Abstract
Background
Epidemiologic studies of the carcinogenic stomach bacterium Helicobacter pylori have been limited by the lack of non-invasive detection and genotyping methods. We developed a new stool-based method for detection, quantification, and partial genotyping of H. pylori using droplet digital PCR (ddPCR), which allows for increased sensitivity and absolute quantification by PCR partitioning.
Materials and Methods
Stool-based ddPCR assays for H. pylori 16S gene detection and cagA virulence gene typing were tested using a collection of 50 matched stool and serum samples from Costa Rican volunteers and 29 H. pylori stool antigen-tested stool samples collected at a U.S. hospital.
Results
The stool-based H. pylori 16S ddPCR assay had a sensitivity of 84% and 100% and a specificity of 100% and 71% compared to serology and stool antigen tests, respectively. The stool-based cagA genotyping assay detected cagA in 22 (88%) of 25 stools from CagA antibody-positive individuals and 4 (16%) of 25 stools from CagA antibody-negative individuals from Costa Rica. All 26 of these samples had a Western-type cagA allele. Presence of serum CagA antibodies was correlated with a significantly higher load of H. pylori in the stool.
Conclusions
The stool-based ddPCR assays are a sensitive, non-invasive method for detection, quantification, and partial genotyping of H. pylori. The quantitative nature of ddPCR-based H. pylori detection revealed significant variation in bacterial load among individuals that correlates with presence of the cagA virulence gene. These stool-based ddPCR assays will facilitate future population-based epidemiologic studies of this important human pathogen.
INTRODUCTION
Helicobacter pylori chronically infects over half the world’s population and is the only bacterium to be classified as a carcinogen by the WHO because of its role in gastric cancer development. Gastric cancer is the third leading cause of cancer deaths worldwide, and 89% of gastric cancer cases are attributable to H. pylori infection [1, 2]. H. pylori is in fact responsible for a range of disease outcomes from asymptomatic gastritis (inflammation of the gastric mucosa) to peptic ulcer and gastric cancers, but may be protective against other diseases including esophageal cancer [3, 4] and asthma [5, 6]. Differences in disease outcome are due in part to genetic differences among H. pylori strains. H. pylori exhibits extensive inter-strain genetic diversity as well as intra-strain genetic diversification during the course of infection [7].
Presence of the strain-variable cagA gene and specific cagA allelic variations are associated with increased risk of peptic ulcers and gastric cancer [8–11]. The cagA gene is contained within the cag pathogenicity island that encodes a Type IV secretion system that delivers the CagA protein into host gastric epithelial cells [12]. Inside the host cell, the CagA protein is tyrosine-phosphorylated at EPIYA (Glu-Pro-Ile-Tyr-Ala) sites and deregulates host SHP-2 tyrosine phosphatase, a bona fide oncoprotein, resulting in changes to the host cell morphology [13–16]. The cagA gene is grouped into two different allele types based on whether it encodes an EPIYA-C or EPIYA-D motif. Phosphorylated EPIYA-D motifs interact more strongly with host SHP-2 than phosphorylated EPIYA-C motifs [17]. cagA alleles encoding an EPIYA-D motif are predominantly found in H. pylori strains circulating in East Asian countries and are associated with an increased risk of gastric cancer development [11].
Molecular epidemiologic studies of H. pylori investigating the role of bacterial genetic diversity in disease outcome have been limited by the lack of non-invasive methods for detection and genotyping of H. pylori. Current H. pylori genotyping methods to assess the presence of known virulence genes or alleles rely on upper gastrointestinal endoscopy for subsequent culturing from a gastric biopsy, which precludes sampling from the majority of H. pylori-infected individuals who are asymptomatic. Non-invasive tests for H. pylori infection, including the urea breath test and stool antigen test, detect the presence of H. pylori infection but do not differentiate between strains or provide H. pylori genetic information. For this reason, there has been great interest in development of non-invasive stool-based tests for detecting and genotyping H. pylori. H. pylori is present in stool at low abundance and is generally not culturable. PCR-based methods (either multiple rounds of conventional PCR or real-time PCR) for detecting H. pylori DNA in stool have been reported with sensitivities varying between 25% and 92% [18–22]. However, PCR-based tests for detecting and genotyping H. pylori from stool have not been adopted for use in epidemiologic studies or diagnostics likely because of issues with false positives from contamination and low sensitivity and reproducibility.
A recent development in PCR technology, droplet digital PCR (ddPCR), addresses many challenges of detection and genotyping of H. pylori in stool. ddPCR partitions a single PCR reaction into 20,000 droplets, allowing increased sensitivity for detection of rare targets and absolute quantitation by analysis of the frequency of positive droplets [23]. Reaction partitioning also increases tolerance of PCR inhibitors, further improving assay sensitivity for inhibition-prone samples such as stool [24, 25]. ddPCR assays can include fluorescent hydrolysis probes to increase the specificity of the assay, and multiple probes can be used to differentiate between different alleles of a gene. Because of the unique advantages that ddPCR offers, this new technology has been adopted for use in detection of infectious agents in a variety of samples types [26, 27] as well as a method for detection of known cancer biomarkers [28–30].
