Typing Method for the QUB11a Locus of Mycobacterium tuberculosis: IS6110 Insertions and Tandem Repeat Analysis

QUB11a is used as a locus for variable number of tandem repeats (VNTR) analysis of Mycobacterium tuberculosis Beijing lineage. However, amplification of QUB11a occasionally produces large fragments (>1,400 bp) that are not easily measured by capillary electrophoresis because of a lack of the typical stutter peak patterns that are used for counting repeat numbers. IS6110 insertion may complicate VNTR analysis of large QUB11a fragments in M. tuberculosis. We established a method for determining both tandem repeat numbers and IS6110 insertion in the QUB11a locus of M. tuberculosis using capillary electrophoresis analysis and BsmBI digestion. All 29 large QUB11a fragments (>1,200 bp) investigated contained IS6110 insertions and varied in the number of repeats (18 patterns) and location of IS6110 insertions. This method allows VNTR analysis with high discrimination.


Introduction
QUB11a, or variable number of tandem repeats (VNTR) locus 2163a, is an analytical locus that has been useful for molecular epidemiological analysis of the Beijing lineage of Mycobacterium tuberculosis [1,2]. Beijing lineage, which is a homogenous strain group [2], is the dominant lineage in Asia [3,4] and Russia [5] and the most prevalent genotype worldwide [6]. Allix-Béguec et al. found that amplification of QUB11a occasionally produced large fragments (>1,400 bp) [7] that were not easily measured by capillary electrophoresis because of a lack of the typical stutter peak patterns, although they also demonstrated the high discriminatory power of QUB11a [7]. Despite this, QUB11a was not included in the international typing scheme based on 24 VNTR loci used for M. tuberculosis [8]. However, Wada et al. [9] and Millet et al. [1] also showed the utility of QUB11a for discrimination of Beijing lineage in Japan, and Vlji et al. recommended the use of hypervariable loci, including QUB11a, for VNTR analysis of Beijing lineage in the United Kingdom [2]. Based on these findings, QUB11a is not an inconsequential locus.
IS6110 (approximately 1,360 bp) [10] is found in the QUB11a locus of Mycobacterium caprae [11], leading us to hypothesize that it may also be present in the corresponding region of M. tuberculosis. Insertion would likely complicate VNTR analysis of large QUB11a fragments in M. tuberculosis, indicating that a method for analyzing these large QUB11a fragments may increase the discriminatory power of M. tuberculosis typing by VNTR. Therefore, the aims of the present study were to determine the presence or absence, frequency, location, and stability of IS6110 insertions in the QUB11a locus of M. tuberculosis isolates and to develop a VNTR typing method for the QUB11a locus.  humans in Fukuoka Prefecture ( = 306) and neighboring prefectures ( = 6), Japan, between 2012 and 2013 was performed as described previously [12,13]. The isolates were obtained from 312 newly diagnosed patients aged from 17 to 102 years (average: 68.3 years; median: 76 years) from eight separate hospitals (127 females and 185 males). In the same period, 944 patients were reported as newly diagnosed in Fukuoka Prefecture as a whole.

Isolates and
Amplification products corresponding to the QUB11a locus were analyzed by capillary electrophoresis (3500 Genetic Analyzer, Life Technologies, Carlsbad, CA, USA) and GeneMapper software (Life Technologies) to determine the fragment sizes (the combination of electrophoresis and measurement is hereafter referred to as capillary electrophoresis analysis). GeneScan 1200 LIZ Size Standard (Life Technologies) was used as an internal size marker. Analysis showed that 29 (9.3%) of the isolates produced large fragments (>1,200 bp) (Figure 1(a)). These large fragments corresponded to a peak located outside of the internal size marker range of GeneMapper software (Figure 1(a)) and therefore could not be measured. These 29 isolates were further analyzed in the following assays.

Stability Test of IS6110.
To confirm the stability of IS6110 in the QUB11a locus, 26 of the 29 isolates were passaged in 2% Ogawa's medium (Kyokuto Pharmaceutical Industrial Co., Tokyo, Japan) every 2 weeks for 20 weeks to give a total of 10 passages (stability test).

IS6110 Detection in QUB11a Using a Simplified Method.
We also aimed to establish a simple method for detection of the IS6110 insertion, along with tandem repeat number(s), in the QUB11a locus based on capillary electrophoresis analysis of BsmBI-digested fragments (Figure 2(o)). BsmBI was predicted to digest the IS6110 region into three segments (approximately 925 bp, 250 bp, and 190 bp) but not cut the QUB11a repeat region. The QUB11a locus was amplified from genomic DNA from the 29 isolates using primers iwamoto-F (forward) and iwamoto-R (reverse), which were labeled with fluorescent dyes fluorescein (FAM) and NED, respectively.

