYLOViZ 2.0 is available online at https://phyloviz.readthedocs.io/en/latest/loading_data.html.

Clinical evaluation with culture-matched sputum samples

A total of 151 patients with pulmonary TB admitted to Guangzhou Chest Hospital (Guangzhou, China) were enrolled in this study. Three sputum samples were collected from each patient: one sample was used for liquid culture, one for the microscopy test, and one for MeltArray spoligotyping. The liquid culture samples were also subjected to MeltArray spoligotyping, and the results were compared with those of matched sputum samples collected from the same patients. Liquid culture, microscopy, and DNA extraction from cultured and sputum samples were performed as previously described (17). This study was reviewed and approved by the Institutional Review Board of Guangzhou Chest Hospital. Informed consent was obtained from all the patients.

The liquid culture samples had higher gDNA concentrations than the matched sputum samples; therefore, spoligotyping would be more easily achieved for the former than for the latter. To ensure the objective of this evaluation, the two types of samples were analyzed separately by two authors (Z.X. and C.T.). The results were separately reported to a third individual (Y.X.), who calculated the concordance rate between the matched samples.

Sanger sequencing

Sanger sequencing was applied to those archived gDNA samples that showed discordance between the recorded spoligotypes and the MeltArray spoligotyping results. The sequencing primers and amplification procedures were performed in accordance with our previous study (13). The amplified products were purified and sequenced by Sangon Inc. (Shanghai, China).

Droplet digital PCR (ddPCR)

ddPCR was used to quantify the MTB copy number in the culture-matched sputum sample. ddPCR was performed in a 30-µL mixture containing 15 µL of 2 × SuperMix (TargetingOne, Beijing, China), 0.96 µM primers (forward, 5′-CACGCCAAGTCGGCCCGGT-3′ and reverse, 5′-AGCGGGTAGCGCAGCGA-3′), 0.3 µM probes (5′-FAM-ATCTACGACACCTGGTGCGC-BHQ1−3′ and 5′-VIC-AGATCGACGCGTCGCCGT-BHQ1−3′) (note: the underlined nucleotides indicate locked nucleic acids [LNAs]), and 5 µL of the gDNA template. Droplets were generated using the Drop Maker (TargetingOne, Beijing, China) and placed in an A300 thermal cycler (LongGene, Hangzhou, China) under the following program: 95°C for 10 min; 40 cycles of 94°C for 30 s, and 61°C for 1 min. The droplets were detected using a Chip Reader (TargetingOne, Beijing, China) according to the manufacturer’s instructions.

RESULTS

Design of single-tube MeltArray spoligotyping

Based on the MeltArray principle (15), two primer/probe sets were designed to accomplish single-tube spoligotyping of MTB (Fig. 1). The first primer/probe set was used to detect gyrB, a conserved gene present as a single copy in the MTB genome, by using modified TaqMan chemistry, where fluorescence was detected during the denaturation stage of the PCR. The obtained Cq values were used to estimate the abundance of MTB. The second primer/probe set, which included a common pair of primers and 43 mediator probes, was used to detect the 43 spacers by using melting curve analysis. The mediator probe consists of a 5′-mediator region and target-binding region. During PCR, the bound probe is cleaved by the 5′-flap endonuclease activity of Taq DNA polymerase, leaving the 5′-mediator free from the mediator probe. The released mediator then hybridizes with and, like a primer, extends along a molecular beacon reporter labeled with a fluorophore and forms a double-stranded DNA with a Tm and fluorophore combination that is unique to that spacer. The Tm value obtained by the melting curve analysis in the corresponding fluorometric detection channel was used to indicate the presence of a spacer. By combining two primer/probe sets in one reaction run in a 6-color real-time PCR thermocycler, simultaneous (single-tube) quantification and spoligotyping can be achieved.

Fig 1.

Fig 1

Open in a new tab

Schematic illustration of the design of MeltArray spoligotyping. (A) The distribution of the primers and probes used for MeltArray spoligotyping. (B) Schematic diagram of MeltArray spoligotyping results. gyrB was quantified by a Cq value obtained by real-time PCR, and 43 spacers were identified by their respective combinations of the fluorophore and Tm value obtained by melting curve analysis.

