Association of pyrazinamide resistance in Mycobacterium tuberculosis with mutations in rpoB, katG, and inhA genes Authors: Kartiki Srivastava, Salman Khan, Rajniti Prasad, Sada Nand Pandey, Shampa Anupurba Published: 2026 ABSTRACT AbstractMultidrug-resistant tuberculosis (MDR-TB) is associated with a high rate of treatment failure. Since the 9 to 11-month shorter oral regimen in MDR-TB treatment includes pyrazinamide (PZA), it is essential to assess the susceptibility of MDR-TB strains to PZA to guide effective therapy. The susceptibility of PZA is highly affected by drug pressure induced by rifampicin (RIF) and isoniazid (INH) resistant strains. This study examines the association between point mutations in the rpoB, katG, inhA, and pncA genes. Resistant patterns of MDR strains were detected through line probe assay. Phenotypic drug susceptibility testing for PZA was performed using MGIT 960 liquid culture for RIF-resistant, INH-resistant, and multidrug-resistant strains. In addition, pncA gene was sequenced to characterize the associated mutation pattern. Out of 780 MTB-positive strains, 195 were found to be resistant to first-line drugs (RIF and/or INH). Among those 195 resistant isolates, 107 (54.87%) were resistant, and 88 (45.12%) were susceptible to PZA. Mutations at S531L in the rpoB gene (RIF), S315T in the katG gene (INH), and C-15T in the inhA gene (INH) were most common (111, 80, and 31, respectively). The highest resistance to PZA was found when all three gene mutations were present in combination (20/23). Out of 24 PZA-resistant strains sequenced pncA gene mutation was found in 23 strains. Significantly high PZA resistance was detected in first-line resistant isolates, and this was much higher in combination mutations. The presence of these mutations predisposes the strain to acquisition of PZA resistance.Keywords: Mycobacterium tuberculosis, Multidrug resistance, Line probe assay, Rifampicin, Isoniazid, PyrazinamideSubject terms: Diseases, Genetics, Microbiology, Molecular biology Introduction Multidrug-resistant Mycobacterium tuberculosis (MDR-TB, resistant to first-line anti-TB drugs, i.e., isoniazid and rifampicin) remains a critical global public health problem due to its high treatment failure rate1–3. According to the 2019 report from the World Health Organization (WHO), there were about 10 million new cases of TB worldwide, and approximately 1.2 million HIV-negative patients died of tuberculosis4. Pyrazinamide (PZA) is one of the drugs recommended for the treatment of MDR-TB, used in both first-line and second-line treatment regimens5. It is an unconventional and paradoxical drug with a unique ability to eradicate tolerant bacilli that are trapped inside macrophages and cannot be eliminated by other drugs6. It is the only proposed drug that has been demonstrated to enhance the treatment of a persistent infection to date. However, it does not impact growing bacilli. PZA supports a crucial idea that it is essential to combine drugs targeting latent persister bacteria with other drugs targeting developing bacteria to reduce treatment duration7. It is well known that if the infecting strain is resistant to PZA, treatment results are worse in high MDR-TB settings, particularly if fluoroquinolone resistance is also present8,9. It has lately been acknowledged that PZA resistance has spread globally, with an estimated one out of every six incident TB cases exhibiting PZA resistance. An estimated 1.4 million new PZA-resistant TB cases occur annually worldwide, and about 270,000 MDR-TB patients also possess PZA resistance10. In order to successfully treat MDR-TB, it is essential to maintain bacterial susceptibility to PZA11. Several studies have indicated that PZA resistance primarily arises from mutations in the pncA gene, which encodes the enzyme pyrazinamidase11–14. The enzyme pyrazinamidase (PZase) converts PZA into its active form, pyrazinoic acid (POA)14. According to studies that have evaluated PZA resistance, 40–50% of primary MDR-TB isolates have PZA resistance15–17, while more than 90% of XDR-TB isolates have PZA resistance18., A multi-centric study found that PZA resistance is higher in previously treated TB cases than in new TB cases19. This research was conducted in India to examine the phenotypic and genotypic drug resistance patterns against PZA from a variety of first-line drug-resistant MTB strains and the correlation between RIF, INH, and PZA resistance. Determining the relationship between PZA, RIF, and INH resistance patterns would help in designing more effective diagnostic tools and tailored therapy regimens, eventually enabling better TB control and management efforts. Materials and methods Ethical statement The ethical approval for this study was