Advertisement

Genetic diversity of the Mycobacterium tuberculosis complex strains from newly diagnosed tuberculosis patients in Northwest Ethiopia reveals a predominance of East-African-Indian and Euro-American lineages

Open AccessPublished:November 13, 2020DOI:https://doi.org/10.1016/j.ijid.2020.11.129

      Highlights

      • Northern Ethiopia hosts a wide diversity of Mycobacterium tuberculosis complex (MTBc) lineages: L1–L4 and L7.
      • East-African-Indian (L3) is the most frequent lineage, followed by Euro-American (L4).
      • MTBc population structure was similar in genotyping of sputa and cultured isolates.
      • Most MTBc strains (94.5%) were susceptible to both rifampicin and isoniazid.

      Abstract

      Objectives

      This study described the population structure of M. tuberculosis complex (MTBc) strains among patients with pulmonary or lymph node tuberculosis (TB) in Northwest Ethiopia and tested the performance of culture isolation and MPT64-based speciation for Lineage 7 (L7).

      Methods

      Patients were recruited between April 2017 and June 2019 in North Gondar, Ethiopia. The MPT64 assay was used to confirm MTBc, and spoligotyping was used to characterize mycobacterial lineages. Line probe assay (LPA) was used to detect resistance to rifampicin and isoniazid.

      Results

      Among 274 MTBc genotyped isolates, there were five MTBc lineages: L1–L4 and L7 were identified, with predominant East-African-Indian (L3) (53.6%) and Euro-American (L4) (40.1%) strains, and low prevalence (2.6%) of Ethiopia L7. The genotypes were similarly distributed between pulmonary and lymph node TB, and all lineages were equally isolated by culture and recognized as MTBc by the MPT64 assay. Additionally, LPA showed that 259 (94.5%) MTBc were susceptible to both rifampicin and isoniazid, and one (0.4%) was multi-drug resistant (resistant to both rifampicin and isoniazid).

      Conclusion

      These findings show that TB in North Gondar, Ethiopia, is mainly caused by L3 and L4 strains, with low rates of L7, confirmed as MTBc by MPT64 assay and with limited resistance to rifampicin and isoniazid.

      Keywords

      Background

      Tuberculosis (TB) remains a major public health threat and the leading infectious cause of morbidity and mortality throughout the world (
      • World Health Organization
      Global tuberculosis report 2019.
      ). According to the World Health Organization (WHO), there were an estimated 10 million incident cases of TB and 1.5 million deaths worldwide, with the African region displaying the highest annual risk of infection, aggravated by high HIV co-infection rates and the emergence of drug-resistant TB (
      • World Health Organization
      Global tuberculosis report 2019.
      ). TB disease, caused by the Mycobacterium tuberculosis (M. tuberculosis) complex (MTBc), has been present in the human population since the beginning of recorded history (
      • Guttierez C.
      • Brisse S.
      • Brosch R.
      • Fabre M.
      • Omaïs B.
      • Marmiesse M.
      • et al.
      Ancient origin and gene mosaicism of the progenitor of Mycobacterium tuberculosis.
      ) and co-evolved with ancient hominids (
      • Gagneux S.
      Host-pathogen coevolution in human tuberculosis.
      ,
      • Comas I.
      • Coscolla M.
      • Luo T.
      • Borrell S.
      • Holt K.E.
      • Kato-Maeda M.
      • et al.
      Out-of-Africa migration and Neolithic coexpansion of Mycobacterium tuberculosis with modern humans.
      ). Today, eight phylogenetic lineages of the MTBc have been identified worldwide, causing TB in humans (
      • Gagneux S.
      • DeRiemer K.
      • Van T.
      • Kato-Maeda M.
      • de Jong B.C.
      • Narayanan S.
      • et al.
      Variable host-pathogen compatibility in Mycobacterium tuberculosis.
      ,
      • Firdessa R.
      • Berg S.
      • Hailu E.
      • Schelling E.
      • Gumi B.
      • Erenso G.
      • et al.
      Mycobacterial lineages causing pulmonary and extrapulmonary Tuberculosis, Ethiopia.
      ,
      • Ngabonziza J.C.S.
      • Loiseau C.
      • Marceau M.
      • Jouet A.
      • Menardo F.
      • Tzfadia O.
      • et al.
      A sister lineage of the Mycobacterium tuberculosis complex discovered in the African Great Lakes region.
      ).
      The distribution of MTBc lineages has shown significant geographical variation (
      • Gagneux S.
      • DeRiemer K.
      • Van T.
      • Kato-Maeda M.
      • de Jong B.C.
      • Narayanan S.
      • et al.
      Variable host-pathogen compatibility in Mycobacterium tuberculosis.
      ,
      • Gagneux S.
      • Small P.M.
      Global phylogeography of Mycobacterium tuberculosis and implications for tuberculosis product development.
      ), with a major impact on disease presentation, drug resistance nature and host adaptation (
      • Ford C.B.
      • Shah R.R.
      • Maeda M.K.
      • Gagneux S.
      • Murray M.B.
      • Cohen T.
      • et al.
      Mycobacterium tuberculosis mutation rate estimates from different lineages predict substantial differences in the emergence of drug-resistant tuberculosis.
      ,
      • Warner D.F.
      • Koch A.
      • Mizrahi V.
      Diversity and disease pathogenesis in mycobacterium tuberculosis.
      ). A high rate of lymph node TB has been reported in Ethiopia (
      • Berg S.
      • Schelling E.
      • Hailu E.
      • Firdessa R.
      • Gumi B.
      • Erenso G.
      • et al.
      Investigation of the high rates of extrapulmonary tuberculosis in Ethiopia reveals no single driving factor and minimal evidence for zoonotic transmission of Mycobacterium bovis infection.
      ,
      • Biadglegne F.
      • Merker M.
      • Sack U.
      • Rodloff A.C.
      • Niemann S.
      Tuberculous lymphadenitis in Ethiopia predominantly caused by strains belonging to the Delhi/CAS lineage and newly identified Ethiopian clades of the mycobacterium tuberculosis complex.
      ,
      • Tadesse M.
      • Abebe G.
      • Bekele A.
      • Bezabih M.
      • de Rijk P.
      • Meehan C.J.
      • et al.
      The predominance of Ethiopian specific Mycobacterium tuberculosis families and minimal contribution of Mycobacterium bovis in tuberculous lymphadenitis patients in Southwest Ethiopia.
      ), and the country uniquely harbors M. tuberculosis Lineage 7 (
      • Firdessa R.
      • Berg S.
      • Hailu E.
      • Schelling E.
      • Gumi B.
      • Erenso G.
      • et al.
      Mycobacterial lineages causing pulmonary and extrapulmonary Tuberculosis, Ethiopia.
      ,
      • Comas I.
      • Hailu E.
      • Kiros T.
      • Bekele S.
      • Mekonnen W.
      • Gumi B.
      • et al.
      Population Genomics of Mycobacterium tuberculosis in Ethiopia Contradicts the Virgin Soil Hypothesis for Human Tuberculosis in Sub-Saharan Africa.
      ), a lineage in-between ‘ancestral’ and ‘modern’ MTBc members. Ethiopia also has a high incidence of TB compared to neighboring countries (
      • Berg S.
      • Schelling E.
      • Hailu E.
      • Firdessa R.
      • Gumi B.
      • Erenso G.
      • et al.
      Investigation of the high rates of extrapulmonary tuberculosis in Ethiopia reveals no single driving factor and minimal evidence for zoonotic transmission of Mycobacterium bovis infection.
      ,
      • World Health Organization
      Global tuberculosis report 2019.
      ), although the reasons remain unclear, justifying further understanding of potential mycobacterial factors. Drug-resistant TB has also been reported as a major problem in different regions of Ethiopia (
      • Mulisa G.
      • Workneh T.
      • Hordofa N.
      • Suaudi M.
      • Abebe G.
      • Jarso G.
      Multidrug-resistant Mycobacterium tuberculosis and associated risk factors in Oromia Region of Ethiopia.
      ,
      • Tesfay K.
      • Tesfay S.
      • Nigus E.
      • Gebreyesus A.
      • Gebreegziabiher D.
      • Adane K.
      More than half of presumptive multidrug-resistant cases referred to a tuberculosis referral laboratory in the Tigray region of Ethiopia are multidrug resistant.
      ,
      • Alene K.A.
      • Viney K.
      • McBryde E.S.
      • Clements A.C.A.
      Spatial patterns of multidrug resistant tuberculosis and relationships to socioeconomic, demographic and household factors in northwest Ethiopia.
      ,
      • Shibabaw A.
      • Gelaw B.
      • Gebreyes W.
      • Robinson R.
      • Wang S.H.
      • Tessema B.
      The burden of pre-extensively and extensively drug-resistant tuberculosis among MDR-TB patients in the Amhara region, Ethiopia.
      ); however, the overall molecular epidemiology of drug-resistant TB is poorly understood in the country.
      Mycobacterial strain genotyping has been used to complement epidemiological studies and provided a helpful understanding of TB transmission dynamics (
      • Kato-Maeda M.
      • Metcalfe J.Z.
      • Flores L.
      Genotyping of Mycobacterium tuberculosis: application in epidemiologic studies.
      ). Genotyping studies have been performed on culture isolates in Ethiopia (
      • Tessema B.
      • Beer J.
      • Merker M.
      • Emmrich F.
      • Sack U.
      • Rodloff A.C.
      • et al.
      Molecular epidemiology and transmission dynamics of Mycobacterium tuberculosis in Northwest Ethiopia: New phylogenetic lineages found in Northwest Ethiopia.
      ,
      • Yimer Sa
      • Norheim G.
      • Namouchi A.
      • Zegeye Ed
      • Kinander W.
      • Tønjum T.
      • et al.
      Mycobacterium tuberculosis lineage 7 strains are associated with prolonged patient delay in seeking treatment for pulmonary tuberculosis in Amhara region, Ethiopia.
      ). It has recently been shown that culture selects for modern M. tuberculosis strains (
      • Sanoussi C.N.
      • Affolabi D.
      • Rigouts L.
      • Anagonou S.
      • de Jong B.
      Genotypic characterization directly applied to sputum improves the detection of Mycobacterium africanum West African 1, under-represented in positive cultures.
      ), thus introducing culture bias in the estimation of the proportion of the different lineages (
      • Metcalfe J.Z.
      • Streicher E.
      • Theron G.
      • Colman R.E.
      • Penaloza R.
      • Allender C.
      • et al.
      Mycobacterium tuberculosis subculture results in loss of potentially clinically relevant heteroresistance.
      ,
      • Sanoussi C.N.
      • Affolabi D.
      • Rigouts L.
      • Anagonou S.
      • de Jong B.
      Genotypic characterization directly applied to sputum improves the detection of Mycobacterium africanum West African 1, under-represented in positive cultures.
      ). Also, data on the association between circulating MTBc lineages and drug resistance in Northwest Ethiopia, one of the high TB burden sub-regions (
      • Alene K.A.
      • Viney K.
      • McBryde E.S.
      • Clements A.C.A.
      Spatial patterns of multidrug resistant tuberculosis and relationships to socioeconomic, demographic and household factors in northwest Ethiopia.
      ,
      • Shibabaw A.
      • Gelaw B.
      • Gebreyes W.
      • Robinson R.
      • Wang S.H.
      • Tessema B.
      The burden of pre-extensively and extensively drug-resistant tuberculosis among MDR-TB patients in the Amhara region, Ethiopia.
      ), remains inadequate.
      This study describes the population structure of MTBc among culture isolates and specimens from TB patients with pulmonary or lymph node presentations in Northwest Ethiopia, and the performance of culture isolation and MPT64-based speciation for Lineage 7.