Establishment of MeltArray spoligotyping

Single-tube MeltArray spoligotyping was carried out by simultaneous real-time PCR detection and melting curve analysis. While the former is easy to accomplish, similar to a singleplex real-time PCR assay, the latter is more complex because up to 43 spacers must be identified in one reaction. To achieve this, mediators were first screened using a library comprising mediators and reporters with predefined Tm values (15). Candidate mediators were chosen to design mediator probes, which were then subjected to PCR using gDNA from both MTB H37Rv and M. bovis BCG. This process was repeated until all spacers were correctly identified using the MeltPro software of the thermocycler. During this process, appropriate modifications, such as LNAs, thiophosphates, and mismatched bases, were introduced to the mediator for successful extension along the reporter. Special attention was paid to the polymorphic variants that occurred within the spacers. The introduction of LNAs, degenerate bases, or multiple binding sequences to the target-binding region was used to overcome the effect of such variants on probe binding. Moreover, spacer distribution in the six fluorometric detection channels was optimized for better resolution of adjacent melting peaks and prevention of nonspecific melting peaks.

The final MeltArray spoligotyping reaction included the real-time PCR detection of gyrB in one channel and melting curve analysis of 43 spacers in six channels (Fig. 2). The spoligotyping of the four representative strains showed that both real-time PCR detection and melting curve analysis produced the expected results. Using the thermocycling program described in the Materials and Methods section, MeltArray spoligotyping of 93 samples, two positive controls (H37Rv and BCG), and one no-template control was completed in one batch within 2.5 hours following the addition of MTB gDNA into the reaction.

Fig 2.

Fig 2

Open in a new tab

MeltArray spoligotyping results of four representative MTB strains. The upper panel displays the presence (solid square) and absence (empty square) of the different spacers for the four studied strains. The lower panel shows the MeltArray spoligotyping output results according to the detection channels. NTC, no-template control. IPC, internal positive control.

Analytical evaluation

Of the four representative strains studied, the Beijing strain exhibited the lowest detectable copy number (five copies/reaction), whereas Manu had the highest minimum detectable copy number (20 copies/reaction), in proportion to their respective spacer numbers (9 vs. 43) (Fig. S1). To ensure reproducible typing, the lowest detectable concentration was defined as 20 copies/reaction, and this concentration corresponded to a mean Cq of 35.96 cycles (95% CI, 35.58 to 36.36), and thus 35 cycles was chosen as the cut-off Cq (Fig. S2A).

Given that Xpert TB Ultra is the major PCR platform globally applied for MTB detection, and users would benefit from knowing if the bacillary load is sufficient to proceed with genotyping using MeltArray. We determined the category in the Xpert TB platform to which the cut-off Cq corresponds to. To do so, we prepared a series of dilutions containing 4,000, 400, 40, and four colony-forming units per milliliter (CFU/mL) of the reference strain H37Rv. To ensure accuracy, each dilution was extracted in triplicate and each extracted DNA template was tested in triplicate, thus yielding nine repeated results for each dilution. The results showed that the qualified gDNA template (Cq <35) could be consistently obtained when the initial input of H37Rv was 400 CFU/mL (Fig. S3). This input concentration was categorized as “low” by Xpert TB Ultra, which corresponded with 102–103 CFU/mL H37Rv according to a previous study (19). It is worth noting that the aforementioned results were obtained using our routinely used extraction protocol. By changing to those extraction protocols that can accommodate larger sputum volume and/or yield smaller DNA elution volume, it is possible to reach 100 CFU/mL H37Rv DNA template and thus promote our assay to samples categorized as "very low" according to Xpert TB Ultra.

The reproducibility study yielded SDs between 0.12°C and 0.39°C for all the 43 Tm values, and no interferences between adjacent melting peaks occurred (Fig. S2B). Specificity evaluation of 23 non-tuberculous mycobacterial species and 16 non-mycobacterial species showed that neither Cq values nor melting curves were generated, confirming the specificity of MeltArray spoligotyping.

Blind evaluation using archived gDNA samples

All 318 gDNA samples had Cq values less than 35 cycles, thus qualifying for spoligotyping. Compared to the recorded spoligotype patterns, 312 (98.1%, 312/318) samples showed fully consistent patterns, and six had one additional spacer detected by MeltArray. The additionally detected spacers were found to harbor a variant in either the spacer or the DR region, as revealed by Sanger sequencing, except for one sample (Table S3). Taken together, the blind evaluation confirmed the accuracy of the MeltArray spoligotyping results.