granted by the Institute’s Ethics Committee of the Institute of Medical Sciences (ECR/526/Inst/UP/2014), Banaras Hindu University, Varanasi. Study design This study was conducted at the Intermediate Reference Laboratory, Department of Microbiology, Institute of Medical Sciences, Banaras Hindu University, Varanasi, India, from November 2024 to April 2025. It was a part of the routine diagnostic workflow under NTEP (National TB Elimination Program). Specimen processing, inoculation, and DST were performed in well-equipped BSL II & III facilities under standard biosafety precautions. All the presumptive TB samples that were detected positive for MTB by GeneXpert were included in this study. The positive samples were subjected to Line Probe Assay (GenoType MTBDRplus), whereas phenotypic DST for PZA was performed for the cases that were resistant to either RIF, INH, or both. The samples that were indeterminate in GeneXpert were excluded from this study. (Fig. 1) Diagrammatic workflow of the testing. GeneXpert The MTB/RIF assay was performed according to the manufacturer’s instructions. Briefly, a sample reagent was added at a 2:1 ratio to clinical specimens. The closed specimen container was manually agitated twice for 15 min at room temperature before 2 ml of the inactivated material was transferred to the test cartridge, then the cartridge was securely closed. The loaded cartridge was inserted into the GeneXpert system20. Specimen processing Specimen processing was done by the standard NALC-NaOH method for digestion, decontamination, and concentration. The concentrated sediments were resuspended in about 2 to 3 ml of phosphate buffer (pH 6.8)21. DNA extraction Extraction of DNA was performed by using the GenoLyse kit (Hain Lifescience, Nehren, Germany) according to the manufacturer’s instructions. After decontamination, 1000 µL sample was centrifuged for 15 min at 10,000×g. After discarding the supernatant, the pellet was re-suspended in 100 µL of lysis buffer and vortexed for 30 s. This suspension was incubated at 95 °C in a water bath for 5 min. After that, 100 µL of neutralizing buffer was added, vortexed for 30 s, then centrifuged for 5 min at 13,000×g. For future usage, the DNA-containing supernatant was moved to a separate tube22. GenoType MTBDRplus The extracted DNA was subjected to the GenoType MTBDRplus assay (Hain Lifescience, Nehren, Germany) for the detection of MTBC and resistance to RIF and INH according to the manufacturer’s instructions. Multiplex PCR was carried out using 45µL amplification mix consisting of 10 µl AM-A and 35µL AM-B. The DNA template (5µL) was added to each tube in a different amplification room, and amplification was performed with the final volume of 50 µl using a thermal cycler (Applied Biosystems) and amplification protocol provided by Hain Lifescience. Hybridization was performed using GT Blot-48 (Hain Lifescience, Nehren, Germany). Each strip of GenoType MTBDRplus had 27 reaction zones (bands), which includes 6 controls that are conjugate control, amplification control, M. tuberculosis complex (TUB) control, rpoB, katG and inhA controls, 8 rpoB wild-type (WT1 to WT8) and 4 mutant probes (rpoB MUT D516V, rpoB MUT H526Y, rpoB MUT H526D, and rpoB MUT S531L), 1 katG wild-type and two mutant probes (katG MUT S315T1 and katG MUT S315T2), and 2 inhA wild-type and 4 mutant probes (inhA MUT1 C15T, inhA MUT2 A16G, inhA MUT3A T8C, inhA MUT3B T8A)23(Figs. 2, 3). The absence of a wild-type band or the presence of a mutant band was taken as an indication of a resistant strain. Incomplete amplification of RIF and/ or INH genes or absence of TUB with an evaluable resistance pattern was considered invalid, and the test samples were repeated. Nitrocellulose strips showing band patterns obtained using the GenoType MTBDRplus assay. Strip A: Susceptible to rifampicin (RIF) and isoniazid (INH); Strip B: rifampicin resistant (rpoB S531L mutation); Strip C: isoniazid mono-resistant (katG S315T1 mutation); Strip D: (isoniazid mono-resistant inhA C-15T mutation); Strip E: MDR- TB (rpoB S531L mutation, katG S315T1, and inhA C-15T mutation). PZA sensitivity across various mutation patterns. Drug susceptibility testing The decontaminated and concentrated sediments were inoculated into mycobacterium growth indicator tubes (MGIT) and inserted into the automated liquid culture system BACTEC MGIT 960 (Becton Dickinson, USA). Simultaneously, a loopful of liquid culture was inoculated in the brain heart infusion (BHI) agar and incubated for 24 h to check the sterility of the positive culture tube. To confirm MTBC, an immunochromatography-based MPT64 rapid antigen detection kit (SD BIOLINE, South Korea) was used24,25. The confirmed MTBC isolates were further subjected to DST by the