OUP user menu

Vertical distribution and diversity of sulfate-reducing prokaryotes in the Pearl River estuarine sediments, Southern China

Lijing Jiang, Yanping Zheng, Xiaotong Peng, Huaiyang Zhou, Chuanlun Zhang, Xiang Xiao, Fengping Wang
DOI: http://dx.doi.org/10.1111/j.1574-6941.2009.00758.x 249-262 First published online: 8 October 2009

Abstract

The vertical distribution and diversity of sulfate-reducing prokaryotes (SRPs) in a sediment core from the Pearl River Estuary was reported for the first time. The profiles of methane and sulfate concentrations along the sediment core indicated processes of methane production/oxidation and sulfate reduction. Phospholipid fatty acids analysis suggested that sulfur-oxidizing bacteria (SOB) might be abundant in the upper layers, while SRPs might be distributed throughout the sediment core. Quantitative competitive-PCR analysis indicated that the ratios of SRPs to total bacteria in the sediment core varied from around 2–20%. Four dissimilatory sulfite reductase (dsrAB) gene libraries were constructed and analyzed for the top layer (0–6 cm), middle layer (18–24 cm), bottom layer (44–50 cm) and the sulfate-methane transition zone (32–42 cm) sediments. Most of the retrieved dsrAB sequences (80.9%) had low sequence similarity with known SRP sequences and formed deeply branching dsrAB lineages. Meanwhile, bacterial 16S rRNA gene analysis revealed that members of the Proteobacteria were predominant in these sediments. Putative SRPs within Desulfobacteriaceae, Syntrophaceae and Desulfobulbaceae of Deltaproteobacteria, and putative SOB within Epsilonproteobacteria were detected by the 16S rRNA gene analysis. Results of this study suggested a variety of novel SRPs in the Pearl River Estuary sediments.

Keywords
  • dsrAB gene
  • 16S rRNA gene
  • QC-PCR
  • phylogenetic analysis
  • phospholipid fatty acids
  • estuarine sediments

Introduction

The Pearl River Estuary (also called Ling Ding Yang) is located in the subtropical region in Guangdong Province, Southern China. The Pearl River Estuary represents one of the most important ecosystems linking the highly developing land area and the South China Sea (SCS) (Cai et al., 2004). It is the second largest estuary in China and is heavily populated. In recent years, the Pearl River has a high load of anthropogenic nutrients from increased agricultural activities, fish dike farming and wastewater runoff due to the increase in population and economic development in Southern China and the Pearl River delta region (Huang et al., 2003). A wide variety of chemical compounds, such as chlorinated and polycyclic aromatic hydrocarbons (polybrominated diphenyl ethers, alkylphenols and organochlorine pesticides), have been detected, generating serious environmental problems (Mai et al., 2002; Zhang et al., 2002b; Zheng et al., 2004; Chen et al., 2005a; Chau, 2006). Many of these chemicals ultimately reside in sediments, where sequestration decreases their bioavailability for microbial transformation and anaerobic processes become the main route for remediation (Eggleton & Thomas, 2004). Dissimilatory sulfate reduction is considered to be the main anaerobic process in the biomineralization of organic matter in the anoxic environment, accounting for up to 50% of its degradation in marine sediments (Jorgensen, 1982). Sulfate-reducing prokaryotes (SRPs) are capable of oxidizing a wide variety of organic carbon compounds through sulfate reduction. These compounds may include short-chain fatty acids (such as acetate, lactate, pyruvate and ethanol), long-chain fatty acids and aromatic compounds (such as benzoate, phenol) that are prevalent in anaerobic environments. In anoxic environments, such as estuarine sediments, SRPs are major contributors to carbon and sulfur cycles (Nealson, 1997). Therefore, the diversity, distribution and physiology of SRPs in various environments have been intensively investigated (Dhillon et al., 2003; Nakagawa et al., 2004; Leloup et al., 2006; Foti et al., 2007; Leloup et al., 2007; Liu et al., 2009). In recent years, the genes encoding the α and β subunits of the siroheme dissimilatory (bi)sulfite reductase (dsrAB) have been found in all known SRPs, including five bacterial and two archaeal phyla. The overall contents of the 16S rRNA gene and dsrAB gene phylogenetic tree suggested vertical transmission of the genes in the evolution of SRPs, but lateral gene transfer events have also been detected (Klein et al., 2001; Zverlov et al., 2005).

In the Pearl River Estuary, geochemical investigations have indicated strong sulfate reduction, methane oxidation and methanogenesis in the sediments (Wu et al., 2006); however, the microorganisms involved in these processes have not been identified or investigated yet and their roles in the biogeochemical cycling of carbon and sulfur in this large estuary and the adjacent SCS are largely unknown. The major objectives of the present study were: (1) to reveal the diversity and abundance of SRPs in the Pearl River Estuary environment; (2) to resolve the vertical distribution profile of SRPs by targeting the functional dsrAB gene; (3) to find putative novel SRPs from the Pearl River Estuary; (4) to reveal the bacterial diversity in the sediments by 16S rRNA gene analysis; and (5) to infer the putative roles of SRPs and related bacteria in the biogeochemical transformation of carbon and sulfur in the estuarine environment. This paper is an important first step toward revealing the bacterial community and its roles in the Pearl River Estuary environments.

