OUP user menu

Rapid molecular identification and subtyping of Helicobacter pylori by pyrosequencing of the 16S rDNA variable V1 and V3 regions

Hans-Jürg Monstein , Shohreh Nikpour-Badr , Jon Jonasson
DOI: http://dx.doi.org/10.1111/j.1574-6968.2001.tb10658.x 103-107 First published online: 1 May 2001

Abstract

We describe here the use of real-time DNA sequence analysis of Helicobacter pylori 16S rRNA gene fragments by pyrosequencing™ for rapid molecular identification and subtyping of clinical isolates based on DNA sequence heterogeneity within the variable V1 and V3 regions. Six individual 16S rDNA V1 alleles (position 75–100) were identified in 23 clinical isolates obtained from gastric biopsy specimens. Eleven of these revealed sequence identities with H. pylori 26695 and one was identical with the rrn genes in strain J99. The other V1 alleles showing single or double nucleotide mutations or single nucleotide insertions could be divided into four groups with 5, 4, 1, and 1 isolates each. Two out of 25 isolates demonstrated single C to T transitions in the V3 region (position 990–1020). The present findings show that subtle DNA sequence variation occurs sufficiently often in the 16S rDNA variable V1 and V3 regions of H. pylori to provide a consistent system for subtyping.

Keywords
  • 16S rDNA PCR
  • Heterogeneity
  • Bacterial typing
  • Pyrosequencing
  • Helicobacter pylori

1 Introduction

Helicobacter pylori is a micro-aerophilic Gram-negative bacterium associated with chronic gastritis [1], peptic ulcer disease [2], and gastric cancer [3]. Initial classification of this bacterium was based on its morphology, biochemical characteristics and growth requirements [1]. In recent years, researchers have tended to use genetic criteria including virulence gene and 16S rRNA gene sequencing for distinguishing it from other curved Gram-negative rods, many of which are difficult to isolate and propagate in culture [4]. Extensive genetic variation has been reported, and by partial 16S rDNA sequence analysis we and others have found sequence variation in the variable V3 and V4 regions in Helicobacter isolates from gastric biopsy specimens as compared with the H. pylori CCUG 17874T type strain [5,6]. Further investigation on this matter seems warranted. However, even if DNA sequence analysis is increasingly used for research purposes it is still considered too costly and time-consuming for use in large-scale molecular identification of microorganisms in a routine clinical diagnostic laboratory setting.

Recently, the principle of pyrosequencing™, a real-time DNA sequence analysis of short DNA stretches, was described [7]. A commercially available instrument has shown to be a powerful tool in human genetics for the identification of single nucleotide polymorphism (SNP) [8]. It has also been applied for the analysis of viral drug resistance genes [9]. We describe here the possible use of pyrosequencing as a means for extremely rapid large-scale molecular identification and subtyping of H. pylori isolates in a clinical setting. The technique could also be suitable for the detection and epidemiological typing of H. pylori in environmental samples.

2 Materials and methods

2.1 Bacterial strains and DNA extraction

The H. pylori reference collection of clinical isolates used in this study (HP-HJM 1–25) had originally been obtained from routine clinical dyspeptic gastric biopsy specimens (mixed age and gender) at the University Hospital, Linköping, Sweden. Reference strains H. pylori 26695 (CCUG 41936) and H. pylori J99 were obtained from the Culture Collection University of Gothenburg, Sweden, and Dr. L. Engstrand, SMI Stockholm, Sweden, respectively. Bacteria were cultured as described elsewhere [10]. Genomic DNA from the H. pylori strains was prepared using a commercially available DNA extraction kit (QIAamp tissue kit, Qiagen, KEBO, Stockholm, Sweden) as described [11].

2.2 In vitro amplification of the 16S rRNA gene

Primers used in this study (Table 1) were obtained from Amersham-Pharmacia Biotech Norden (Sollentuna, Sweden) or Scandinavian Gene Synthesis (Köping, Sweden). Sequential amplification of the 16S rRNA gene was performed using two sets of primer pairs and Ready-To-Go® PCR beads (Amersham-Pharmacia Biotech). The 16S rDNA variable V1 region was amplified using primers bio-pBR-5′/se (10 pmol) and pBR-V1/as (10 pmol), and the variable V3 region was amplified using primers bio-pJB-se and HP-V3T/as, respectively. PCR amplification was carried out in a thermal controller PTC-100™ (MJ Research Inc., SDS-Falkenberg) using 2 μl of DNA extract and a final volume of 25 μl as follows: denaturation step at 94°C for 2 min (1 cycle); followed by denaturation at 94°C for 40 s, annealing at 55°C for 40 s, extension at 72°C for 1 min (25 cycles) and a final extension step at 72°C for 10 min. Subsequently, PCR-amplified products (5 μl) were analyzed by agarose gel electrophoresis [10]. The expected sizes for the V1 and V3 amplicons were approximately 110 bp and 85 bp, respectively.

