Analysis of transduction in wastewater bacterial populations by targeting the phage-derived 16S rRNA gene sequences

Authors


  • Editor: Julian Marchesi

Correspondence: Leonid A. Kulakov, School of Biological Sciences, Medical Biology Centre, The Queen's University of Belfast, 97 Lisburn Road, Belfast BT9 7BL, Northern Ireland. Tel.: +44 289 097 2799; fax: +44 289 097 5877; e-mail: l.kulakov@qub.ac.uk

Abstract

Bacterial 16S rRNA genes transduced by bacteriophages were identified and analyzed in order to estimate the extent of the bacteriophage-mediated horizontal gene transfer in the wastewater environment. For this purpose, phage and bacterial DNA was isolated from the oxidation tank of a municipal wastewater treatment plant. Phylogenetic analysis of the 16S rRNA gene sequences cloned from a phage metagenome revealed that bacteriophages transduce genetic material in several major groups of bacteria. The groups identified were as follows: Betaproteobacteria, Gammaproteobacteria, Alphaproteobacteria, Actinomycetales and Firmicutes. Analysis of the 16S rRNA gene sequences in the total bacterial DNA from the same sample revealed that several bacterial groups found in the oxidation tank were not present in the phage metagenome (e.g. Deltaproteobacteria, Nitrospira, Planctomycetes and many Actinobacteria genera). These results suggest that transduction in a wastewater environment occurs in several bacterial groups; however, not all species are equally involved into this process. The data also showed that a number of distinctive bacterial strains participate in transduction-mediated gene transfer within identified bacterial groupings. Denaturing gradient gel electrophoresis analysis confirmed that profiles of the transduced 16S rRNA gene sequences and those present in the whole microbial community show significant differences.

Introduction

Microbial communities inhabiting wastewater environments are of significant interest for applied as well as basic microbiology. These populations have been extensively studied for a number of years. However, only with the development of molecular and metagenomic approaches has it become possible to assess the true diversity of wastewater communities (Snaidr et al., 1997). These studies revealed the dominant bacterial groups (e.g. Betaproteobacteria) as well as the microbial diversity found in waste water (Wagner et al., 1993; Manz et al., 1994; Snaidr et al., 1997). Bacteriophages are the most abundant biological entities and virtually every bacterial strain isolated would have at least one phage strain that is able to attack it (Moebius & Nattkemper, 1981). Phages strongly influence bacterial populations by controlling their size, establishing lysogeny (temperate phages) and introducing genetic exchange via transduction processes. The crucial role of bacteriophages in the marine environment has been studied during the last decade (e.g. Williamson et al., 2002). The abundance of bacteriophages in the wastewater environment is also well-documented (Otawa et al., 2007) and their applied aspects, i.e. their use as bacterial indicators, has been addressed (Mandilara et al., 2006). However, despite their obvious importance, the interaction of phages in waste water with their bacterial hosts has not been studied sufficiently. The role of bacterial viruses in the evolution of bacterial communities has been demonstrated; Joo et al. (2006) observed and quantified the competitive advantage of lysogenic Bordetella strain. Transduction of genetic material by bacteriophages is a well-documented process in laboratory experiments. Generalized transduction, when phage erroneously package host's DNA, is suspected to be essential for a phage-mediated lateral gene transfer (Canchaya et al., 2003). Several recent studies have indicated a crucial role for bacteriophages in the dissemination of important genes in bacterial populations. The work of Tobe et al. (2006) demonstrated that ‘type III secretion in Escherichia coli is linked to a vast phage metagenome, acting as a crucible for the evolution of pathogenicity’. Recent analyses of viral metagenomes revealed that up to half of their contents originated from transduced bacterial and archaeal DNA sequences (Breitbart et al., 2002, 2004, 2008). Studies conducted 5 years ago have also demonstrated that 16S rRNA genes can be found in generalized transducing phages SN-T, UT1 and D3112; the bacteriophage SN-T can transduce ribosomal genes of Pseudomonas aeruginosa and Sphaeroticus natans (Beumer & Robinson, 2005). Temperate bacteriophage preparations obtained from soil samples appear to have a diverse set of 16S rRNA genes as well as genes responsible for dehalogenation of atrazine (Ghosh et al., 2008). It was indicated (Sander & Schmieger, 2001) that because a generalized transducing phage preparation can contain a library representing the entire bacterial host genome, it is possible to study transduction by analyzing 16S rRNA genes derived from bacteriophage preparations. This host-independent approach was used to demonstrate that Aeromonas and Acinetobacter species participate in transduction in samples obtained from a wastewater treatment plant (Sander & Schmieger, 2001). The most recent analysis of viral metagenome isolated from an activated sludge also confirmed the abundance of transduced genes, including 16S rRNA gene sequences (Parsley et al., 2010).