Materials and methods

Sample collection and DNA extraction

Sediment samples were collected from the top 50 cm using a single-core sampler at the Qi'ao Island (Pearl River Estuary) (113°38′07.3E, 22°27′21.4N) in the Guangdong province, China (Fig. 1). The water depth was 0.5 m. The bottom water temperature was 21.5 °C and the salinity at the sediment surface was 26‰. The sediment core was sectioned into 2-cm slices to a depth of 50 cm and then transferred to sterile Falcon tubes on a clean bench. Samples were stored aseptically at −20 °C until analysis.

1

Map of the sampling site in the Pearl River Estuary. The sampling station for sediment core QA0504-7 is indicated by (▪).

The sediment core was separated into four zones: top (0–6 cm), middle (18–24 cm), bottom (44–50 cm) and the sulfate-methane transition zone (SMTZ, 32–42 cm). Similar amounts of sediments in each 2-cm slice were collected from each interval and used for DNA extraction as described earlier (Xu et al., 2007).

Sulfate and methane concentrations

For sulfate analysis, sediment pore water was collected by centrifugation (6000 g, 15 min) followed by pressure filtration through a 0.45-μm membrane filter. Filtered pore water samples were preserved in a 20% (w/v) zinc acetate solution and either analyzed immediately or frozen for later analysis. Sulfate was first purified through an IonPac AS19 Analytical Column (4 × 250 mm) using sodium carbonate (9 mM) as an eluent. Its concentration was then determined by ion chromatography (Dionex DX-600). Methane concentrations were measured according to the methods of Jorgensen et al. (2001). Briefly, 3-mL sediments were transferred using a syringe to 24-mL Bellco anaerobic tubes (Bellco Glass Inc., Vineland, NJ) containing 6 mL of 1 M NaOH solution to inhibit methanogenesis. The vials were capped immediately with butyl rubber stoppers and shaken vigorously for 1 min. Then, a 0.5-mL gas sample from the headspace was withdrawn and analyzed using a Porapak Q column (2 × 3 mm) on a gas chromatograph (Agilent 6820) equipped with a flame ionization detector. N2 was the carrier gas (flow rate of 30 mL min−1). The CH4 peak was recorded on a strip-chart recorder and quantified by comparison with a methane standard curve.

Lipid extraction

Because surface sediments may experience physical–chemical changes caused by bottom currents, lipid analysis was performed using samples collected at a finer scale (2-cm intervals) in the top 0–6 cm than in the deeper sediments (6–12-cm intervals). Lipid extraction was conducted according to the procedure of White et al. (1979). Briefly, frozen (−70 °C) sediment samples were lyophilized, and 8 g of the dried material was used for lipid extraction using a single-phase organic solvent system containing chloroform, methanol and aqueous 50 mM phosphate buffer (pH 7.4) in a ratio of 1 : 2 : 0.8 (by volume). After overnight extraction, chloroform and nanopure water were added to the extract in equal volumes. The lipids confined to the lower phase were collected and separated into neutral lipids, glycolipids and polar lipids on a silicic acid column (Guckert et al., 1985). The polar lipids were treated by mild alkaline methanolysis to produce fatty acid methyl esters (FAMEs). The FAMEs were identified using an Agilent 6890 series GC interfaced with an Agilent 5973 mass selective detector as described previously (Zhang et al., 2005). Methyl heneicosanoate was used as the internal standard. The FAMEs were expressed as equivalent peaks against the internal standard. Double-bond positions of monounsaturated FAMEs were determined by GC-MS analysis of the dimethyl disulfide adducts (Nichols et al., 1986). Cis and trans isomers of compounds were identified based on known standards.

Quantification of 16S rRNA and dsrAB genes by competitive PCR

The quantitative competitive (QC)-PCR method for quantification of 16S rRNA and dsrAB genes in the sediments followed Wang et al. (2005). The plasmid pMD18-T-dsrAB containing the dsrAB gene fragment from the dsrAB clone library was extracted and digested with HindIII. The linear plasmid pMD18-T-dsrAB was used as the standard dsrAB gene fragment template to evaluate the efficiency and accuracy of the competitive template DNA. The plasmid pMD18-T-dsrAB was digested with BsmI to get a 96-bp deletion in the region between the PCR primer pair DSR1F [AC(C/G)CACTGGAAGCACG] and DSR4R (GTGTAGCAGTTACCGCA) (Wagner et al., 1998), self-ligated and then digested with HindIII. The linear plasmid pMD18-TΔ96bp-dsrAB was used as competitive template DNA. The quantities of pMD18-T-dsrAB and pMD18-TΔ96bp-dsrAB were determined using an ND-1000 spectrophotometer (NanoDrop). The internal competitors for bacteria were constructed by Wang et al. (2005). Primer pairs EUB f933 (GCACAAGCGGTGGAGCATGTGG) and EUB r1387 (GCCCGGGAACGTATTCACCG) for bacteria (Iwamoto et al., 2000) were used in this study.

