Analysis of antimicrobial resistance patterns and genetic mutations in Helicobacter pylori from West Bengal, India depicting escalating clarithromycin and high levofloxacin resistance

Patients seeking diagnosis at Manipal Hospitals Broadway, Kolkata, with gastroduodenal diseases like DU, PU, functional dyspepsia, and GC, during years 2018–2020 were included in this study.

They were from both sex groups; aged between 14–72 yrs. Last date of sample collection was 18^th March, 2020, before COVID-19 cases were reported from this study population. With informed consent, two tissues each from antrum and corpus were collected. Rapid urease test (RUT) was done with one tissue, another part stored for culturing, in ice cold transport media [Glycerol(15%) + Brucella broth] and stored at – 80 ℃, within 2–3 hrs of collection. This study has been approved by ICMR-NICED ethical committee. Identity of patients were not disclosed during experimental procedures.

The thawed tissue samples were vortexed for 15 mins following which 100–150 µl of transport media was plated onto Brain Heart Infusion Agar (BHIA) + charcoal (0.2%) agar plates and incubated in microaerobic condition for 3–7 days. The water droplet like transparent H. pylori colonies were identified and further confirmed via urease, catalase and oxidase reactions. Each time H. pylori reference strain 26695 was used as an experimental positive control. From each biopsy specimen, 10 single and one pooled colonies were picked, and sub-cultured on alternate days onto fresh BHIA plates. One loopful of bacterial culture was then stored at – 80 ℃ in BHI broth + glycerol (20%), as mentioned earlier [2].

Loopful bacterial culture was harvested and DNA was isolated using CTAB (Cetyltrimethylammonium bromide)—method with phenol–chloroform as mentioned previously [19] and stored at – 20 ℃.

PCR amplification of ureB using gene-specific primers was performed for further confirmation of H. pylori infection. Simultaneously, PCR amplification of cagA and vacA (s and m alleles) were carried out using gene-specific primers (Table 1) to characterize the genotypes of the infecting strains. This also provided insight into the possibility of multiple strain infections within a single host [2, 20].

Additionally, strains which tested negative for cagA by PCR were further screened using cag-empty site PCR with specific primers (Luni1 and cagAR5280) to confirm the absence of the entire cag pathogenicity island (cagPAI). The primers as mentioned in Table 1. PCR was done in 20µl reaction volume, with respective conditions for each gene [2, 18] in a Master Cycler apparatus (Eppendorf, Hamburg, Germany). The amplified products were analyzed by 2% agarose gel electrophoresis against molecular markers (NEB) and imaging with analysis was done using Bio-Rad GelDoc Go.

All the antibiotics used in this study (Supplementary file 1) were purchased from Sigma Chemicals, USA. Minimum inhibitory concentrations (MICs) for all antibiotics were determined by Agar-dilution method. All the dosages of antibiotics based upon which this study was conducted were in accordance with previous study [17] and EUCAST guidelines [20]. Isolates were considered to be resistant when the MIC was > 8 µg/ml for metronidazole, > 1 µg/ml for tetracycline, > 0.5 µg/ml for clarithromycin, > 1 µg/mlfor levofloxacin and > 0.125 µg/ml for amoxicillin as per EUCAST guidelines [21]. Whereas for furazolidone isolates were considered to be resistant when the MIC was > 2 µg/ml [17]. BHIA plates containing different concentrations of each antibiotic were prepared by adding the appropriate amount of the drug from their respective stock solutions. Frozen H. pylori cultures were streaked on a antibiotic—free BHIA plate and incubated under microaerobic conditions at 37 ℃ for 2 days. Then bacterial growth was spread on fresh antibiotic—free BHIA plate and incubated for 1 day. The resulting young exponentially growing cells were suspended in Phosphate Buffered Saline (PBS); serial 10- fold dilutions of these suspensions were prepared and 10 µl of each dilution was spotted on freshly prepared antibiotic supplemented as well as on drug free plate used as a control. All plates were incubated for 72 hrs. A strain was considered to be susceptible to concentrations of drugs that decreased its efficiency of colony formation at least 10 folds for metronidazole [24] and 1000-folds for rest of the tested antibiotics, as compared with the control plate [18].