In this study, we analyzed and compared phage and total bacterial metagenomic DNA isolated from the aerobic part of a municipal wastewater treatment plant. We found the presence of a wide variety of 16S rRNA genes that define bacterial species participating in transduction-mediated horizontal gene transfer (HGT). We also demonstrate that this transfer involves a subpopulation of bacteria inhabiting the wastewater environment.

Materials and methods

Sample collection and purification of viral particles

Water samples were collected in August and September 2009 and in January 2010 from the oxidation tank of a municipal wastewater treatment plant located in Northern Ireland. The design population (capacity measured in units of human users) of this plant is 40 000 with a dry weather flow of 83 L s−1 and it serves one of the major cities in Ireland. Samples (15 L) were collected and transported to laboratory within 1 h, stored at 4 °C and processed within 24 h. Isolation and purification of viral (bacteriophage) particles was essentially performed according to the protocol described by Thurber et al. (2009). The samples collected did not require prefiltering. The Centramate tangential flow filtration (TFF) system (Pall Corporation) was used with a 0.2-μm TFF membrane for the removal of bacteria and a 100-kDa cut-off TFF membrane for concentration of viral particles. The pressure in the TFF system was kept below 10 p.s.i. Phage preparations were concentrated in a final volume of 15 mL and were further filtered using 0.2-μm syringe filters and stored at 4 °C. The purity from bacterial contamination was monitored at all stages by epifluorescence microscopy of the SYBR Gold (Invitrogen)-stained samples. Slides were prepared using 500 μL of samples from 0.2-μm filtered waste water and concentrated phage preparations. Staining was performed as described by Chen et al. (2001). Stained samples (20 fields) were examined using a Leica DMR-HR fluorescent microscope (Leica Microsystems UK Ltd., Milton Keynes, UK) within 12 h (stored at room temperature during this period).

Free bacterial DNA that could contaminate bacteriophage preparations was destroyed by DNase treatment (5 U of the enzyme per 1 mL of samples) as described by Thurber et al. (2009). The presence of free bacterial DNA was monitored by PCR with universal 16S rRNA gene primers (pA and pH, Table 1). DNase treatment was continued until no 16S rRNA gene sequences were detected in phage preparations. In control experiments, phage preparations were spiked with DNA of P. aeruginosa lytic phage phikF77 (accession number NC_012418; Kulakov et al., 2009) that is not present in waste water, and DNase treatment was performed on these samples followed by PCR detection with sequence-specific primers.

Table 1.   16S rRNA gene primers used in this work
PrimerSequence (5′–3′)DescriptionPosition*Reference
  • *

    Escherichia coli numbering for pA, pH, LK274, LK275, LK272, LK257, P1, P2 and P3.