Construction of clone libraries and DNA sequencing

The bacterial 16S rRNA and dsrAB gene fragments were PCR amplified from total DNA extracted from sediments in the top (0–6 cm), middle (18–24 cm), bottom (44–50 cm) layers and the SMTZ (32–42 cm) using primer pairs Eubac27F (AGAGTTTGATCCTGGCTCAG)/Eubac1492R (GGTTACCTTGTTACGACTT) and DSR1F/DSR4R, respectively (Lane, 1991; Wagner et al., 1998). All reactions were carried out in 50-μL mixtures containing approximately 50–100 ng DNA, 10 × reaction buffer, 200 μM dNTP, 10 pmol of each primer, 1.5 mM MgC12 and 5 U Taq polymerase. Negative controls were performed using water instead of the DNA template. The reactions were performed in a T3 thermocycler (Biometra, Germany) using 30 cycles at 94 °C for 1 min, 55 °C for 1 min and 72 °C for 1.5 min, followed by a final extension step at 72 °C for 10 min. Reaction products were checked by agarose gel electrophoresis. PCR products were cloned into the pMD18-T vector supplied with the TA cloning kit (TaKaRa, Dalian, China) following the manufacturer's instructions. Ligation mixtures were used to transform competent cells of Escherichia coliΔH5α according to the suggestions of the manufacturer (TaKaRa). Positive clones containing the 16S rRNA or dsrAB gene fragments were picked randomly for sequencing (Sangon Inc., Shanghai, China).

Phylogenetic analyses

The 16S rRNA gene sequences were submitted to the chimera-check program at the Ribosomal Database Project II (http://rdp8.cme.msu.edu/cgis/chimera.cgi?su=SSU) to detect the presence of chimeric artifacts. The dsrAB gene sequence analysis included the full-length translated region of the dsrAB gene (approximately 1900 bp), excluding the regions between the stop and the start codons of the α, β subunits. A blastn search (http://www.ncbi.nlm.nih.gov/BLAST) of the GenBank database was conducted to identify the phylogenetic relationships of the insert sequences, including 16S rRNA gene sequences and the deduced amino acid sequences of α, β subunits of the DSR gene. 16S rRNA gene sequences were assigned to bacterial phyla based on comparisons to database sequences (http://simo.marsci.uga.edu/public_db/rdp_query.htm). For further phylogenetic analyses, sequence alignment with portions of the 16S rRNA gene and deduced DSR amino acid sequences from the GenBank was performed using the clustal x program (version 1.83), and matrices of evolutionary distance were constructed by the neighbor-joining method (Saitou & Nei, 1987). Bootstrap resampling analysis of 1000 replicates was performed to estimate the confidence of tree topologies. The phylogenetic tree was constructed with mega (molecular evolutionary genetics analysis; http://www.megasoftware.net/index.html) software, version 3.1 (Kumar et al., 2004).

Nucleotide sequence accession numbers

The nucleotide and amino acid sequences obtained in this study were deposited in the NCBI GenBank database under accession numbers EF999343EF999360, EF999362EF999377, EF999379EF999386, EF999388EF999402, EF999406 and FJ748768FJ748822 for the bacterial 16S rRNA genes, and FJ748823FJ748857 for the dsrAB genes.

Results

Sulfate and methane profiles

The sediments were soft silt, changing from gray in the surface layer to black only several centimeters below. The black sediment had a light hydrogen sulfide smell. Consistent with the physical observation, the sulfate concentration was highest at the sediment surface (10.9 mM) and decreased with depth to <1 mM below 42 cm (Fig. 2a). The methane concentration was <10 μM in the top 32 cm but increased gradually below 32 cm to a maximum value of 469 μM at 50 cm. An SMTZ was defined at a depth interval of 32–42 cm, in which sulfate reduction might be coupled with anaerobic oxidation of methane (Fig. 2a).

2

Depth distributions of selected geochemical and biological variables in the sediment core QA0504-7. (a) Concentrations of sulfate and methane. (b) Bacterial 16S rRNA and dsrAB gene abundances (copies g−1 wet weight) as measured by QC-PCR. (c) Depth profile of the relative contribution of SRPs to the total number of bacterial cells. The average copy number of 16S rRNA gene per cell was 3.6 for bacteria. The copy numbers of the dsrAB gene was estimated to be 1 per cell. The ratio of SRPs vs. total bacteria was calculated based on cell numbers converted from gene copies.

