Abstract
This study shows that the addition of a consensus 4-locus set of hypervariable mycobacterial interspersed repetitive-unit–variable-number tandem repeat (MIRU-VNTR) loci to the spoligotyping-24-locus MIRU-VNTR typing strategy is a well-standardized approach that can contribute to an improvement of the true cluster definition while retaining high typeability in non-Beijing strains.
TEXT
The worldwide implementation of 24-locus mycobacterial interspersed repetitive-unit–variable-number tandem-repeat (MIRU-VNTR) analysis has deeply contributed to the development of molecular epidemiology in the tuberculosis (TB) field (1, 2). It is considered the international standard for molecular typing (3, 4), sometimes supplemented by spoligotyping (5). Despite its remarkable advantages in terms of technical simplicity, reproducibility, and portability compared with IS6110 restriction fragment length polymorphism (RFLP) analysis (6), several studies have highlighted the relative lack of resolution power of MIRU-VNTR analysis (7), particularly for highly clonal strain groups, such as the Beijing lineage (8, 9).
The introduction of whole-genome sequencing (WGS) for Mycobacterium tuberculosis strain typing, due to its ability to discriminate closely related strains, appears to be the optimal solution for achieving the highest possible resolution (10, 11). However, the wider use of WGS for routine epidemiology is still hampered by cost, especially in resource-limited settings, and by the lack of standardization and established quality assurance programs (1) (12).
A possible transitional strategy might involve the inclusion of additional MIRU-VNTR markers to the standard set. In this regard, four novel MIRU-VNTR loci, i.e., 1982, 3232, 3820, and 4120, have recently been suggested for the subtyping of clustered samples belonging to the highly homogeneous Beijing lineage (8). Nevertheless, scarce information is available on the possible contribution of these markers to the improvement of discrimination among non-Beijing M. tuberculosis lineages.
Our study aimed at assessing the epidemiological value and technical feasibility of the inclusion of this 4-locus set of hypervariable loci for subtyping non-Beijing M. tuberculosis strains identified as clusters by the standard spoligotyping-24-locus MIRU-VNTR typing strategy.
A total of 220 M. tuberculosis samples collected in Albania during the 2010 National TB Drug Resistance Survey (DRS), were genotyped by 43-spacer spoligotyping, as previously described (13), and by 24-locus MIRU-VNTR analysis using the GenoScreen MIRU typing kit (GenoScreen, Lille, France) (2).
To infer lineages and clusters, a combined spoligotyping-VNTR analysis was carried out using the MIRU-VNTRplus Web application (http://www.miru-vntrplus.org) (14), requiring 100% identity (both in spoligotype and MIRU profile) for grouping samples.
The recently developed consensus 4-locus set of hypervariable MIRU-VNTR markers, based on the 1982 (alias QUB-18), 3232 (alias QUB-3232), 3820, and 4120 loci, was used for further analysis on all clustered strains, regardless of lineage. The subtyping was performed by using previously published primers in single PCRs and running on a 3% agarose gel. For peak calling, previously outlined conventional rules were applied (8).
The recent transmission index (RTI) was calculated by using the n − 1 method (15). Detailed demographic data were collected by the Albanian TB laboratory network for all patients and sent to the National Reference Laboratory (NRL) in Tirana, Albania, for a contact tracing study. This study was approved by the local ethics committee.
Ghana was the most represented lineage among our M. tuberculosis strains (27.3%), followed by Uganda I (18.2%) and Haarlem (14.6%). Noteworthy, no strain belonging to the Beijing lineage was isolated.
As expected, the highest resolution was obtained by combining standard spoligotyping and 24-locus MIRU-VNTR analysis in a polyphasic analysis: 122 out of 220 samples were included in 39 clusters, resulting in an RTI of 37.7% (Table 1). In fact, this value is lower than both the 75.9% obtained using spoligotyping alone and the 39.5% measured using data only from the 24-locus MIRU-VNTR analysis.
TABLE 1.