We have developed a new, non-invasive method for detection, quantification, and cagA genotyping of H. pylori from stool samples that uses ddPCR. We tested this method using a collection of matched serum and stool samples from volunteers in Costa Rica, a country with an H. pylori prevalence of 78% [31] and one of the highest incidence and mortality rates of stomach cancer [32]. Development of non-invasive methods to genotype H. pylori will facilitate challenging molecular epidemiology studies to investigate prevalence and track transmission of specific H. pylori strains in populations and to examine H. pylori-host genetic interactions in disease progression.
METHODS
Study populations and specimen collection
Samples from Costa Rica
Stool and serum samples were collected from 50 volunteers as part of a larger population-based study of the human microbiome that included 150 adult volunteers from Hojancha, Guanacaste, chosen by random selection of household number using census data. The 50 volunteers whose samples were analyzed in this study were selected to include 25 men and 25 women under the age of 60 years to minimize the likelihood of H. pylori clearance due to H. pylori-related disease progression. The 50 volunteers had a median age of 35 years (range 19 to 56 years). Stool samples were collected at home and placed by participants into a vial containing 5 ml RNAlater nucleic acid preservative (Ambion) and immediately stored in a cooler with dry ice prior to delivery to the study clinic. At the clinic, samples were transferred to liquid nitrogen for storage and shipment to a biorepository in the U.S. Blood was collected at the study clinic and the serum was separated and frozen in aliquots. Samples were collected within 6 days of each other and only from individuals who had not been on antibiotic treatment for at least six weeks. The study protocols and procedures for the protection of human subjects were approved by the Institutional Review Boards of the U.S. National Cancer Institute and the National University of Costa Rica. Informed consent was obtained from each participant.
Samples from the United States
Stool samples were collected from 29 individuals seen at outpatient clinics at Harborview Medical Center in Seattle, Washington for a variety of medical reasons. Stool samples were collected at home by the patient and brought to the clinic on the same day. Stools were tested for H. pylori antigen because of clinical indications and deidentified leftover specimens were provided for this study. Patients ranged in age from 1 to 68 years (median 27 years). Specimens were held at 4°C before division into ~1ml aliquots and frozen at −75°C. Formed stools were emulsified in a sufficient quantity of molecular grade water to facilitate aliquot distribution.
H. pylori and CagA IgG ELISA
Serum samples from the Costa Rican participants were analyzed for H. pylori antibodies using the Wampole Helicobacter pylori IgG ELISA II Test System and for CagA antibodies using the CagA IgG ELISA Kit (Alpco). Individuals were considered H. pylori-positive by serum test if they were positive for H. pylori antibodies and/or CagA antibodies.
Stool DNA extraction
Effects of stool sample storage conditions and DNA extraction methods on yield of H. pylori DNA was tested using a stool sample from one H. pylori-positive volunteer (Supplementary Materials and Methods, Supplementary Table S4). Based on the results, stool DNA of all other samples were extracted using the QIAamp Stool DNA Mini Kit (Qiagen) according to the manufacturer’s instructions, with the lysis step performed at 95°C. For samples stored in RNAlater, the sample was first transferred to a microcentrifuge tube and centrifuged to remove the RNAlater.
Primer and probe design
The primer and probe sequences for all assays are listed in Table 1 and the design and testing of the primers and probes are described in the Supplementary Materials and Methods.
Table 1.
Primers and probes used for droplet digital PCR (ddPCR) assays to detect and genotype H. pylori from stool
| ddPCR assay | Primer or probe name | Sequence and chemical modifications1 | Reference |
|---|---|---|---|
| H. pylori 16S | HPF | 5′ – GCGACCTGCTGGAACATTAC – 3′ | Gramley, 1999 |
| HPR | 5′ – CGTTAGCTGCATTACTGGAGA – 3′ | Gramley, 1999 | |
| Hp16S_HEX | 5′ HEX - AAGCCCTCCAACAACTAGCATCCAT - BHQ1 3′ | this study | |
| cagA detection | cagA_F | 5′ – TGGCTCAAGCTCGTGAAT – 3′ | this study |
| cagA_R | 5′ – TGGAAAACTTGAACGAATCAGA – 3′ | this study | |
| cagA_FAM | 5′ FAM – CTTCCYACATTATGYGCAACKATC – BHQ1 3′ | this study | |
| cagA EPIYA typing | EPIYA_F1 | 5′ – TCAGTTAGCCCTGAACC – 3′ | this study |
| EPIYA_F2 | 5′ – TCAACTAGCCCTGAACC – 3′ | this study | |
| EPIYA_R1 | 5′ – GCCCTACCTTACTGAGAT – 3′ | this study | |
| EPIYA_R2 | 5′ – GAAAGCCCTACTTTACTGAG – 3′ | this study | |
| EPIYA-C_HEX | 5′ HEX – TCCGCCGAGATCATCAATCGTAGC - BHQ1 3′ | this study | |
| EPIYA-D_FAM | 5′ FAM – AAGCCTGCTTGATTTGCCTCATCAAA – BHQ1 3′ | this study |
HEX: hexachloro-fluorescein, FAM: 6-carboxyfluorescein, BHQ: black hole quencher
Droplet Digital PCR
Droplet digital PCR was performed according to the manufacturer’s instructions with each 20 μl reaction containing 1x ddPCR Supermix for Probes (BioRad), 900 nM of each primer (Table 1), 250 nM of each probe (Table 1), and 10 μl stool DNA. For the Costa Rican stool DNA samples, the concentration was adjusted to 100 ng/μl and 1 μg stool DNA was analyzed per reaction. Droplets were generated using the QX200 Droplet Generator (BioRad). Reactions were subject to thermal cycling with the following conditions: 95°C for 10 minutes, 45 cycles of 94°C for 30 seconds and 55°C for 1 minute, and 98°C for 10 minutes. Droplets were then analyzed for fluorescent amplitude using the QX200 Droplet Reader (BioRad). Data was analyzed using the QuantaSoft software version 1.6.6 (BioRad) and thresholds were set manually. Placement of the threshold was set for each assay by visually inspecting the amplitude plot for a positive control sample and placing the threshold between the cluster of positive droplets and the cluster of negative droplets. The thresholds were set at 4500 for the H. pylori 16S assay, 6500 for the cagA assay, and 2000 for the EPIYA assay. To increase the total amount of stool DNA being assayed, two reactions were run for each stool sample and the results were combined by averaging the number of copies per μl. A positive control (stool DNA from an H. pylori-positive volunteer) and negative control (molecular grade water) were included in each batch of samples analyzed.