pA (f)AGAGTTTGATCCTGGCTCAGUniversal8–27Pascual et al. (1995)
pH (r)AAGGAGGTGATCCAGCCGCAUniversal1541–1522Pascual et al. (1995)
LK274 (f)AGCAAACAGGATTAGATACCSequencing776–795Kulakov et al. (2002)
LK275GGCGTGGACTACCAGGGTASequencing810–792Kulakov et al. (2002)
LK272TGCCAGCAGCCGCGGTASequencing515–531Kulakov et al. (2002)
LK257TCGTTGCGGGACTTAACCSequencing1104–1087Kulakov et al. (2002)
P1 (f)CCTACGGGAGGCAGCAGDGGE341–358Muyzer et al. (1993)
P2 (r)ATTACCGCGGCTGCTGGDGGE516–534Muyzer et al. (1993)
P3 (f)CGCCCGCCGCGCGCGGCGGGCGGGGCGGGGGCACGGGGGGCCTACGGGAGGC AGCAGDGGE341–358Muyzer et al. (1993)
Cf1AACGGGTGGTAATGCCGAAA4 specific (R)1318–1300This work
Cr4ACCTTCCTCCGGCTTAGCACA4 specific (F)337–356This work
Ag2fCGAAGGGAAAACTGCATCTCB33 specific (F)509–528This work
Ag5rCATCCCTGCGGAACAACTCCB33 specific (R)1364–1345This work
AfGACAACGTTCCGAAAGGAGCA16 specific (F)134–153This work
A7rCCATCTCTGGAAAGTTCTCTGA16 specific (R)1006–986This work
A44fGGGTATCTAATCCTGTTCGCA44 specific (F)749–768This work
A44rCGGATATGACCAAAGGCTGA44 specific (R)1333–1351This work
A28FGCACTCCTCAATCTCTCAAGA28 specific (F)516–535This work
A28RTCGGAACGTACCCAGTCGTGA28 specific (R)1400–1419This work
A43FCTTAGAGTGCCCACCATAACA43 specific (F)390–409This work
A43RCAGTAAATTAATACTTTGCTGA43 specific (R)1068–1088This work
TA13FAGCCATGTGAACCCTACTAGTA13 specific (F)779–798This work
TA13RACGGGTAATACTCGATGATGTA13 specific (R)1339–1358This work
T35FGTTCGGAGGGAAAGCCACACT35 specific (F)174–193This work
T35RTCCGAACAACGCTTGCAACT35 specific (R)517–535This work
TA5FGCTTACTCTCACGAGCTTGTA5 specific (F)260–278This work
TA5RCTTAACCCAAAGCCTGCATCTA5 specific (R)901–920This work
TA15FTACTTGCTGGCAACATACGATA15 specific (F)408–427This work
TA15RAGTTGGTGAGGTAATGGCTCTA15 specific (R)1256–1275This work
T80FAGAGTGCCCAACTTAATGCTT80 specific (F)390–409This work
T80RCTGTGAGGCTCGAGTATGGAT80 specific (R)879–898This work

Extraction of DNA and nucleic acid manipulations

Phage metagenomic DNA was isolated from samples free of bacterial cells and external bacterial DNA. DNA extraction was performed using formamide and CTAB/NaCl as described by Casas & Rohwer (2007). Total wastewater DNA was isolated as described in Griffiths et al. (2000). Standard methods of DNA manipulation were used throughout this work (Sambrook et al., 1989). GFX PCR DNA and Gel Band Purification Kit (GE Healthcare, UK) were used for the recovery and purification of DNA fragments from agarose. PCR amplification of 16S rRNA gene sequences was performed as described previously (Kulakov et al., 2002). PCR and sequencing primers used in this work are listed in Table 1. DNA sequences of the 16S rRNA gene fragments were obtained using the University of Dundee Sequencing Service (http://www.dnaseq.co.uk/).

