Novel Mutations Conferring Amoxicillin Resistance in Helicobacter pylori in South Korea

Helicobacter pylori is the primary causative agent of gastritis, gastric ulcers, duodenal ulcers, gastric cancer, and peripheral B-cell lymphoma. H. pylori eradication often fails due to elevated antibiotic resistance. However, no previous studies have thoroughly examined amoxicillin resistance. Here, the objective was to identify clinical strains of H. pylori with amoxicillin resistance and to analyze single-nucleotide polymorphisms (SNPs) associated with amoxicillin resistance. From March 2015 to June 2019, genotypic and phenotypic amoxicillin resistance was analyzed using an E-test and whole-genome sequencing (WGS). Analysis of 368 clinical strains confirmed amoxicillin resistance in 31 strains (resistance rate of 8.7%). The genomes were extracted from nine resistant (<0.125 mg/L) strains, and WGS was performed for genetic analysis. WGS analysis identified SNPs present in pbp1a, pbp2, nhaC, hofH, hofC, and hefC in all nine isolates. Some of these genes may be related to amoxicillin resistance. A total of six SNPs (A69V, V374L, S414R, T503I, A592D, and R435Q) were identified in PBP2 of H-8, the most resistant strain. We predict that these six SNPs are associated with high amoxicillin resistance. Amoxicillin resistance should be considered in the clinical setting for the treatment failure of H. pylori eradication.


Introduction
Helicobacter pylori was first discovered in 1983, and H. pylori infection is associated with the pathogenesis of gastric ulcers, duodenal ulcers, gastric cancer, and peripheral B-cell lymphoma [1]. After H. pylori was first successfully cultured in 1982, numerous studies tried to determine the pathogenesis and treatment of H. pylori [1]. A report from 2013 indicated that the H. pylori prevalence rate in South Korea was reported to be 54% [2]. Since 1998, triple therapy containing clarithromycin and amoxicillin has been widely used to treat H. pylori infection. However, the success rate of H. pylori eradication treatment is decreasing owing to an increase in antibiotic resistance [3]. Eradication rates of H. pylori infection with a proton-pump inhibitor (PPI)-based triple therapy have reported a success rate of less than 80% worldwide. Clarithromycin resistance has been reported to be 32.3% in the United States and 17.8% in South Korea [4,5]. Such a high resistance rate is associated with treatment failures in patients and, thus, requires clinicians' attention. In addition, amoxicillin resistance in PPI-based triple therapy has recently shown a rapid increase by approximately 15% [6]. Therefore, it is necessary to develop new strategies to assess antibiotic resistance and select an appropriate treatment regimen [7][8][9]. To prevent treatment failure owing to antibiotic resistance, patient-tailored antimicrobial therapy through antibiotic resistance testing of clinical samples from patients has been proposed [10]. Some studies have reported that eradication therapy optimized for the H. pylori strain derived from patient samples is likely to have a higher success rate than that of conventional methods, considering treatment duration and cost [11]. However, because H. pylori is difficult to cultivate in a laboratory environment and antibiotic resistance testing requires additional time and cost, customized eradication therapy is difficult to use in real-world clinical practice [10,11].
In the clinical practice, a rapid diagnostic kit using polymerase chain reaction (PCR) is used to rapidly identify single-nucleotide polymorphisms (SNPs) (A2142G and A2143G) which are associated with clarithromycin resistance [12]. However, the PCR method cannot be applied to identify amoxicillin resistance because no specific SNPs have been reported.
This study's purpose was to prospectively collect clinical strains, investigate the amoxicillin resistance rate, and identify SNPs associated with amoxicillin resistance using whole-genome sequencing (WGS) to provide a basis for antibiotic resistance testing to aid H. pylori eradication therapy.