Statistical Analysis
Differences in H. pylori load as measured by the H. pylori 16S ddPCR assay were compared using the Wilcoxon rank-sum test. Statistical analyses were performed using SAS version 9.4 (SAS Institute Inc.)
RESULTS
Development and optimization of stool-based ddPCR assay to detect H. pylori
To first test the performance of the ddPCR assay to detect the H. pylori 16S gene amongst a complex background of stool DNA, we spiked 10-fold dilutions of known quantities of H. pylori strain G27 genomic DNA, as determined by a NanoDrop spectrophotometer (Thermo Scientific), into stool DNA from an H. pylori-negative volunteer. As shown in Figure 1, we observed a clear separation between droplets containing H. pylori 16S and droplets negative for H. pylori 16S, and the assay correctly quantified the concentration of H. pylori genomes spiked down to the lowest amount tested (one H. pylori genome per μg stool DNA).
Figure 1.
H. pylori 16S ddPCR amplitude plot for H. pylori-negative stool DNA spiked with three different biologically relevant concentrations of H. pylori strain G27 genomic DNA. Each dot represents one droplet. Dots above the threshold (black line set at fluorescence amplitude of 4500) are positive for the H. pylori 16S gene and dots below are negative. Expected number of H. pylori genome copies and measured number of H. pylori 16S gene copies with Poisson 95% confidence intervals in parentheses are shown above the amplitude plots. Two 20 μl reactions were run for each sample and the results were combined.
Development of stool-based ddPCR assay to detect and EPIYA type the H. pylori cagA gene
Unlike the H. pylori 16S gene, the cagA gene exhibits extensive nucleotide diversity among H. pylori strains making the design of universal primers and probes for this gene more challenging. To assess the compatibility of the cagA ddPCR primers and probes with a globally representative sample of H. pylori strains, we examined the polymorphisms present within the primer and probe regions for 37 H. pylori strains described in Olbermann et al. [33] and tested the ability of the cagA ddPCR assays to detect cagA and distinguish between EPIYA types in strains having some of these polymorphisms (Supplementary Materials and Methods, Supplementary Tables S1–S3). We spiked genomic DNA of six North American strains (EPIYA-C) [34] and eight Japanese H. pylori strains (EPIYA-D) [11], as well as two fully sequenced European strains (EPIYA-C), 26695 [35] and G27 [36], into H. pylori-negative stool DNA. The cagA detection ddPCR assay detected the cagA gene for 15 (94%) of 16 strains tested. The cagA gene was not detected for strain Em7-1 even though this strain has the same sequence at the cagA primer and probe sites as strain 26695, which was positive by the cagA ddPCR assay (Supplementary Table S3). The cagA EPIYA type was correctly determined for 15 (94%) of 16 strains tested. Neither EPIYA motif was detected for Japanese strain Oki633, which had two polymorphisms within the EPIYA-D probe site (Supplementary Table S3). The ddPCR amplitude plots for the cagA detection and EPIYA typing assays for two of the spiked strains, Em47-1 (EPIYA-C) and Oki573 (EPIYA-D), are shown in Supplementary Figure S1.
Validation of ddPCR assays using population-based and clinical samples
The stool-based ddPCR assays were tested using a collection of 50 matched serum and stool samples from Costa Rican volunteers. Of these 50, 32 (64%) were H. pylori-positive based on the H. pylori serum test and 25 (50%) were positive for H. pylori having a cagA gene based on the serum test to detect CagA antibodies. Of the 25 volunteers positive for CagA antibodies, 20 were also positive for H. pylori antibodies and five were negative for H. pylori antibodies. The stool-based H. pylori 16S ddPCR assay detected H. pylori in the stool of 27 (84%) of the 32 participants positive for H. pylori serum antibodies and 31(84%) of the 37 participants positive for H. pylori or CagA antibodies by the serum test. The H. pylori 16S and cagA ddPCR assays were positive for four of the five participants who were positive for only CagA antibodies and the H. pylori loads in the stool were relatively high in these samples (17, 34, 48, and 452 H. pylori 16S copies per μg stool DNA). Of the 13 participants negative for both H. pylori and CagA antibodies, all 13 (100%) were negative for the H. pylori 16S ddPCR assay (Table 2).