      Methods

      Study design and patients

      A cross-sectional study of consecutive patients with presumptive TB was conducted between April 2017 and June 2019 in North Gondar at the University of Gondar Hospital (UoGH), Ethiopia. The hospital serves as a comprehensive specialized hospital for people living in North Gondar and the neighboring cities of South Gondar and the Tigray region. It serves as a treatment-initiating center for multi-drug resistant TB (MDR-TB), with screening facilities for patients with presumptive resistant TB. All patients aged ≥15 years who directly presented to UoGH or were referred from other health centers and who were diagnosed with new pulmonary TB or TB lymphadenitis were asked for informed consent before initiating treatment. The sample size was calculated assuming the expected prevalence of Lineage 7 to be 15.6% among pulmonary TB (
      • Yimer Sa
      • Norheim G.
      • Namouchi A.
      • Zegeye Ed
      • Kinander W.
      • Tønjum T.
      • et al.
      Mycobacterium tuberculosis lineage 7 strains are associated with prolonged patient delay in seeking treatment for pulmonary tuberculosis in Amhara region, Ethiopia.
      ) and 9.8% (
      • Biadglegne F.
      • Merker M.
      • Sack U.
      • Rodloff A.C.
      • Niemann S.
      Tuberculous lymphadenitis in Ethiopia predominantly caused by strains belonging to the Delhi/CAS lineage and newly identified Ethiopian clades of the mycobacterium tuberculosis complex.
      ) among TB lymphadenitis, using a 95% confidence interval (CI) and 5% absolute precision (
      • Fenn Buderer N.M.
      Statistical methodology: I. Incorporating the prevalence of disease into the sample size calculation for sensitivity and specificity.
      ). The sample size was 532 for pulmonary TB and 816 for TB lymphadenitis. The computed sample size was not achieved, as only 244 pulmonary TB and 161 TB lymphadenitis patients were enrolled in this study. A structured and pre-tested questionnaire was used to collect sociodemographic and clinical data.

      Culture

      Sputum samples from presumed pulmonary TB patients and lymph node aspirates from patients with presumed TB lymphadenitis were collected based on standard protocols, as described by the WHO and Global Laboratory Initiative (GLI) (
      • Global Laboratory Initiative
      Mycobacteriology laboratory manual, Global Laboratory Initiative.
      ). In the TB culture laboratory, the standard cultivation method on Löwenstein-Jensen (LJ) medium (
      • Global Laboratory Initiative
      Mycobacteriology laboratory manual, Global Laboratory Initiative.
      ) was carried out for isolating mycobacterial species. Before decontamination, an aliquot of the sputum was preserved in 95% ethanol, as previously described (
      • Williams D.L.
      • Gillis T.P.
      • Dupree W.G.
      Ethanol fixation of sputum sediments for DNA-based detection of Mycobacterium tuberculosis.
      ). Briefly, an overnight stand and liquefied sputum sample was added into 1 mL of 95% ethanol in a 2 mL screw-capped cryovial tube (final concentration of 50%), vortexed, and kept at ambient temperature. Subsequently, the mixtures were centrifuged at 13,000 xg for 5 min, the supernatant was discarded, and the sediment (250 μL) was transported to ITM, Antwerp, Belgium. All aliquots were stored at 2–8 °C in Antwerp until DNA extraction.

      Confirmation of Mycobacterium tuberculosis complex

      Preliminary identification of the isolates obtained from positive cultures as being mycobacterial species was performed based on morphological characteristics on LJ media, and detection of acid-fast bacilli (AFB) (with cording) using Ziehl-Neelsen (ZN) smear microscopy (
      • Global Laboratory Initiative
      Mycobacteriology laboratory manual, Global Laboratory Initiative.
      ). The MPT64-based SD Bioline rapid test (BD Diagnostic System, Sparks, MD) confirmed mycobacterial isolates with MTBc, as per manufacturer’s instructions. Also, spoligotyping (
      • Kamerbeek J.
      • Schouls L.
      • Kolk A.
      • van Agterveld M.
      • van Soolingen D.
      • Kuijper S.
      • et al.
      Simultaneous detection and strain differentiation of Mycobacterium tuberculosis for diagnosis and epidemiology.
      ) was used as a reference standard identification for further confirmation of the MPT64-positive and MPT64-negative isolates. Isolates were considered as MTBc positive by spoligotyping when any spoligotype pattern was seen, and negative when the absence of all spacers was observed.

      DNA extraction

      Deoxyribonucleic acid (DNA) was extracted from LJ cultures using the GenoLyse® DNA extraction kit (Hain Life Science, Nehren, Germany) and stored at −20 °C until shipment for spoligotyping in Antwerp. Also, DNA from sputum specimens was extracted by the Promega Maxwell®16 DNA extraction kit (Promega, USA) following the adapted ITM protocol with pre-treatment, and previously published protocols (
      • Eddyani M.
      • Vandelannoote K.
      • Meehan C.J.
      • Bhuju S.
      • Porter J.L.
      • Aguiar J.
      • et al.
      A Genomic Approach to Resolving Relapse versus Reinfection among Four Cases of Buruli Ulcer.
      ). Briefly, each sputum sample (200 μL) was pretreated with 20 μL proteinase K and 200 μL in-house lysis buffer, and homogenized overnight at 60 °C in a shaking incubator (200 rpm). Subsequently, 330 μL of Maxwell®16 lysis buffer was added to the pre-treated sputum, vortexed, and transferred to the cartridge well (maximum 750 μL); the DNA was then eluted in 50 μL elution buffer in a Maxwell®16 LEV device (model AS3000). A Mycobacterium bovis (M. bovis) BCG inactivated suspension and molecular biological (Milli-Q) water were included as positive and negative controls, respectively.

      Genotypic drug susceptibility testing

      The MTBDRplus LPAs, including PCR amplification and hybridization procedures, were conducted according to the manufacturer’s guidelines (
      • Hain Lifescience
      GenoType MTBDRplus products.
      ). The resulting strips were taped to LPA worksheets and interpreted using WHO guidelines (
      • World Health Organization
      The use of molecular line probe assay for the detection of resistance to isoniazid and rifampicin: policy update.
      ). Internal quality control was ensured by using H37Rv DNA as positive and distilled water as a negative control in each run.

      Molecular genotyping

      Spoligotype analysis was performed on in-house prepared membranes after PCR amplification for the direct repeat regions (
      • Kamerbeek J.
      • Schouls L.
      • Kolk A.
      • van Agterveld M.
      • van Soolingen D.
      • Kuijper S.
      • et al.
      Simultaneous detection and strain differentiation of Mycobacterium tuberculosis for diagnosis and epidemiology.
      ). Both M. tuberculosis H37Rv and M. bovis BCG reference strains were included as positive controls, and distilled water as a negative control in each run.

