The isolates came from a nationally representative survey covering 40 districts and cities across 12 provinces, with 70 healthcare facilities contributing smear-positive tuberculosis cases. Samples were initially screened with the Xpert MTB/RIF assay, and phenotypic drug susceptibility testing was carried out on the BACTEC MGIT 960 system for all rifampicin-resistant specimens plus a random 10 percent of rifampicin-susceptible ones. For sequencing, the team selected all 160 rifampicin-resistant isolates that yielded cultures, together with rifampicin-susceptible isolates from retreatment cases and a 35 percent sample of new cases, producing 1,598 eligible specimens. After excluding cultures that failed, samples with insufficient DNA, and 148 specimens that missed quality-control thresholds of at least 30-fold sequencing depth and 90 percent genome coverage, 1,281 isolates, or 80.2 percent of those attempted, produced analyzable whole-genome data.

The sequencing itself was performed on the Illumina MiSeq platform, generating paired-end reads of 2 by 300 base pairs from libraries prepared with the Nextera XT kit. Genomic DNA was extracted using the CTAB protocol, and bioinformatic processing relied on the MTBseq pipeline: reads were trimmed with Trimmomatic, aligned to the H37Rv reference genome with the BWA-MEM algorithm, and variants were called with Samtools mpileup under stringent quality thresholds. Notably, the pipeline was tuned to detect low-frequency variants, capturing mutations supported by as few as two reads at a minimum allele frequency of 5 percent, a design choice intended to flag resistant subpopulations within apparently mixed infections.

The lineage results reveal a striking geographic mosaic. Four of the seven globally recognized Mycobacterium tuberculosis lineages were detected. Lineage 4, the Euro-American lineage, dominated at 43.1 percent, followed by Lineage 2 at 30.4 percent and the ancient Lineage 1 at 26.3 percent. Lineage 3 was vanishingly rare, with just two isolates, yet its detection is itself a first: this is the first report of Lineage 3 strains sequenced in Indonesia, found in two unrelated patients on islands roughly 1,400 kilometers apart, in Central Java and South Sulawesi. Both Lineage 3 isolates remained fully susceptible to first- and second-line drugs, but their appearance hints at a lineage possibly entering the country through international travel in a nation that serves as a major maritime hub.

At the sub-lineage level, the most prevalent were L2.2.1 and its variants at 18.5 percent, L1.2.1 at 11.3 percent, and L4.1.2.1 at 9.7 percent. The regional contrasts are telling. In Sumatra, the East African-Indian lineage predominates, a pattern the authors attribute to centuries of trade and population movement across the Bay of Bengal and the Strait of Malacca. Java-Bali, the most densely populated part of the country, shows far greater heterogeneity, with East African-Indian, Beijing, and Euro-American strains including Haarlem, LAM, X-type, and T-type co-circulating, consistent with intense internal migration and international connectivity. Eastern Indonesia skews toward Euro-American strains, a distribution the authors link to historical Dutch colonial-era trade networks and the region’s position along old spice routes. This west-to-east gradient suggests that multiple independent transmission networks coexist within a single national setting.

On the resistance front, first-line drug resistance was the most frequent finding, and the molecular culprits were the classic ones. Isoniazid resistance was driven chiefly by the katG p.Ser315Thr mutation, observed in 7.5 percent of isolates, while rifampicin resistance was mediated predominantly by rpoB mutations in codons 426 through 452, most notably p.Ser450Leu at 5 percent. The rpoB gene encodes the beta-subunit of bacterial RNA polymerase, and mutations there impose fitness costs on the bacterium. Tellingly, the team also detected two compensatory mutations in rpoC, which encodes the beta-prime subunit: p.Ile885Val and p.Leu527Val. These substitutions sit at the interface between the two subunits and are thought to restore RNA polymerase activity and bacterial fitness, a well-established signature of clonal adaptation that can help resistant strains persist and transmit.

Weighted to reflect the survey’s stratified cluster sampling design, the resistance proportions showed a familiar but concerning pattern. Isoniazid resistance reached 9.51 percent among new cases and 17.84 percent among retreated cases, while rifampicin resistance stood at 7.90 percent and 16.70 percent respectively. The consistently higher resistance among retreated cases replicates the global finding that prior treatment history is the strongest risk factor for drug resistance, reflecting selective pressure from incomplete adherence or inadequate regimens. Resistance to ethambutol, by contrast, was nearly identical between new and retreated cases, at 5.06 and 5.32 percent, which the authors suggest may indicate primary transmission of resistant strains in the community rather than treatment-acquired resistance.