All the clarithromycin resistant strains and representative sensitive strains were PCR-amplified using primers 23SrRNA F/R generating the 617 bp product as done previously, [19] with few modifications. This purified PCR product was subjected to restriction enzyme digestion as per manufacturer’s protocol with BsaI-HFv2 and BbsI (NEB). Digested products were electrophoresed in 2.5–3% agarose gel and analysed as mentioned before.

All the PCRs (50 µl reaction volume) were performed using TaKaRa Taq polymerase with respective primers as mentioned in Table 1.

gyrA F/R were used for amplifying the 554 bp Quinolone Resistance Determining Region (QRDR) segment of gyrA (GenBank accession no. CP003904.1; locus tag: C694_03615) to detect point mutations responsible for levofloxacin resistance [22].

rdxA (oxygen-insensitive NAD(P)H nitroreductase gene GenBank accession no. AE000511.1:complement 1013553.1014185; locus tag: HP_0954) and frxA, were PCR amplified for metronidazole-resistant strains.

Amplified PCR product were then purified, using purification kit (GeneJET PCR purification Kit, ThermoFisher Scientific) and sequenced as mentioned earlier [1].

The sequenced data was assembled for each gene from respective forward and reverse sequence reads using FinchTV (Geospiza, Inc.) and were compared against reference strain 26695 (GenBank accession: CP003904.1), based on which all the primers were designed. DNA sequences were aligned with strain 26695 using the ClustalW Multiple Sequence Alignment tool to identify single nucleotide mutations. Corresponding protein sequences were generated using the ExPASy Translate Tool and the derived protein sequences were further analyzed using NCBI protein BLAST (BLASTp) to confirm protein identity and to access any amino acid changes resulting from non-synonymous mutations.

The correlation between the respective mutation and its association in development of antibiotic resistance were statistically justified using Cohen’s Kappa co-efficient analysis through the GraphPad (quantify interrater agreement with Kappa) online tool.

The observed discordance between RUT and culture results may be attributed to the patchy distribution of H. pylori within the gastric mucosa. It is likely that the biopsy sample used for RUT contained an insufficient bacterial load to produce a detectable urease reaction, whereas the parallel specimen used for culture harbored a higher concentration of viable organisms, resulting in successful isolation. Another possible explanation involves the frequent use of over-the-counter proton pump inhibitors (PPIs) by patients experiencing gastroduodenal symptoms before clinical consultation. PPI use is known to reduce the sensitivity of urease-based diagnostics [13]. Consequently, biopsy specimens obtained during endoscopy from such patients may yield false-negative RUT results, even though cultures from the same patients may show positivity.

Assessment of multiple strain infections within individual hosts was conducted using a combination of virulence gene profiling (cagA presence/absence and vacA allelic subtypes: m1/m2, s1/s2) and antimicrobial susceptibility testing. Out of 210, 71 cases were found to have H. pylori infection with single strain. One of the patients, exhibited infection with 3 distinct strains by virtue of having varying antibiotic susceptibility profile and genetic background from its antrum and corpus biopsy tissues. Dual infections were identified in 7 cases yielding 14 strains. Among these 7, 4 cases showed clear evidence of co-infection with genetically distinct H. pylori strains, as indicated by differences in both virulence gene profiles and antibiotic susceptibility patterns. While the remaining 3 cases exhibited heterogeneity solely in their antibiotic resistance profiles, suggestive of mixed-strain colonization. In total, 88 (71 + 3 + 14 = 88) H. pylori strains were isolated from the 79 positive biopsy samples and included in this study. Of 88 strains, 8 strains (9%) were susceptible to all the antibiotics tested. A comprehensive summary of the findings is provided in Table 2 and illustrated in Figure 1.

Bild vergrößern

All H. pylori isolates (n = 88) included in the present study were found to be sensitive to both tetracycline and furazolidone. Among the tetracycline-sensitive strains, a majority (83 out of 88; 94.3%) exhibited a MIC range between 0.125 µg/ml and ≤ 0.5 µg/ml. The remaining 5 strains (5.7%) were also sensitive, with MIC values ranging from 0.5 µg/ml to ≤ 1.0 µg/ml. Similarly, out of the 88 furazolidone-sensitive strains, 71 isolates (80.7%) demonstrated MIC values in the range of 0.0625 µg/ml to ≤ 0.125 µg/ml. The remaining 17 strains (19.3%) had MIC values between 0.125 µg/ml and ≤ 0.25 µg/ml, as illustrated in Fig 1.