      Genotyping and phylogenetic analyses

      The publicly available international multimarker database of the Pasteur Institute of Guadeloupe (SITVIT2) (http://www.pasteur-guadeloupe.fr:8081/SITVIT2/) (
      • Couvin D.
      • David A.
      • Zozio T.
      • Rastogi N.
      Macro-geographical specificities of the prevailing tuberculosis epidemic as seen through SITVIT2, an updated version of the Mycobacterium tuberculosis genotyping database.
      ) was used to assign genotypes. Mycobacterial lineages and families were defined according to signatures provided in SITVIT2 (
      • Demay C.
      • Liens B.
      • Burguière T.
      • Hill V.
      • Couvin D.
      • Millet J.
      • et al.
      SITVITWEB - A publicly available international multimarker database for studying Mycobacterium tuberculosis genetic diversity and molecular epidemiology.
      ,
      • Couvin D.
      • David A.
      • Zozio T.
      • Rastogi N.
      Macro-geographical specificities of the prevailing tuberculosis epidemic as seen through SITVIT2, an updated version of the Mycobacterium tuberculosis genotyping database.
      ) and based on TBlineage, Spotclust and MIRU-VNTRplus websites. In addition, spolTools (http://spoltools.emi.unsw.edu.au/) were used to analyze MTBc spoligotype data for probable relationships among genotypes using Spoligoforests (spolTools; hierarchical layout) (
      • Tang C.
      • Reyes J.F.
      • Luciani F.
      • Francis A.R.
      Tanaka MM. spolTools: Online utilities for analyzing spoligotypes of the Mycobacterium tuberculosis complex.
      ).

      Statistical analysis

      Descriptive statistics like frequency distributions and percentages were conducted using STATA® 15.1 (StataCorp, USA). Proportions of LPA and spoligotype results for both MTBc lineages were compared using the two-sample proportion test analysis. Odds ratios (OR) and 95% confidence levels (CIs) were calculated to measure the association of MTBc lineages or families with patients’ sociodemographics and disease presentation using binary logistic regression. Additionally, two-sided Pearson’s Chi-square test was used to assess associations of geographic MTBc lineage distribution in study districts. A p-value of < 0.05 was considered statistically significant.

      Ethical review

      Institutional permission to perform this study was obtained from the Institutional Research and Ethical review committees of the University of Gondar, Ethiopia, and also from the Institute of Tropical Medicine (ITM, Antwerp) and University of Antwerp (UZA), Belgium before the commencement of research activities. Voluntarily signed informed consent was obtained from each patient and no patient personal identifiers were used in this study.

      Results

      Patients and MTBc characteristics

      A total of 405 patients were recruited, including 244 with new pulmonary TB and 161 with TB lymphadenitis (Figure 1). Most patients were male (249/405, 61.5%) and the median age was 30.0 years (IQR 24.0–40.0). The median body mass index (BMI) was 18.7 kg/m2 (range 11.3–26.3, IQR 17.2–20.5). Among 211 TB patients with known HIV test results, 30 (14.2%) were HIV-positive. Few patients (6.7%) had been in prison or had a smoking history (2.5%). The majority of patients (69.1%) were from rural areas of the districts (Table 1).
      Figure 1
      Figure 1TB patients included in spoligotyping and line probe assay (LPA) analysis.
      Table 1Patients’ characteristics by type of TB presentation.
      Total (n = 405)Pulmonary TB (n = 244)TB lymphadenitis (n = 161)OR[95% CI]p-value
      Gender, n (%)
      Male249 (61.5)160 (65.6)89 (55.3)0.730.43–1.240.250
      Age, median (IQR)30.0 (24–40)28.5 (23–38)30.0 (24–40)1.260.74–2.160.395
      BMI, median (IQR)18.7 (17.2–20.2)18.4 (16.8–20.0)19.5 (18.0–20.9)2.061.24–3.430.005
      HIV status, n (%)
      Positive30 (7.4)23 (9.4)7 (4.4)0.670.21–2.140.495
      Negative181 (44.7)100 (41.0)81 (50.3)1.410.83–2.390.201
      unknown194 (47.9)121 (49.6)73 (45.3)Ref.
      Prison history, n (%)
      Yes27 (6.7)12 (4.9)15 (9.3)2.650.93–7.570.068
      Smoking history, n (%)
      Yes10 (2.5)6 (2.5)4 (2.5)1.100.27–4.490.898
      Patient address, n (%)
      Rural280 (69.1)155 (63.5)125 (77.6)2.291.26–4.170.006
      Urban125 (30.9)89 (36.5)36 (22.4)Ref.
      Smear microscopy
      Positive176 (43.5)148 (60.7)28 (17.4)0.140.08–0.220.000
      BMI, body mass index; CI, confidence interval; HIV, human immunodeficiency virus; IQR, interquartile range; OR, odds ratio; Ref., reference; TB, tuberculosis.
      In total, the study showed smear-positive samples among 60.7% (148/244) of pulmonary TB- and 17.4% (28/161) of TB lymphadenitis patients (p <  0. 001, Table 1). Among 405 culture processed samples, a total of 280 (69.1%) were positive: 79.5% (194/244) of pulmonary specimens and 53.4% (86/161) of TB lymphadenitis patients. Most samples that did not yield an isolate had no growth in culture (45/244 from pulmonary TB and 74/161 from TB lymphadenitis), while only a few were contaminated (Figure 1). Of the 280 isolates, 274 (97.9%) were identified as MTBc by the MPT64 Ag test (Figure 1).

      Genotypes of the Mycobacterium tuberculosis complex

      Spoligotyping on isolates yielded 274/280 interpretable results (Figure 1). The six MPT64-negative pulmonary isolates included four likely non-TB mycobacteria (negative spoligotypes) and two MPT64 false-negative MTBc isolates (one isolate from L1 and the other from L3). All L7 isolates were correctly classified as MTBc by MPT64 analysis.
      Among the 274 spoligotype results, 68 different spoligopatterns were identified: 44 known in the SITVIT2 database and 24 orphan patterns (Table 1 and Supplementary Table 1). Amongst the total 68 spoligopatterns, 38 distinct spoligotypes were identified from pulmonary TB, nine patterns from TB lymphadenitis patients, and 21 patterns were shared by both pulmonary TB and TB lymphadenitis. Five major MTBc lineages (Ls) were identified: eight (2.9%) Indo-Oceanic (L1), two (0.7%) East Asian (Beijing, L2), 147 (53.6%) East-African-Indian (L3), 110 (40.1%) Euro-American (L4), and seven (2.6%) Ethiopian lineage (L7) (Table 2). Within the predominant East-African-Indian L3, the CAS1-Delhi (140, 51.1%) was the most common family, with few isolates classified as CAS1-Kili (4, 1.5%) and CAS (3, 1.1%). Euro-American L4 comprised T- (42, 15.3%), Haarlem- (28, 10.2%), T3-ETH- (20, 7.3%), LAM- (16, 5.8%), S- (2, 0.7%) and X- (2, 0.7%) families (Table 2 and Supplementary Table 1). Fifty orphan spoligotypes (18.2%) were further analyzed to the nearest lineages and/or families by the TBlineage and Spotclust database and added into those identified MTBc families (Table 2).
      Table 2Distribution of Mycobacterium tuberculosis complex (MTBc) lineages and families among isolates, by TB presentation.
      Lineages (Ls)Total, n (%)Family (n)SIT nrNo. of total isolates, n/274 (%)TB presentation
      PTB, (%) n = 188TB LN, (%) n = 86
      Indo-Oceanic (L1)8 (2.9)Family 36 (1)SIT41 (0.4)1 (0.5)0 (0.0)
      Manu (2)SIT542 (0.7)0 (0.0)2 (2.3)
      Family 34 (5)SIT564 (1.4)3 (1.4)1 (1.2)
      orphan1 (0.4)1 (0.5)0 (0.0)
      East Asian (L2)2 (0.7)Beijing (2)SIT12 (0.7)1 (0.5)1 (1.2)
      East-African-Indian (L3)147 (53.6)CAS1-Delhi (140)SIT11984 (1.4)3 (1.4)1 (1.2)
      SIT11992 (0.7)1 (0.5)1 (1.2)
      SIT13431 (0.4)1 (0.5)0 (0.0)
      SIT1411 (0.4)1 (0.5)0 (0.0)
      SIT1422 (0.7)2 (1.1)0 (0.0)
      SIT2473 (1.1)1 (0.5)2 (2.3)
      SIT2589 (32.5)56 (29.8)33 (38.4)
      SIT2610 (3.6)9 (4.8)1 (1.2)
      SIT9523 (1.1)2 (1.1)1 (1.2)
      orphan25 (9.1)22 (11.7)3 (3.5)
      CAS1-Kili (n = 4)SIT214 (1.4)2 (1.1)2 (2.3)
      CAS (n = 3)SIT17891 (0.4)1 (0.5)0 (0.0)
      orphan2 (0.7)2 (1.1)0 (0.0)
      Euro-American (L4)110 (40.1)Haarlem (n = 28)SIT1214 (1.4)4 (2.1)0 (0.0)
      SIT13412 (4.4)10 (5.3)2 (2.3)
      SIT351 (0.4)0 (0.0)1 (1.2)
      SIT471 (0.4)0 (0.0)1 (1.2)
      SIT504 (1.4)1 (0.5)3 (3.5)
      SIT7504 (1.4)3 (1.4)1 (1.2)
      SIT7642 (0.7)0 (0.0)2 (2.3)
      T-family (42)SIT16261 (0.4)1 (0.5)0 (0.0)
      SIT16881 (0.4)1 (0.5)0 (0.0)
      SIT17451 (0.4)0 (0.0)1 (1.2)
      SIT1961 (0.4)1 (0.5)0 (0.0)
      SIT2051 (0.4)1 (0.5)0 (0.0)
      SIT3581 (0.4)1 (0.5)0 (0.0)
      SIT441 (0.4)1 (0.5)0 (0.0)
      SIT523 (1.1)1 (0.5)2 (2.3)
      SIT5321 (7.7)13 (6.9)8 (9.3)
      orphan11 (4.0)8 (4.3)3 (3.5)
      T3-ETH (20)SIT14919 (6.9)10 (5.3)9 (10.5)
      orphan1 (0.4)1 (0.5)0 (0.0)
      LAM (16)SIT331 (0.4)1 (0.5)0 (0.0)
      SIT412 (0.7)1 (0.5)1 (1.2)
      SIT421 (0.4)1 (0.5)0 (0.0)
      SIT591 (0.4)1 (0.5)0 (0.0)
      SIT2301 (0.4)1 (0.5)0 (0.0)
      orphan10 (3.6)9 (4.8)1 (1.2)
      S-family (2)SIT341 (0.4)1 (0.5)0 (0.0)
      SIT1561 (0.4)1 (0.5)0 (0.0)
      X-family (2)SIT1191 (0.4)1 (0.5)0 (0.0)
      SIT3361 (0.4)1 (0.5)0 (0.0)
      Ethiopian lineage (L7)7 (2.6)ETH1 family (7)SIT3431 (0.4)1 (0.5)0 (0.0)
      SIT9101 (0.4)0 (0.0)1 (1.2)
      SIT17295 (1.8)3 (1.6)2 (2.3)
      Mycobacterium tuberculosis complex lineages and families from 274 isolates from pulmonary TB (PTB) and TB lymphadenitis (TB LN) patients in Northwest Ethiopia.
      L, lineage; CAS, Central Asian; LAM family, Latin-American-Mediterranean.
      Overall, 234 (85.4%) MTBc isolates were clustered by spoligotyping, with 21 clusters comprising 2–89 isolates per cluster, while 40 isolates were singletons. The most predominant cluster was largely designated by the SIT25 (identified as CAS1-Delhi) with 89 MTBc isolates, followed by SIT53 (n = 21 isolates). The majority of MTBc clusters were from pulmonary TB (66.2%; n = 155) and rural areas (67.5%; n = 158) (Supplementary Table 1).