View this table:
1

Primers used for PCR amplification and pyrosequencing

2.3 Pyrosequencing

Twenty μl of biotinylated V1 and V3 amplicons, respectively, were mixed with 25 μl of 2×BW buffer (10 mM Tris–HCl, 2 M NaCl, 1 mM EDTA and 0.1% Tween 20, pH 7.6) and 10 μl Dynabeads (Dynabeads® M280–streptavidin), and immobilized by incubation at 65°C for 15 min (shaking). Single stranded DNA was obtained by incubation (1 min) of the captured biotin–streptavidin complex (magnetic beads) in 50 μl of 0.50 M NaOH (each well), using a PSQ 96 Sample Prep Tool (Pyrosequencing AB, Uppsala, Sweden). Subsequently, each sample (well) was washed with 100 μl 1×annealing buffer (200 mM Tris–acetate and 50 mM Mg–acetate). pBR-V1/as (V1 region) and HP-V3T/as (V3 region of the type strain H. pylori CCUG 17874T), respectively, were also used as sequencing primers and hybridized to the single stranded PCR products. For that purpose, 1 μl of sequencing primer (15 pmol) was incubated in 44 μl of annealing buffer (each well) at 80°C for 2 min, followed by cooling to room temperature. Pyrosequencing was performed using a SNP Reagent Kit (enzyme and substrate mixture, dATP-S, dCTP, dGTP, and dTTP) as provided by the manufacturer (Pyrosequencing AB, Uppsala, Sweden).

3 Results and discussion

Our current picture of the phylogenetic relationships of bacteria derives to a large extent from DNA sequence analysis of the ribosomal small subunit 16S rRNA genes. Tens of thousands such molecules have been catalogued with sequences, structures and taxonomy in public molecular databases, e.g. GenBank at NCBI (http://www.ncbi.nlm.nih.gov/). These data can advantageously be used for identifying unknown bacteria by 16S rRNA gene sequencing. The conventional DNA sequencing procedure involves extraction of nucleic acids, PCR-mediated amplification of a 16S rRNA gene fragment, a cycle sequencing reaction, and electrophoresis or chromatography. The time and effort associated with such analysis and the cost are major limitations.

The present study describes a new approach for rapid molecular identification and subtyping of H. pylori isolates by pyrosequencing and signature matching of PCR-amplified variable regions within the 16S rDNA. The main reasons for using the pyrosequencing method instead of conventional DNA sequencing for this study were capacity, speed and convenience.

Partial sequences within the variable V1 and V3 regions were obtained from 25 strains of a H. pylori reference collection of clinical isolates and two reference strains (H. pylori 26695 and J99, respectively). One set of two primers was used for each locus (Table 1). Based on nucleotide sequences within the variable V1 region between positions 75 and 100, the 25 clinical isolates could be divided into six different lineages (Fig. 1). The corresponding pyrograms are shown in Fig. 2. Lineage A comprising 11 isolates (HP-HJM 2, 3, 5, 7, 8, 9, 13, 19, 20, 21, 25) had a sequence that was identical with that of H. pylori 26695 (Fig. 1). Single or double nucleotide mutations were observed in lineages B (HP-HJM 1, 4, 14, 18, 22), C (HP-HJM 11, 15, 17, 23) and D (HP-HJM 24) as compared with the H. pylori 26695 sequence (Fig. 1). A single nucleotide insertion was present in lineage E (HP-HJM 10). Lineage F (HP-HJM 6), which differed significantly in the V1 region from the other isolates, demonstrated DNA sequence identity with the corresponding region of reference strain H. pylori J99 (Fig. 1).