Denaturing gradient gel electrophoresis (DGGE) analysis of 16S rRNA genes

DGGE was performed as described previously (Myers et al., 1985) with the CBS Scientific system. Polyacrylamide gradient gels (6%) were formed as described by Muyzer et al. (1993) and a urea gradient range of 20–100% (100% denaturant contained 7 M urea and 40% of deionized formamide) was used. Electrophoresis was performed at a constant voltage of 90 V and a temperature of 60 °C in TAE buffer. After electrophoresis, the gels were stained in 1 × SYBR Gold (Invitrogen) for 40 min, visualized by UV transillumination (Versadoc, Biorad) and scanned. DGGE primers used in this work (Table 1) were described by Muyzer et al. (1993). For sequencing analysis, DGGE bands were cut out of the gel and incubated in 40 μL of water at 4 °C for 48 h. Samples (1 μL) were used as templates for PCR. Numerical analysis of the DGGE profiles has been performed in order to compare profiles obtained from different samples. For the DGGE profile analysis, we used several approaches to quantify the similarity between lanes: the Jaccard coefficient, which counts the number of matching bands between two lanes and divides it by the number of lane bands; the Dice similarity, similar to Jaccard with bands weighted twice; and the Ochiai coefficient, which minimizes the effect of different numbers of bands in various lanes. Analysis was performed using phoretix 1D v10.3 (TotalLab Ltd) software package.

Construction of 16S rRNA gene libraries and analysis

16S rRNA gene sequences were amplified from bacteriophage and total wastewater DNA preparations using universal 16S rRNA gene primers (Pascual et al., 1995). Amplification was performed with Pfu DNA polymerase (Fermentas) as described earlier (Kulakov et al., 2002). The following temperature profile was used: denaturation at 95 °C for 3 min, followed by 30 cycles of 94 °C for 40 s, 60 °C for 30 s and 72 °C for 1 min. Libraries of 16S rRNA genes were constructed by cloning PCR fragments into pJET1.2 vector (Fermentas). Cloning was performed using CloneJET PCR Cloning Kit (Fermentas) according to the manufacturers' instructions. Clones were isolated from each library and restriction fragment length polymorphism (RFLP) analysis was conducted using RsaI and HaeIII restriction enzymes (in separate experiments) as described by Scholten et al. (2005). Restriction profiles were analyzed by 1.5% agarose gel electrophoresis. Clones showing identical profiles were assigned to the same RFLP groups. At least 30% of the clones from each RFLP group were partially sequenced using pJET1.2 sequencing primers and complete nucleotide sequences of both strands were obtained for at least one clone of each type.

Phylogenetic analysis of the microbial community

Initial computer analysis of the sequences was performed using the dnasis (Hitachi) software package and sequences were assembled with codoncode aligner (Codon Code Corp.). Searches for nucleotide sequence similarities were carried out using the blast program (Altschul et al., 1990) in GenBank (blastn and megablast algorithms were used); the reference genomic sequences, nucleotide collection and NCBI Genome databases were used. Alignments of the sequences (>1350 nt) were performed using the clustalw algorithm (Thompson et al., 1994). Phylogenetic analysis of alignments and construction of trees were conducted with mega software version 4 (Tamura et al., 2007). Trees were generated by neighbor joining using the maximum composite likelihood model (Tamura & Nei, 1993). To evaluate tree phylogenies, these were also constructed by maximum parsimony and unweighted pair group method with arithmetic averages (UPGMA) methods. Phylogenies obtained by these methods were similar and the trees obtained by neighbor joining are presented (Fig. 1a and b). For bootstrap analysis, 1000 data sets were generated. To apply phylogenetic classification to the analyzed clones, closely related 16S rRNA gene reference sequences of characterized strains from GenBank were used; these were identified in blast searches.

Figure 1.

 Epifluorescence micrograph demonstrating phage particles from wastewater sample (September 2009) that are free from bacteria. Staining with SYBR Gold was conducted as described in Materials and methods.