Clustering before and after considering the 4 hypervariable loci
| Sample IDa | Clustering results by approach |
|||||
|---|---|---|---|---|---|---|
| Spoligotyping + MIRU-24 (lineage)b | Hypervariable loci |
Spoligotyping + MIRU-24 + 4 hypervariable locib | ||||
| 1982 | 3232 | 3820 | 4120 | |||
| 1 | 1 (Haarlem) | 6 | >22c | 1 | 4 | 1 |
| 175 | 6 | >22c | 1 | 4 | ||
| 217 | 6 | >22c | 1 | 4 | ||
| 136 | 2 (Haarlem) | 6 | >22c | 1 | 4 | 2 |
| 162 | 6 | >22c | 1 | 4 | ||
| 21 | 3 (Haarlem) | 6 | 20 | 4 | 4 | 3A |
| 74 | 6 | 20 | 4 | 4 | ||
| 118 | 6 | 0(+20)d | 4 | 4 | ||
| 143 | 3 | 22 | 4 | 4 | Not clustered | |
| 65 | 4 (Haarlem) | 6 | 19 | 3 | 4 | 4 |
| 172 | 6 | 19 | 3 | 4 | ||
| 180 | 5 (Haarlem) | 2 | 18 | 3 | 5 | 5 |
| 210 | 2 | 18 | 3 | 5 | ||
| 32 | 6 (Haarlem) | 2 | 1 | 3 | 6 | 6 |
| 86 | 2 | 1 | 3 | 6 | ||
| 98 | 7 (LAM) | 2 | 5 | 3 | 5 | 7 |
| 190 | 2 | 5 | 3 | 5 | ||
| 75 | 8 (Haarlem) | 6 | 6 | 3 | 5 | Not clustered |
| 160 | 6 | 8 | 3 | 5 | Not clustered | |
| 50 | 9 (URAL) | 9 | 5 | 3 | 4 | 9 |
| 139 | 9 | 5 | 3 | 4 | ||
| 100 | 10 (Uganda I) | 12 | 6 | 14 | 4 | Not clustered |
| 122 | 12 | 5 | 14 | 3 | Not clustered | |
| 164 | 11 (Ghana) | Xe | 6 | 12 | 4(+5) | 11 |
| 165 | 12 | 6 | 12 | 4(+5) | ||
| 8 | 12 (TUR) | 12 | 6 | 6 | 4 | 12A |
| 76 | 12 | 6 | 6 | 4 | ||
| 110 | 12 | 6 | 6 | 4 | ||
| 34 | 12 | 6 | 8 | 4 | Not clustered | |
| 36 | 13 (URAL) | 5 | 7 | 3 | 3 | 13A |
| 101 | 5 | 7 | 3 | 3 | ||
| 144 | 5 | 7 | 3 | 3 | ||
| 168 | 5 | 7 | 3 | 3 | ||
| 201 | 5 | 7 | 3 | 3 | ||
| 93 | 5 | 5 | 3 | 3 | Not clustered | |
| 28 | 14 (URAL) | 5 | 7 | 3 | 3 | 14 |
| 47 | 5 | 7 | 3 | 3 | ||
| 200 | 5 | 7 | 3 | 3 | ||
| 7 | 15 (Orphan) | 5 | 9 | 3 | 3 | 15 |
| 26 | 5 | 9 | 3 | 3 | ||
| 63 | 5 | Xe | 3 | 3 | ||
| 137 | 5 | 9 | 3 | 3 | ||
| 6 | 16 (Ghana) | 5 | 6 | 5 | 2 | 16 |
| 39 | 5 | 6 | 5 | 2 | ||
| 22 | 17 (Ghana) | 5 | 6 | 5 | 2 | 17 |
| 46 | 5 | 6 | 5 | 2 | ||
| 85 | 5 | 6 | 5 | 2 | ||
| 64 | 18 (Ghana) | 5 | 6 | 5 | 2 | 18 |
| 94 | 5 | 6 | 5 | 2 | ||
| 167 | 5 | 6 | 5 | 2 | ||
| 43 | 19 (Uganda I) | 2 | 2 | 5 | 2 | 19 |
| 170 | 2 | 2 | 5 | 2 | ||
| 24 | 20 (Uganda I) | 2 | 2 | 5 | 2 | Not clustered |
| 40 | 5 | 6 | 5 | 2 | 20A | |
| 142 | 5 | 6 | 5 | 2 | ||