1

Upper panel: Sequence alignment of the 16S rDNA variable V1 region of H. pylori isolates HP-HJM 1–25 and reference strains H. pylori 26695 and J99. Gaps indicate deletions, and dashes indicate positions at which the sequences were homologous to that of reference strain H. pylori 26695. Lineages A to F indicate six individual 16S rDNA V1 alleles (signature sequences) at positions 75–99 (Escherichia coli nomenclature [12]). The 16S rDNA broad-range sequencing primer pBR-V1/as corresponds to a consensus sequence between positions 120 and 100 of many clinically important bacteria. Lower panel: Sequence alignment of the variable V3 region of H. pylori isolates HP-HJM 1–25, reference strain H. pylori 26695 (AE000620/644), H. pylori J99 (AE001534/56), and the type strain H. pylori CCUG 17878T (U01331 and [14]). Gaps indicate deletions, and dashes indicate DNA sequence homologies compared to the type strain. The HP-V3T/as sequencing primer corresponds to the sequence of type strain H. pylori CCUG 17874T. For clarity, the corresponding sequences of H. pylori-related strains Helicobacter heilmanii (Y18028), H. bilis (AF047847), H. hepaticus (L39122) and Helicobacter cholecystus (U46129) are included.

2

Pyrosequencing of the 16S rDNA variable V1 region of H. pylori isolates performed as described in the text with cyclic dispensation of the nucleotides (dispensation order: ACGT). Each pyrogram represents an individual H. pylori lineage (A–F). The corresponding nucleotide signature sequences as interpreted by a custom-made application program are shown in Fig. 1 (upper panel).

All isolates, except HP-HJM 10 and HP-HJM 21, revealed sequence identity in the V3 region (pyrograms not shown) with H. pylori CCUG 17874T, H. pylori 26695, and H. pylori J99 (Fig. 1). HP-HJM 10 and HP-HJM 21 (lineages B and A, respectively, in the V1 region) demonstrated a single C to T transition (Fig. 1).

The short 25–30 nt DNA sequence obtained for each isolate and region was used as a ‘signature’ of the 16S rDNA of the particular isolate, which thus gained multiple signature attributes. The uniqueness of each signature was investigated by matching it against a ‘signature template’ consisting of all catalogued bacterial 16S rDNA sequences available at NCBI using the BLAST advanced option tools including taxonomy and lineage reports (details not given). Considering the information content of each signature, the fact that only certain residues are variable and therefore informative implies a straightforward exclusion analysis if only perfect matches are taken into account.

The primer HP-V3T/as used for sequencing between position 990 and 1020 of the V3 region was designed based on the H. pylori type strain CCUG 17874T sequence (U01331). The Tax BLAST Lineage Report indicated specificity for Helicobacter group (epsilon subdivision of proteobacteria). Therefore, when HP-V3T/as is used as a primer in PCR, DNA from other microorganisms should in all likelihood not yield a PCR product under stringent conditions. Verification of the actual strain being a member of the species H. pylori was obtained for 23/25 isolates through the criteria of signature matching in the V3 region, disregarding the non-human Helicobacter nemestrinae (Fig. 1).

The primer pBR-V1/as used for sequencing between position 75 and 100 of the V1 region was designed as a broad-range primer based on conserved residues appearing in most clinically important eubacteria. The sequencing of the V1 segment was primarily aimed at allocating the actual strain to a certain lineage. However, despite the DNA sequence variation in this region, lineages A to E were tentatively identified as H. pylori also by signature matching of the V1 region allowing for one or two mismatches in those cases where the signature was unknown to the database. The H. pylori J99 and lineage F signatures of the V1 region matched with Helicobacter spp. such as Helicobacter bilis, Helicobacter hepaticus, Helicobacter canadensis, Helicobacter cinaedi, Helicobacter rappini, Helicobacter mustelae, and also with Campylobacter jejuni (Fig. 1).

In conclusion, the present findings show that subtle DNA sequence variation does occur in the 16S rDNA variable V1 and V3 regions of H. pylori, which provides a consistent system for subtyping. The PSQ96 automated system allows for rapid (ca. 30 min) determination of 20–30 nt of target sequences dispensed in 96-well microtiter plates. From the system output, information on nucleotide sequences could easily be extracted for automatic evaluation using a simple algorithm and a local 16S rDNA position-based database (not shown). An important feature of pyrosequencing is that the technique is robust, very rapid and allows for the simultaneous analysis of a large number of quite different samples, which is often needed in clinical microbiology laboratories.

Acknowledgements

The LMÖ-Molecular Biology Program, University Hospital, Linköping supported this work. The technical assistance of Margaretha Olofsson is gratefully acknowledged.

References

  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
View Abstract