A minimum spanning tree showed that 226 of the 318 spoligotypes were classified into 14 known lineages (Fig. 3; Table S4) and the remaining 92 spoligotypes, mostly unidentified, were classified into 10 predicted lineages. These results indicate the comprehensive lineage coverage of the studied samples.

Fig 3.

Fig 3

Open in a new tab

Lineage distribution of the 318 archived MTB gDNA samples. (A) Minimum spanning tree of 318 archived MTB gDNA. Each circle indicates one spoligotype pattern. (B) Spoligotype distribution in different lineages. “Unknown” represents unfound lineage in the SITVIT 2 database. The outer circle indicates the predicted lineages of isolates from the TB lineage website (http://tbinsight.cs.rpi.edu/run_tb_lineage.html).

Direct MeltArray spoligotyping of sputum samples

Among the 151 liquid culture-sputum paired samples, all liquid culture samples showed Cq <35, 122 sputum samples showed Cq <35 (qualified), and 29 sputum samples showed Cq >35 (unqualified). A comparison of 122 qualified sputum samples with their matched liquid culture samples showed a full concordance in the spoligotyping pattern. Of the 29 unqualified sputum samples, 18 were consistent with their matched liquid culture samples, and 11 provided partial or no spacer information. All 18 correctly typed samples belonged to the Beijing lineage, which is consistent with a previous observation that the Beijing strain had a lower detectable copy number per reaction. These results demonstrate that all qualified sputum samples were successfully typed while the assay may yield inaccurate results when the sample exhibits a Cq >35.

Interestingly, the three qualified samples contained two mixed spoligotypes according to their uneven melting peaks (Fig. 4). From the higher peaks, a dominant Beijing strain was derived based on the characteristic deletion of 1–34 spacers. The lower peaks failed to produce a definite spoligotype because of the possible existence of undetected spacers, which could either merge with those of the Beijing strain or remain undetected because of their low abundance. These results are in line with the clinical observations that some patients may have co-infection with two MTB strains (20, 21). For simplicity, these samples were identified as Beijing strains in subsequent analyses.

Fig 4.

Fig 4

Open in a new tab

MeltArray spoligotyping results of sputum–culture paired samples. Sample #002 has a typical Beijing lineage strain (red squares and stars), whereas samples #058, #171, and #173 all contain mixed strains, where one is Beijing lineage strain and the other is unknown (blue squares and stars). IPC, internal positive control.

Minimum spanning trees were constructed for both liquid culture (151) and sputum samples (122) (Fig. S4). Thirty-five spoligotypes were obtained from the liquid culture samples, 28 of which were known SITs (144 samples) and the other seven were novel (seven samples) (Table S5). The distribution of the MTB lineages was similar to that obtained from a separate sample set from the same hospital, as previously reported (13). A nearly identical distribution of MTB lineages was also obtained from the sputum samples, despite the smaller sample size (122 vs 151).

Cq is critical in MeltArray spoligotyping; therefore, we further verified its quantitative performance in detecting the MTB genome. The results of 75 randomly chosen sputum samples were compared using the ddPCR assay targeting gyrA. Both correlation analysis (r = 0.993) and Bland–Altman analysis (mean ratio = 1.061) confirmed the consistency of the MTB genome copy number between the two assays (Fig. 5A and B). We further compared the Cq values with the positive values obtained from the microscopy test, which is common practice in sputum examination. An inversely proportional relationship between the two values was observed in 141 sputum samples. However, exceptional cases were often found in each positive order, and some smear-negative samples had lower Cq values than smear-positive samples (Fig. 5C). These results demonstrate that the Cq value is more reliable than positive values for estimating MTB abundance.

Fig 5.

Fig 5

Open in a new tab

Quantitative ability of MeltArray spoligotyping. (A) Pearson’s correlation analysis of 75 sputum samples quantified using ddPCR and MeltArray assays. The 95% confidence band and 95% prediction band are represented by dark red shade and light red shade, respectively. (B) Bland–Altman plot of 75 sputum samples quantified using ddPCR and MeltArray assays. The mean ratio of the concentration quantified by ddPCR and MeltArray is indicated by the red dotted line. The limit of agreement (SD times 1.96) of the mean ratio is indicated by the black dotted lines. (C) Distribution of the Cq value of sputum samples among different positive values obtained by microscopic examination and liquid culture samples. Neg, smear-negative. Pos, smear-positive. The cut-off Cq value is 35 (red dotted line). For box plots, centerlines indicate medians, box limits are the first and third quartiles, and whisker ends represent medians ± 2 SD.