Table 2.
Comparison of results of stool-based H. pylori ddPCR assay detection and H. pylori load with results of H. pylori serum test (samples from 50 Costa Rican volunteers) and H. pylori stool antigen test (samples from 29 patients at a U.S. hospital)
| H. pylori 16S ddPCR | Costa Rican Samples H. pylori or CagA serum testa | U.S. Samples H. pylori stool antigen test | ||
|---|---|---|---|---|
| positive n=37 |
negative n=13 |
positive n=12 |
negative n=17 |
|
| positive | 31 (84%) | 0 | 12 (100%) | 5 (29%) |
| median copies H. pylori 16S per μg stool DNA (range) | 8.4 (1–452) | N/A | 22.3 (1.1–121) | 1.1 (0.5–3.5) |
| negative | 6 (16%) | 13 (100%) | 0 | 12 (71%) |
Positive= positive for H. pylori or CagA serum antibody test, Negative=negative for both
The stool-based H. pylori 16S ddPCR assay was further tested using 29 stool samples from patients at a U.S. hospital for which the H. pylori stool antigen test was performed. Of the 12 that were H. pylori-positive, all 12 (100%) were positive for the H. pylori 16S ddPCR assay. Of the 17 samples that were H. pylori negative by stool antigen test, 12 (71%) were negative by H. pylori 16S ddPCR assay and five were positive (Table 2). The five samples that were positive for the H. pylori 16S ddPCR assay but negative for the H. pylori stool antigen assay had H. pylori loads in the lower half of the range, which may have been below the detection limit for the stool antigen assay. Furthermore, three of these five patients had a previous diagnosis of H. pylori infection indicated in their medical records.
The stool-based cagA ddPCR assays were compared to the CagA serum antibody test using the samples from 50 Costa Rican volunteers. Of the 25 participants serum positive for CagA antibodies, 21 (84%) were positive for the cagA detection ddPCR assay and 22 (88%) were positive for the cagA EPIYA typing ddPCR assay, with all samples having a cagA gene encoding the EPIYA-C motif. Of the 25 participants serum negative for CagA antibodies, 16 (64%) were negative for the cagA detection ddPCR assay and 21 (84%) were negative for the cagA EPIYA typing ddPCR assay (Table 3).
Table 3.
Comparison of stool-based cagA gene ddPCR and CagA serum antibody assays for 50 Costa Rican volunteers
| cagA ddPCR assay | CagA serum test | |
|---|---|---|
| positive n=25 |
negative n=25 |
|
| cagA gene ddPCR | ||
| positive | 21 (84%) | 9 (36%) |
| negative | 4 (16%) | 16 (64%) |
| cagA EPIYA typing ddPCR | ||
| positive | 22 (88%) | 4 (16%) |
| negative | 3 (12%) | 21 (84%) |
| both cagA ddPCR assays | ||
| positive for both | 20 (80%) | 3 (12%) |
| positive for cagA gene ddPCR, negative for EPIYA typing ddPCR | 1 (4%) | 6 (24%) |
| negative for cagA gene ddPCR, positive for EPIYA typing ddPCR | 2 (8%) | 1 (4%) |
| negative for both | 2 (8%) | 15 (60%) |
Load of H. pylori in the stool varies and is correlated with serum CagA antibody status
For the 31 Costa Rican stool samples positive for H. pylori by ddPCR, there was a two log range of H. pylori load in the stool (median 8.4, range 1 – 452 H. pylori copies per μg stool DNA). Of the 17 U.S. stool samples that were positive by H. pylori 16S ddPCR assay, the load of H. pylori in the stool varied between 0.5 to 120.9 copies H. pylori 16S per μg stool DNA, with a median of 13.6 (Table 2). The H. pylori 16S ddPCR assay was performed a second time for the 50 Costa Rican samples, and the copy number data was highly reproducible, with a correlation coefficient of 0.98 (Figure 2A). For 35 of the 50 samples having 10 or fewer H. pylori 16S copies per μg stool DNA, the correlation coefficient was 0.74 (Figure 2B).
Figure 2.
Analysis of H. pylori load in the stool by quantification of H. pylori 16S gene copy number per μg stool DNA using ddPCR. A–B) Comparison of H. pylori 16S copy number of 50 Costa Rican stool samples (A) and 35 of the 50 Costa Rican stool samples having H. pylori 16S copy numbers ≤ 10 (B) for two separate ddPCR runs. A trend line and the corresponding linear equation and correlation coefficient (R2) are indicated. C) Boxplots (box indicates 25th–75th percentile and median, whiskers indicate the minimum and maximum data points) of the H. pylori 16S copy number of stool samples from H. pylori-positive Costa Rican volunteers with positive (n=25) and negative (n=12) serum CagA antibody test. The H. pylori 16S copy number per μg stool DNA was compared using the Wilcoxon rank-sum test (p=0.009).