      Distribution of genotypes among direct (sputum) and indirect (isolate) spoligotyping

      Direct spoligotyping was successful for 180 (73.8%) of 244 sputum specimens, while 55 (22.5%) were negative (missing all spacers; of which 47 samples were microscopy smear-negative) and nine (3.7%) had non-interpretable spoligotype results (too weak spots; of which four were smear-negative). The majority (93.3%; 168/180) of sputa with successful direct spoligotype results were also positive in culture. Of the 168 patients with both direct (sputa) and indirect (isolates) spoligotype results available, spoligotypes were concordant in 91.1% (153/168) (Supplementary Table 2). The 15 discrepant profiles differed by one spacer (n = 5), two (n = 3), three (n = 3), five (n = 3), or 28 spacers (n = 1) (Supplementary Table 2). Those with ≤5 spacers difference were identified without any change between families and/or lineage pairs, while the 28 spacer difference led to an interfamily and interlineage change (L4 T3-ETH in sputum and L3 CAS1_Delhi in culture) (Supplementary Table 3). The proportion of lineages did not significantly differ between spoligotype analysis directly on sputum versus on cultured isolates (73.8% versus 77.0%; 95% CI [-0.12–0.56]; p = 0.476) (Table 3). Notably, the proportion of L7 was similar in sputum and isolates (Table 3).
      Table 3Mycobacterium tuberculosis complex (MTBc) lineages and families obtained by direct (on sputum) and indirect (on isolates) spoligotyping.
      Lineages (Ls)FamilySpoligotyping resultsp -value*[95% CI]
      p-value and 95% CI were calculated using the two sample proportion test.
      Direct (sputa), N = 180 (%)Indirect (isolates), N = 188 (%)
      Indo-Oceanic (L1)5 (2.8)5 (2.6)0.984[-0.19–0.20]
      Family 361 (0.6)1 (0.5)
      Family 344 (2.2)4 (2.1)
      East Asian (L2)1 (0.6)1 (0.5)
      Beijing1 (0.6)1 (0.5)
      East African Indian (L3)101 (56.1)104 (55.3)0.908[-0.13–0.14]
      CAS1-Delhi98 (54.4)99 (52.6)
      CAS1-Killi1 (0.6)2 (1.1)
      CAS2 (1.1)3 (1.6)
      Euro-American (L4)69 (38.3)74 (39.4)0.893[-0.17–0.15]
      Haarlem16 (8.9)18 (9.6)
      T-family24 (13.3)28 (14.9)
      T3-ETH11 (6.1)10 (5.3)
      LAM13 (7.2)14 (7.4)
      S-family2 (1.1)2 (1.1)
      X-family3 (1.7)2 (1.1)
      Ethiopian lineage (L7)4 (2.2)4 (2.1)0.992[-0.20–0.20]
      ETH1 family4 (2.2)4 (2.1)
      CAS, Central Asian; CI, confidence interval; L, lineage; LAM family, Latin-American-Mediterranean.
      * p-value and 95% CI were calculated using the two sample proportion test.

      Spoligoforests among spoligotypes and MTBc genotypes

      The spoligoforest tree was drawn by using the spoligotype data in http://www.emi.unsw.edu.au/spolTools/ (
      • Tang C.
      • Reyes J.F.
      • Luciani F.
      • Francis A.R.
      Tanaka MM. spolTools: Online utilities for analyzing spoligotypes of the Mycobacterium tuberculosis complex.
      ) and presented with a Hierarchical layout in Figure 2, which shows four parent trees with linked nodes and 13 distinct nodes. The tree shows SIT26, SIT54, SIT910, and SIT35 spoligotypes as a root (i.e. these spoligotypes represent the parent spoligotypes). Six spoligotypes descended from SIT26, of which five spoligotypes consisted of small clusters (SIT952 (n = 3), SIT247 (n = 3), SIT142 (n = 2), SIT143 (n = 1), and SIT1343 (n = 1)), and one spoligotype (SIT25, named as CAS1-Delhi) with a big cluster (n = 89), whereas SIT54 (Manu family) descended into one big cluster (SIT53, named as T-family). The spoligoforest tree also indicates that SIT910 (ETH-L7) provided the SIT1720 and SIT343 spoligotypes, which were designated as the genotypes of the Ethiopian L7. It was also observed that the tree visualizes the locations of the predominant SITs such as SIT26, SIT25, SIT53, SIT149, and SIT134 in the hierarchical chain. The majority (n = 9) of the 13 distinct nodes represented orphan spoligotypes.
      Figure 2
      Figure 2The spoligoforest trees were drawn using http://www.emi.unsw.edu.au/spolTools/ in Hierarchical Layout. Each spoligotype pattern represents a node in the tree with the number of isolates. Loss of spacers is represented by the edges between nodes, with the arrowheads pointing to descendant spoligotypes. Solid black lines link patterns with a maximum weight of distance. Dashed lines represent a link of weight comprised between 0.5 and 1, and dotted lines represent a link of weight <0.5 (
      • Tang C.
      • Reyes J.F.
      • Luciani F.
      • Francis A.R.
      Tanaka MM. spolTools: Online utilities for analyzing spoligotypes of the Mycobacterium tuberculosis complex.
      ).

      Geographical distribution of MTBc lineages

      The majority of the 274 MTBc isolates were identified from four districts: Gondar (74, 27.0%), Dembia (50, 18.3%), Gondar Zuria (31, 11.3%), and Wegera (31, 11.3%) (Figure 3 and Supplementary Table 4). Genotypes belonging to L3 and L4 were widespread in all districts, while the ancient L1 was found only in Alefa, Belesa, Dembia, Gondar, and Wegera with one or two isolates each. The only two L2 (Beijing) isolates were identified in Gondar and Belesa districts. The Ethiopian L7 was only identified in Belesa, Chilga, Dembia, Gondar, and Wegera districts with one or two isolates each. The geographic distribution of lineages did not significantly differ by district (p >  0.05, data not shown).
      Figure 3
      Figure 3Distribution of TB patients and MTBc lineages in the different districts of the study area of Northwest Ethiopia. The map was created using a QGIS Desktop 3.12.0 (https://qgis.org/en/site/forusers/download.html) (
      • Team QD
      QGIS Geographic Information System.
      ) for MTBc lineage distribution. Free shapefiles were downloaded from the DIVA-GIS website: http://www.diva-gis.org/gdata at free spatial data’ using country level and administrative areas.

      Distribution of MTBc lineages amongst the pulmonary TB and TB lymphadenitis

      Most isolates (68.6%; n = 188) were from pulmonary TB patients. In both disease presentations, L3 and L4 strains were most common, and the lineage distribution did not significantly associate with TB disease presentation (OR = 1.1, p =  0.724, 95% CI [0.78–1.41]) (Supplementary Table 5).

