Human tuberculosis (TB) is a communicable disease caused by some members of the Mycobacterium tuberculosis complex (MTBC), mainly; Mycobacterium tuberculosis (Mtb), Mycobacterium africanum (Maf) and Mycobacterium bovis (Mbo). It is one of the leading causes of death from a single infectious organism, infecting about a quarter of the world’s population (WHO, 2020). It remains a global pandemic, despite the availability of interventional control measures such as the use of a live attenuated vaccine (BCG) and multi-drug therapy. The situation has been further aggravated by the lack of rapid and reliable, point-of-care diagnostic methods for low-resource areas, and the use of various forms of insufficient treatment procedures among poor resource countries (WHO, 2020).

TB in an individual is often assumed to be caused by a single clonal MTBC lineage, although mixed infections have been previously noted (Hingley-Wilson et al., 2013). Advances in molecular-based approaches in TB studies also demonstrated multiple lineages causing TB in the same patient (Van Rie et al., 2005; Huyen et al., 2012; Zetola et al., 2014) and the occurrence of mixed-lineage TB infections in high TB endemic regions has been reported (Cohen et al., 2011). In TB management, mixed-lineage TB infections have been strongly associated with poor treatment outcome (Zetola et al., 2014).

West-Africa has one of the highest incidences of TB world-wide with a unique set of circulating MTBC species namely: M. tuberculosis, M. africanum and M. bovis. While M. tuberculosis is generally the predominant pathogen for human TB, unusually almost 50% of all TB cases in West Africa are caused by M. africanum (Mostowy et al., 2004). In The Gambia, 39% of TB cases are caused by M. africanum (de Jong et al., 2010a). In Ghana, M. africanum rates remain stable at around 20%, with one of the highest rates of infections in the Northern part of Ghana (De Jong et al., 2009). While the reservoir of infection for M. tuberculosis is the latently infected human population, a non-human reservoir of infection for M. africanum in Ghana has been postulated, likely to be more concentrated in Northern Ghana (Otchere et al., 2018).

The gold standard of TB diagnosis is the isolation of MTBC by culture and the use of biochemical tests (Gholoobi et al., 2014). However, these methods are very laborious and time-consuming which further risk aggravating the condition of patients due to delayed treatment. In addition, with culturing-based techniques in a mixed infection, the fastest growing is often noted as a single infection (Hingley-Wilson et al., 2013). Species differentiation is often challenged with misidentification. For instance, M. africanum Lineage 5 (MafL5) and Lineage 6 (MafL6) exhibit growth characteristics which are intermediates of both M. tuberculosis and M. bovis (de Jong et al., 2010b). Since 2010, WHO recommended the use of GeneXpert assay in diagnostic facilities as a first-line TB diagnostic tool (Goig et al., 2019). It detects MTBC through the identification of insertion sequence (IS6110) as well as identifying rifampicin resistant genes. Although an improved modified GeneXpert Ultra version has been produced with high sensitivity and specificity, it is unable to differentiate the individual MTBC lineages to inform selection of appropriate medication.

In low-resource regions, MTBC lineages are often not differentiated prior to treatment due to reasons such as unavailability of high cost, non-portable genome sequencing machines and length of time for culture results. This can lead to inappropriate treatment regimens, for example, M. bovis is intrinsically resistant to pyrazinamide, one of the frontline drugs used collectively for standard TB treatment (Oryan et al., 2022). Indeed, patient exposure to prolonged pyrazinamide treatment can result in hepatotoxicity and polyarthralgia (Papastavros et al., 2002) and should therefore be avoided if not required. Additionally, antibiotic treatment duration of M. bovis infections is recommended for 9 months (rather than the standard 6 months) because of the absence of pyrazinamide efficacy (Lan et al., 2016). In general, TB treatment durations shorter than recommendation may lead to incomplete sterilization of an infection and increase the risk of the development of antibiotic resistance (Khalif Ali et al., 2017; Ali et al., 2019). It is therefore important to investigate and identify lineage-specific TB molecular markers for designing diagnostic assays with high level of sensitivity and specificity to inform selection of appropriate medication to limit morbidity and drug resistance.

Using the comparative genomics workflow previously described by Akwani et al., 2022, MTBC lineage-specific genes identified were transferred into the development of multiplex PCR assay for TB lineage differentiation. This will enhance precise disease diagnosis, improve epidemiological surveillance studies and help inform selection of appropriate TB drug regimens at early time point especially in low resource settings with high TB incidence.

The ability to differentiate between the lineages of the MBTC is very important in TB management because it provides reliable information for epidemiological surveillance and treatment choice. In this study, MTBC have been phylogenetically classified leading to the identification of lineage-specific genes. These lineage-specific genes have been explored for the development of a multiplex PCR assay which distinguishes between members of the MTBC.