It is not known whether the load in the stool is correlated with the load in the stomach or whether other factors such as level of inflammation, disease state, or bacterial genetic factors are involved. To assess the potential role of the cagA virulence gene in load of H. pylori in the stool, the H. pylori 16S copy number per μg stool DNA was compared between Costa Rican volunteers positive for CagA antibodies and negative for CagA antibodies. There was a statistically significant increase in H. pylori load in the stool of volunteers positive for CagA antibodies compared with volunteers negative for CagA antibodies [median H. pylori 16S copy number per μg stool DNA (range) CagA positive: 11.2 (0 – 443), CagA negative: 1.6 (0 – 30); Wilcoxon rank-sum test p-value: 0.009] (Figure 2C).
DISCUSSION
The interplay between environment, host genotype, and H. pylori strain differences that lead to a wide range of disease outcomes (gastritis, peptic ulcers, and gastric cancer) as well as possible protective effects against esophageal cancer and asthma are not well understood. Non-invasive tests to detect and genotype H. pylori will facilitate molecular epidemiology studies to examine the distribution of specific H. pylori genes and alleles in populations and their role in disease outcome. While there are well established non-invasive tests for detecting H. pylori infection (urea breath test, serology test, stool antigen test), these tests do not provide information about the genotype of the H. pylori strain or the bacterial load. H. pylori genotype can be determined by first obtaining a culture from a gastric biopsy, but this involves an expensive and invasive procedure. We have developed a new non-invasive method for detection, quantification, and genotyping of H. pylori from stool samples using ddPCR that is sensitive and reproducible.
This study takes advantage of two different collections of samples to compare these new, non-invasive ddPCR assays to two different H. pylori diagnostic tests, the serum test and stool antigen test, which are widely used for epidemiologic studies and clinical diagnosis. The sensitivity and specificity of the stool-based ddPCR assay to detect H. pylori infection depended on the H. pylori test that was used for the comparison, either serology or stool antigen test. The serology test has a sensitivity of 85% and specificity of 79% and the stool antigen test has a sensitivity of 94% and specificity of 92% [37]. These two tests detect different biological processes of H. pylori infection (antibody response to H. pylori and H. pylori shedding into the stool) so it is not surprising that the stool-based H. pylori ddPCR assay would compare differently to these two tests. Serology tests for H. pylori can produce false positives due to circulating antibodies still present after clearance of infection and false negatives due to low antibody response to the H. pylori antigen employed by the test. Of the six individuals who were positive for either H. pylori or CagA antibodies but were negative for H. pylori 16S ddPCR stool assay, two were positive by ddPCR stool assay upon repeat testing and both had low copy numbers of H. pylori 16S, indicating that they were near the limit of detection for this assay. For the samples from the U.S. hospital, the higher sensitivity is likely due to both tests detecting H. pylori shed into the stool. The five samples that were positive for the H. pylori 16S ddPCR assay but negative for the H. pylori stool antigen assay had low H. pylori loads, which may have been below the detection limit for the stool antigen assay. Three of these five patients also had a previous diagnosis of H. pylori infection indicated in their medical records, supporting the positive results by H. pylori 16S ddPCR assay.
It is also possible that differences in storage conditions between the Costa Rican and the U.S. stool samples could have contributed to differences in sensitivity of the assay. We used the ddPCR assays to quantitatively assess different stool DNA extraction methods and stool sample storage conditions (see Supplementary Materials and Methods and Supplementary Table S4). We found that the H. pylori DNA was robust to cycles of freezing and thawing and storage at ambient temperatures in RNAlater preservative, making it possible to conduct epidemiologic studies in which immediate freezing of the sample is not feasible. We also found that stool DNA extraction methods that employ bead beating to lyse bacterial cells yielded less H. pylori 16S copies per μg stool DNA than those without bead beating. H. pylori lyses fairly easily, so the decrease in the relative proportion of H. pylori in samples that underwent bead beating was likely due to an increase in the yield of DNA from difficult to lyse bacterial species in the stool. Stool samples were collected from one H. pylori-positive volunteer at two different time points that were 20 months apart. H. pylori 16S was detected by the stool-based ddPCR assay at both time points although the load was slightly higher at the second time point (81 vs. 49 16S copies per μg stool DNA, Supplementary Table S4). Future studies that include more H. pylori-positive individuals would be needed to further examine fluctuations in H. pylori load in the stool over time.
Aside from detection of H. pylori infection, we have also developed ddPCR assays to detect the H. pylori cagA gene and distinguish between different alleles of this gene, which are associated with differences in gastric cancer risk. The cagA detection ddPCR assay and the cagA EPIYA typing ddPCR assay have similar sensitivities for detection of the cagA gene (84% and 88%, respectively), but the specificity is higher for the cagA EPIYA typing ddPCR assay (84% compared to 64% for the cagA detection ddPCR assay). The cagA EPIYA typing ddPCR assay can be used for both detection and allele typing of cagA although it will not detect cagA allele types that encode neither an EPIYA-C nor an EPIYA-D motif. However, these allele types make up less than 3% of cagA allele types [38].
All 26 of the Costa Rican stool samples that were positive for the cagA gene by the cagA EPIYA typing ddPCR assay had the Western type of the gene encoding an EPIYA-C motif. While this result cannot be confirmed by the serum CagA antibody test, which does not distinguish between cagA allele types, this result is expected based on analysis of the cagA allele types of 33 Costa Rican H. pylori strains, all of which had Western type cagA gene [38]. Furthermore, an analysis of the geographical origin of 24 Costa Rican H. pylori strains characterized 21 as belonging to the hpEurope group and three as belonging to the hspWAfrica group [39], which would be expected to have cagA genes encoding an EPIYA-C motif. H. pylori strains of Amerindian origin have an EPIYA-DC motif with characteristics of both EPIYA-C and EPIYA-D motifs [40]. The cagA EPIYA typing assay would characterize the cagA gene of these strains as encoding an EPIYA-C motif.