      Drug susceptibility profiles

      Based on the GenoType LPA results of 274 MTBc isolates, 259 (94.5%) MTBc isolates were susceptible to both rifampicin (RIF) and isoniazid (INH), while 13 (4.7%) showed INH mono-resistance, one (0.4%) RIF mono-resistance, and one (0.4%) isolate was resistant to both INH and RIF (MDR-TB) (Supplementary Table 6). All 13 INH mono-resistant isolates had the same mutation in the katG gene (katG MUT1), the globally widespread S315 T mutation. The rifampicin-resistant isolate showed a missing rpoBWT8 band only on the LPA strip, presumably mutation L452 P, while the MDR isolate showed the most common rpoB S450 L and katG S315 T mutations. Isoniazid mono-resistant isolates belonged to diverse families within L3 and L4, without association between resistance and any specific genotype (data not shown).

      Discussion

      In Ethiopia, home to a distinct MTBc population structure that includes global Lineages 1–4 and the Ethiopia-specific Lineage 7, knowledge of the genotypic distribution of mycobacterial strains within the country is currently limited, with prior studies suggesting that L7 is more predominant in the north of the country. Ethiopia is furthermore unique in the high proportion of TB presenting as lymphadenitis, which is not explained by pastoralist populations infected with M. bovis (
      • Berg S.
      • Schelling E.
      • Hailu E.
      • Firdessa R.
      • Gumi B.
      • Erenso G.
      • et al.
      Investigation of the high rates of extrapulmonary tuberculosis in Ethiopia reveals no single driving factor and minimal evidence for zoonotic transmission of Mycobacterium bovis infection.
      ). This study, therefore, described the population structure of circulating MTBc strains among TB patients in Northwest Ethiopia. The predominance of the East-African-Indian L3 and Euro-American L4 was consistent with other studies in Ethiopia (
      • Tessema B.
      • Beer J.
      • Merker M.
      • Emmrich F.
      • Sack U.
      • Rodloff A.C.
      • et al.
      Molecular epidemiology and transmission dynamics of Mycobacterium tuberculosis in Northwest Ethiopia: New phylogenetic lineages found in Northwest Ethiopia.
      ,
      • Biadglegne F.
      • Merker M.
      • Sack U.
      • Rodloff A.C.
      • Niemann S.
      Tuberculous lymphadenitis in Ethiopia predominantly caused by strains belonging to the Delhi/CAS lineage and newly identified Ethiopian clades of the mycobacterium tuberculosis complex.
      ,
      • Tadesse M.
      • Abebe G.
      • Bekele A.
      • Bezabih M.
      • de Rijk P.
      • Meehan C.J.
      • et al.
      The predominance of Ethiopian specific Mycobacterium tuberculosis families and minimal contribution of Mycobacterium bovis in tuberculous lymphadenitis patients in Southwest Ethiopia.
      ,
      • Damena D.
      • Tolosa S.
      • Hailemariam M.
      • Zewude A.
      • Worku A.
      • Mekonnen B.
      • et al.
      Genetic diversity and drug susceptibility profiles of Mycobacterium tuberculosis obtained from Saint Peter’s TB specialized Hospital, Ethiopia.
      ), albeit with less L7 than the expected prevalence of 9–15% (
      • Biadglegne F.
      • Merker M.
      • Sack U.
      • Rodloff A.C.
      • Niemann S.
      Tuberculous lymphadenitis in Ethiopia predominantly caused by strains belonging to the Delhi/CAS lineage and newly identified Ethiopian clades of the mycobacterium tuberculosis complex.
      ,
      • Yimer Sa
      • Norheim G.
      • Namouchi A.
      • Zegeye Ed
      • Kinander W.
      • Tønjum T.
      • et al.
      Mycobacterium tuberculosis lineage 7 strains are associated with prolonged patient delay in seeking treatment for pulmonary tuberculosis in Amhara region, Ethiopia.
      ). L3 and L4 have also been reported in other African countries, particularly in neighboring Sudan (
      • Mbugi E.V.
      • Katale B.Z.
      • Streicher E.M.
      • Keyyu J.D.
      • Kendall S.L.
      • Dockrell H.M.
      • et al.
      Mapping of Mycobacterium tuberculosis Complex Genetic Diversity Profiles in Tanzania and Other African Countries.
      ,
      • Chihota V.N.
      • Niehaus A.
      • Streicher E.M.
      • Wang X.
      • Sampson S.L.
      • Mason P.
      • et al.
      Geospatial distribution of Mycobacterium tuberculosis genotypes in Africa.
      ,
      • Couvin D.
      • Reynaud Y.
      • Rastogi N.
      Two tales: Worldwide distribution of Central Asian (CAS) versus ancestral East-African Indian (EAI) lineages of Mycobacterium tuberculosis underlines a remarkable cleavage for phylogeographical, epidemiological and demographical characteristics.
      ,
      • Shuaib Y.A.
      • Khalil E.A.G.
      • Wieler L.H.
      • Schaible U.E.
      • Bakheit M.A.
      • Mohamed-Noor S.E.
      • et al.
      Mycobacterium tuberculosis Complex Lineage 3 as Causative Agent of Pulmonary Tuberculosis, Eastern Sudan.
      ). The current findings confirm that L3 and L4 are endemic TB strains in this study area, possibly due to co-adaptation with its human host (
      • Gagneux S.
      Host-pathogen coevolution in human tuberculosis.
      ,
      • Couvin D.
      • Reynaud Y.
      • Rastogi N.
      Two tales: Worldwide distribution of Central Asian (CAS) versus ancestral East-African Indian (EAI) lineages of Mycobacterium tuberculosis underlines a remarkable cleavage for phylogeographical, epidemiological and demographical characteristics.
      ), and capable of evading the host immune response and progress to rapid TB disease and transmission (
      • de Jong B.C.
      • Hill P.C.
      • Aiken A.
      • Awine T.
      • Antonio M.
      • Adetifa I.M.
      • et al.
      Progression to active tuberculosis, but not transmission, varies by Mycobacterium tuberculosis lineage in The Gambia.
      ,
      • Tanveer M.
      • Hasan Z.
      • Kanji A.
      • Hussain R.
      • Hasan R.
      Reduced TNF-α and IFN-γ responses to Central Asian strain 1 and Beijing isolates of Mycobacterium tuberculosis in comparison with H37Rv strain.
      ,
      • Portevin D.
      • Gagneux S.
      • Comas I.
      • Young D.
      Human macrophage responses to clinical isolates from the Mycobacterium tuberculosis complex discriminate between ancient and modern lineages.
      ). It has been also shown that strains of evolutionarily modern L3 and L4 lineages are successful human pathogens with increased virulence, enabling their worldwide distribution (Hershberg et al., 2008;
      • Gagneux S.
      Host-pathogen coevolution in human tuberculosis.
      ). The current study area, North Gondar, is one of the most touristic places in Ethiopia and is also known for its mobility between the neighboring regions and bordering Sudan, which could play a role in the TB transmission dynamics in the area.
      Even though this study showed a lower prevalence of L7 (2.6%) than expected, the MPT64-based rapid test efficiently identified L7 strains in culture, as compared with the lower sensitivity of this assay to detect the West African specific lineages of the MTBc strain (
      • Ofori-Anyinam B.
      • Kanuteh F.
      • Agbla S.C.
      • Adetifa I.
      • Okoi C.
      • Dolganov G.
      • et al.
      Impact of the Mycobaterium africanum West Africa 2 Lineage on TB Diagnostics in West Africa.
      ,
      • N’Dira Sanoussi C.
      • de Jong B.C.
      • Odoun M.
      • Arekpa K.
      • Ligali M.A.
      • Bodi O.
      • et al.
      Low sensitivity of the MPT64 identification test to detect lineage 5 of the Mycobacterium tuberculosis complex.
      ). L7 was initially identified from patients in Woldia districts of the Amhara region, Ethiopia, at a prevalence of 13% (
      • Firdessa R.
      • Berg S.
      • Hailu E.
      • Schelling E.
      • Gumi B.
      • Erenso G.
      • et al.
      Mycobacterial lineages causing pulmonary and extrapulmonary Tuberculosis, Ethiopia.
      ), and later in other places in Southwest and Northwest Ethiopia, with a prevalence ranging 2–15% (
      • Biadglegne F.
      • Merker M.
      • Sack U.
      • Rodloff A.C.
      • Niemann S.
      Tuberculous lymphadenitis in Ethiopia predominantly caused by strains belonging to the Delhi/CAS lineage and newly identified Ethiopian clades of the mycobacterium tuberculosis complex.
      ,
      • Yimer Sa
      • Norheim G.
      • Namouchi A.
      • Zegeye Ed
      • Kinander W.
      • Tønjum T.
      • et al.
      Mycobacterium tuberculosis lineage 7 strains are associated with prolonged patient delay in seeking treatment for pulmonary tuberculosis in Amhara region, Ethiopia.
      ,
      • Tadesse M.
      • Abebe G.
      • Bekele A.
      • Bezabih M.
      • de Rijk P.
      • Meehan C.J.
      • et al.
      The predominance of Ethiopian specific Mycobacterium tuberculosis families and minimal contribution of Mycobacterium bovis in tuberculous lymphadenitis patients in Southwest Ethiopia.