conventional 1% proportion method using the MGIT 960 system according to the NTEP guidelines for PZA (100 µg/ml). The growth control tube was diluted in 1:100 with normal saline and the DST of PZA was set up with a 21-day protocol. Sequencing of pncA gene A total of 24 samples were sent for Sanger sequencing, from which 8 samples had a mutation in the rpoB & katG gene, 8 had a mutation in the katG & inhA gene, and 8 had a mutation in the rpoB, katG & inhA gene. The genomic DNA was isolated with the CTAB-chloroform method. The quality and quantity of DNA were analysed with a spectrophotometer (Thermo Scientific NanoDrop 2000). The primer was taken from the study by Tam et al.26. The pncA gene was amplified with the help of primers, and the product size was confirmed by agarose gel electrophoresis. The genomic DNA was then sent for Sanger sequencing to Barcode Bioscience, Bangalore, India. The obtained pncA gene sequence was aligned with the M. tuberculosis reference H37Rv sequence using Clustal Omega (EMBL-EBI) using default parameters. Mutation positions were identified using the alignment output. Quality control strain and growth control Mycobacterium tuberculosis H37Rv strain was used for quality control (QC) testing in Line Probe Assay and DST. The tubes containing only media, growth supplement, and inoculum were taken as growth control. Statistical analysis Data were entered and organized using Microsoft Excel, and statistical analyses were performed using Jeffreys’s Amazing Statistics Program (JASP) version 0.19.3.0. Descriptive statistics were used to summarize frequencies and proportions of PZA resistance across various mutation patterns. A Chi-square test of independence was conducted using a 5 × 2 contingency table to evaluate the association between specific mutation patterns and PZA resistance. Mutation patterns included were S531L (rpoB), S315T1 (katG), C-15T (inhA), and their combinations. The PZA susceptibility status was taken as the categorically dependent variable. Cramér’s V test was performed to evaluate the strength of association. Results A total of 4990 presumptive TB cases were tested using GeneXpert MTB/RIF assay between November 2024 and April 2025. Of these, 780 were MTB-positive, wherein RIF resistance was detected in 118 isolates, and 641 were sensitive to RIF. The remaining 21 isolates were indeterminate for RIF resistance and were excluded from this study. Upon further testing of 759 MTB-positive samples with the GenoType MTBDRplus assay, 7 were found to be resistant to only RIF, 77 to only INH, and 111 were MDR. There were no hetero-resistant strains. Drug susceptibility testing of these 195 first-line resistant cases using the MGIT 960 system revealed that 107 (54.87%) were resistant and 88 (45.12%) were susceptible to PZA. Upon Sanger sequencing of the pncA gene, 23 out of 24 strains had synonymous and non-synonymous mutations, and one had no mutation in the pncA gene, which showed resistance to INH. Band patterns in Geno Type MTBDRplus assay Among the 195 first-line drug-resistant strains, 7 showed mutations exclusively in the S531L region (MUT3 band) of rpoB. Additionally, one strain had a mutation in the H526D region (MUT2B band) of rpoB along with concurrent mutations in S315T1 (MUT1 band) in katG and C-15T (MUT1 band) in inhA. The katG mutation was dominant (52/77) S315T1 in INH-resistant strains. Most of the MDR-TB and INH mono-resistant strains had mutations in rpoB S531L (111/114) (MUT3 band) and katG S315T1 (80/81) (MUT1 band). In inhA most common mutation was at position C-15T (31/36) (MUT1 band). T-8 C (5/36) (MUT3A band) mutation was also found in the inhA gene. Overall, concurrent rpoB, katG, and inhA mutations were found in 23 MDR-TB strains (Table 1). Table 1Band patterns and mutation profiles identified using the GenoType MDRplus assay.GeneBandGene mutation/RegionRIF Resistance PatternINH resistance PatternMDRTB Resistance Pattern rpoB WT1506–509195195WT2510–513195195WT3513–517195195WT4516–519195195WT5518–522195195WT6521–525195195WT7526–529195195WT8530–53318884MUT1D516V0000MUT2AH526Y0000MUT2BH256D0003MUT3S531L07111 katG WT1315143115MUT1S315T15280MUT2S315T20001 inhA WT1-15/-16170164WT2-8194190MUT1C-15T2531MUT2A-16G0000MUT3AT-8 C0105MUT3BT-8 A0000 Band patterns and mutation profiles identified using the GenoType MDRplus assay. Drug susceptibility testing and resistance pattern in PZA All 195 isolates were confirmed to be MTBC by using MPT64 rapid antigen detection kits after growth in MGIT960. Drug susceptibility testing of PZA revealed that 107 isolates (54.87%) were resistant, while 88 isolates (45.12%) were sensitive. Among the PZA-resistant isolates, resistance-associated mutations were detected in 7 isolates (3.58%) in the rpoB gene, 