Results and discussion

Detection of bacterial 16S rRNA genes in phage metagenomic DNA

Water samples were collected and total bacterial DNA was extracted as described in the Materials and methods. For the preparation of phage samples from waste water, removal of microbial cells was performed as described above. This was monitored by fluorescence microscopy and samples free from bacteria were obtained for further purification and analysis (Fig. 1). It was shown previously that decay of viral samples could occur at 4 °C (Helton et al., 2006). To minimize the possibility of decay, samples were processed within 24 h and phage preparations were stored in phage buffer, which significantly reduces possibility of phage decay/inactivation. Epifluorescence microscopy analysis of the final preparation did not detect noticeable changes in phage during the 2-month period. As it was essential to completely remove free bacterial DNA present in these samples, a DNase treatment was conducted and the absence of contaminating DNA was checked by PCR using universal 16S rRNA gene primers. No 16S rRNA gene amplification was detected after 10 min of treatment. Spiking of viral preparations with unrelated phage DNA has been used successfully before to evaluate their suitability for PCR studies (Allander et al., 2001). Treatment of the control samples with DNase after spiking with phikF77 DNA (Materials and methods) also confirmed that external DNA was successfully removed from the phage preparations. Accordingly, phage preparations that were free from external DNA were obtained. DNA was extracted from these preparations. PCR reactions with universal bacterial 16S rRNA gene primers were carried out using this DNA, and 16S rRNA gene fragments of 1.4 kbp were confirmed to be present in the phage metagenome. These fragments were isolated and used for cloning.

Construction and analysis of 16S rRNA gene libraries

16S rRNA gene fragments of 1.4 kbp were used for cloning into pJET1.2 plasmid. Two types of libraries were constructed from samples collected in September 2009: wastewater 16S rRNA gene phage-transduced library (WWP) and wastewater 16S rRNA gene total bacterial library (WWTB). Plasmids were isolated from 100 clones of WWP and 90 clones of WWTB libraries and all these clones were analyzed with respect to their RFLP profiles. At least one representative from each group was completely sequenced and one strand's sequence (with universal primers) was obtained for 30–50% of the clones. blastn searches were carried out for all sequences obtained. The results of these analyses are summarized in Table 2. Bacterial diversity detected in the analyzed wastewater sample generally corresponded to those reported for other sewage treatment plants, i.e. 16S rRNA gene sequences related to Proteobacteria, Actinobacteria and other groups usually found in this environment were present (Wagner & Loy, 2002; Miura et al., 2007). However, it was rather unexpected that Gammaproteobacteria were not found to be one of the major groups (WWTB library). It is worth noting that the wastewater community structure strongly depends on local factors such as influent wastewater composition (Miura et al., 2007).

Table 2.   Summary of analyses of wastewater 16S rRNA gene libraries
 Transduced 16S rRNA gene library (WWP)Total bacterial 16S rRNA gene library (WWTB)
ClonesRFLP typesSequenced
(complete sequence)
ClonesRFLP typesSequenced
(complete sequence)
  1. ND, not detected; although corresponding sequence were not detected among clones, PCR probing of WWTB identified sequences identical to those found in the WWP library.

Alphaproteobacteria2879 (4)1134 (3)
Betaproteobacteria1613 (3)2468 (4)
Gammaproteobacteria291317 (4)ND00
DeltaproteobacteriaND00312 (1)
Actinobacteria723 (1)4189 (8)
Firmicutes1424 (2)ND00
NitrospiraND00111 (1)
PlanctomycetesND00612 (1)

Using the results obtained from sequencing of almost complete 16S rRNA genes from the WWP, it became possible to infer a phylogenetic tree for bacterial species participating in HGT mediated by bacteriophages (Fig. 2a). As seen in this figure, there are three major groups of Proteobacteria (Alpha-, Beta- and Gamma-) present in the analyzed environment that have bacteriophages carrying out transduction. Gram-positive bacteria also have transducing bacteriophages in at least two groups: Actinobacteria species related to Rhodococcus and Firmicutes species related to Lysinibacillus.

Figure 2.