BioMed Research International
BsmBI-treated fragments were then analyzed simultaneously by capillary electrophoresis, as described above.

IS6110 Detection in QUB11a
Using Sanger Analysis. All large QUB11a fragments (29/29) contained IS6110 insertions and varied in the number of repeats (18 patterns) and the locations of IS6110 insertions ( Figure 2). IS6110 was inserted within the repeat region of QUB11a in all isolates except one (Figure 2(d)), in which IS6110 was localized to a region flanking the tandem repeats.

Stability of IS6110.
The number of repeats and the presence and location of IS6110 in QUB11a did not change in any of the 26 isolates during the 10 passages.

IS6110 Detection in QUB11a Using the Simplified Method.
When the large fragment of QUB11a (Figure 1(a)) was digested with BsmBI, two of the expected fragments were detected by capillary electrophoresis (Figures 1(b) and 1(c)), and agarose gel electrophoresis (2% agarose) confirmed the presence of the third fragment (Figure 1(d)). Subsequently, each amplicon produced two labeled fragments and an intervening unlabeled fragment (Figure 2(o)). Therefore, based on the size of the fragments detected by capillary electrophoresis analysis, the number of tandem repeats in the region can be calculated, in addition to determining the presence of the IS6110 insertion. For example, fragments (a)-(c) shown in Figure 1 correspond to fragment (o) in Figure 2, where the detected product in Figure 1 used detects repeats (including the IS6110 region) as well as contiguous sequences (flanking sequences, 171 bp) in the QUB11a locus.

Comparison of the Two Analysis
Methods. The repeat numbers estimated by the capillary electrophoresis analysis method were then compared with those estimated by Sanger analysis for all 29 isolates ( Figure 2). Results showed that there were some discrepancies between the two methods ( Figure 3). Regression analysis was then carried out on the results obtained using the two different methods. A simple linear regression formula was obtained using KaleidaGraph (Synergy Software, Reading, PA, USA): = 0.0814 + 1.0016 , = 0.9998 (where and are the numbers of tandem repeats calculated from capillary electrophoresis results and Sanger analysis results, resp.). Using this regression equation, the isolate that produced fragments (a)-(c) in Figure 1 was estimated to contain 4.99 repeats, which agreed with the results of DNA sequencing of fragment (o) shown in Figure 2 (five repeats). Reproducibility of the experiment was confirmed by triplicate capillary electrophoresis analysis of BsmBI-digested PCR products from eight amplified QUB11a loci. Incidentally, the amplified QUB11a fragments from four isolates were much smaller (<1,200 bp) than the rest of the isolates and were not cleaved by BsmBI (3a, 3b, 4a, and 4b in Figure 1(d)). These novel QUB11a sequences were deposited in the DNA Data Bank of Japan under accession numbers LC005454-LC005482.

Discussion
In the present study, we established a method to detect IS6110 and count the number of tandem repeats in large QUB11a fragments in M. tuberculosis. Information regarding the presence or absence of an IS6110 insertion in the locus will provide an additional way to distinguish between isolates, along with the number of repeats. This new method takes less than 7 h (including amplification of template DNA, digestion of PCR product by BsmBI, separation by capillary electrophoresis, and calculation of the number of tandem repeats) and increases the accuracy and discriminatory power of VNTR typing at the QUB11a locus of M. tuberculosis.
Discrimination between strains of Beijing lineage is important for countries and areas where Beijing lineage is dominant [1,2]. Using our method, two isolates in group A exhibited the same VNTR pattern when 24 loci were examined: Mtub4, MIRU10, Mtub21, Mtub24, QUB11b, VNTR2372, MIRU26, QUB15, MIRU31, QUB3336, QUB26, QUB4156, ETR A, QUB18, QUB3232, VNTR3820, VNTR4120, MIRU04, MIRU16, MIRU40, ETR C, Mtub30, Mtub39, and QUB11a (nine repeats + IS6110) (data not shown). The present method typed QUB11a regions correctly, and the 24-locus typing assay revealed the possibility of epidemiological linkage of these isolates. Although our method might not be applicable for all other strains that harbor large QUB11a fragments, it may contribute to expanding the discriminatory power of VNTR for M. tuberculosis Beijing lineage.