      ) and representing SIT910 and SIT1729 families (designated as Ethiopia_1 in MIRU-VNTRplus database). This variability might be due to chance or possible human host susceptibility differences in the study area, as a previous study (
      • Yimer Sa
      • Norheim G.
      • Namouchi A.
      • Zegeye Ed
      • Kinander W.
      • Tønjum T.
      • et al.
      Mycobacterium tuberculosis lineage 7 strains are associated with prolonged patient delay in seeking treatment for pulmonary tuberculosis in Amhara region, Ethiopia.
      ,
      • Yimer S.A.
      • Namouchi A.
      • Zegeye E.D.
      • Holm-Hansen C.
      • Norheim G.
      • Abebe M.
      • et al.
      Deciphering the recent phylogenetic expansion of the originally deeply rooted Mycobacterium tuberculosis lineage 7.
      ) reported that the L7 genotype was associated with slow in vitro growth and reflected lower fitness and virulence in TB patients than other lineages. The current study only included patients diagnosed with pulmonary TB and clinically diagnosed TB lymphadenitis patients who had direct access and/or were referred to a specialized university hospital. They might not be representative of all TB patients in the remote rural and general population of North Gondar, and may hence reflect the lower occurrence of L7 in this study. Furthermore, the small sample size of study participants might have impacted the prevalence estimates of L7. Future studies might be needed to explain the relationship between L7 virulence and clinical presentation in unbiased sampling at the first point of contact for TB diagnostics.
      Genotypes of M. tuberculosis infection in pulmonary disease and lymph nodes were similar, with a predominance of the CAS1-Dehli; this agrees with other studies (
      • Srilohasin P.
      • Chaiprasert A.
      • Tokunaga K.
      • Nishida N.
      • Prammananan T.
      • Smittipat N.
      • et al.
      Genetic diversity and dynamic distribution of mycobacterium tuberculosis isolates causing pulmonary and extrapulmonary tuberculosis in Thailand.
      ,
      • Hadifar S.
      • Shamkhali L.
      • Kargarpour Kamakoli M.
      • Mostafaei S.
      • Khanipour S.
      • Mansoori N.
      • et al.
      Genetic diversity of Mycobacterium tuberculosis isolates causing pulmonary and extrapulmonary tuberculosis in the capital of Iran.
      ). In Ethiopia, like most developing countries, people living in areas with uncontrolled TB transmission (
      • Kebede A.H.
      • Alebachew Wagaw Z.
      • Tsegaye F.
      • Lemma E.
      • Abebe A.
      • Agonafir M.
      • et al.
      The first population-based national tuberculosis prevalence survey in Ethiopia, 2010-2011.
      ) are more likely to have mixed MTBc infections, in which the strains might be different in lineages and/or families, including heterogeneous drug susceptibility profiles (
      • Shamputa I.C.
      • Rigouts L.
      • Eyongeta L.A.
      • El Aila N.A.
      • van Deun A.
      • Salim A.H.
      • et al.
      Genotypic and Phenotypic Heterogeneity among Mycobacterium tuberculosis Isolates from Pulmonary Tuberculosis Patients.
      ,
      • Cohen T.
      • van Helden P.D.
      • Wilson D.
      • Colijn C.
      • McLaughlin M.M.
      • Abubakar I.
      • et al.
      Mixed-strain Mycobacterium tuberculosis infections and the implications for tuberculosis treatment and control.
      ). While the current methodology was not focused on the detection of mixed infection, the patient with L3 in sputum and L4 in the derived culture may indeed have mixed infection.
      Otherwise, direct (sputum) spoligotyping showed a similar distribution of MTBc families and lineages compared to using indirect (isolate) spoligotyping in this study. However, the yield of spoligotypes was slightly lower than previously reported (
      • Sanoussi C.N.
      • Affolabi D.
      • Rigouts L.
      • Anagonou S.
      • de Jong B.
      Genotypic characterization directly applied to sputum improves the detection of Mycobacterium africanum West African 1, under-represented in positive cultures.
      ,
      • Kargarpour Kamakoli M.
      • Khanipour S.
      • Hadifar S.
      • Ghajavand H.
      • Farmanfarmaei G.
      • Fateh A.
      • et al.
      Challenge in direct Spoligotyping of Mycobacterium tuberculosis: a problematic issue in the region with high prevalence of polyclonal infections.
      ). This difference might be due to inclusion of smear-negative sputum specimens, below the limit of detection of direct spoligotype analysis, and DNA extraction of preserved sputa pellets (aliquots) after a long period of storage in the current study. Studies indicated that direct genotyping, such as spoligotyping and MIRU-VNTRplus, is more successful using baseline samples with high smear grades (
      • Bidovec-Stojkovič U.
      • Seme K.
      • Žolnir-Dovč M.
      • Supply P.
      Prospective genotyping of mycobacterium tuberculosis from fresh clinical samples.
      ,
      • Sanoussi C.N.
      • Affolabi D.
      • Rigouts L.
      • Anagonou S.
      • de Jong B.
      Genotypic characterization directly applied to sputum improves the detection of Mycobacterium africanum West African 1, under-represented in positive cultures.
      ), although the nature and quality of the sample may influence the extraction method, DNA contents and interpretability of spoligotyping results (
      • Tønjum T.
      • Klintz L.
      • Bergan T.
      • Baann J.
      • Furuberg G.
      • Cristea M.
      • et al.
      Direct detection of Mycobacterium tuberculosis in respiratory samples from patients in Scandinavia by polymerase chain reaction.
      ,
      • Pathak D.
      • Chakravorty S.
      • Hanif M.
      • Tyagi J.S.
      Lysis of tubercle bacilli in fresh and stored sputum specimens: Implications for diagnosing tuberculosis in stored and paucibacillary specimens by PCR.
      ). Previous studies conducted using direct spoligotyping have also reported lower detection rates of MTBc strains in patients with lower bacillary load specimens (
      • Suresh N.
      • Arora J.
      • Pant H.
      • Rana T.
      • Singh U.B.
      Spoligotyping of Mycobacterium tuberculosis DNA from Archival Ziehl-Neelsen-stained sputum smears.
      ,
      • Cafrune P.I.
      • Possuelo L.G.
      • Ribeiro A.W.
      • Ribeiro M.O.
      • Unis G.
      • Jarczewski C.A.
      • et al.
      Prospective study applying spoligotyping directly to DNA from sputum samples of patients suspected of having tuberculosis.
      ), with insufficient DNA to be detected by spoligotyping.
      In addition to the general dominance of L3 and L4 strains, it was also found that all drug-resistant TB was identified among L3 (4.8%, 7/147) and L4 (7.3%, 8/110) strains. Despite the population including referral patients, only one (0.4%) MDR isolate was identified in contrast to earlier reports in Ethiopia, with a range of 1.4–5.2% from newly diagnosed TB patients (
      • Damena D.
      • Tolosa S.
      • Hailemariam M.
      • Zewude A.
      • Worku A.
      • Mekonnen B.
      • et al.
      Genetic diversity and drug susceptibility profiles of Mycobacterium tuberculosis obtained from Saint Peter’s TB specialized Hospital, Ethiopia.
      ,
      • Diriba G.
      • Kebede A.
      • Tola H.H.
      • Alemu A.
      • Tadesse M.
      • Tesfaye E.
      • et al.
      Surveillance of drug resistance tuberculosis based on reference laboratory data in Ethiopia.
      ,
      • World Health Organization
      Global tuberculosis report 2019.
      ). However, 4.7% INH resistance with the katGS315 T mutation was found in the current study; these patients are at higher risk of acquiring RIF resistance. The current study, however, might not reflect the actual drug-resistant TB prevalence in the general population, since data were collected from the referral hospital, which was likely enriched with TB patients suspected for complicated diseases such as drug resistance. LPA might also have a lower resolution of resistant strains compared to phenotypic drug susceptibility testing, as rpoB mutations outside the 81 bp target region might have been missed.
      This study had limitations. It only used spoligotyping to characterize MTBc lineages and families, and this could have resulted in limited resolution of genotypes, precluding identification of chains of transmission. Due to resource constraints, the study only included patients attending a specialized university hospital. It also did not explore statistical analysis for clustered spoligotypes due to the unequal proportion between pulmonary and lymph node TB. Furthermore, because of the small quantity of lymph node aspirates available for only culture, direct spoligotyping from lymph node TB was unable to be compared. Phenotypic drug susceptibility testing was unable to be performed due to logistic limitations. This study provides the population structure of MTBc lineages and families circulating in Northwest Ethiopia, with limited drug resistance profiles, to inform clinicians and public health professionals.