Phospholipid fatty acids (PLFAs) analysis

The PLFAs were mainly composed of saturated, terminally branched saturated and monounsaturated fatty acids (Supporting Information, Table S1). Saturated straight-chain fatty acids dominated the depth profile (27.1–53.9%) and had an increasing trend down the core with a maximum at the 30–36 cm (Fig. 3). Terminally branched saturated fatty acids (iso- and anteiso-15:0, 16:0 and 17:0) were the second predominant components (16.9–35.1%) and followed the same trend of the saturated fatty acids (Fig. 3); these biomarkers are characteristic of SRPs (Taylor & Parkes, 1983; Vainshtein et al., 1992; Zhang et al., 2002a; Londry et al., 2004). Monounsaturated fatty acids represented by 16:1ω7c and 18:1ω9t/18:1ω7c accounted for approximately 12.6–17.1% in the top 12 cm and decreased considerably below 18 cm (Fig. 3); these may be used as biomarkers for sulfur-oxidizing bacteria (SOB) (Zhang et al., 2005).

3

Distribution of saturated, terminally branched saturated, monounsaturated fatty acids along the sediment core from the Pear River Estuary. Saturated fatty acids included straight-chain C14–C24 fatty acids. Terminally branched saturated fatty acids included both iso and anteiso compounds. Monounsaturated fatty acids only included 16:1ω7c and 18:1ω9t/18:1ω7c here.

Bacteria and SRP abundance

Total bacteria estimated by QC-PCR varied from 5.2±0.7 × 108 to 1.5±0.3 × 109 16S rRNA gene copies g−1 (wet weight). The abundance increased with depth in the top 24 cm followed by a decreasing trend with depth toward the bottom (Fig. 2b). The dsrAB gene copies varied from 3.4±0.5 × 106 to 7.0±0.1 × 107 copies g−1 (wet weight) and had a similar pattern as the total bacteria (Fig. 2b). Assuming that one bacterial cell contains 3.6 copies of 16S rRNA gene and SRPs contain one dsrAB gene per cell, the proportion of SRPs in the bacterial community was 2.0–20.3%, with the highest proportion at the 12–18-cm sediment core (Fig. 2c). These data implied that SRPs constituted a significant proportion of the bacterial community.

Phylogenetic diversity and vertical zonation of dsrAB genes

The four dsrAB gene libraries from the top (0–6 cm), middle (18–24 cm), bottom (44–50 cm) layers and the SMTZ (32–42 cm) resulted in a total of 126 dsrAB gene clones (see Table 1 for details). These dsrAB gene sequences could be assigned into 35 OTUs based on a 5% cutoff; their closest reference dsrAB gene sequences in the GenBank are shown in Table 1. Most of the retrieved dsrAB gene sequences from the Pearl River Estuary were novel, as most of them had <70% amino acid identity with those from cultivated strains, and <90% identity with those from environmental clones (Table 1). The coverage of these dsrAB clone libraries was 64%, 91%, 80% and 100% in the top, middle, SMTZ and bottom layer, respectively. The diversity indexes of these libraries suggested a decreasing diversity of the SRPs community with increasing depth (Table 1); however, the SMTZ had the highest diversity in the core. Rarefaction curve suggested a similar trend (data not shown).

View this table:
1

Distribution of different dsrAB gene subclusters identified in the sediment samples from the Pearl River Estuary