DISCUSSION

We have developed a single-tube, one-step MeltArray spoligotyping that can be performed on a real-time PCR thermocycler, which is the mainstream instrument in standard molecular laboratories worldwide in the post-COVID-19 era (22, 23). This method is rapid, reproducible, easy to use, and has a high throughput. The inclusion of real-time PCR detection of gyrB in the reaction resulted in the acquisition of a Cq for estimation of MTB abundance. MeltArray spoligotyping can detect MTB gDNA at as low as 20 copies/reaction and is thus sufficient for analyzing all the liquid culture samples. The eligibility of a sputum sample for spoligotyping can be determined by the Cq value, and as shown in our study, most sputum samples can be directly used for spoligotyping without culture.

Compared with the traditional reverse dot blot hybridization method and other “off-line” detection techniques such as microbead-based suspension array (2, 24), mass spectrometry (25), and microarray analysis (26, 27), MeltArray spoligotyping combined the PCR and detection into a single-tube format, reaching a one-step “on-line” detection. In addition to saving labor and time, the carry-over contamination of amplicons can be avoided by closed-tube detection. The advantage of single-tube detection is distinct from that of our previously reported ligation-assisted PCR melting curve analysis (14) and multicolor melting curve analysis-based McSpoligotyping (13). The former requires an extra ligation step before PCR, whereas the latter requires three reactions, despite its one-step nature. Moreover, neither approach has the quantitative ability to define a cut-off Cq for qualified spoligotyping. In addition, McSpoligotyping is sensitive to the influence of variants in the probe-binding region, which can cause a Tm shift that may lead to incorrect typing results.

We demonstrated that MeltArray spoligotyping can be used to simultaneously quantify and type MTB in sputum samples. Direct spoligotyping of sputum samples is extremely attractive for the real-time surveillance of MTB transmission (28). However, a challenge lies in the wide distribution of MTB in these samples, as shown in this study and in our previous studies (17, 29). Given the possible loss of certain spacers at low TB abundance, reliable typing is difficult, if not impossible, to achieve without knowing the TB abundance in the sputum sample. Thus far, spoligotyping has mostly been used for cultured isolates rather than for clinical samples such as sputum, despite its successful use in different clinical samples in the first report of spoligotyping (1). The use of cultured samples not only requires a time-consuming and labor-intensive culture step but also makes real-time genotyping during an outbreak impossible (28). The introduction of a Cq cut-off overcomes this issue. In our study, all qualified sputum samples (Cq <35) yielded typing results consistent with their matched liquid culture samples. To the best of our knowledge, this is the first time that such a cutoff value of Cq has been set for spoligotyping. Well-controlled spoligotyping directly from sputum samples opens new avenues for large-scale epidemiological studies.

In the context of next-generation sequencing, genotyping schemes other than whole-genome sequencing (WGS) are being replaced because of their low discriminative power. Indeed, WGS has been implemented in high-income countries and has profoundly transformed TB diagnosis (30–32). Unfortunately, all countries with a high TB burden worldwide are LMICs, and their limited resources heavily reduce their access to WGS. In these countries, a multi-tier genotyping scheme may be a better choice: spoligotyping as the first-line, followed by mycobacterial interspersed repetitive unit-variable number tandem repeats as the second-line, and WGS as the third (33–38). Such a three-tiered genotyping framework is more competitive than any of the aforementioned individual schemes in terms of cost-effectiveness, accessibility, and discriminatory power. Within this framework, MeltArray spoligotyping can play a more active role than current approaches for epidemiological purposes, MTB complex species identification, and even building a database of spoligotypes representative of LMICs.

In conclusion, MeltArray spoligotyping achieved single-tube, one-step, simultaneous MTB quantification and spoligotyping using a widely available real-time PCR thermocycler. Successful extension to sputum samples has further boosted its application in epidemiological studies, either individually or jointly with other genotyping assays.

ACKNOWLEDGMENTS

This work was supported by the 2023 Fujian Province Technology Innovation Key Research and Industrialization Project (2023 G051), Xiamen Major Science and Technology Project (3502Z20191007), Basic Research Project of Shenzhen Basic Research Program (JCYJ20180306170526435), Xiamen Industry-University-Research Project Subsidy Project (2022CXY0113), 2021 Major Science and Technology Project of Inner Mongolia Autonomous Region (2021ZD0006), and Industry-University-Research Joint Innovation Project of University of Fujian Provincial (2021Y4001).