 Phylogenetic trees of 16S rRNA gene sequences amplified from the oxidation tank of a wastewater treatment plant. Trees were constructed by neighbor joining as described in Materials and methods and major groupings (e.g. Alphaproteobacteria) are highlighted by shading. GenBank accession numbers for clones TA25, T22, TA18, T10, TA15, TA20, TA5, T64, TA13, A21, B24, B4, A43, A32, A16, B33, A23, A10, A44, A7, TA28, T3, T80, T17, T35 and T30 are HM007524HM007549. Only one sequence was deposited when 16S rRNA gene sequence identity was >99% (e.g. clones A43 and A36; A44 and B33; A10 and A19, A21 and A28). 16S rRNA gene sequences of these clones were aligned with representatives of AlphaproteobacteriaRhizobium sp PF-M (accession number DQ202284), Methylobacterium populi BJ001 (NR_029082), Brevundimonas diminuta DSM1635 (X87274), Sphingomonas sp. B18 (AF410927); BetaproteobacteriaLeptothrix cholodnii SP-6 (CP001013), Azospira oryzae strain N1 (DQ863512), Zooglea ramigera 106 (NR_026130); GammaproteobacteriaPseudoxanthomonas mexicana AMX26B (NR_025105), Escherichia coli DH1 (P001637), Pseudomonas migulae AE2 (EF528261), Acinetobacter junii S33 (AB101444); Deltaproteobacteria, ActinobacteriaIlumatobacter fluminis (AB360343), Candidatus Microthrix parvicella (X89560), Geodermatophilus obscurus DSM 43160 (CP001867), Rhodococcus sp NSA6 (AB177885); FirmicutesLysinibacillus fusiformis WH22 (FJ418643); PlanctomycetesGemmata obscuriglobus (X56305); DeltaproteobacteriaHaliangium ochraceum DSM 14365 (CP001804). (a) Phylogenetic tree of sequences from bacteriophage metagenome. (b) Phylogenetic tree of sequences from total bacterial metagenome. Clones with identical sequences are shown in brackets. Scale bar at the bottom of figures shows estimated sequence divergence. Bootstrap values over 50% are shown at the nodes of the trees.

It is important to note that A21, A28 and B24 sequences (all related to Zooglea ramigera; Fig. 2a) have variability in three nucleotide positions. Two clones (A17 and A7) related to Lysinibacillus fusiformis differed by two deletions of one nucleotide each. Similarly, Brevundimonas-related A10 and A4 differed by one nucleotide in each position. These results suggest that different strains of the same species may be involved in transduction. Our current data, however, do not allow a conclusion to be drawn about whether these transductions are carried out by the same or different bacteriophages. Similarly, we are not able to identify at the moment whether transduction between different bacterial species or even distant strains of the same species takes place in the wastewater environment. A complete sequence of the WWP metagenome may be useful for answering these questions, especially if the specialized transducing phages (sequences) could be identified.

Bacterial 16S rRNA genes (WWTB) were analyzed similarly. A phylogenetic tree of the bacterial species found in wastewater is presented in Fig. 2b. It is notable that three out of eight major groups are present in both WWP and WWTB libraries as revealed by the analysis of the sequenced clones (Alpha- and Betaproteobacteria and Actinobacteria). It is also notable that Z. ramigera-related sequences are present in both libraries (Fig. 2a and b). Direct alignment of TA25, TA2 (WWTB), A21 and B24 (WWP) revealed a high level of sequence identity (over 99% identity; variability was found only in seven nucleotide positions) among sequences from two metagenomic libraries.

The results obtained indicate that transduction occurs in several major bacterial groups inhabiting wastewater; however, it is apparent that not all species are equally involved in transduction. In particular, Deltaproteobacteria, Firmicutes, Nitrospira and Planctomycetes were not found among the sequenced clones of the phage metagenomic library and Gammaproteobacteria among the WWTB clones. This result could be explained by the lower numbers of generalized transducing phages in these groups as the transduction of 16S rRNA genes is most likely to be associated with the availability of either temperate or virulent generalized transducing phages. To the best of our knowledge, there were no comparative studies of generalized transducing phage distribution for various bacterial groups. However, one approach for the assessment of transduction was suggested in the work of Sander & Schmieger (2001). The approach was based on the analysis of the transduced 16S rRNA gene sequences. It is interesting that although Actinobacteria was presented in the total wastewater library by five types of sequences (Fig. 2b), only Rhodococcus-related sequences were found in a phage-derived library (Fig. 2a). Similarly, several groupings belonging to Gammaproteobacteria that were found in WWP library (Pseudomonas, Acinetobacter, Rhizobium and Brevundimonas) were not detected among the analyzed clones of WWTB (Fig. 2a and b and Table 2).