Phylogenetic group*No. of clones in library from samplesClosest environmental sequence (% amino acid identity)Closest cultured relative (% amino acid identity)
Top layer (0–6 cm)Middle layer (18–24 cm)Bottom layer (44–50 cm)SMTZ (32–42 cm)
Cluster DSR-A
SMTZDsrH170001Deep-sea sediment clone: NTd-I26 (89%)Desulfohalobium retbaense (80%)
SMTZDsr320001Deep-sea sediment clone: NTd-I26 (88%)Desulfohalobium retbaense (80%)
TopDsr178120Deep-sea sediment clone: NTd-I23 (92%)Desulfohalobium retbaense (82%)
SMTZDsrH450002Deep-sea sediment clone: NTd-I18 (91%)Desulfohalobium retbaense (83%)
Cluster DSR-B
SMTZDsr250001Deep-sea sediment clone: NTd-I20 (77%)Desulfohalobium retbaense (74%)
Cluster DSR-C
SMTZDsr170003Deep-sea sediment clone: NTd-IV04 (90%)Desulfobacterium aniline (84%)
SMTZDsr290001Deep-sea sediment clone: NTd-IV04 (89%)Desulfobacterium aniline (85%)
SMTZDsrH440001Estuary mudflat clone VO6 (89%)Desulfobacterium aniline (85%)
Cluster DSR-D
SMTZDsr300002Deep-sea sediment clone: NTd-VII01 (73%)Desulfohalobium retbaense (80%)
Cluster DSR-E
SMTZDsr440004Desulfobacca acetoxidans (76%)
Cluster DSR-F
TopDsr251200Estuary mudflat clone VN10 (81%)Desulfotomaculum kuznetsovii (62%)
TopDsr71000Marine sediment fosmid ws39f7 (82%)Desulfotomaculum alkaliphilum (61%)
SMTZDsr260001Deep-sea sediment clone: NTd-V06 (89%)Desulfotomaculum Kuznetsovii (58%)
TopDsr351000Marine sediment fosmid ws7f8 (75%)
TopDsr361000Hydrothermal vent sediment clone B04P037 (83%)Pelotomaculum sp. MGP (64%)
TopDsr131000Hydrothermal vent sediment clone B04P026 (79%)Desulfotalea psychrophila (57%)
TopDsr591000Estuary sediment clone PIMO2D05 (85%)Desulfobacca acetoxidans (61%)
TopDsr85223Hydrothermal vent sediment clone B04P037 (91%)Pelotomaculum sp. MGP (65%)
TopDsr611000Hydrothermal vent sediment clone B04P026 (83%)Desulfotomaculum kuznetsovii (59%)
TopDsr781000Estuary sediment clone PIMO2D05 (88%)Desulfotalea psychrophila (58%)
BotDsr730044Estuary sediment clone PIMO2D05 (88%)Thermodesulforhabdus norvegica (63%)
TopDsr24001Estuary sediment clone PIMO2D05 (88%)Thermodesulforhabdus norvegica (63%)
MidDsr710200Estuary sediment clone PIMO2D05 (88%)Desulfobacca acetoxidans (61%)
Cluster DSR-G
MidDsr750100Estuary mudflat clone VO10 (88%)Pelotomaculum sp. MGP (67%)
TopDsr711244Estuary mudflat clone VO10 (87%)Pelotomaculum sp. MGP (68%)
TopDsr11566Hot spring clone MS3.103 (85%)Pelotomaculum sp. MGP (68%)
Cluster DSR-H
TopDsr141003Deep-sea sediment clone: NTd-VI01 (87%)Carboxydothermus hydrogenoformans (66%)
TopDsr40100Deep-sea sediment clone: NTd-VI01 (86%)Carboxydothermus hydrogenoformans (67%)
SMTZDsr130001Deep-sea sediment clone: NTd-VI01 (85%)Anaerobic bacterium sk.prop8 (67%)
Cluster DSR-I
TopDsr42340Estuary sediment clone PIMO1F10 (82%)Desulfotomaculum alkaliphilum (67%)
TopDsr222200Estuary sediment clone PIMO1F10 (83%)Desulfotomaculum alkaliphilum (66%)
SMTZDsr150001Estuary sediment clone PIMO1F10 (81%)Carboxydothermus hydrogenoformans (71%)
MidDsr670200Estuary sediment clone PIMO1F10 (81%)Carboxydothermus hydrogenoformans (70%)
Cluster DSR-J
SMTZDsr210008Low sulfate, acidic Fens clone dsrSbII-39 (80%)Desulfotomaculum alkaliphilum (63%)
SMTZDsr160001Low sulfate, acidic Fens clone dsrSbII-39 (80%)Desulfotomaculum alkaliphilum (64%)
n/N17/3310/226/2220/49
Coverage (%)649110080
Shannon–Wiener index3.63.172.483.95
Reciprocal of Simpson's index8.718.075.2612.4
Evenness0.8810.9540.960.913
  • * Representative dsrAB gene sequences used in phylogenetic analysis; n/N, the number (n) of OTU identified in the library composed of N clones. For the dsrAB gene, a 5% cutoff is defined as an OTU. Coverage was calculated using the following formula: C=1−(n1/N), where n1, number of phylotypes that occurred only once in the clone library; N, is the total number of clones analyzed (Mullins et al., 1995). The Shannon–Wiener index, Reciprocal of Simpson's index and evenness (equitability) were calculated using the equations from Krebs (Pielou, 1969; Krebs, 1989; Brown & Bowman, 2001).

The phylogenetic analysis indicated that the 35 dsrAB OTUs could be classified into 10 subclusters (DSR-A to DSR-J) (Fig. 4). The subcluster DSR-A belongs to the family Desulfobacteriaceae and the subcluster E belongs to the family Syntrophaceae; the other clusters do not show clear affiliation with known SRP families (Fig. 4). The distribution and proportion of the subclusters in the sediment layers are presented in Table 1. The dsrAB sequences were dominated by subclusters A, F and I in the top layer and dominated by subclusters F, G and I in both the middle and the bottom layers. The dsrAB sequences in the SMTZ were dominated by subclusters C, F, G and J and were significantly different from those in other layers. Subclusters B, C, D, E, J were only detected in the SMTZ; however, dsrAB sequences associated with methane-oxidizing archaea (such as within genera Desulfosarcina, Desulfococcus and Desulfobulbus) were not discovered in the SMTZ.