Contributor Information

Ye Xu, Email: xuye@xmu.edu.cn.

Qingge Li, Email: qgli@xmu.edu.cn.

Melissa B. Miller, The University of North Carolina at Chapel Hill School of Medicine, Chapel Hill, North Carolina, USA

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/jcm.01183-23.

Fig. S1 to S4, Table S1 to S5. jcm.01183-23-s0001.pdf.

Fig. S1 (Analytical sensitivity study on MeltArray-based spoligotyping using H37Rv, BCG, Beijing SIT 1, and Manu SIT 523 strains), S2 (Analytical evaluation of the MeltArray-based spoligotyping), S3 (Relationship between initial input and Cq value of the MeltArray spoligotyping assay), and S4 (Minimum spanning tree of all 151 liquid culture samples and 122 qualified sputum samples) and Tables S1 (Primers and probes used in the MeltArray assay for spoligotyping), S2 (Information on 23 non-tuberculous mycobacteria and 16 non-mycobacterial), S3 (Discrepant spacers in six isolates), S4 (Information on 318 spoligotypes) and S5 (Spoligotyping results of 151 sputum-culture pairs).

jcm.01183-23-s0001.pdf (1.4MB, pdf)
DOI: 10.1128/jcm.01183-23.SuF1

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.

REFERENCES

  • 1. Kamerbeek J, Schouls L, Kolk A, van Agterveld M, van Soolingen D, Kuijper S, Bunschoten A, Molhuizen H, Shaw R, Goyal M, van Embden J. 1997. Simultaneous detection and strain differentiation of Mycobacterium tuberculosis for diagnosis and epidemiology. J Clin Microbiol 35:907–914. doi: 10.1128/jcm.35.4.907-914.1997 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Cowan LS, Diem L, Brake MC, Crawford JT. 2004. Transfer of a Mycobacterium tuberculosis genotyping method, spoligotyping, from a reverse line-blot hybridization, membrane-based assay to the luminex multianalyte profiling system. J Clin Microbiol 42:474–477. doi: 10.1128/JCM.42.1.474-477.2004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Fabre L, Zhang J, Guigon G, Le Hello S, Guibert V, Accou-Demartin M, de Romans S, Lim C, Roux C, Passet V, Diancourt L, Guibourdenche M, Issenhuth-Jeanjean S, Achtman M, Brisse S, Sola C, Weill F-X. 2012. CRISPR typing and subtyping for improved laboratory surveillance of salmonella infections. PLoS One 7:e36995. doi: 10.1371/journal.pone.0036995 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Amlerova J, Bitar I, Hrabak J. 2018. Genotyping of Mycobacterium tuberculosis using whole genome sequencing. Folia Microbiol (Praha) 63:537–545. doi: 10.1007/s12223-018-0599-y [DOI] [PubMed] [Google Scholar]
  • 5. Muzyukina P, Soutourina O. 2023. CRISPR genotyping methods: tracing the evolution from spoligotyping to machine learning. Biochimie:S0300-9084(23)00178-5. doi: 10.1016/j.biochi.2023.07.017 [DOI] [PubMed] [Google Scholar]
  • 6. Tulu B, Ameni G. 2018. Spoligotyping based genetic diversity of Mycobacterium tuberculosis in Ethiopia: a systematic review. BMC Infect Dis 18:140. doi: 10.1186/s12879-018-3046-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Ismail S, Al Amry K, Aggor G, El Naggar H, Selim S. 2019. Spoligotyping with pncA sequencing strategy conferring the transmission of multidrug-resistant tuberculosis in Egypt. Int J Mycobacteriol 8:211–217. doi: 10.4103/ijmy.ijmy_111_19 [DOI] [PubMed] [Google Scholar]
  • 8. Ramazanzadeh R, Shakib P, Rouhi S, Mohammadi B, Mohajeri P, Borji S. 2020. Molecular epidemiology of Mycobacterium tuberculosis isolates in Iran using spoligotyping. New Microbes New Infect 38:100767. doi: 10.1016/j.nmni.2020.100767 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Mekonnen D, Derbie A, Abeje A, Shumet A, Kassahun Y, Nibret E, Munshea A, Bobosha K, Wassie L, Biadglegne F, Aseffa A, Sack U. 2019. Genomic diversity and transmission dynamics of M. tuberculosis in Africa: a systematic review and meta-analysis. Int J Tuberc Lung Dis 23:1314–1326. doi: 10.5588/ijtld.19.0127 [DOI] [PubMed] [Google Scholar]