Probing the corresponding metagenomes with strain-specific primers was used to identify 16S rRNA gene sequences not detected by the universal 16S rRNA gene primers. We designed sets of primers specific to clones B33 (related to Rhizobium), A16 (related to Acinetobacter) and A10 (Brevundimonas), A44 (Rhodococcus), A43 (Pseudomonas) and A28 (Zooglea) (Fig. 2a; Table 1). PCR analyses of WWP and WWTB metagenomes followed by the sequencing of the isolated fragments identified corresponding sequences in both metagenomes. The results indicate that although Rhizobium-, Acinetobacter- and Brevundimonas-related bacteria were not represented in the WWTB library, their bacteriophages are certainly active in transductional gene transfer. We also designed primers specific to strains found in WWTB, but not in WWP library; these were TA13 (Gemmata), T35 (Nitrospira), TA5 (Illumatobacter), TA15 (Geodematophilus) and T80 (Haliangium) (Table 1, Fig. 2b). PCR analysis of the WWP metagenomic DNA with these primers has not detected the corresponding sequences. This confirms results obtained from the analysis of WWTB and WWP libraries that above strains indeed do not participate in HGT by transduction. We were not able to perform similar analysis for Firmicutes as primers based on A7 and A17 strains produced products apparently containing multiple sequences.

The strain-specific primers (Table 1) were also used for the PCR analysis of the samples collected in August 2009 and January 2010. The results of these experiments (Table 3) indicate that the same bacterial strains participated in transductional HGT in both August and September 2009 water samples. Analysis of WWTB and WWP DNA isolated from the January 2010 sample detected only two sequences homologous to those found in September 2010 (B33, Rhizobium and A28, Zooglea; Table 3).

Table 3.   PCR analysis of metagenomic DNA using strain-specific primers
Primer pairStrainMetagenome
TB (A)PH (A)TB (S)PH (S)TB (J)PH (J)
  1. TB, bacterial metagenomes isolated August 2009 (A), September 2009 (S) and January 2010 (J); PH, phage metagenomes isolated August 2009 (A), September 2009 (S) and January 2010 (J); Designation of primers and strains are as in Table 1 and Fig. 1a and b.

  2. +, PCR product of the corresponding gene is detected; −, no specific product.

Cf1; Cr4A4++++
Ag2f; Ag5rB33++++++
Af; A7rA16++++
A44f; A44rA44++++
A28f; A28rA28++++++
A43f; A43rA43++++
TA13F;TA13rTA13++
T35F; T35RT35++
TA5F; TA5rTA5++
TA15f;TA15rTA15++
T80F; T80rT80++

DGGE analysis of 16S rRNA gene transduction in wastewater community

DGGE analyses of 16S rRNA genes from both WWP and WWTB metagenomes were conducted to further analyze transduction in wastewater environment. Cloned 16S rRNA gene sequences from WWP and WWTB libraries were used as reference markers in these experiments. A resulting DGGE gel is presented in Fig. 3. The profiles obtained show significant differences between 16S rRNA gene sequences in WWP and WWTB metagenomes. Bands A, B and C corresponding to clones A7 (Lysinibacillus), A36 (Pseudomonas) and A16 (Acinetobacter) are visible only in the WWP lane. Bands corresponding to A7 and A16 were isolated and analyzed as described in Materials and methods, and sequences identical to those of A7 and A16 clones were confirmed. It is important to note that these sequences were identified in the WWP clone library but not in the WWTB library, which suggested active transductional gene transfer among the corresponding bacterial species. However, the A16 sequence was detected in the bacterial metagenome by PCR with strain-specific primers. Therefore, we assume that these species, although minor in wastewater environment, have a significant population of transducing bacteriophages that makes their 16S rRNA genes more abundant in the WWP metagenome. Bands E and F corresponding to Actinobacteria sequences, found only among WWTB clones, are similarly visible in the WWTB lane only. We suggest that these species either have a very limited population of generalized transducing phages or perhaps do not have such phages at all. It is important to note that several common bands are still clearly detectable in both metagenomes (Fig. 3). In particular, band D corresponds to Zooglea-related sequences (clone B24) that were found in both metagenomes. The analysis of this band confirmed that indeed it has a DNA fragment with the sequence identical to the B24 clone. Clones with Zooglea-related sequences were found in both cloning libraries (Fig. 2a and b). In general, DGGE data strongly corroborate results obtained from the sequencing analysis of the two libraries and corroborates our conclusion that transduction in wastewater is biased towards some bacterial species.