| 27 | 21 (Ghana) | 5 | 5 | 5 | 1 | 21A |
| 73 | 5 | 5 | 5 | 1 | ||
| 67 | 5 | 6 | 5 | 2 | Not clustered | |
| 97 | 22 (Ghana) | 5 | 8 | 5 | 2 | 22 |
| 147 | 5 | 8 | 5 | 2 | ||
| 151 | 5 | 8 | 5 | 2 | ||
| 152 | 5 | 8 | 5 | 2 | ||
| 166 | 5 | 8 | 5 | 2 | ||
| 198 | 5 | 8 | 5 | 2 | ||
| 202 | 5 | 8 | 5 | 2 | ||
| 205 | 23 (Ghana) | 5 | 8 | 5 | 2 | 23 |
| 215 | 5 | 8 | 5 | 2 | ||
| 105 | 24 (Ghana) | 5 | 6 | 5 | 2 | 24 |
| 203 | 5 | 6 | 5 | 2 | ||
| 80 | 25 (Cameroon) | 5 | 3 | 4 | 4 | 25 |
| 121 | 5 | 3 | 4 | 4 | ||
| 171 | 5 | 3 | 4 | 4 | ||
| 216 | 5 | 3 | 4 | 4 | ||
| 38 | 26 (Orphan) | 5 | 7 | 5 | 2 | 26 |
| 72 | 5 | 7 | 5 | 2 | ||
| 31 | 27 (Uganda I) | 5 | 7 | 3 | 2 | 27 |
| 77 | 5 | 7 | 3 | 2 | ||
| 157 | 5 | 7 | 3 | 2 | ||
| 55 | 28 (Uganda I) | 5 | 7 | 3 | 2 | 28 |
| 69 | 5 | 7 | 3 | 2 | ||
| 71 | 5 | 7 | 3 | 2 | ||
| 178 | 5 | 7 | 3 | 2 | ||
| 199 | 5 | 7 | 3 | 2 | ||
| 42 | 29 (Uganda I) | 5 | 7 | 3 | 2 | 29 |
| 44 | 5 | 7 | 3 | 2 | ||
| 141 | 30 (Uganda I) | 5 | 7 | 3 | 2 | 30 |
| 173 | 5 | 7 | 3 | 2 | ||
| 186 | 5 | 7 | 3 | 2 | ||
| 9 | 31 (LAM) | 2 | 2 | 5 | 4 | 31 |
| 188 | 2 | 2 | 5 | 4 | ||
| 213 | 2 | 2 | 5 | 4 | ||
| 13 | 32 (LAM) | 2 | 2 | 5 | 4 | 32 |
| 23 | 2 | 2 | 5 | 4 | ||
| 30 | 2 | 2 | 5 | 4 | ||
| 35 | 2 | 2 | 5 | 4 | ||
| 49 | 2 | 2 | 5 | 4 | ||
| 54 | 2 | 2 | 5 | 4 | ||
| 78 | 2 | 2 | 5 | 4 | ||
| 89 | 2 | 2 | 5 | 4 | ||
| 163 | 2 | 2 | 5 | 4 | ||
| 207 | 2 | 2 | 5 | 4 | ||
| 208 | 2 | 2 | 5 | 4 | ||
| 211 | 2 | 2 | 5 | 4 | ||
| 3 | 33 (Orphan) | 5 | 6 | 2 | 2 | 33 |
| 155 | 5 | 6 | 2 | 2 | ||
| 159 | 5 | 6 | 2 | 2 | ||
| 204 | 5 | 6 | 2 | 2 | ||
| 220 | 5 | 6 | 2 | 2 | ||
| 37 | 34 (Haarlem) | 5 | 6 | 2 | 2 | 34 |
| 169 | 5 | 6 | 2 | 2 | ||
| 184 | 5 | 6 | 2 | 2 | ||
| 61 | 35 (Orphan) | 5 | 0 | 5 | 2 | 35 |
| 114 | 5 | 0 | 5 | 2 | ||
| 117 | 5 | 0 | 5 | 2 | ||
| 107 | 36 (Orphan) | 5 | 0 | 5 | 2 | 36 |
| 108 | 5 | 0 | 5 | 2 | ||
| 4 | 37 (Uganda I) | 6 | 6 | 5 | 6 | 37 |
| 196 | 6 | 6 | 5 | 6 | ||
| 130 | 38 (Uganda I) | 5 | 6 | 3 | 2 | 38 |
| 209 | 5 | 6 | 3 | 2 | ||
| 81 | 39 (Uganda I) | 5 | 8 | 5 | 2 | 39 |
| 140 | 5 | 8 | 5 | 2 | ||
In bold type are the strains excluded from clusters by classic epidemiological investigation. ID, identification.