52 isolates (26.66%) in katG, 25 isolates (12.82%) in inhA, 68 isolates (34.87%) in rpoB + katG, 20 isolates (10.28%) in rpoB + inhA, and 23 (11.79%) isolates showed mutations in all three genes. Association between mutation regions of rpoB, inhA, katG, and PZA resistance Among the isolates with a single mutation at rpoB S531L, katG S315T1, or inhA C-15T, the phenotypic PZA resistance rate was 28.6% (2/7), 23.2% (13/56), and 13.8% (4/29), respectively. When the katG S315T1 mutation co-occurred with the rpoB S531L mutation, 85% (68/80) of the strains showed PZA resistance. Of the 23 strains with concurrent mutations at rpoB S531L, katG S315T1, and inhA C-15T, 87% (20/23) were resistant to PZA (Table 2 and Graph). The chi-square test revealed a statistically significant association between mutation patterns and PZA resistance (X2= 83.27, df = 4, p < 0.001), indicating that the distribution of the PZA resistance was not independent of the mutation type. Cramér’s V test yielded a value of 0.653, indicating a strong effect size. Table 2Association of RIF and INH mutation regions with PZA susceptibility. Chi-square = 83.27, p < 0.001 (significant association across mutation patterns).Mutation regionPZAResistantSensitiveS531L (7) (rpoB)0205S315T1 (52) (katG)1139C-15T (25) (inhA)0421S531L+S315T1 (68) (rpoB+katG)5810S531L + C-15T (20) (rpoB+inhA)1010S531L+S315T1 + C-15T (23) (rpoB+katG+inhA)2003 Association of RIF and INH mutation regions with PZA susceptibility. Chi-square = 83.27, p < 0.001 (significant association across mutation patterns). Discussion Tuberculosis cases infected with PZA-resistant MTBC strains have been associated with a higher risk of treatment failure compared to those with PZA-sensitive strains17, possibly due to increased acquisition of resistance mechanisms. The present study found that 54.87% of isolates resistant to RIF and/or INH were also resistant to PZA. This is similar to a study in Ningbo, China, by Che et al., wherein 59.1% of such Isolates were resistant to PZA26. A study conducted in Thailand reported PZA resistance rate of 49.0% among MDR-TB cases, which is slightly lower than the rate observed in the present study27. The findings are in agreement with a previous study that assessed twelve Sub-Saharan African countries and found that PZA resistance rate among RIF-resistant TB cases varied significantly by country, with Burkina Faso having 80% and Togo having 21.0%28. Furthermore, studies from China and the United States show that PZA resistance was 38.5%29 and 38.0%30 in MDR-TB patients, respectively. This result confirms earlier findings that PZA resistance was low in pan-susceptible isolates but high in MDR-TB isolates31,32. PZA resistance may be acquired either by transfer of a PZA-resistant MTB strain or by antibiotic selection pressure. This could be influenced by prior treatment history, as prolonged anti-tubercular therapy in existing MDR isolates may exert drug pressure, leading to potential cross-resistance among PZA and other drugs. Consequently, in RIF-resistant cases, especially those exposed to prior treatment, this should raise clinical suspicion for concomitant PZA resistance10. In this study, we found the most common mutations at S531L in the rpoB gene, S315T1 in the katG gene, and C-15T in the inhA gene. Mutations were detected in 52 isolates (26.66%) in katG, 25 isolates (12.82%) in inhA, 7 isolates (3.58%) in the rpoB gene. Combined mutation was seen in 68 isolates (34.87%) in rpoB + katG, 20 isolates (10.28%) in rpoB + inhA, and 23 (11.79%) isolates showed mutations in all three genes. Similar mutation patterns were observed in a study from South India, where the combined mutation in rpoB + katG (50.98%) had the highest prevalence, followed by rpoB + inhA (34.96%). The katG (25.17%) gene had the highest percentage among single gene mutations, followed by inhA (9.31%) and rpoB (8.20%). However, they did not report mutations in all three genes33. A study in Southeast Mexico found that the rpoB S531L + katG S315T combination occurred in approximately 29% of MDR isolates, highlighting the frequent co-emergence of both mutations in high-burden locations34. Similarly, in Thailand, researchers discovered that the same three hotspots, katG S315T (70%) and rpoB S531L (81%), predominated among MDR-TB isolates35. In this study single mutations at rpoB S531L, katG S315T1, and inhA C-15T were linked to low rates of phenotypic PZA resistance (28.6%, 23.2%, and 13.8%, respectively). However, the co-occurrence of katG S315T1 and rpoB S531L increased the PZA resistance rate to 85%, whereas strains with all three mutations had the greatest resistance rate (87%).Our findings on increased PZA resistance in isolates bearing combined mutations in RIF and