In low resource regions, Ziehl-Neelsen acid-fast staining microscopy is the most common technique used to diagnose TB (Denkinger et al., 2013). It requires about 5,000 – 10,000 bacilli per ml of sputum for successful detection (Ausina Ruiz et al., 2013). Thus, its limitations are low sensitivity as well as the inability to differentiate between different mycobacterial species. Although culture, biochemical tests and sequencing are considered gold standard for identification and differentiation (Gholoobi et al., 2014), these are expensive, laborious and time-consuming. Additionally, obtaining results from these methods are sometimes unreliable due to difficulty in identification of some lineages (M. africanum lineages exhibit growth characteristics which are intermediates of M. tuberculosis and M. bovis) (de Jong et al., 2010a). The advent of genome sequencing techniques has provided relevant data for performing extensive genomic analyses. As a result, several molecular-based assays have been designed to detect MTBCs. These methods are highly sensitive and specific because unique gene sequences are targeted for amplification. Researchers have discovered gene markers such as IS6110, hsp65, dnaJ, psbA, lepA and MPT64 to detect MTBCs against other respiratory pathogens such NTMs (Chin et al., 2018). A recent multiplex PCR assay (Akwani et al., 2022) demonstrated successful separation of Mycobacterium abscessus complex subspecies from other NTMs as well as M. tuberculosis, although evaluation of assay performance in clinical samples needs to be carried out. Since 2010, WHO has recommended the use of GeneXpert MTB/RIF assay as the first-line diagnostic tool which detects MTBC together with rifampicin resistance (Goig et al., 2019). This is a molecular approach based on detection of the repetitive elements IS6110 and IS1081 and rifampicin resistance region. However, misdiagnoses of TB using the GeneXpert assay have been observed in NTM species at a high bacterial load (Pang et al., 2017). In TB endemic areas with infections caused by a diversity of MTBC species, a suitable differential diagnostic approach will be required since GeneXpert lacks the ability to distinguish between MTBC lineages.

In West Africa, MTBC classification has been achieved using spoligotyping technique which involves the amplification of direct repeat copies, followed by hybridization into intergenic spacers experiments (De Jong et al., 2009; de Jong et al., 2010b; Ofori-Anyinam et al., 2016; Otchere et al., 2018; Otchere et al., 2019). This is a two-step approach which is expensive, laborious and time-consuming. In Ghana, a single multiplex PCR experiment was conducted on the MTBC differentiation using primers from spacer regions 33 and 34 of the DR copies of MTBC, IS6110 and the hsp65 (Yeboah-Manu et al., 2001). Although this assay is not successful in separation of M. tuberculosis from M. africanum L6, the assay could still be used to complement biochemical testing.

However, the present study introduces a successful differentiation of MTBCs via a single multiplex PCR method which is rapid, cost effective and has a short turnaround time. The different PCR product sizes can be easily used to distinguish between lineages without the need for sequencing. One advantage of this PCR assay is the ease of adapting it to the available hardware as it will work on any PCR platform. This new assay provides a reliable solution to misdiagnoses with other NTM infections reported in some endemic regions (Brown-Elliott et al., 2012; Yılmaz et al., 2017; He et al., 2022). Indeed, our assay did not cross-react with a range of NTMs, respiratory bacterial and viral pathogens ( Supplementary Tables S4 , S5 ). We tested our assay using clinical samples ( Table 4 ; Supplementary Figure S1 ) to demonstrate its utility at revealing the diversity of MTBC lineages in Ghana. The highest number of cases (24.9%) was caused by M. tuberculosis, followed by M. africanum L5 & L6 (23.5%). M. bovis recorded 1.8% which is comparable to 1.5% observed by Otchere et al., 2019. Negative PCR results (27.0%) and unspeciated lineages (17.0%) may require further confirmation via genome sequencing, although samples have been previously detected as MTBCs by liquid cultures, followed by confirmations using purity tests (on blood agar) and rapid test kit (TB cID) (Sylverken et al., 2021). However, since these are retrospective samples stored over time, sample integrity may have been compromised through repeated freeze/thaw cycles which were beyond our control. Following the reports of mixed MTBC infections among high TB burden settings (Van Rie et al., 2005; Huyen et al., 2012; Zetola et al., 2014), this study detected 20 (5.9%) cases of mixed-lineage TB infections. Poor treatment outcomes have been strongly associated with mixed-lineage TB infections (Zetola et al., 2014). Therefore, the effect of mixed-lineage TB infections in TB management cannot be overlooked as treatment failures are often observed in various regions of Ghana (Agyare et al., 2021).

In summary, this assay is not an alternate replacement for GeneXpert which is currently the first-line TB diagnostic tool recommended by WHO. However, it will be beneficial to low-resource regions where TB is caused by diverse members of the MTBC providing rapid diagnosis to inform appropriate TB drug selection, reduce treatment relapse and the development of antimicrobial resistance. It will also be useful in epidemiological surveillance studies providing reliable information on TB lineage prevalence as well as identifying cases of mixed-lineage tuberculosis infections.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding authors.

The concept and study design were established by SH-W, AS, NR, and GS. Bioinformatic analyses was performed by AvV. Assay optimization by WO, WA, and SH-W. Assay validation using clinical samples was performed in Ghana by RA, SA, and AS. Manuscript writing and editing were done by WO, AS, AvV, GS, and SH-W. All authors contributed to the article and approved the submitted version.