4

Phylogenetic tree based on the translated amino acid sequences of PCR-amplified dsrAB gene. The tree was constructed using the neighbor-joining method with Thermodesulfovibrio islandicus and Thermodesulfovibrio yellowstonii as the outgroups. Sequences determined in this study are in bold. The clones from the top, middle, bottom layers and the SMTZ are differentiated as TopDsr, MidDsr, BotDsr or SMTZDsr, respectively. Numbers in parentheses indicate the number of clones found in the clone libraries. Bootstrap values (in %) are based on 1000 replicates each and are shown at the nodes with >50% bootstrap support. OTUs with bootstrap value either >80% or inter-OTU DsrAB identities >70% were utilized to designate the DsrAB phylogenetic clusters. The scale bar represents 5% sequence divergence. LA-dsrAB Firmicutes is a group of low-G+C-content gram-positive bacteria with laterally acquired dsrAB genes.

The dsrAB gene sequences in subcluster DSR-F dominated in all the four clone libraries with 51%, 27%, 27% and 18% in the top, middle, bottom and the SMTZ, respectively. The Pearl River Estuary subcluster DSR-F could be assigned to a previously annotated SRP lineage Group IV (Dhillon et al., 2003). The Group IV dsr lineage is a deeply branching lineage, not affiliated with any cultured SRP. Cluster DSR-A was the second dominant group in the surface layer (0–6 cm), accounting for 24%, but not so abundant in the other layers (<10%). DSR-A sequences had high sequence similarity (88–92% identity) with those of some environmental clones from Nankai Trough deep-sea sediment (Kaneko et al., 2007). Clusters DSR-C and DSR-E were only detected in the SMTZ, accounting for 10% and 8% of total sequences, respectively. DSR-C was affiliated with those from Desulfobacterium of Firmicutes, whose dsrAB genes were found to be laterally derived from Deltaproteobacteria (Klein et al., 2001). DSR-E was closely affiliated with dsrAB sequence from Desulfobacca acetoxidans, which was an acetate-degrading sulfate-reducing bacterium belonging to Syntrophaceae.

Sequences from other dsrAB subclusters (DSR-B, D, G, H, I, J) had very low similarities with known dsrAB sequences and formed separate, deeply branching dsrAB lineages in the dsrAB gene phylogenetic tree (Fig. 4).

Bacterial 16S rRNA gene analysis

A total of 135 bacterial clones were randomly selected from the bacterial clone libraries for sequencing, which resulted in 112 OTUs based on a 3% cutoff. Members of the Proteobacteria (45.5%) were most dominant, which consisted of 21.3%Deltaproteobacteria, 11%Gammaproteobacteria, 9.6%Betaproteobacteria, 2.9%Epsilonproteobacteria and 0.7%Alphaproteobacteria.

Phylogenetically, a total of 25 OTUs were affiliated with Deltaproteobacteria, among which 14.5% had close relationships with known SRPs within Syntrophobacteraceae, Desulfobulbaceae and Desulfobacteriaceae. Most of these SRP sequences were closely affiliated (87–94% similarity) with Desulfatibacillum and Desulfobacterium within the Desulfobacteriaceae. Two OTUs were closely related (86–90% identity) to Syntrophobacter sulfatireducens within Syntrophobacteraceae, which could degrade propionate in syntrophic association with methanogens (Chen et al., 2005b). A few (1.6% of total Deltaproteobacteria) sequences clustered within the family Desulfobulbaceae and were most closely related to Desulfobulbus sp. RPf35L17, isolated from fluidized-bed reactors treating acidic, metal-containing wastewater (Kaksonen et al., 2004). The rest of the deltaproteobacterial sequences could be assigned into different subgroups, such as Nitrospinaceae, Geobacteraceae, Enhygromyxa, Sorangiineae and unknown subgroups (Fig. 5).

5

Phylogenetic tree of proteobacterial sequences. The phylogenetic relationships of bacterial 16S rRNA gene sequences retrieved from this study and related reference sequences are shown. Sequences determined in this study are in bold. The clones from the top, middle and bottom layers are differentiated as TopBa, MidBa and BotBa. Numbers in parentheses indicate the number of clones found in the clone libraries. The tree was constructed by the neighbor-joining method using the nearly full-length aligned nucleotides sequences with Escherichia coli as the outgroup. Bootstrap values (in %) are based on 1000 replicates each (distance and minimum evolution) and are shown at the nodes with >50% bootstrap support. The scale bar represents 5% sequence divergence.

Most of the sequences within Gammaproteobacteria were closely affiliated with Acinetobacter sp. PAMU-1.11, which was a strain capable of degradation of phenylacetic acid (Abe-Yoshizumi et al., 2004); the others either clustered with Pseudomonas or classified as unknown gammaproteobacterial lineage (Fig. 5). Many of the sequences within Betaproteobacteria were clustered with contaminant degraders such as phenanthrene-degrading bacteria, Alcaligenaceae bacterium, Denitratisoma oestradiolicum or Sterolibacterium denitrificans. Sequences within Epsilonproteobacteria were clustered with free-living and symbiotic sulfur-oxidizing bacterial strains (Fig. 5).