INH are consistent with previous research, underscoring the risks presented by cumulative resistance-conferring genotypes. The isolates which were sequenced in our study showed mutations across ten different positions (3,6,9,12,15,27,30,45,51, and 60). The non-synonymous changes, such as Gly3Ser, Gly12Arg, Ser45Tyr, Gly51Ser, and Ile60Ser, were most likely responsible for the alteration of the pyrazinamidase enzyme, and even more disruptive was a stop mutation at position 27 that cut the protein short, likely killing its enzymatic activity. Some changes at positions 6,9,15, and 30 were synonymous silent mutations that did not alter the protein sequence but can potentially alter methylation patterns that can affect protein function or expression (Table 3). The recurrence of Gly3Ser and Gly12Arg was across multiple isolates, while some of the isolates had unique mutations, highlighting the genetic heterogeneity. Table 3Summary of sequence variations detected in the pncA gene.PositionMutationCodon changeAA changeMutation type3G → AGGC → AGCGly → SerNon-synonymous6C → TGGC → GGTGly → GlySynonymous9T → CATT → ATCIle → IleSynonymous12G → AGGA → AGAGly → ArgNon-synonymous15C → TGAC → GATAsp → AspSynonymous27G → TGGA → TGAGly → STOPStop30A → GGCA → GCGAla → AlaSynonymous45C → ATCC → TACSer → TyrNon-synonymous51G → AGGT → AGTGly → SerNon-synonymous60T → GATT → AGTIle → SerNon-synonymous Summary of sequence variations detected in the pncA gene. Although mutations in pncA, rpoB, and katG/inhA are frequently observed together in the clinical isolates, these genes do not share genetic or biochemical pathways. All the genes contribute to drug resistance with independent processes. Despite this independence, a strong correlation in mutation pattern is commonly observed, the possible reason could be selective pressure due to combination therapy which include all three drugs. Mtb mutates in multiple drug target genes as a survival strategy. Additionally physiological factors such as stress response pathway and efflux pump activity may contribute to correlation in mutation. One phenotypically resistant strain did not show any mutation in pncA. The probable reason for this could be either a mutation in some other gene, rather than pncA, or overexpression of an efflux pump. There are some limitations to this study; we employed sequencing in the pncA gene for only 24 isolates out of 107, and no PZA susceptible strains due to lack of funding resources. This may not be fully representative of the strains, resulting in bias in the observed mutation frequencies. Additionally, the sequencing was limited to the pncA gene, which may overlook other genetic mutations such as rpsA, panD, and clpC1 genes, which have been reported in other studies36. Future research could benefit from whole-genome sequencing to capture additional resistance. Despite these drawbacks, this is the first study in the field that examines the correlation of point mutation in first-line drug resistance to PZA resistance. This highlights the importance of comprehensive molecular profiling in MDR-TB cases, not only for RIF and INH resistance but also as a predictive marker for PZA resistance, which is crucial for optimizing effective treatment regimens, thereby improving TB management planning. Conclusion Based on our findings, PZA resistance was prevalent in RIF and INH-resistant isolates. This suggests a strong association between various point mutation patterns in rpoB, katG, inhA, and pncA genes; these mutations could be prioritized for targeted pyrazinamide susceptibility testing, enabling early modification of the patient treatment regimen if necessary. Acknowledgements We thank the National TB Elimination Program (NTEP), Ministry of Health and Family Welfare, Govt of India, for their valuable support. Author contributions K.S. performed the methodology and wrote the main manuscript, S.K reviewed and gave suggestions, R.P and S.N.P gave inputs during analysis, S.A. conceptualized, supervised, reviewed, and edited. Data availability The sequencing data have been deposited in the GenBank repository under the accession numbers PX985174, PX985175, PX985176, PX985177, PX985178, PX985179, PX985180, PX985181, PX985182, PX985183, PX985184, PX985185, PX985186, PX985187, PX985188, PX985189, PX985190, PX985191, PX985192, PX985193, PX985194, PX985195, PX985196, PX985197. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The sequencing data have been deposited in the GenBank repository under the accession numbers PX985174, PX985175, PX985176, PX985177, PX985178, PX985179, PX985180, PX985181, PX985182, PX985183, PX985184, PX985185, PX985186, PX985187, PX985188, PX985189, PX985190, PX985191, PX985192, PX985193, PX985194, PX985195, PX985196, PX985197.