Spoligotyping of Mycobacterium africanum, Burkina Faso

Using Ziehl-Neelsen–positive slides collected from tuberculosis diagnostic centers in Burkina Faso, we showed that 20% of 80 spoligotyping-positive DNA samples had a characteristic Mycobacterium africanum–specific genomic signature. This result suggests that M. africanum is still present in Burkina Faso at almost the same prevalence as 15–20 years ago.

In the neighboring country of Ghana (which has 200 km of common borders with Burkina Faso), another study suggested that the population structure of M. tuberculosis complex comprises 1) 34% spoligo-international type (SIT) 61 (named the Cameroon clade, also present in Burkina Faso); 2) 30% M. africanum (including M. africanum West African 1 and West African 2); and 3) 36% principal genetic group 2 and 3 modern strains (e.g., T, U [unknown], Haarlem, X, LAM [Latino-American and Mediterranean]), with minor prevalence of other principal genetic groups, i.e., the East-African Indian, Beijing, and M. bovis clades (4,5). These observations-and their congruence to estimates by Ledru

The Study
The study, which we conducted during March-September 2010, had 3 goals. First, we wanted to determine whether we could extract DNA and perform high-throughput spoligotyping on a Luminex 200 device (Luminex, Austin, TX, USA) on acid-fast bacillus-positive slides (6). Second, we wanted to reestimate the prevalence of M. africanum in Burkina Faso from a recent and random sample of slides. Third, we wanted to further analyze the relative proportion of M. africanum West African I and West African 2 strains in Burkina Faso because this country is part of central western Africa, where the 2 M. africanum West African 1 and 2 strains are present at various relative rates (2). We report on all the goals of this project, even though goal 3 remains to be confi rmed because of the small sample size.
From within 14 geographically independent centers in Burkina Faso (Figure), we recruited a random sample of 186 Ziehl-Neelsen (ZN) slides that had been included in a national study on drug resistance, as approved by the ethical committee for health research in Burkina Faso (2007-031; June 28, 2009). Of 186 DNA samples extracted from as many ZN slides, 143 sputum samples had been scored 3+, 18 were scored 2+, 10 were scored 1+, 5 had 1-9 bacilli total (±), totaling 176 positive slides from as many sputum samples. In addition, test results were negative for 9 and unknown for 1. In a preliminary trial of 9 independent 3+ positive slides, DNA extraction was attempted by 2 methods: an enzymatic method (7) and a classical thermic lysis in a Chelex suspension (InstaGene; Bio-Rad, Hercules, CA, USA) (8). In our study, only the Chelex method produced good results, i.e., enabled us to obtain DNA that was successfully PCR amplifi ed and produced a full spoligotyping pattern (results not shown). The quantity of DNA extracted was superior for all tests by the enzymatic lysis (n = 3) as by the Chelex (n = 6), as estimated by spectrophotometry (NanoDrop ND-1000; LabTech, Ringmer, UK). Thus, DNA can be successfully extracted by the enzymatic method for many human or bacterial cells but not for M. tuberculosis complex because no spoligotype could be obtained. We therefore analyzed the 176 experimental slides by using the Chelex extraction procedure.
The origins of all ZN slides assessed in this study are shown in the Figure, and genotyping results are shown in the  Ledru et al. in 1996 (5). Third, the T and Haarlem strains represented 16 (22%) and 10 (14%), respectively, of the patterns; other genotypes were rare (5 CAS [Central Asian], 4 X, 1 M. bovis, 1 Beijing, 1 LAM). Finally, the relative prevalence of M. africanum West African 1 from M. africanum West African 2 could fi rst be assessed by the spoligotyping signature (2 vs. 14; online Technical Appendix). Specifi c single nucleotide polymorphism detection could constitute another classifi cation tool for M. africanum sublineages (10,11). Unfortunately, detection of katG203 single nucleotide polymorphism failed on the slide-extracted DNAs (results not shown), and our study is limited by a suboptimal yield in positive spoligotyping results (80 [43%] of 186), an issue that should be improved.

Conclusions
The results of our study diverge on the M. africanum prevalence in Burkina Faso from results from Godreuil et al. (2) (1.9% vs. 20%). These authors were intrigued to not identify more M. africanum isolates and suggested that their fi nding might refl ect "a decrease in M. africanum prevalence in these countries," referring to a similar decrease in Cameroon during 1971-2003 (12,13). We believe that in the study by Godreuil et al., an unintentional bias was introduced against M. africanum, given the diffi culty of isolating this genotypic variant in routine practice in mycobacteriologic laboratories. Differences in M. africanum prevalence in culture-based and sputum-based studies might refl ect the diffi culties of growing and isolating M. africanum in some national TB reference laboratories in western Africa. M. africanum, which is closely related to M. bovis, has peculiar growing requirements that are not always satisfi ed. Supplementation of Löwenstein-Jensen medium with pyruvate is mandatory and not standardized (from 0.1% to 0.4%).
The pyruvate requirements of some members of the M. tuberculosis complex were recently shown to be caused by a mutation creating an inactive pyruvate kinase (14). This specifi c mutation of M. africanum has major implications for its metabolism and growth.
Implementation of adequate culture and molecular identifi cation facilities in Burkina Faso are needed. A potential solution to avoiding the bias from culture and from DNA extraction from slides could be to extract DNA directly from sputum, e.g., by storing surplus sputum prospectively in 70% ethanol. Additional work also is needed to improve analytical methods for ZN slides to refi ne description of M. tuberculosis genetic diversity and eventually to provide predictive genetic drug susceptibility testing. Introduction of newer and faster TB diagnostic methods are urgently needed in this area of western Africa.