Figure 3.

 DGGE analysis of 16S rRNA gene sequences amplified from wastewater samples. PCR amplification was performed as described in Materials and methods. Samples were applied onto 6% polyacrylamide denaturing gradient gels. Gradients were formed and electrophoresis was performed at a constant voltage of 90 V and a temperature of 60°C in TAE buffer. Lanes: 1, WWTB clone T17, Actinobacteria; 2, WWTB clone TA5, Actinobacteria; 3, WWTB DNA (total bacterial DNA from wastewater); 4, WWP DNA (phage metagenome DNA from the same wastewater sample); 5, WWP clone A7, Lysinibacillus; 6, WWP clone A16, Acinetobacter; 7, WWP clone A32, Escherichia coli; 8, WWP clone A4, Brevundimonas; 9, WWP clone A36, Pseudomonas; 10, WWP clone B24, Zooglea. A-F, indicate bands analyzed by sequencing.

To further investigate the above conclusions, 16S rRNA gene DGGE profiles of total bacterial and phage metagenomic DNA from three different wastewater samples (August and September 2009 and January 2010) were compared. All dendrograms produced showed the same clustering with very high cophenetic correlation coefficients, which measure the consistency of a cluster compared with the original similarity matrix. The results of analysis using UPGMA clustering and Ochiai Coefficient similarity is presented in Fig. 4. This clearly showed that DGGE banding patterns resulted in two clusters representing transduced 16S rRNA gene sequences and 16S rRNA gene sequences of total bacterial metagenome. It is also notable that patterns obtained from August and September samples are much more closely related to each other than to those from the January 2010 sample.

Figure 4.

 Cluster analysis of 16S rRNA gene DGGE profiles. Analysis was performed using the UPGMA clustering and Ochiai coefficient-based similarities with cophenetic correlation coefficients shown at the nodes. The pairwise comparison is based on 22 matching bands.

In summary, in this work, we investigated the 16S rRNA gene transfer mediated by bacteriophage transduction in wastewater treatment processes. We exploited the fact that generalized transducing phages are able to include any part of the host genome into their capsid. Analysis of 16S rRNA genes derived from bacteriophage particles and their comparison with the total bacterial DNA metagenome allowed identification of bacteria involved in transduction. The results presented demonstrate that transduction occurs in several major bacterial groups inhabiting wastewater (Proteobacteria, Firmicutes and Rhodococcus). At the same time, many bacterial groups found in the aeration tank were not found in the phage metagenome (e.g. Deltaproteobacteria, Nitrospira, Planctomycetes and most of Actinobacteria genera). The DGGE profiles of the metagenomes compared also had significant differences. Our results suggest that transduction is widespread in wastewater. A defined subpopulation of bacteria inhabiting this environment apparently participates in this type of HGT. It is worth noting that the analysis of viral metagenome could be biased by such factors as phage purification/concentration techniques and PCR primers used. In this work, we used widely accepted techniques of phage purification that were successfully applied to many environments (Thurber et al., 2009). The same 16S rRNA gene primers were used to analyze both WWP and WWTB libraries, which ensures the consistency of the results obtained.

Acknowledgements

This work was supported by the QUESTOR Centre (Queens University of Belfast). We are especially grateful to Mr Ciaran Prunty for the expert help in collecting wastewater samples and to Dr Dave Lipscomb for the help in conducting some experiments.

Ancillary