Discussion
H. pylori has reportedly infected approximately 4.4 billion people worldwide, and its prevalence varies from 18.9% to 87.7% in different regions [10]. H. pylori causes various diseases, such as chronic progressive gastritis, gastric ulcer, duodenal ulcer, and gastric cancer, and eradication of H. pylori reduces the incidence of gastric cancer [13]. Triple therapy with clarithromycin and PPIs is the treatment for H. pylori infection [14]. However, due to increasing clarithromycin resistance, the effectiveness of standard therapy is declining, and the World Health Organization is also aware of this problem [13]. Amoxicillin, another essential antibiotic in the H. pylori treatment regimen, also exhibits an increasing resistance rate of approximately 10% [13]. Selecting antibiotic combinations according to antibiotic susceptibility will enhance the success rate of H. pylori eradication [15]. However, H. pylori antibiotic susceptibility testing remains challenging due to the H. pylori cultivation difficulty and the time-consuming nature of H. pylori culture [16]. Clarithromycin resistance can be rapidly and readily identified using known resistance-associated SNPs through PCR. Using multiplex PCR (GenoType ® HelicoDR kit), the positive rate of clarithromycin resistance identification was 94.3%, which is higher than that of conventional methods, such as culture (77.1%) and histology (71.4%) [16]. A similar approach can be applied to identify amoxicillin resistance, another vital cause of H. pylori eradication failure. However, only a few SNPs associated with amoxicillin resistance have been reported because of a lack of molecular testing. The point mutation in PBP1A is the only reported SNP to date in amoxicillin resistance [17].
From the WGS analysis in this study, SNPs associated with amoxicillin resistance were identified in pbp1a, pbp2, hofH, and hofC [17]. F125L, D479E, D535N, S589G, K648Q, and R649K of PBP1A have been reported in previous studies [18][19][20]. This finding suggests that pbp1 and pbp2 SNPs induce antibiotic resistance by alternating the protein binding site, while hofH and hofC SNPs are associated with changes in the composition of the outer membrane.
In addition, 16 consecutive SNPs were observed in hofC. Therefore, it was considered a genetic change caused by a community acquisition mechanism rather than a SNP owing to a single mutation. This is because the strains used in the analysis were clinical strains isolated from patients treated at local hospitals.
The H-8 strain, with the highest resistance at the minimum inhibitory concentration of 32 mg/L, presented mutations of A69V, V374L, S414R, T503I, and A592DP in PBP1A and R435Q in PBP2. The A69V, V374L, and S414R SNPs in PBP1A were identified in a previous study [6,21,22], but T503I and A592D in PBP1A, and R435Q in PBP2 were newly discovered SNPs from this study. In subsequent studies, additional amoxicillin-resistant clinical strains will need to be collected and analyzed to determine whether T503I, A592D, and R435Q in PBP1A are related to amoxicillin resistance, which has not been previously reported.
This study had several limitations. Since H. pylori is difficult to culture under laboratory conditions, the isolation rate of H. pylori strains is only approximately 19% using culture. Therefore, cultural negativity does not ensure successful H. pylori eradication. M17V and R656P of PBP1A, which were first discovered in this study, require additional studies using other analytical methods. Furthermore, studies are required to analyze SNPs and MIC values with amoxicillin-resistance mechanisms. In the WGS, amoxicillin-sensitive H. pylori was not included in this analysis; therefore, it was difficult to ascertain whether these SNPs were directly related to resistance.
In our study, the resistance rate of amoxicillin and tetracycline was low compared to previous studies. This low resistance rate may be attributed to regional, national, or racial differences. Antibiotic resistance rates vary in different regions. Our test results indicate characteristics of antibiotic resistance in South Korea, especially Seoul, the capital region. We can add MIC resistance data from previous H. pylori studies.
Furthermore, we did not discuss tetracycline or levofloxacin in this study. We can discuss triple or gyrA mutations through additional analysis since we have WGS data. In addition, it would have been better to select amoxicillin-sensitive strains and compare them with amoxicillin-resistant strains to confirm that the SNPs we found contribute to amoxicillin resistance. However, we could only study strains with amoxicillin resistance and a reference ATCC 43504 strain due to funding limitations.
In future studies, we can confirm that SNPs are associated with high amoxicillin resistance through mutagenesis. New mutations of PBP2 can be studied as well. Furthermore, amoxicillin-sensitive strains can be selected and compared.

H. pylori Isolates and Culture Conditions
Clinical H. pylori strains were isolated from the gastric corpus and antrum specimens of patients who underwent an upper endoscopy at Gangnam Severance Hospital, Seoul, between March 2015 and June 2019. Detailed methods can be found in our institution's published paper [14]. The specimens were transferred to the lab using a transport medium (Noble Biosciences Inc., Hwaseong, Republic of Korea). The patient samples were inoculated in egg yolk emulsion (EYE) agar (Yuhan LabTech, Seoul, Republic of Korea) and grown in a microaerophilic incubator (10% carbon dioxide, 5% oxygen, and 85% nitrogen at 37 • C) for 3-5 days. Although the EYE medium used in this study has a high contamination risk, it has a fast growth rate and easy visual identification due to the color change of the medium because it contains triphenyl tetrazolium chloride [23]. The grown strains were confirmed via visual inspection and identified using MALDI-TOF-MS (MALDI Biotyper™, Bruker, Billerica, MA, USA) and MBT Compass IVD software (Bruker, Middlesex County, MA, USA). H. pylori strain ATCC 43504 was used as the quality control. The H. pylori strain was inoculated into a Microbank™ (Pro-Lab Diagnostics, Richmond Hill, ON, Canada) and stored at −70 • C.