A total of 9 out of 88 isolates (10.2%) were identified as multidrug-resistant (MDR), defined as resistant to at least three classes of antibiotics. Among these, 8 strains exhibited simultaneous resistance to metronidazole, clarithromycin, and levofloxacin, while a single strain showed combined resistance to metronidazole, amoxicillin, and levofloxacin. In addition to the MDR isolates, dual-drug resistance was also observed in several strains. The most common dual resistance pattern involved metronidazole and levofloxacin, identified in 29 out of 88 isolates (33%). Furthermore, dual resistance to clarithromycin and levofloxacin was detected in 6 isolates (6.8%). Only one isolate showed dual resistance to both metronidazole and clarithromycin. A detailed summary of the resistance profiles is provided in Table 2

As shown in Table 4, the distribution of cagA and vacA genotypes among the 88 H. pylori isolates demonstrated that four genotype combinations were represented across the study population, irrespective of their antibiotic susceptibility profiles. The combinations included cagA-positive and -negative strains along with varying vacA allelic types (s1m1, s1m2, s2m2). This uniform distribution suggests no exclusive association between any specific cagA/vacA genotype and overall antibiotic resistance. However, a noteworthy observation was made within the subset of strains harboring the cagA-positive/vacA s1m1 genotype. Out of the 50 isolates belonging to this group, 20 strains (40%) exhibited dual resistance to both metronidazole and levofloxacin. Despite this, statistical analysis did not indicate a significant association between this genotype and dual-drug resistance, and no specific cagA/vacA combination was consistently linked to resistance against any single antibiotic.

In our study, resistance rate to clarithromycin was observed in 19.3% of the H. pylori isolates, posing a significant threat. This figure not only exceeds the empirical resistance threshold [12] but also marks a stark rise from the 0% resistance reported earlier in Kolkata [17]. A similar upward trend has been documented in other regions such as Korea, where resistance rate increased from 37% in 2012 [29] to a troubling 71.1% by 2019 [30]. Across India, clarithromycin resistance is increasingly prevalent, with reported rates in Karnataka (20.4%), Odisha (34.4%), Tamil Nadu (8.7%), North East India (6.5%), and Delhi (11.8%) [10, 31‐34]. Comparable data from neighboring countries reflect this growing resistance: Myanmar (7.7%), Bangladesh (39.3%), Nepal (21.4%) and Pakistan (5.4%) [35‐37]. Globally, resistance rates are similarly high: China reports 32.4% [38], Africa, 29.2%, Europe, 34%; and the Americas, 14% resistance [13, 39]. These figures collectively validate the World Health Organization’s classification of H. pylori as a high-priority pathogen.

Our study identified the A2143G mutation in the 23SrRNA gene as the predominant genetic determinant of clarithromycin resistance, found in 15 out of 17 (88%) resistant strains. This finding aligns with previous data from China [40]. This strong correlation between clarithromycin resistance and the A2143G mutation was confirmed using a combination of novel PCR assays (including PCR-RFLP), and Sanger sequencing, all showing 100% concordance. The remaining 2 (12%) resistant strains did not show any positive band in this PCR which implies resistance in these strains may be mediated by alternative mutations in the 23SrRNA gene or through other, as-yet-uncharacterized mechanisms, highlighting the complexity and evolving nature of antibiotic resistance in H. pylori. Other mutations, including A2182C, A1821G, G1826A, and T1830C, were detected in both resistant and sensitive strains, corroborating the previous findings [41‐44], suggesting they may not be directly involved in clarithromycin resistance but rather represent naturally occurring polymorphisms within the H. pylori 23SrRNA gene. Thus, controversy lies regarding the roles of these polymorphic forms in clarithromycin resistance development. The almost perfect agreement, quantified by a Cohen’s Kappa analysis value of 1, underscores the reliability of these diagnostic tools. Furthermore, transformation experiments demonstrated that the A2143G mutation alone is sufficient to confer clarithromycin resistance, indicating the potential for rapid and widespread dissemination of resistance in natural populations. Considering the limited availability of sequencing facilities in many parts of India, these alternative molecular diagnostic methods such as restriction enzyme digestion and MAMA-PCR offer feasible, specific, and cost-effective options for the detection of clarithromycin resistance-associated mutations.