  • 10. Cerezo-Cortés MI, Rodríguez-Castillo JG, Hernández-Pando R, Murcia MI. 2019. Circulation of M. tuberculosis Beijing genotype in Latin America and the Caribbean. Pathog Glob Health 113:336–351. doi: 10.1080/20477724.2019.1710066 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Medeiros TF, Nogueira CL, Prim RI, Scheffer MC, Alves EV, Rovaris DB, Zozio T, Rastogi N, Bazzo ML. 2018. Molecular epidemiology of Mycobacterium tuberculosis strains from prison populations in Santa Catarina, Southern Brazil. Infect Genet Evol 58:34–39. doi: 10.1016/j.meegid.2017.12.010 [DOI] [PubMed] [Google Scholar]
  • 12. Ririe KM, Rasmussen RP, Wittwer CT. 1997. Product differentiation by analysis of DNA melting curves during the polymerase chain reaction. Anal Biochem 245:154–160. doi: 10.1006/abio.1996.9916 [DOI] [PubMed] [Google Scholar]
  • 13. Zeng X, Xu Y, Zhou Y, Li H, Zheng R, Tan Y, Sun B, Zhao Y, Li Q. 2018. Mcspoligotyping, a one-step melting curve analysis-based protocol for spoligotyping of Mycobacterium tuberculosis. J Clin Microbiol 56:e00539-18. doi: 10.1128/JCM.00539-18 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Zeng X, Li H, Zheng R, Kurepina N, Kreiswirth BN, Zhao X, Xu Y, Li Q. 2016. Spoligotyping of Mycobacterium tuberculosis complex isolates by use of ligation-based amplification and melting curve analysis. J Clin Microbiol 54:2384–2387. doi: 10.1128/JCM.00857-16 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Huang Q, Chen D, Du C, Liu Q, Lin S, Liang L, Xu Y, Liao Y, Li Q. 2022. Highly multiplex PCR assays by coupling the 5'-Flap endonuclease activity of Taq DNA polymerase and molecular beacon reporters. Proc Natl Acad Sci USA 119:e2110672119. doi: 10.1073/pnas.2110672119 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Couvin D, David A, Zozio T, Rastogi N. 2019. Macro-geographical specificities of the prevailing tuberculosis epidemic as seen through SITVIT2, an updated version of the Mycobacterium tuberculosis genotyping database. Infect Genet Evol 72:31–43. doi: 10.1016/j.meegid.2018.12.030 [DOI] [PubMed] [Google Scholar]
  • 17. Liang B, Tan Y, Li Z, Tian X, Du C, Li H, Li G, Yao X, Wang Z, Xu Y, Li Q. 2018. Highly sensitive detection of isoniazid heteroresistance in Mycobacterium tuberculosis by deepmelt assay. J Clin Microbiol 56:e01239-17. doi: 10.1128/JCM.01239-17 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Nascimento M, Sousa A, Ramirez M, Francisco AP, Carriço JA, Vaz C. 2017. PHYLOViZ 2.0: providing scalable data integration and visualization for multiple phylogenetic inference methods. Bioinformatics 33:128–129. doi: 10.1093/bioinformatics/btw582 [DOI] [PubMed] [Google Scholar]
  • 19. Chakravorty S, Simmons AM, Rowneki M, Parmar H, Cao Y, Ryan J, Banada PP, Deshpande S, Shenai S, Gall A, Glass J, Krieswirth B, Schumacher SG, Nabeta P, Tukvadze N, Rodrigues C, Skrahina A, Tagliani E, Cirillo DM, Davidow A, Denkinger CM, Persing D, Kwiatkowski R, Jones M, Alland D. 2017. The new Xpert MTB/RIF ultra: improving detection of Mycobacterium tuberculosis and resistance to rifampin in an assay suitable for point-of-care testing. mBio 8:e00812-17. doi: 10.1128/mBio.00812-17 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Chen Y, Jiang Q, Liu Q, Gan M, Takiff HE, Gao Q, Jacobs JL. 2023. Whole-genome sequencing exhibits better diagnostic performance than variable-number tandem repeats for identifying mixed infections of Mycobacterium tuberculosis. Microbiol Spectr 11:e0357022. doi: 10.1128/spectrum.03570-22 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Hajimiri ES, Masoomi M, Ebrahimzadeh N, Fateh A, Hadizadeh Tasbiti A, Rahimi Jamnani F, Bahrmand AR, Mirsaeidi M, Vaziri F, Siadat SD. 2016. High prevalence of Mycobacterium tuberculosis mixed infection in the capital of moderate tuberculosis incidence country. Microb Pathog 93:213–218. doi: 10.1016/j.micpath.2016.02.015 [DOI] [PubMed] [Google Scholar]