Using ddPCR also allows for absolute quantitation of H. pylori load in the stool, information not obtained by other stool-based methods. We observed a two log range of H. pylori loads in the stool and the load is significantly higher in those with a cagA-positive strain. Density of H. pylori in the stomach also varies among infected individuals, and a higher H. pylori load in the stomach among individuals infected with a cagA-positive strain has been previously reported [41]. The relationship between H. pylori load in the stool and density of infection in the stomach, as well as other possible factors such as level of inflammation, disease state, or bacterial genetic factors, is not presently understood. Future studies that use matched gastric biopsy and stool samples will need to be conducted to address these questions.
It is a limitation of this study that the stool-based ddPCR assays were not compared to histology and culturing from gastric biopsy, which are considered as the gold standard for H. pylori diagnosis. Further validation of the stool-based ddPCR assays should be conducted on matched gastric biopsy and stool samples from various populations. These studies could also include comparison of the H. pylori genotypic information obtained from stomach samples to that from stool samples since individuals can have a mixture of H. pylori genotypes due to co-infecting strains of H. pylori or within-strain genotypic variation.
In conclusion, we have developed a sensitive and reproducible method to non-invasively detect and genotype H. pylori and measure bacterial load from stool samples. This non-invasive method to genotype H. pylori will allow future studies of H. pylori biomarkers of transmission and pathogenesis. Further understanding of the host and bacterial factors involved in H. pylori-associated disease development can inform strategies for tailored treatment decisions to reduce disease burden. This is particularly relevant to H. pylori since infection confers risk for some diseases and possible protection against others.
Supplementary Material
Supplementary Figure S1. ddPCR amplitude plots for the cagA gene detection assay (probe: cagA_FAM) and the cagA EPIYA typing assay (probes: EPIYA-C_HEX and EPIYA-D_FAM). H. pylori genomic DNA was spiked in to H. pylori-negative stool DNA. The H. pylori 16S copy number is indicated in parentheses after the name of each spiked strain (note: the H. pylori genome has two copies of the 16S gene and a single copy of the cagA gene). On the top are the amplitude plots for a North American H. pylori strain, Em47-1, which has a Western type cagA gene encoding an EPIYA-C motif. On the bottom are the amplitude plots for a Japanese H. pylori strain, Oki573, which has an East Asian cagA gene encoding an EPIYA-D motif. Each dot represents one droplet. Dots above the threshold (black line set at fluorescence amplitude of 6500 for the cagA gene detection assay and 2000 for the cagA EPIYA typing assay) are positive for the cagA gene. Number of cagA gene copies is indicated above each amplitude plot. Two 20 μl reactions were run for each sample and the results of each reaction were combined.
Acknowledgments
This work was supported by grants K01DK090103 and R01AI054423 from the NIH and by the NIH Intramural Research Program.
We would like to thank Yoshio Yamaoka for providing Japanese H. pylori strains for testing of the cagA assays. We would also like to thank Tina Gall and Jennifer Taylor for their assistance in constructing the figures.
Footnotes
Competing interests: the authors have no competing interests.
References
- 1.Ferlay J, Soerjomataram I, Dikshit R, et al. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015;136:E359–86. doi: 10.1002/ijc.29210. [DOI] [PubMed] [Google Scholar]
- 2.Plummer M, Franceschi S, Vignat J, Forman D, de Martel C. Global burden of gastric cancer attributable to Helicobacter pylori. Int J Cancer. 2015;136:487–90. doi: 10.1002/ijc.28999. [DOI] [PubMed] [Google Scholar]
- 3.Chow WH, Blaser MJ, Blot WJ, et al. An inverse relation between cagA+ strains of Helicobacter pylori infection and risk of esophageal and gastric cardia adenocarcinoma. Cancer Res. 1998;58:588–90. [PubMed] [Google Scholar]
- 4.de Martel C, Llosa AE, Farr SM, et al. Helicobacter pylori infection and the risk of development of esophageal adenocarcinoma. J Infect Dis. 2005;191:761–7. doi: 10.1086/427659. [DOI] [PubMed] [Google Scholar]