In contrast to clarithromycin, metronidazole, although highly variable across regions, remains consistently elevated in economically disadvantaged countries [15]. Our current findings indicate a 61.4% metronidazole resistance rate in WB, a substantial decline compared to earlier reports from 2005 [1, 17], likely reflecting reduced clinical use for parasitic and anaerobic infections.

Nonetheless, very high metronidazole resistance has been reported from other parts of India, such as Odisha (81.2%) and Karnataka (81.4%) [31, 32], and from neighbouring countries including Myanmar (80%), Bangladesh (94.6%), Nepal (88.1%), and Pakistan (89%) [35, 36, 45]. High resistance is also reported in China (90.1%) and Korea (67.7%), as well as across Africa (75.8%) [30, 38, 39]. In contrast, metronidazole resistance rates are relatively lower in developed regions such as Europe (38%) and the Americas (27%) [13]. Despite a reduction in resistance among WB isolates, the levels remain well above the 15% threshold, rendering the drug unsuitable for empirical inclusion in eradication regimens.

Consistent with this, our sequencing data revealed that two highly resistant H. pylori strains with an intact frxA harbored mutations in the rdxA, reaffirming its primary role in metronidazole resistance.

Metronidazole resistance poses a major obstacle in the efficacy of standard triple therapy for H. pylori eradication, increasing the risk of treatment failure by approximately 2.5-fold [15]. Furthermore, metronidazole resistance undermines the success of sequential therapy and may even contribute to the emergence of primary clarithromycin resistance, compounding the challenge of effective treatment [19]. However, traditional bismuth-containing quadruple therapies (BQTs), which use higher doses of metronidazole (1500–1600 mg/day), can effectively address this issue [29].

Second-line regimens combining PPIs, bismuth, amoxicillin, and levofloxacin can achieve up to 90% eradication after triple therapy failure [11]. However, rising levofloxacin resistance threatens their effectiveness. In our study, 69.3% of WB isolates were resistant. Similar increases have been reported in Northeast India, where resistance rose from 73.2% in 2016 to 89.1% in 2019 [34]. High rates are also seen in Karnataka (54.9%) and Odisha (65.6%) [31, 32], while Gujarat reports a lower rate of 13.8% [46]. Neighboring countries, including Bangladesh (66%), Nepal (43%) [47], Myanmar (33.8%) [35], and Pakistan (76%) [45], also face substantial resistance. These findings emphasize the need for routine susceptibility testing and judicious fluoroquinolone use to preserve second-line therapies.

The global surge in fluoroquinolone consumption, an estimated 64% increase between 2000 and 2010, is likely attributable to the widespread use of fluoroquinolone-monotherapy as an alternative first-line treatment for community-acquired pneumonia [47]. In our local population, levofloxacin is among the most commonly used over-the-counter antibiotics, frequently prescribed for upper respiratory tract infections such as coughs and colds, contributing to escalating resistance levels.

Molecular analysis revealed two common GyrA mutations, occurring at amino acid positions 87 and 91. Similar to findings in Japan [22], the mutation at position 91 in WB isolates was associated with low-level resistance, with MIC values ranging from 2 µg/ml to ≤ 4–8 µg/ml. However, additional mutations at position 87 led to significantly higher MICs, ranging from 16 µg/ml to ≤32 µg/ml. Strains with a single mutation at position 87 exhibited MICs as high as 8–32 µg/ml, suggesting that the 87^th amino acid substitution plays a critical role in not only conferring levofloxacin resistance but also elevating the resistance level beyond that observed with the 91^st position mutation.