Noteworthy, 24.3% of the 16S rRNA gene sequences fell into the Acidobacteria, which was the second dominant group in the Pearl River estuarine sediment. The remaining sequences clustered with the Lentispheaerae, WS3, Planctomycetes, Gemmatimonadetes, Actinobacteria, Firmicutes, Nitrospirae, Chloroflexi, Spirochetes, Bacteroidetes, candidate division OP10, OP11 and an unclassified bacterial group, respectively (Fig. 6).

6

Phylogenetic tree of other bacterial sequences except those of Proteobacteria. For details see Fig. 5 legend.

Discussion

The Pearl River Estuary represents an important and unique system linking the highly developing inland area and the SCS. A variety of organic substances including contaminants have been detected in the sediments of the estuary. Sulfate reduction is known as an important process for the degradation of these organic contaminants and a dominant anaerobic organic carbon transformation process in anoxic environments (Jorgensen, 1982). In this study, the diversity and distribution of SRPs in a sediment core from the Pearl River Estuary were reported. Sulfate and methane profiles clearly indicated sulfate reduction and methane oxidation in the sediment core; molecular phylogenetic analysis demonstrated that several different novel lineages of SRPs reside in the Pearl River Estuary sediment. This study provides the phylogenetic basis for understanding the roles of these SRPs in the contaminant degradation and in the carbon cycling of the large estuary and its adjacent continental shelf in the SCS.

Amplification of dsrAB revealed the presence of SRPs throughout the sediment core, and the ratios of SRPs vs. bacteria indicated the predominance of SRPs in the middle depths. PLFA analysis supported the presence of SRPs throughout the sediment core, as characteristic SRP PLFA were abundant in all the sediment layers analyzed; in particular, their ratios were significantly higher in the layers at 24–36 cm where the abundance of SRP cells was also high. The dsrAB sequences obtained from the Pearl River Estuary sediment could be divided into 10 lineages, and most of these sequences had low similarity with dsrAB sequences from cultured SRPs, which formed deeply branching clades of SRPs containing no cultured representatives. The data strongly suggest that novel lineages of SRPs inhabit the Pearl River Estuary.

Large variations in the dsrAB gene compositions were observed among the top, middle, bottom layers and the SMTZ. In the top layer (0–6 cm), lineage DSR-A, F, I were most dominant, and lineage F, G, I dominated in both the middle and bottom layer, while in the SMTZ, lineage C, F, G, J dominated. The variation of dsr phylotypes along the sediment core may therefore be a function of the different physiochemical conditions present at different depths in the Pearl River estuarine sediment, especially with regard to oxygen, sulfide, sulfate and methane concentrations. Lineage DSR-F belonged to a deeply branching dsr lineage Group IV annotated previously, and was dominant in all the layers analyzed, implying that it may have a flexible lifestyle and be insensitive to the physical–chemical changes along the sediment core. As there is no cultivated representative in the Group IV SRPs, no direct physiological or metabolic information could be obtained for this group. However, it has been noticed that Group IV dsr were frequently obtained from a variety of organic-rich environments such as the Guaymas Basin hydrothermal vent sediments (Dhillon et al., 2003), marine sediment (Thomsen et al., 2001; Mussmann et al., 2005), Seine River Estuary (Leloup et al., 2006) and New England salt marsh (Bahr et al., 2005). It is therefore possible that the Group IV SRPs play an important role in the degradation of various hydrocarbons. Lineage DSR-A was only dominant in the top layer, suggesting that they were more oxygen-tolerant than other lineages or preferred microaerobic environments. DSR-A sequences could be assigned into the Deltaproteobacteria family Desulfobacteriaceae, which are capable of complete oxidation of organic carbon.

The lineage DSR-I dominated in the top, middle and bottom layers, but not in the SMTZ; DSR-G dominated in all the layers except the top layer; DSR-J, C were dominant in the SMTZ, but absent in other layers. These lineages represent novel, deeply branching dsr lineages not known before. Sequences in DSR-I clustered with Clone PIMO1F10 from New England salt marsh, which has not been assigned into any known dsr lineages (Bahr et al., 2005). DSR-G sequences clustered with environmental clone VO10 from mudflat of Seine River Estuary (Leloup et al., 2006) and clone MS3.103 from a low-sulfate hot spring microbial mat (Dillon et al., 2007). The lesser abundance of DSR-G in the top layer of the sediment core suggested that this group of SRPs may be sensitive to oxygen. The presence of only DSR-J, C in the SMTZ suggested that they may have some roles or relationships with anaerobic methane oxidization. The low similarity of our retrieved dsrAB sequences with those from cultured SRPs strongly suggested that the sources of these dsrAB genes may be phylogenetically distinct organisms. In addition, a variety of chemical compounds, such as chlorinated pesticides, polychlorinated biphenyls, polycyclic aromatic hydrocarbons, alkylphenol ethoxylates and polybrominated diphenyl ethers, have been detected in the Pearl River Estuary (Mai et al., 2002; Zhang et al., 2002b; Zheng et al., 2004; Chen et al., 2005a; Chau, 2006). It is very likely that at least some of these lineages may turn out to be contaminant-degrading species of SRPs.