- 5.Chen Y, Blaser MJ. Helicobacter pylori colonization is inversely associated with childhood asthma. J Infect Dis. 2008;198:553–60. doi: 10.1086/590158. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Arnold IC, Dehzad N, Reuter S, et al. Helicobacter pylori infection prevents allergic asthma in mouse models through the induction of regulatory T cells. J Clin Invest. 2011;121:3088–93. doi: 10.1172/JCI45041. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Dorer MS, Talarico S, Salama NR. Helicobacter pylori’s unconventional role in health and disease. PLoS Pathog. 2009;5:e1000544. doi: 10.1371/journal.ppat.1000544. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Covacci A, Censini S, Bugnoli M, et al. Molecular characterization of the 128-kDa immunodominant antigen of Helicobacter pylori associated with cytotoxicity and duodenal ulcer. Proc Natl Acad Sci U S A. 1993;90:5791–5. doi: 10.1073/pnas.90.12.5791. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Nomura AM, Perez-Perez GI, Lee J, Stemmermann G, Blaser MJ. Relation between Helicobacter pylori cagA status and risk of peptic ulcer disease. Am J Epidemiol. 2002;155:1054–9. doi: 10.1093/aje/155.11.1054. [DOI] [PubMed] [Google Scholar]
- 10.Blaser MJ, Perez-Perez GI, Kleanthous H, et al. Infection with Helicobacter pylori strains possessing cagA is associated with an increased risk of developing adenocarcinoma of the stomach. Cancer Res. 1995;55:2111–5. [PubMed] [Google Scholar]
- 11.Matsunari O, Shiota S, Suzuki R, et al. Association between Helicobacter pylori virulence factors and gastroduodenal diseases in Okinawa, Japan. J Clin Microbiol. 2012;50:876–83. doi: 10.1128/JCM.05562-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Censini S, Lange C, Xiang Z, et al. cag, a pathogenicity island of Helicobacter pylori, encodes type I-specific and disease-associated virulence factors. Proc Natl Acad Sci U S A. 1996;93:14648–53. doi: 10.1073/pnas.93.25.14648. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Segal ED, Cha J, Lo J, Falkow S, Tompkins LS. Altered states: involvement of phosphorylated CagA in the induction of host cellular growth changes by Helicobacter pylori. Proc Natl Acad Sci U S A. 1999;96:14559–64. doi: 10.1073/pnas.96.25.14559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Stein M, Bagnoli F, Halenbeck R, Rappuoli R, Fantl WJ, Covacci A. c-Src/Lyn kinases activate Helicobacter pylori CagA through tyrosine phosphorylation of the EPIYA motifs. Mol Microbiol. 2002;43:971–80. doi: 10.1046/j.1365-2958.2002.02781.x. [DOI] [PubMed] [Google Scholar]
- 15.Selbach M, Moese S, Hauck CR, Meyer TF, Backert S. Src is the kinase of the Helicobacter pylori CagA protein in vitro and in vivo. J Biol Chem. 2002;277:6775–8. doi: 10.1074/jbc.C100754200. [DOI] [PubMed] [Google Scholar]
- 16.Higashi H, Tsutsumi R, Muto S, et al. SHP-2 tyrosine phosphatase as an intracellular target of Helicobacter pylori CagA protein. Science. 2002;295:683–6. doi: 10.1126/science.1067147. [DOI] [PubMed] [Google Scholar]
- 17.Higashi H, Tsutsumi R, Fujita A, et al. Biological activity of the Helicobacter pylori virulence factor CagA is determined by variation in the tyrosine phosphorylation sites. Proc Natl Acad Sci U S A. 2002;99:14428–33. doi: 10.1073/pnas.222375399. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sicinschi LA, Correa P, Bravo LE, Schneider BG. Detection and typing of Helicobacter pylori cagA/vacA genes by radioactive, one-step polymerase chain reaction in stool samples from children. J Microbiol Methods. 2003;52:197–207. doi: 10.1016/s0167-7012(02)00158-6. [DOI] [PubMed] [Google Scholar]
- 19.Sicinschi LA, Correa P, Bravo LE, et al. Non-invasive genotyping of Helicobacter pylori cagA, vacA, and hopQ from asymptomatic children. Helicobacter. 2012;17:96–106. doi: 10.1111/j.1523-5378.2011.00919.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Puz S, Innerhofer A, Ramharter M, et al. A novel noninvasive genotyping method of Helicobacter pylori using stool specimens. Gastroenterology. 2008;135:1543–51. doi: 10.1053/j.gastro.2008.08.006. [DOI] [PubMed] [Google Scholar]
- 21.Falsafi T, Favaedi R, Mahjoub F, Najafi M. Application of stool-PCR test for diagnosis of Helicobacter pylori infection in children. World J Gastroenterol. 2009;15:484–8. doi: 10.3748/wjg.15.484. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Hirai I, Sasaki T, Fujimoto S, Moriyama T, Azuma T, Yamamoto Y. A method for assessment of Helicobacter pylori genotype using stool specimens. FEMS Immunol Med Microbiol. 2009;56:63–6. doi: 10.1111/j.1574-695X.2009.00549.x. [DOI] [PubMed] [Google Scholar]
- 23.Hindson BJ, Ness KD, Masquelier DA, et al. High-throughput droplet digital PCR system for absolute quantitation of DNA copy number. Anal Chem. 2011;83:8604–10. doi: 10.1021/ac202028g. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Dingle TC, Sedlak RH, Cook L, Jerome KR. Tolerance of droplet-digital PCR vs real-time quantitative PCR to inhibitory substances. Clin Chem. 2013;59:1670–2. doi: 10.1373/clinchem.2013.211045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Sedlak RH, Kuypers J, Jerome KR. A multiplexed droplet digital PCR assay performs better than qPCR on inhibition prone samples. Diagn Microbiol Infect Dis. 2014;80:285–6. doi: 10.1016/j.diagmicrobio.2014.09.004. [DOI] [PubMed] [Google Scholar]