The RTI for spoligotyping plus MIRU-24 was 37.7%, and that for spoligotyping plus MIRU-24 plus the 4 hypervariable loci was 35.5%. LAM, Latin American Mediterranean; TUR, Turkish.
Nonresolvable high-molecular-weight amplicon (>1,500 bp).
Ambiguous double pattern.
Noninterpretable multiband pattern.
Interestingly, for some of the clusters (n = 10), we did not confirm any possible epidemiological link for at least one of the samples in the cluster (Table 1).
The results obtained by subtyping all clustered samples (n = 122) after the polyphasic analysis using the 4 hypervariable loci are summarized in Table 1. A total of 9 strains (4.1%) were excluded from 7 clusters (3, 8, 10, 12, 13, 20, and 21) belonging to five lineages, namely, Haarlem, Uganda I, Turkish (TUR), Ural, and Ghana, by the additional use of the 4 hypervariable loci, decreasing the RTI to 35.5%. Crucially, the revised clusters reflected the collected demographic data. All epidemiologically confirmed or very probable cases of recent transmission showed the same 28-locus MIRU code. In our study, 100% of the genotypic identity was in perfect concordance with the contact tracing investigation, without any case of erroneous exclusion from clusters due to either 24- or 28-locus MIRU-VNTR single-locus variation (SLV).
The four M. tuberculosis isolates (10, 131, 191, and 194) sharing a 24-locus MIRU genotype with one or more strains from our collection but presenting a unique spoligotype were not further discriminated by the use of the additional four hypervariable loci (Table 2).
TABLE 2.
Results obtained by subtyping the samples clustered by 24-locus MIRU but not by spoligotyping
Altogether, the four hypervariable loci were tested on 126 samples. The total typeability rate was as high as 93.7%. No cases of a lack of amplification were detected. Seven out of eight noninterpretable results were confined to locus 3232: in five cases, we observed nonresolvable and very high amplicon sizes (>1,500 bp) from Haarlem lineage samples, and in two cases, we observed a doubtful double and multiband pattern. Only one noninterpretable result (multiband pattern) was detected in locus 1982 (Ghana lineage). Overall, Haarlem was the most problematic lineage in terms of typeability (92.9%).
High-resolution molecular typing tools are needed for TB control (1), and, to the best of our knowledge, this study is the first attempt to introduce the additional 4-locus set of hypervariable loci for subtyping M. tuberculosis strains clustered by a conventional genotyping strategy, regardless of their lineage.
Even though for other clusters in our collection, a suspicion of false clustering remains and is probably resolvable only by whole-genome sequencing analysis, the use of these hypervariable loci allowed us to solve 40.9% (9/22) of the confirmed unlinked cases, with a high typeability.
The use of the already available commercial kit (GenoScreen, Lille, France) for this 4-locus set of MIRU-VNTR loci might allow the maintenance of optimal levels of standardization and reproducibility.
In addition, this work indicates that spoligotyping can still provide some epidemiologically significant contribution to the total discriminatory power, even when considering a 28-locus MIRU-VNTR scheme.
Although a larger typing study comprising samples from different settings is needed in order to evaluate the real contribution of these hypervariable loci to an increase in the discriminatory power of MIRU-VNTR analysis, we have shown that the addition of the 4-locus set could be used to improve transmission discrimination when molecular results cannot be confirmed by classic epidemiological investigation.
