ERIC-PCR fingerprinting-based community DNA hybridization to pinpoint genome-specific fragments as molecular markers to identify and track populations common to healthy human guts

Abstract

Bacterial populations common to healthy human guts may play important roles in human health. A new strategy for discovering genomic sequences as markers for these bacteria was developed using Enterobacterial Repetitive Intergenic Consensus (ERIC)-PCR fingerprinting. Structural features within microbial communities are compared with ERIC-PCR followed by DNA hybridization to identify genomic fragments shared by samples from healthy human individuals. ERIC-PCR profiles of fecal samples from 12 diseased or healthy human and piglet subjects demonstrated stable, unique banding patterns for each individual tested. Sequence homology of DNA fragments in bands of identical size was examined between samples by hybridization under high stringency conditions with DIG-labeled ERIC-PCR products derived from the fecal sample of one healthy child. Comparative analysis of the hybridization profiles with the original agarose fingerprints identified three predominant bands as signatures for populations associated with healthy human guts with sizes of 500, 800 and 1000 bp. Clone library profiling of the three bands produced 17 genome fragments, three of which showed high similarity only with regions of the Bacteroides thetaiotaomicron genome, while the remainder were orphan sequences. Association of these sequences with healthy guts was validated by sequence-selective PCR experiments, which showed that a single fragment was present in all 32 healthy humans and 13 healthy piglets tested. Two fragments were present in the healthy human group and in 18 children with non-infectious diarrhea but not in eight children with infectious diarrhea. Genome fragments identified with this novel strategy may be used as genome-specific markers for dynamic monitoring and sequence-guided isolation of functionally important bacterial populations in complex communities such as human gut microflora.

Introduction

A large number of microbial species exist in the human gastrointestinal tract. Intestinal microflora play important roles in digestive function, immunity and disease resistance. However, the relationship between gut flora and human health remains largely unknown due to the complexity of the system (Hooper and Gordon, 2001, Mackie et al., 2000). The structure of any microbial community is central to the system's function (Dahllof, 2002). One important strategy for elucidating the structure–function relationship of microbial communities involves long-term, systematic monitoring of changes associated with functional dynamics of a community, thus leading to a better understanding of a community's ecological function(s) (Bond et al., 1995, Borneman and Triplett, 1997, Watanabe and Hino, 1996, Watanabe et al., 1998). This can be achieved by comparing structural features of communities to identify functionally important populations. However, traditional culture-based techniques are inappropriate in studying the structure–function relationship of microbial communities due to their highly selective, laborious and time-consuming nature. Only a relatively small fraction of population members in a complex natural community, such as human gastrointestinal system (Amann et al., 1995, Langendijk et al., 1995, Muyzer et al., 1993), can be recovered with culture-based techniques.

DNA-based technology has advanced the characterization of microbial structural features. Microbial communities can be regarded as a mixture of microbial genomes. Genomic DNA sequences and their copy numbers are a faithful reflection of community structure. Community structure has been defined as the amount and distribution of (genomic) information in a particular habitat (Torsvik et al., 2002). Phylogenetically meaningful sequences, such as small subunit ribosomal RNA genes, conserved functional genes and randomly amplified genome fragments, have been used as molecular markers in analyzing microbial communities with different technical strategies, such as clone library profiling, genetic fingerprinting and molecular hybridization (Franklin et al., 1999, Hill et al., 2002, Rotthauwe et al., 1997, Suau et al., 1999, Wikstroim et al., 1999, Wu et al., 2001). The genomic DNA-based analysis has overcome limitations of conventional culture-based technology and provided the most powerful tool for researching structure–function relationships of microbial communities.

The structural dynamics of microbial communities have been monitored using amplified sequences or total genomic DNA and a variety of genetic fingerprinting techniques, such as Terminal Fragment Length Polymorphism (T-RFLP), Amplified Ribosome DNA Restriction Analysis (ARDRA), Denaturing Gradient Gel Electrophoresis/Temperature Gradient Gel Electrophoresis (DGGE/TGGE), of amplified partial 16S/18S rDNAs or fragments of conserved functional genes and Random Amplified Polymorphic DNA (RAPD), or Arbitrarily Primed PCR (AP-PCR) (Liu et al., 1997, Marsh, 1999, Muyzer et al., 1993, Norris et al., 2002, Simpson et al., 1999, Tannock, 2002, Weidner et al., 1996, Zoetendal et al., 1998). These techniques provide genomic patterns or profiles of microbial communities, in which the number of bands reflects the number of predominant community members, while the intensity of bands theoretically reflects population levels. However, bands with identical positions in TGGE/DGGE or RAPD gels may contain different DNA sequences (Jackson et al., 2000, Sekiguchi et al., 2001), making it difficult to compare structural features of different community samples based solely on the banding patterns. Thus, the scoring of different samples using PCR-TGGE/DGGE or RAPD profiles may exaggerate the similarity between the samples.

In this study, we developed a new strategy to simultaneously compare sequence-based structural features of large numbers of community samples. Enterobacterial Repetitive Intergenic Consensus (ERIC)-PCR was used to fingerprint the microbial community of fecal samples of research subjects. ERIC-PCR profiles were transfer blotted onto nylon film to form an array-like organization made of amplified genomic DNA fragments distributed to reflect community structural differences. All ERIC-PCR amplicons from one community sample were DIG-labeled to hybridize with the community DNA arrays. DNA bands in the fingerprints sharing sequence homology with the probes would develop signals, while bands with no sequence homology in the probes would be “erased” from the fingerprints. This led to a straightforward identification of the common bands containing homologous sequences among all the samples. If the appearance of these bands in the community samples were associated with a certain functional status, cloning and sequencing of DNA fragments in these bands would provide the basis for designing specific primers or probes for dynamic monitoring and sequence-guided isolation of the corresponding functional populations.