H. pylori Isolates and Culture Conditions
Primary isolates of H. pylori colonies were sub-cultured for 2-4 days. Individual colonies were collected using aseptic needles and smeared into MSP 96 target-polished steel BC (Bruker, Middlesex County, MA, USA), and the colony was put into 1.5 µL of an α-cyano-4-hydroxycinnamic acid matrix (Bruker, Middlesex County, MA, USA).

H. pylori Antibiotic Susceptibility Test
We performed an antimicrobial susceptibility test for amoxicillin on isolated H. pylori strains using the gradient diffusion method with E-test ® (bioMérieux, Marcy-l'Étoile, France). After inoculation at a concentration of 1.5 ± 0.5 McFarland using DensiCHEK-Plus (bioMérieux, Marcy-l'Étoile, France), the samples were placed in an Anoxomat™ AN2CTS system (MART Microbiology, B.V., Lichtenvoorde, Netherlands), replaced with microaerophilic conditions, and incubated at 37 • C for 48-72 h. According to the European Committee on Antimicrobial Susceptibility Testing (EUCAST) guidelines, strains were classified as resistant to amoxicillin when the MIC was >0.125 µg/mL.

Whole-Genome Sequencing
For DNA extraction, a Gram-negative bacterial kit was used (GenElute™ Bacterial Genomic DNA Kit; G1N350, Sigma-Aldrich, St. Louis, MA, USA), according to the manufacturer's instructions. DNA digestion was performed using g-tubes™ (Covaris, Woburn, MA, USA), so the average length of the extracted gDNA was 14-16 kb. A DNA library was prepared using the Preparing Multiplexed Microbial Libraries Using the SMRTbell ® Express Template Prep Kit 2.0 protocol method of the SMRTbell ® Express Template Prep Kit 2.0 (Pacific Biosciences, Menlo Park, CA, USA). The Microbial Multiplexing Calculator was used for pooling, and the size of 12.5-15 kb was selected. Quality control was performed using Genomic DNA ScreenTape Analysis and a Qubit™ dsDNA HS Assay Kit. For the final preparation, the volume, size, concentration, product version, and on-plate loading concentration of the sample were entered using the Sequel + New Calculation of SMRT ® Link Software (Pacific Biosciences). The SMRTbell™ template Sequel Binding and Internal Ctrl Kit 3.0 (Pacific Biosciences) was used for annealing and binding. WGS was performed with the SMRT Sequel ® II system (Pacific Biosciences). Movie time per SMRT cell (hours) was set to 10 h, and the pre-extension time was set to 2 h. The final sample (85 µL/well) was dispensed into a 96-well PCR plate (Pacific Biosciences), and each cell was sequenced for 12 h in the Sequel II System. The other experimental methods were performed according to the manufacturer's instructions.

Multilocus Sequencing Typing (MLST)
A total of thirteen clarithromycin-resistant strains and three sensitive strains were selected from H. pylori clinical strains, incubated, and cultured on EYE agar plates. After culturing, H. pylori colonies were added to 1 mL of sterile water in an E-tube. After that, it was inactivated by heating at 95 • C in a heat block for 10 min. DNA was extracted from the inactivated samples using an InstaGene matrix (BIO-RAD), PCR was performed using a DNA Engine Tetrad 2 Peltier Thermal Cycler (BIO-RAD), and product purification was performed using a Multiscreen filter plate (Millipore Corp.). Sequencing was performed using an ABI PRISM 3730XL Analyzer (96 capillary type). The genes sequenced for MLST analysis were atpA, efp, mutY, ppa, trpC, ureI, and yphC, and allele and sequence types were determined.

Conclusions
We verified the previously reported SNPs related to amoxicillin resistance using WGS and discovered additional SNPs likely associated with amoxicillin resistance. The discovery of SNPs related to amoxicillin resistance will help understand treatment failures and guide antibiotic selection for H. pylori eradication.