- 26.Cao Y, Raith MR, Griffith JF. Droplet digital PCR for simultaneous quantification of general and human-associated fecal indicators for water quality assessment. Water Res. 2015;70:337–49. doi: 10.1016/j.watres.2014.12.008. [DOI] [PubMed] [Google Scholar]
- 27.Wilson M, Glaser KC, Adams-Fish D, Boley M, Mayda M, Molestina RE. Development of droplet digital PCR for the detection of Babesia microti and Babesia duncani. Exp Parasitol. 2015;149:24–31. doi: 10.1016/j.exppara.2014.12.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Reid AL, Freeman JB, Millward M, Ziman M, Gray ES. Detection of BRAF-V600E and V600K in melanoma circulating tumour cells by droplet digital PCR. Clin Biochem. 2014 doi: 10.1016/j.clinbiochem.2014.12.007. [DOI] [PubMed] [Google Scholar]
- 29.Heredia NJ, Belgrader P, Wang S, et al. Droplet Digital PCR quantitation of HER2 expression in FFPE breast cancer samples. Methods. 2013;59:S20–3. doi: 10.1016/j.ymeth.2012.09.012. [DOI] [PubMed] [Google Scholar]
- 30.Cochran RL, Cravero K, Chu D, et al. Analysis of BRCA2 loss of heterozygosity in tumor tissue using droplet digital polymerase chain reaction. Hum Pathol. 2014;45:1546–50. doi: 10.1016/j.humpath.2014.03.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Porras C, Nodora J, Sexton R, et al. Epidemiology of Helicobacter pylori infection in six Latin American countries (SWOG Trial S0701) Cancer Causes Control. 2013;24:209–15. doi: 10.1007/s10552-012-0117-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Bertuccio P, Chatenoud L, Levi F, et al. Recent patterns in gastric cancer: a global overview. Int J Cancer. 2009;125:666–73. doi: 10.1002/ijc.24290. [DOI] [PubMed] [Google Scholar]
- 33.Olbermann P, Josenhans C, Moodley Y, et al. A global overview of the genetic and functional diversity in the Helicobacter pylori cag pathogenicity island. PLoS Genet. 2010;6:e1001069. doi: 10.1371/journal.pgen.1001069. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Talarico S, Gold BD, Fero J, et al. Pediatric Helicobacter pylori isolates display distinct gene coding capacities and virulence gene marker profiles. J Clin Microbiol. 2009;47:1680–8. doi: 10.1128/JCM.00273-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Tomb JF, White O, Kerlavage AR, et al. The complete genome sequence of the gastric pathogen Helicobacter pylori. Nature. 1997;388:539–47. doi: 10.1038/41483. [DOI] [PubMed] [Google Scholar]
- 36.Baltrus DA, Amieva MR, Covacci A, et al. The complete genome sequence of Helicobacter pylori strain G27. J Bacteriol. 2009;191:447–8. doi: 10.1128/JB.01416-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Testerman TL, Morris J. Beyond the stomach: an updated view of Helicobacter pylori pathogenesis, diagnosis, and treatment. World J Gastroenterol. 2014;20:12781–808. doi: 10.3748/wjg.v20.i36.12781. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Xia Y, Yamaoka Y, Zhu Q, Matha I, Gao X. A comprehensive sequence and disease correlation analyses for the C-terminal region of CagA protein of Helicobacter pylori. PLoS One. 2009;4:e7736. doi: 10.1371/journal.pone.0007736. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Molina-Castro SE, Herrera D, Malespin-Bendana W, Ramirez V, Une C. The geographic origin of Helicobacter pylori isolated from Costa Rican patients. Gut Microbes. 2014;5:517–21. doi: 10.4161/gmic.32148. [DOI] [PubMed] [Google Scholar]
- 40.Duncan SS, Valk PL, Shaffer CL, Bordenstein SR, Cover TL. J-Western forms of Helicobacter pylori cagA constitute a distinct phylogenetic group with a widespread geographic distribution. J Bacteriol. 2012;194:1593–604. doi: 10.1128/JB.06340-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Belda S, Saez J, Santibanez M, et al. Relationship between bacterial load, morbidity and cagA gene in patients infected by Helicobacter pylori. Clin Microbiol Infect. 2012;18:E251–3. doi: 10.1111/j.1469-0691.2012.03884.x. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Figure S1. ddPCR amplitude plots for the cagA gene detection assay (probe: cagA_FAM) and the cagA EPIYA typing assay (probes: EPIYA-C_HEX and EPIYA-D_FAM). H. pylori genomic DNA was spiked in to H. pylori-negative stool DNA. The H. pylori 16S copy number is indicated in parentheses after the name of each spiked strain (note: the H. pylori genome has two copies of the 16S gene and a single copy of the cagA gene). On the top are the amplitude plots for a North American H. pylori strain, Em47-1, which has a Western type cagA gene encoding an EPIYA-C motif. On the bottom are the amplitude plots for a Japanese H. pylori strain, Oki573, which has an East Asian cagA gene encoding an EPIYA-D motif. Each dot represents one droplet. Dots above the threshold (black line set at fluorescence amplitude of 6500 for the cagA gene detection assay and 2000 for the cagA EPIYA typing assay) are positive for the cagA gene. Number of cagA gene copies is indicated above each amplitude plot. Two 20 μl reactions were run for each sample and the results of each reaction were combined.