In terms of amoxicillin resistance, only one resistant strain was identified in our study, marking a slight change from 0% resistance reported in 2005 [17]. In contrast, resistance rates were higher in Odisha (3.1%), Tamil Nadu (24%) and Delhi (17.6%) [10, 31, 33, 48]. Other regions such as Lucknow and Chandigarh reported 100% sensitivity [48], while neighboring countries like Bangladesh (3.6%) and Myanmar (4.6%) also showed low resistance [35, 36]. Several countries including those in Europe, as well as the USA, Japan, and Turkey, continue to report high rates of amoxicillin sensitivity [29]. The consistently low amoxicillin resistance in WB underscores its continued efficacy, though regular monitoring remains essential. Sequencing of the pbp1A gene identified mutations consistently present in all 10 naturally transformed resistant colonies, suggesting a potential role in resistance development. Additionally, transformation-based validation supports the functional involvement of the four identified pbp1A point mutations in mediating amoxicillin resistance and helps distinguish true resistance-associated variants from incidental polymorphisms. However, further functional studies are needed to conclusively determine the significance of these mutations.

Furazolidone is often considered as an alternative treatment in regions with high metronidazole resistance. Encouragingly, our WB isolates exhibited 100% sensitivity to furazolidone. Similar results have been reported in Bangladesh, Nepal [36], Korea [30] and China [38]. However, resistance has been noted in northern India (22.1%) [10] and Gujarat (13.8%) [46]. Despite noting good sensitivity in WB, the commercial unavailability of furazolidone acts as the rate limiting step in overcoming this therapeutic failure. Tetracycline sensitivity was 100% in our WB isolates, aligning with reports from several other Indian states [48, 49] and neighboring countries [35, 36]. Notably, resistance in Gujarat was reported with an unusually high rate of 50% [46]. The improvement in WB is particularly noteworthy, as resistance has decreased from 7.5% in 2005 [17] to 0% in the current study.

MDR H. pylori represents one of the most pressing challenges in eradication therapy, with resistance patterns varying by region, time, and host-related factors. MDR development is influenced by mechanisms such as coccoid transformation, efflux pump activation, and biofilm formation. The most common global MDR profile involves resistance to clarithromycin, metronidazole, and quinolones—mirrored in our findings, with 88.9% of MDR strains exhibiting resistance to this combination [50]. Similar trends were noted in Karnataka, where 86.6% of MDR strains were resistant to at least metronidazole and levofloxacin [32]. In comparison, primary MDR rates remain ≤ 10% in many European countries, but exceed 40% in Peru, highlighting a global disparity [50].

Virulence factor screening (cagA, vacA) of the 88 isolates revealed no significant association with resistance phenotypes, indicating that the presence of these genes may not directly influence antimicrobial resistance in our population. These findings align with the conclusions of Lee et al., [7], suggesting that epidemiological profiling alone is insufficient for guiding treatment strategies. Instead, region-specific antibiotic susceptibility testing should be a prerequisite for effective eradication therapy. In addition, we identified cases in which multiple H. pylori strains co-infected a single host, each exhibiting distinct antibiotic susceptibility profiles. This phenomenon significantly complicates treatment planning and highlights the importance of comprehensive diagnostics. If not properly monitored and managed with judicious antibiotic use, cases such as this are likely to accelerate disease progression and contribute to the further spread of resistance.

Hence, its high time to buckle up and address this resistance crisis with great care both by formulating therapeutic regimen in correct combination and by ensuring rapid identification of drug resistance with a constant monitoring and sustained surveillance. The global surge in macrolide and fluoroquinolone usage has significantly contributed to resistance trends, underscoring the urgent need for judicious antibiotic stewardship. Alternatives such as bismuth-containing quadruple-therapy must be considered more routinely, particularly in high-resistance regions. Finally, combating the emergence and spread of drug-resistant H. pylori demands a coordinated and multi-sectoral approach integrating healthcare providers, policymakers, researchers, and public health awareness initiatives. In parallel, continued research into the molecular mechanisms and genetic basis of resistance will be critical for developing targeted, next-generation therapies and ensuring sustained success in H. pylori eradication efforts.

While this study provides valuable insights into the antimicrobial resistance (AMR) patterns and genetic mutations in H. pylori strains isolated from patients in WB, Eastern India, several limitations must be acknowledged. The isolates analyzed were obtained from a specific region, and therefore, the findings may not be representative of other regions in India or globally. Regional variations in antibiotic usage and genetic diversity of H. pylori may influence resistance patterns. Resistance mechanisms beyond point mutations, such as efflux pump activity or epigenetic modifications, were not explored and may contribute to the observed resistance. Finally, the study did not integrate patient clinical data, such as treatment history or therapeutic outcomes, which limits the clinical interpretability of these findings.