Among second-line agents, resistance was highest for ethionamide and prothionamide, each at 3.9 percent of isolates, followed by the fluoroquinolone class at 2.03 percent, with 26 resistant specimens, mostly carrying gyrA mutations, the most common being p.Asp94Ala. Resistance to the aminoglycosides amikacin and kanamycin was detected in 20 specimens, or 1.56 percent, mostly through the rrs r.1401a>g mutation. The genuinely reassuring news concerns the newest drugs: no resistance mutations were found for bedaquiline, delamanid, or para-aminosalicylic acid anywhere in the 1,281 isolates. Bedaquiline, introduced by WHO in 2013 for multidrug-resistant tuberculosis, inhibits the bacterium’s ATP synthase, while delamanid blocks mycolic acid cell-wall synthesis, and both remain critical Group A components of modern MDR-TB regimens.

Yet the study’s most consequential finding is the identification of four pre-extensively drug-resistant, or pre-XDR, strains among the survey isolates, designated Mtb_R475, Mtb_R737, Mtb_R3618, and Mtb_R3619. Under WHO definitions, pre-XDR tuberculosis is MDR or rifampicin-resistant tuberculosis that is also resistant to any fluoroquinolone. These four strains, representing about 1.1 percent of the analyzed sample, were confirmed by both MGIT 960 phenotypic testing and Illumina sequencing, and were found across two islands, Java-Bali and Sumatra, in both new and retreated patients. All harbored critical gyrA mutations at codon 94, either Asp94Ala or Asp94Gly, alongside rpoB mutations including p.Asp435Tyr, p.Asn437Asp, p.Ser450Leu, and p.Ser450Trp. Mechanistically, the Asp94 substitutions replace a negatively charged aspartic acid that normally helps magnesium-fluoroquinolone complexes dock into the DNA gyrase enzyme; swapping in neutral alanine or glycine blocks that binding and confers resistance. Notably, the Asp94Gly mutation can extend resistance to fourth-generation fluoroquinolones such as moxifloxacin, not just levofloxacin.

Because fluoroquinolones are a core component of current MDR-TB regimens, resistance to this class substantially narrows treatment options and raises the risk of progression to fully XDR tuberculosis. The authors argue that this low but real prevalence should be read as an early warning rather than reassurance, particularly given that three of the four pre-XDR patients had never smoked and one was in the 15-to-24-year age group, suggesting transmission is not confined to classical risk groups and may already be occurring in the community. The study has limitations the authors acknowledge: budget constraints meant whole-genome sequencing was not performed for all rifampicin-susceptible participants, and phenotypic susceptibility testing was not done for everyone. Even so, the work represents the largest genomic collection of drug-susceptible and drug-resistant strains from Indonesia and the first time sequencing results have been folded into a national drug resistance survey. The team’s conclusion is direct: integrating whole-genome sequencing into Indonesia’s national tuberculosis control program is now essential to track emerging lineages, catch resistance before it spreads, and keep newer drugs effective for as long as possible.

Subject of Research: Whole-genome sequencing of Mycobacterium tuberculosis isolates from Indonesia's 2017-2018 national drug resistance survey to characterize lineage diversity and drug resistance mutations

Article Title: Genomic profiling and drug resistance patterns of Mycobacterium tuberculosis in Indonesia: whole-genome sequencing of isolates from the 2017–2018 national Tuberculosis Drug Resistance Survey

Article References: Rukminiati, Y., Pangaribuan, L., Adam, K., Muna, F., Perwitasari, D., Tobing, K. L., Wibowo, H., Rahayu, S., Farid, M. N., Christian, M. R. L., Marbun, J. M. M., Cabibbe, A. M., Ghodousi, A., Sudarmono, P. P., Mesak, F., & Lolong, D. B. (2026). Genomic profiling and drug resistance patterns of Mycobacterium tuberculosis in Indonesia: whole-genome sequencing of isolates from the 2017–2018 national Tuberculosis Drug Resistance Survey. Frontiers in Public Health, 14, 1844150. https://doi.org/10.3389/fpubh.2026.1844150

Image Credits: AI Generated

DOI: 10.3389/fpubh.2026.1844150

Keywords: tuberculosis, whole-genome sequencing, Indonesia, drug resistance, pre-XDR-TB, Mycobacterium tuberculosis, lineage 3, fluoroquinolone resistance, MDR-TB, public health surveillance, katG, rpoB