      Conclusion

      This study demonstrated that the TB epidemic in Northwest Ethiopia is caused by a wide diversity of MTBc strains belonging to lineages L1–L4 and L7, with a predominance of L3-CAS1-Delhi, L4-T-family, L4-Haarlem, and L4-T3-ETH families, and low prevalence (2.6%) of Ethiopia-specific L7. The genotypes were similarly distributed between pulmonary and lymph node TB, and all lineages were equally likely to be detectable in culture and be recognized as MTBc by the MPT64 antigen test. Genotypic drug-susceptibility testing revealed relatively low INH and/or RIF resistance levels. Therefore, these findings provide new insights into phylogeographic diversity, with lower rates of L7, yet no overt sign of culture and laboratory diagnosis bias, and contribute information to scientists and public health researchers for understanding the population structure and resistance profiles of MTBc strains.

      Authors' contributions

      Mebrat Ejo, Ermias Diro, Florian Gehre, Leen Rigouts, and Bouke C. de Jong conceived and designed the experiments. Mebrat Ejo, Meseret Kassa, Yilak Girma, and Cecile Uwizeye performed the experiments and molecular works. Tiruzer Mekonnen and Yilkal Abebe performed a clinical diagnosis of TB lymphadenitis and taken fine-needle aspirates. Ermias Diro, Gabriela Torrea, Leen Rigouts, and Bouke C. de Jong supervised the research work. Mebrat Ejo and Bouke C. de Jong analyzed the data. Gabriela Torrea, Leen Rigouts, and Bouke C. de Jong contributed reagents/materials/analysis tools. Mebrat Ejo, Gabriela Torrea, Florian Gehre, Ermias Diro, Leen Rigouts, and Bouke C. de Jong wrote the manuscript. All authors read and approved the final manuscript before submission.

      Competing interests

      The authors declare that they have no competing interests.

      Acknowledgments

      We would like to acknowledge the Belgian Directorate-General for Development (DGD) and the Institute of Tropical Medicine (ITM), Belgium for financial support for this research work. We are grateful to the staff members of the DOTs clinic, TB culture, and Pathology labs at the University of Gondar, Ethiopia, and also the staff in the Mycobacteriology Unit of ITM, Antwerp, Belgium. In addition, we are also very thankful for patients for their cooperation and participation in this study.

      Appendix A. Supplementary data

      The following is Supplementary data to this article:

      References

        • Alene K.A.
        • Viney K.
        • McBryde E.S.
        • Clements A.C.A.
        Spatial patterns of multidrug resistant tuberculosis and relationships to socioeconomic, demographic and household factors in northwest Ethiopia.
        PLoS One. 2017; 12e0171800
        • Berg S.
        • Schelling E.
        • Hailu E.
        • Firdessa R.
        • Gumi B.
        • Erenso G.
        • et al.
        Investigation of the high rates of extrapulmonary tuberculosis in Ethiopia reveals no single driving factor and minimal evidence for zoonotic transmission of Mycobacterium bovis infection.
        BMC Infect Dis. 2015; 15: 112
        • Biadglegne F.
        • Merker M.
        • Sack U.
        • Rodloff A.C.
        • Niemann S.
        Tuberculous lymphadenitis in Ethiopia predominantly caused by strains belonging to the Delhi/CAS lineage and newly identified Ethiopian clades of the mycobacterium tuberculosis complex.
        PLoS One. 2015; 10e0137865
        • Bidovec-Stojkovič U.
        • Seme K.
        • Žolnir-Dovč M.
        • Supply P.
        Prospective genotyping of mycobacterium tuberculosis from fresh clinical samples.
        PLoS One. 2014; 9e109547
        • Cafrune P.I.
        • Possuelo L.G.
        • Ribeiro A.W.
        • Ribeiro M.O.
        • Unis G.
        • Jarczewski C.A.
        • et al.
        Prospective study applying spoligotyping directly to DNA from sputum samples of patients suspected of having tuberculosis.
        Can J Microbiol. 2009; 55: 895-900
        • Chihota V.N.
        • Niehaus A.
        • Streicher E.M.
        • Wang X.
        • Sampson S.L.
        • Mason P.
        • et al.
        Geospatial distribution of Mycobacterium tuberculosis genotypes in Africa.
        PLoS One. 2018; 13e0200632
        • Cohen T.
        • van Helden P.D.
        • Wilson D.
        • Colijn C.
        • McLaughlin M.M.
        • Abubakar I.
        • et al.
        Mixed-strain Mycobacterium tuberculosis infections and the implications for tuberculosis treatment and control.
        Clin Microbiol Rev. 2012; 25: 708-719
        • Comas I.
        • Coscolla M.
        • Luo T.
        • Borrell S.
        • Holt K.E.
        • Kato-Maeda M.
        • et al.
        Out-of-Africa migration and Neolithic coexpansion of Mycobacterium tuberculosis with modern humans.
        Nat Genet. 2013; 45: 1176-1182
        • Comas I.
        • Hailu E.
        • Kiros T.
        • Bekele S.
        • Mekonnen W.
        • Gumi B.
        • et al.
        Population Genomics of Mycobacterium tuberculosis in Ethiopia Contradicts the Virgin Soil Hypothesis for Human Tuberculosis in Sub-Saharan Africa.
        Curr Biol. 2015; 25: 3260-3266
        • Couvin D.
        • David A.
        • Zozio T.
        • Rastogi N.
        Macro-geographical specificities of the prevailing tuberculosis epidemic as seen through SITVIT2, an updated version of the Mycobacterium tuberculosis genotyping database.
        Infect Genet Evol. 2019; 72: 31-43
        • Couvin D.
        • Reynaud Y.
        • Rastogi N.
        Two tales: Worldwide distribution of Central Asian (CAS) versus ancestral East-African Indian (EAI) lineages of Mycobacterium tuberculosis underlines a remarkable cleavage for phylogeographical, epidemiological and demographical characteristics.
        PLoS One. 2019; 14e0219706
        • Damena D.
        • Tolosa S.
        • Hailemariam M.
        • Zewude A.
        • Worku A.
        • Mekonnen B.
        • et al.
        Genetic diversity and drug susceptibility profiles of Mycobacterium tuberculosis obtained from Saint Peter’s TB specialized Hospital, Ethiopia.
        PLoS One. 2019; 14e0218545
        • Demay C.
        • Liens B.
        • Burguière T.
        • Hill V.
        • Couvin D.
        • Millet J.
        • et al.
        SITVITWEB - A publicly available international multimarker database for studying Mycobacterium tuberculosis genetic diversity and molecular epidemiology.
        Infect Genet Evol. 2012; 12: 755-766
        • Diriba G.
        • Kebede A.
        • Tola H.H.
        • Alemu A.
        • Tadesse M.
        • Tesfaye E.
        • et al.
        Surveillance of drug resistance tuberculosis based on reference laboratory data in Ethiopia.
        Infect Dis Poverty. 2019; 8: 54
        • Eddyani M.
        • Vandelannoote K.
        • Meehan C.J.
        • Bhuju S.
        • Porter J.L.
        • Aguiar J.
        • et al.
        A Genomic Approach to Resolving Relapse versus Reinfection among Four Cases of Buruli Ulcer.
        PLoS Negl Trop Dis. 2015; 9e0004158
        • Fenn Buderer N.M.
        Statistical methodology: I. Incorporating the prevalence of disease into the sample size calculation for sensitivity and specificity.
        Acad Emerg Med. 1996; 3: 895-900
        • Firdessa R.
        • Berg S.
        • Hailu E.
        • Schelling E.
        • Gumi B.
        • Erenso G.
        • et al.
        Mycobacterial lineages causing pulmonary and extrapulmonary Tuberculosis, Ethiopia.
        Emerg Infect Dis. 2013; 19: 460-463
        • Ford C.B.
        • Shah R.R.
        • Maeda M.K.
        • Gagneux S.
        • Murray M.B.
        • Cohen T.
        • et al.
        Mycobacterium tuberculosis mutation rate estimates from different lineages predict substantial differences in the emergence of drug-resistant tuberculosis.
        Nat Genet. 2013; 45: 784-790
        • Gagneux S.
        Host-pathogen coevolution in human tuberculosis.
        Philos Trans R Soc B Biol Sci. 2012; 367: 850-859
        • Gagneux S.
        • DeRiemer K.
        • Van T.
        • Kato-Maeda M.
        • de Jong B.C.
        • Narayanan S.
        • et al.
        Variable host-pathogen compatibility in Mycobacterium tuberculosis.
        Proc Natl Acad Sci. 2006; 103: 2869-2873
        • Gagneux S.
        • Small P.M.
        Global phylogeography of Mycobacterium tuberculosis and implications for tuberculosis product development.
        Lancet Infect Dis. 2007; 7: 328-337
        • Global Laboratory Initiative
        Mycobacteriology laboratory manual, Global Laboratory Initiative.
        World Health Organization, Geneva, Switzerland2014: 1-146
        • Guttierez C.
        • Brisse S.
        • Brosch R.
        • Fabre M.
        • Omaïs B.
        • Marmiesse M.
        • et al.
        Ancient origin and gene mosaicism of the progenitor of Mycobacterium tuberculosis.
        PLoS Pathog. 2005; 1: e5
        • Hadifar S.
        • Shamkhali L.
        • Kargarpour Kamakoli M.
        • Mostafaei S.
        • Khanipour S.
        • Mansoori N.
        • et al.
        Genetic diversity of Mycobacterium tuberculosis isolates causing pulmonary and extrapulmonary tuberculosis in the capital of Iran.
        Mol Phylogenet Evol. 2019; 132: 46-52
        • Hain Lifescience
        GenoType MTBDRplus products.
        Mycobacteria Diagnostics: Hain Lifesciences. GmbH, Nehren, Germany2013
        • de Jong B.C.
        • Hill P.C.
        • Aiken A.
        • Awine T.
        • Antonio M.
        • Adetifa I.M.
        • et al.
        Progression to active tuberculosis, but not transmission, varies by Mycobacterium tuberculosis lineage in The Gambia.
        J Infect Dis. 2008; 198: 1037-1043
        • Kamerbeek J.
        • Schouls L.
        • Kolk A.
        • van Agterveld M.
        • van Soolingen D.
        • Kuijper S.
        • et al.
        Simultaneous detection and strain differentiation of Mycobacterium tuberculosis for diagnosis and epidemiology.
        J Clin Microbiol. 1997; 35: 907-914
        • Kargarpour Kamakoli M.
        • Khanipour S.
        • Hadifar S.
        • Ghajavand H.
        • Farmanfarmaei G.
        • Fateh A.
        • et al.
        Challenge in direct Spoligotyping of Mycobacterium tuberculosis: a problematic issue in the region with high prevalence of polyclonal infections.
        BMC Res Notes. 2018; 11: 486
        • Kato-Maeda M.
        • Metcalfe J.Z.
        • Flores L.
        Genotyping of Mycobacterium tuberculosis: application in epidemiologic studies.
        Future Microbiol. 2011; 6: 203-216
        • Kebede A.H.
        • Alebachew Wagaw Z.
        • Tsegaye F.
        • Lemma E.
        • Abebe A.
        • Agonafir M.
        • et al.
        The first population-based national tuberculosis prevalence survey in Ethiopia, 2010-2011.
        Int J Tuberc Lung Dis. 2014; 18: 635-639
        • Mbugi E.V.
        • Katale B.Z.
        • Streicher E.M.
        • Keyyu J.D.
        • Kendall S.L.
        • Dockrell H.M.
        • et al.
        Mapping of Mycobacterium tuberculosis Complex Genetic Diversity Profiles in Tanzania and Other African Countries.
        PLoS One. 2016; 11e0154571
        • Metcalfe J.Z.
        • Streicher E.
        • Theron G.
        • Colman R.E.
        • Penaloza R.
        • Allender C.
        • et al.
        Mycobacterium tuberculosis subculture results in loss of potentially clinically relevant heteroresistance.
        Antimicrob Agents Chemother. 2017; 61: e00888-17
        • Mulisa G.
        • Workneh T.
        • Hordofa N.
        • Suaudi M.
        • Abebe G.
        • Jarso G.
        Multidrug-resistant Mycobacterium tuberculosis and associated risk factors in Oromia Region of Ethiopia.
        Int J Infect Dis. 2015; 39: 57-61
        • N’Dira Sanoussi C.
        • de Jong B.C.
        • Odoun M.
        • Arekpa K.
        • Ligali M.A.
        • Bodi O.
        • et al.
        Low sensitivity of the MPT64 identification test to detect lineage 5 of the Mycobacterium tuberculosis complex.
        J Med Microbiol. 2018; 67: 1718-1727
        • Ngabonziza J.C.S.
        • Loiseau C.
        • Marceau M.
        • Jouet A.
        • Menardo F.
        • Tzfadia O.
        • et al.
        A sister lineage of the Mycobacterium tuberculosis complex discovered in the African Great Lakes region.
        bioRxiv. 2020; 11: 2917
        • Ofori-Anyinam B.
        • Kanuteh F.
        • Agbla S.C.
        • Adetifa I.
        • Okoi C.
        • Dolganov G.
        • et al.
        Impact of the Mycobaterium africanum West Africa 2 Lineage on TB Diagnostics in West Africa.
        PLoS Negl Trop Dis. 2016; 10e0004801
        • Pathak D.
        • Chakravorty S.
        • Hanif M.
        • Tyagi J.S.
        Lysis of tubercle bacilli in fresh and stored sputum specimens: Implications for diagnosing tuberculosis in stored and paucibacillary specimens by PCR.
        BMC Microbiol. 2007; 7: 83
        • Portevin D.
        • Gagneux S.
        • Comas I.
        • Young D.
        Human macrophage responses to clinical isolates from the Mycobacterium tuberculosis complex discriminate between ancient and modern lineages.
        PLoS Pathog. 2011; 7e1001307
        • Sanoussi C.N.
        • Affolabi D.
        • Rigouts L.
        • Anagonou S.
        • de Jong B.
        Genotypic characterization directly applied to sputum improves the detection of Mycobacterium africanum West African 1, under-represented in positive cultures.
        PLoS Negl Trop Dis. 2017; 11e0005900
        • Shamputa I.C.
        • Rigouts L.
        • Eyongeta L.A.
        • El Aila N.A.
        • van Deun A.
        • Salim A.H.
        • et al.
        Genotypic and Phenotypic Heterogeneity among Mycobacterium tuberculosis Isolates from Pulmonary Tuberculosis Patients.
        J Clin Microbiol. 2005; 42: 5528-5536
        • Shibabaw A.
        • Gelaw B.
        • Gebreyes W.
        • Robinson R.
        • Wang S.H.
        • Tessema B.
        The burden of pre-extensively and extensively drug-resistant tuberculosis among MDR-TB patients in the Amhara region, Ethiopia.
        PLoS One. 2020; 15e0229040
        • Shuaib Y.A.
        • Khalil E.A.G.
        • Wieler L.H.
        • Schaible U.E.
        • Bakheit M.A.
        • Mohamed-Noor S.E.
        • et al.
        Mycobacterium tuberculosis Complex Lineage 3 as Causative Agent of Pulmonary Tuberculosis, Eastern Sudan.
        Emerg Infect Dis. 2020; 26: 427-436
        • Srilohasin P.
        • Chaiprasert A.
        • Tokunaga K.
        • Nishida N.
        • Prammananan T.
        • Smittipat N.
        • et al.
        Genetic diversity and dynamic distribution of mycobacterium tuberculosis isolates causing pulmonary and extrapulmonary tuberculosis in Thailand.
        J Clin Microbiol. 2014; 52: 4267-4274
        • Suresh N.
        • Arora J.
        • Pant H.
        • Rana T.
        • Singh U.B.
        Spoligotyping of Mycobacterium tuberculosis DNA from Archival Ziehl-Neelsen-stained sputum smears.
        J Microbiol Methods. 2007; 68: 291-295
        • Tadesse M.
        • Abebe G.
        • Bekele A.
        • Bezabih M.
        • de Rijk P.
        • Meehan C.J.
        • et al.
        The predominance of Ethiopian specific Mycobacterium tuberculosis families and minimal contribution of Mycobacterium bovis in tuberculous lymphadenitis patients in Southwest Ethiopia.
        Infect Genet Evol. 2017; 55: 251-259
        • Tang C.
        • Reyes J.F.
        • Luciani F.
        • Francis A.R.
        Tanaka MM. spolTools: Online utilities for analyzing spoligotypes of the Mycobacterium tuberculosis complex.
        Bioinformatics. 2008; 24: 2414-2415
        • Tanveer M.
        • Hasan Z.
        • Kanji A.
        • Hussain R.
        • Hasan R.
        Reduced TNF-α and IFN-γ responses to Central Asian strain 1 and Beijing isolates of Mycobacterium tuberculosis in comparison with H37Rv strain.
        Trans R Soc Trop Med Hyg. 2009; 103: 581-587
        • Team QD
        QGIS Geographic Information System.
        Open Source Geospatial Foundation Project, 2019
        • Tesfay K.
        • Tesfay S.
        • Nigus E.
        • Gebreyesus A.
        • Gebreegziabiher D.
        • Adane K.
        More than half of presumptive multidrug-resistant cases referred to a tuberculosis referral laboratory in the Tigray region of Ethiopia are multidrug resistant.
        Int J Mycobacteriol. 2016; 5: 324-327
        • Tessema B.
        • Beer J.
        • Merker M.
        • Emmrich F.
        • Sack U.
        • Rodloff A.C.
        • et al.
        Molecular epidemiology and transmission dynamics of Mycobacterium tuberculosis in Northwest Ethiopia: New phylogenetic lineages found in Northwest Ethiopia.
        BMC Infect Dis. 2013; 13: 131
        • Tønjum T.
        • Klintz L.
        • Bergan T.
        • Baann J.
        • Furuberg G.
        • Cristea M.
        • et al.
        Direct detection of Mycobacterium tuberculosis in respiratory samples from patients in Scandinavia by polymerase chain reaction.
        Clin Microbiol Infect. 1996; 2: 127-131
        • Warner D.F.
        • Koch A.
        • Mizrahi V.
        Diversity and disease pathogenesis in mycobacterium tuberculosis.
        Trends Microbiol. 2015; 23: 14-21
        • Williams D.L.
        • Gillis T.P.
        • Dupree W.G.
        Ethanol fixation of sputum sediments for DNA-based detection of Mycobacterium tuberculosis.
        J Clin Microbiol. 1995; 33: 1558-1561
        • World Health Organization
        The use of molecular line probe assay for the detection of resistance to isoniazid and rifampicin: policy update.
        WHO/HTM/TB/2016.12, Geneva, Switzerland2016: 1-56
        • World Health Organization
        Global tuberculosis report 2019.
        WHO/CDS/TB/2019.15, Geneva, Switzerland2019: 1-283
        • Yimer S.A.
        • Namouchi A.
        • Zegeye E.D.
        • Holm-Hansen C.
        • Norheim G.
        • Abebe M.
        • et al.
        Deciphering the recent phylogenetic expansion of the originally deeply rooted Mycobacterium tuberculosis lineage 7.
        BMC Evol Biol. 2016; 16: 146
        • Yimer Sa
        • Norheim G.
        • Namouchi A.
        • Zegeye Ed
        • Kinander W.
        • Tønjum T.
        • et al.
        Mycobacterium tuberculosis lineage 7 strains are associated with prolonged patient delay in seeking treatment for pulmonary tuberculosis in Amhara region, Ethiopia.
        J Clin Microbiol. 2015; 53: 1301-1309