  • 22. Green DA, Zucker J, Westblade LF, Whittier S, Rennert H, Velu P, Craney A, Cushing M, Liu D, Sobieszczyk ME, Boehme AK, Sepulveda JL. 2020. Clinical performance of SARS-CoV-2 molecular tests. J Clin Microbiol 58:e00995-20. doi: 10.1128/JCM.00995-20 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Ashford F, Best A, Dunn SJ, Ahmed Z, Siddiqui H, Melville J, Wilkinson S, Mirza J, Cumley N, Stockton J, Ferguson J, Wheatley L, Ratcliffe E, Casey A, Plant T, COVID-19 Genomics UK (COG-UK) Consortium, Quick J, Richter A, Loman N, McNally A. 2022. SARS-CoV-2 testing in the community: testing positive samples with the taqman SARS-CoV-2 mutation panel to find variants in real time. J Clin Microbiol 60:e0240821. doi: 10.1128/jcm.02408-21 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Zhang J, Abadia E, Refregier G, Tafaj S, Boschiroli ML, Guillard B, Andremont A, Ruimy R, Sola C. 2010. Mycobacterium tuberculosis complex CRISPR genotyping: improving efficiency, throughput and discriminative power of ‘spoligotyping with new spacers and a microbead-based hybridization assay. J Med Microbiol 59:285–294. doi: 10.1099/jmm.0.016949-0 [DOI] [PubMed] [Google Scholar]
  • 25. Honisch C, Mosko M, Arnold C, Gharbia SE, Diel R, Niemann S. 2010. Replacing reverse line blot hybridization spoligotyping of the Mycobacterium tuberculosis complex. J Clin Microbiol 48:1520–1526. doi: 10.1128/JCM.02299-09 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Song EJ, Jeong HJ, Lee SM, Kim CM, Song ES, Park YK, Bai GH, Lee EY, Chang CL. 2007. A DNA chip-based spoligotyping method for the strain identification of Mycobacterium tuberculosis isolates. J Microbiol Methods 68:430–433. doi: 10.1016/j.mimet.2006.09.005 [DOI] [PubMed] [Google Scholar]
  • 27. Ruettger A, Nieter J, Skrypnyk A, Engelmann I, Ziegler A, Moser I, Monecke S, Ehricht R, Sachse K. 2012. Rapid spoligotyping of Mycobacterium tuberculosis complex bacteria by use of a microarray system with automatic data processing and assignment. J Clin Microbiol 50:2492–2495. doi: 10.1128/JCM.00442-12 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. García De Viedma D, Pérez-Lago L. 2018. The evolution of genotyping strategies to detect, analyze, and control transmission of tuberculosis. Microbiol Spectr 6. doi: 10.1128/microbiolspec.MTBP-0002-2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29. Zhang S, Chen X, Lin Z, Tan Y, Liang B, Pan Y, Huang M, Su B, Hu X, Xu Y, Li Q. 2023. Quantification of isoniazid-heteroresistant Mycobacterium tuberculosis using droplet digital PCR. J Clin Microbiol 61:e0188422. doi: 10.1128/jcm.01884-22 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Genestet C, Hodille E, Bernard A, Vallée M, Lina G, Le Meur A, Refrégier G, Dumitrescua O, Lyon TB study group . 2022. Consistency of Mycobacterium tuberculosis complex spoligotyping between the membrane-based method and in Silico approach. Microbiol Spectr 10:e0022322. doi: 10.1128/spectrum.00223-22 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Cabibbe AM, Trovato A, De Filippo MR, Ghodousi A, Rindi L, Garzelli C, Baretti S, Allodi G, Mannino R, Rossolini GM, Bartoloni A, Tortoli E, Cirillo DM. 2018. Countrywide implementation of whole genome sequencing: an opportunity to improve tuberculosis management, surveillance and contact tracing in low incidence countries. Eur Respir J 51:1800387. doi: 10.1183/13993003.00387-2018 [DOI] [PubMed] [Google Scholar]