The 16S rRNA gene analysis was aimed at providing further information about the total bacterial communities. We constructed three 16S rRNA gene libraries for the top, middle and bottom layers separately. Around 50 clones were randomly selected from each library and sequenced. However, given the low coverage of the clone libraries (data not shown), it may need to sequence at least several hundreds of more clones to adequately represent the community diversity. As the focus of the present study was to investigate the species diversity based on the functional gene dsrAB, we combined all the obtained 16S rRNA gene sequences together (135 sequences) to get an overall impression of the bacterial community in the sediment. The data revealed that the bacteria in the Pearl River Estuary were extremely diverse, and most of the obtained 16S rRNA gene sequences represented as yet uncultivated phylotypes distinct from any other cultivated bacteria. Sequence analysis also found members of SRPs within Deltaproteobacteria (21.3%), including Syntrophobacteraceae, Desulfobulbaceae and Desulfobacteriaceae. These sequences were closely related to those from representative cultures putatively involved in contaminant degradation and sulfur metabolism, including S. sulfatireducens (Chen et al., 2005b), Desulfobulbus sp. RPf35L17 (Kaksonen et al., 2004), Desulfatibacillum alkenivorans (Cravo-Laureau et al., 2004) and Desulfobacterium anilini (Schnell et al., 1989). As the majority of the dsrAB and 16S rRNA genes are novel, we cannot infer the relationship between those genes. To reveal the origin of these novel dsrAB genes, isolation of SRPs from the environment is needed. All the retrieved epsilonproteobacterial sequences were closely related with SOB sequences, suggested a similar function for sulfur oxidation of these bacteria in the Pearl River Estuary sediments.

Except Proteobacteria, Acidobacteria were found to be dominant members in the bacterial community of the environment. Organisms in the Acidobacteria phylum have been detected by 16S rRNA gene-based surveys in a wide variety of environments. In soils and sediments, the Acidobacteria appear to be abundant, comprising 10–50% of 16S rRNA gene sequences in clone libraries from materials that vary greatly in physical, geochemical and biological characteristics (Hugenholtz et al., 1998; Barns et al., 1999; Janssen, 2006). The phylum Acidobacteria has eight currently recognized subdivisions (designated 1–8) that have class-level ranking (Hugenholtz et al., 1998). The subdivisions have been expanded to 11 (Zimmermann et al., 2005) and more recently to 26 (Barns et al., 2007). However, little information is known regarding the metabolic potential of Acidobacteria, and there are only four formally described genera within the phylum Acidobacteria, including subdivision 1 members Acidobacterium capsulatum (Kishimoto et al., 1991) and Terriglobus roseus (Eichorst et al., 2007); and subdivision 8 members Geothrix fermentans (Coates et al., 1999) and Holophaga foetida (Liesack et al., 1994). Here, the Acidobacteria-like clones are affiliated with 12 subdivisions (Fig. S1). These sequences were closely related (>90%) with uncultivated organisms from a variety of environments such as sediments (Sorensen et al., 2007; Li et al., 2008; Zhang et al., 2008), soil samples (Ludwig et al., 1997; LaMontagne et al., 2003; Tringe et al., 2005; de Carcer et al., 2007; Elshahed et al., 2008; Hansel et al., 2008; Tarlera et al., 2008), aquifer environments (Skidmore et al., 2005; Beier et al., 2008; Isenbarger et al., 2008) and sea-floor lavas (Santelli et al., 2008). The highly diverse range of phylogenetically related sequences in many environments may suggest a potential for acidobacteria-like organisms to be physiologically versatile. The distribution and physiology of these Acidobacteria need to be further investigated and the Pearl River Estuary seems to provide a good source for the future investigation.

In summary, this study demonstrated a diversified SRP community in the Pearl River Estuary sediments, which may have participated or is participating in the sulfate reduction. Most of these SRPs appear to be novel and their physiology and ecological roles in the carbon transformation in the environment will be of great interest in ongoing studies.

Supporting Information

Table S1. Relative percentages of the detected PLFA along the sediment core in the Pearl River Estuary.

Fig. S1. The phylogenetic relationship of Acidobacteria phylum 16S rRNA gene sequences retrieved from this study and related reference sequences are shown.

Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

Acknowledgements

This work was partly supported by the National Science Foundation of China (grant number 40830213), and the COMRA Fund (DYXM-115-02-2-01). We would like to thank Hongchen Jiang, Yundan Pi and Noelle Garvin for helping with lipid analysis.

Footnotes

  • Editor: Jizheng He

References

View Abstract