REFERENCES
- 1.Niemann S, Supply P. 2014. Diversity and evolution of Mycobacterium tuberculosis: moving to whole-genome-based approaches. Cold Spring Harb Perspect Med 4:a021188. doi: 10.1101/cshperspect.a021188. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Supply P, Allix C, Lesjean S, Cardoso-Oelemann M, Rusch-Gerdes S, Willery E, Savine E, de Haas P, van Deutekom H, Roring S, Bifani P, Kurepina N, Kreiswirth B, Sola C, Rastogi N, Vatin V, Gutierrez MC, Fauville M, Niemann S, Skuce R, Kremer K, Locht C, van Soolingen D. 2006. Proposal for standardization of optimized mycobacterial interspersed repetitive unit–variable-number tandem repeat typing of Mycobacterium tuberculosis. J Clin Microbiol 44:4498–4510. doi: 10.1128/JCM.01392-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Jagielski T, van Ingen J, Rastogi N, Dziadek J, Mazur PK, Bielecki J. 2014. Current methods in the molecular typing of Mycobacterium tuberculosis and other mycobacteria. Biomed Res Int 2014:645802. doi: 10.1155/2014/645802. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Trovato A, Borroni E, Tafaj S, Hafizi H, Cirillo DM. 2014. Unusual mycobacterial interspersed repetitive-unit alleles in loci 580 and 4348 among Mycobacterium tuberculosis isolates in Albania. J Clin Microbiol 52:4437–4438. doi: 10.1128/JCM.02096-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Oelemann MC, Diel R, Vatin V, Haas W, Rusch-Gerdes S, Locht C, Niemann S, Supply P. 2007. Assessment of an optimized mycobacterial interspersed repetitive-unit–variable-number tandem-repeat typing system combined with spoligotyping for population-based molecular epidemiology studies of tuberculosis. J Clin Microbiol 45:691–697. doi: 10.1128/JCM.01393-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Brossier F, Sola C, Millot G, Jarlier V, Veziris N, Sougakoff W. 2014. Comparison of a semiautomated commercial repetitive-sequence-based PCR method with spoligotyping, 24-locus mycobacterial interspersed repetitive-unit–variable-number tandem-repeat typing, and restriction fragment length polymorphism-based analysis of IS6110 for Mycobacterium tuberculosis typing. J Clin Microbiol 52:4082–4086. doi: 10.1128/JCM.02226-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Roetzer A, Diel R, Kohl TA, Ruckert C, Nübel U, Blom J, Wirth T, Jaenicke S, Schuback S, Rüsch-Gerdes S, Supply P, Kalinowski J, Niemann S. 2013. Whole genome sequencing versus traditional genotyping for investigation of a Mycobacterium tuberculosis outbreak: a longitudinal molecular epidemiological study. PLoS Med 10:e1001387. doi: 10.1371/journal.pmed.1001387. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Allix-Béguec C, Wahl C, Hanekom M, Nikolayevskyy V, Drobniewski F, Maeda S, Campos-Herrero I, Mokrousov I, Niemann S, Kontsevaya I, Rastogi N, Samper S, Sng LH, Warren RM, Supply P. 2014. Proposal of a consensus set of hypervariable mycobacterial interspersed repetitive-unit–variable-number tandem-repeat loci for subtyping of Mycobacterium tuberculosis Beijing isolates. J Clin Microbiol 52:164–172. doi: 10.1128/JCM.02519-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Nikolayevskyy V, Gopaul K, Balabanova Y, Brown T, Fedorin I, Drobniewski F. 2006. Differentiation of tuberculosis strains in a population with mainly Beijing-family strains. Emerg Infect Dis 12:1406–1413. doi: 10.3201/eid1209.041263. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Walker TM, Monk P, Smith EG, Peto TE. 2013. Contact investigations for outbreaks of Mycobacterium tuberculosis: advances through whole genome sequencing. Clin Microbiol Infect 19:796–802. doi: 10.1111/1469-0691.12183. [DOI] [PubMed] [Google Scholar]
- 11.Walker TM, Ip CL, 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 TE. 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]
- 12.Schurch AC, van Soolingen D. 2012. DNA fingerprinting of Mycobacterium tuberculosis: from phage typing to whole-genome sequencing. Infect Genet Evol 12:602–609. doi: 10.1016/j.meegid.2011.08.032. [DOI] [PubMed] [Google Scholar]
- 13.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] [PMC free article] [PubMed] [Google Scholar]
- 14.Weniger T, Krawczyk J, Supply P, Harmsen D, Niemann S. 2012. Online tools for polyphasic analysis of Mycobacterium tuberculosis complex genotyping data: now and next. Infect Genet Evol 12:748–754. doi: 10.1016/j.meegid.2012.01.021. [DOI] [PubMed] [Google Scholar]
- 15.Tanaka MM, Francis AR. 2005. Methods of quantifying and visualising outbreaks of tuberculosis using genotypic information. Infect Genet Evol 5:35–43. doi: 10.1016/j.meegid.2004.06.001. [DOI] [PubMed] [Google Scholar]