  • 32. Walker TM, Ip CLC, Harrell RH, Evans JT, Kapatai G, Dedicoat MJ, Eyre DW, Wilson DJ, Hawkey PM, Crook DW, Parkhill J, Harris D, Walker AS, Bowden R, Monk P, Smith EG, Peto TEA. 2013. Whole-genome sequencing to delineate Mycobacterium tuberculosis outbreaks: a retrospective observational study. Lancet Infect Dis 13:137–146. doi: 10.1016/S1473-3099(12)70277-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Yang C, Lu L, Warren JL, Wu J, Jiang Q, Zuo T, Gan M, Liu M, Liu Q, DeRiemer K, Hong J, Shen X, Colijn C, Guo X, Gao Q, Cohen T. 2018. Internal migration and transmission dynamics of tuberculosis in Shanghai, China: an epidemiological, spatial, genomic analysis. Lancet Infect Dis 18:788–795. doi: 10.1016/S1473-3099(18)30218-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Yang C, Luo T, Shen X, Wu J, Gan M, Xu P, Wu Z, Lin S, Tian J, Liu Q, Yuan Z, Mei J, DeRiemer K, Gao Q. 2017. Transmission of multidrug-resistant Mycobacterium tuberculosis in Shanghai, China: a retrospective observational study using whole-genome sequencing and epidemiological investigation. Lancet Infect Dis 17:275–284. doi: 10.1016/S1473-3099(16)30418-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Yin C, Mijiti X, Liu H, Wang Q, Cao B, Anwaierjiang A, Li M, Liu M, Jiang Y, Xu M, Wan K, Zhao X, Li G, Xiao H. 2023. Molecular epidemiology of clinical Mycobacterium tuberculosis isolates from Southern Xinjiang, China using spoligotyping and 15-locus MIRU-VNTR typing. Infect Drug Resist 16:1313–1326. doi: 10.2147/IDR.S393192 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Demay C, Liens B, Burguière T, Hill V, Couvin D, Millet J, Mokrousov I, Sola C, Zozio T, Rastogi N. 2012. SITVITWEB--a publicly available international multimarker database for studying Mycobacterium tuberculosis genetic diversity and molecular epidemiology. Infect Genet Evol 12:755–766. doi: 10.1016/j.meegid.2012.02.004 [DOI] [PubMed] [Google Scholar]
  • 37. Zhang D, An J, Wang J, Hu C, Wang Z, Zhang R, Wang Y, Pang Y. 2013. Molecular typing and drug susceptibility of Mycobacterium tuberculosis isolates from Chongqing municipality, China. Infect Genet Evol 13:310–316. doi: 10.1016/j.meegid.2012.10.008 [DOI] [PubMed] [Google Scholar]
  • 38. Lazzarini LC, Huard RC, Boechat NL, Gomes HM, Oelemann MC, Kurepina N, Shashkina E, Mello FC, Gibson AL, Virginio MJ, Marsico AG, Butler WR, Kreiswirth BN, Suffys PN, Lapa E Silva JR, Ho JL. 2007. Discovery of a novel Mycobacterium tuberculosis lineage that is a major cause of tuberculosis in Rio de Janeiro, Brazil. J Clin Microbiol 45:3891–3902. doi: 10.1128/JCM.01394-07 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Fig. S1 to S4, Table S1 to S5. jcm.01183-23-s0001.pdf.

Fig. S1 (Analytical sensitivity study on MeltArray-based spoligotyping using H37Rv, BCG, Beijing SIT 1, and Manu SIT 523 strains), S2 (Analytical evaluation of the MeltArray-based spoligotyping), S3 (Relationship between initial input and Cq value of the MeltArray spoligotyping assay), and S4 (Minimum spanning tree of all 151 liquid culture samples and 122 qualified sputum samples) and Tables S1 (Primers and probes used in the MeltArray assay for spoligotyping), S2 (Information on 23 non-tuberculous mycobacteria and 16 non-mycobacterial), S3 (Discrepant spacers in six isolates), S4 (Information on 318 spoligotypes) and S5 (Spoligotyping results of 151 sputum-culture pairs).

jcm.01183-23-s0001.pdf (1.4MB, pdf)
DOI: 10.1128/jcm.01183-23.SuF1

Articles from Journal of Clinical Microbiology are provided here courtesy of American Society for Microbiology (ASM)

RESOURCES