Appl Microbiol Biotechnol (2014) 98:875–884 DOI 10.1007/s00253-013-4854-5

ENVIRONMENTAL BIOTECHNOLOGY

Bacterial polycyclic aromatic hydrocarbon ring-hydroxylating dioxygenases in the sediments from the estuary,

Peng Wu & You-Shao Wang & Fu-Lin Sun & Mei-Lin Wu & Ya-lan Peng

Received: 17 December 2012 /Revised: 10 March 2013 /Accepted: 11 March 2013 /Published online: 5 April 2013 # Springer-Verlag Berlin Heidelberg 2013

Abstract Bacterial community compositions were charac- Introduction terized using denaturing gradient gel electrophoresis anal- ysis of bacterial 16S rRNA gene in the sediments of the Polycyclic aromatic hydrocarbons (PAHs) are a class of Pearl River estuary. Sequencing analyses of the excised ubiquitous hydrophobic organic pollutants with known or bands indicated that Gram-negative , especially potential toxic, mutagenic, or carcinogenic properties Gammaproteobacteria, were dominant in the Pearl River (DeBruyn et al. 2007; Zhou et al. 2006). Due to the low estuary. The diversity of polycyclic aromatic hydrocarbon solubility of PAHs in water, the sediment represents a major ring-hydroxylating dioxygenase (PAH-RHD) gene in this sink for PAHs released to the environment (Pohlman et al. estuary was then assessed by clone library analysis. The 2002). The Pearl River Delta (PRD) region, one of the phylogenetic analyses showed that all PAH-RHD gene economically fastest growing regions in China, has under- sequences of Gram-negative bacteria (PAH-RHD[GN])were gone rapid urbanization and industrialization (Guan et al. closely related to the nagAc gene described for Ralstonia sp. 2009). Eventually, the contamination of PAHs has been U2 or nahAc gene for Pseudomonas sp. 9816–4, while the reported in various environmental media in the Pearl River PAH-RHD gene sequences of Gram-positive bacteria (Liu et al. 2011; Luo et al. 2008; Mai et al. 2001). Based on

(PAH-RHD[GP])atsamplingsiteA1showedhighsequence the ability of particular bacteria to degrade PAHs, bioreme- similarity to the nidA gene from Mycobacterium species. diation is proposed as a convenient tool for the cleanup of Meanwhile, molecular diversity of the two functional genes contaminated sites (Doyle et al. 2008). The diversity infor- was higher at the upstream of this region, while lower at the mation of bacteria involved in PAH decontamination can downstream. Redundancy analysis indicated that environmen- provide the basic knowledge for PAH bioremediation tech- tal factors, such as NH4-N, ∑PAHs, pH, SiO3-Si, and water nology in a natural environment. depth, affected the distribution of the PAH-RHD[GN] gene in The primary step of PAH metabolism commonly occurs via the Pearl River estuary. the incorporation of molecular oxygen into the aromatic nu- cleus by a multicomponent PAH ring-hydroxylating Keywords Polycyclic aromatic hydrocarbons . dioxygenase (PAH-RHD) enzyme (Ding et al. 2010;Penget Ring-hydroxylating dioxygenase . The Pearl River estuary . al. 2010). Iwai et al. (2011) have analyzed all the existing Bacterial community primer systems targeting the PAH-RHD enzymes and found that those enzymes can be divided into two subclades: PAH- : : : : RHD enzyme from Gram-negative bacteria (PAH-RHD[GN]) P. Wu Y.-S. Wang F.-L. Sun M.-L. Wu Y.-l. Peng and PAH-RHD enzyme from Gram-positive bacteria (PAH- State Key Laboratory of Tropical Oceanography, Institute of Oceanology, Chinese Academy of Sciences, RHD[GP]). Many previous studies demonstrated that positive 510301, China amplicons of the PAH-RHD gene can be obtained only after additionally amending PAHs in sediments or soils (Bordenave : * : : : P. Wu Y.-S. Wang ( ) F.-L. Sun M.-L. Wu Y.-l. Peng et al. 2008; DeBruyn et al. 2007;Dingetal.2010;Meynetet Daya Bay Marine Biology Research Station, Chinese Academy of Sciences, 518121, China al. 2012;NiChadhainetal.2006;Pengetal.2010;Zhouetal. e-mail: [email protected] 2006). However, only several studies focused on the diversity 876 Appl Microbiol Biotechnol (2014) 98:875–884 of the PAH-RHD gene in the natural environment (Flocco et al. 2009; Gomes et al. 2007; Jurelevicius et al. 2012). Because the catabolic potential of the PRD region is essentially un- known, culture-independent methods targeting the PAH-RHD gene were employed to provide for a broader estimation of the biodegradation potential in this region. Bacterial community structure can be significantly influenced by environmental variables, and environmental heterogeneity can drive a taxa–area relationship of bacteria (Cao et al. 2012; Hong et al. 2011; Horner-Devine et al. 2004; Meynet et al. 2012; Zhang et al. 2009). For example, Cao et al. (2012) reported that water depth and temperature were major factors shaping the community structure of ammonia-oxidizing β- from the Pearl River Delta to the South China Sea. Our previous study has also shown that NH4-N and SiO3-Si can obviously influence the bacterial community composition in the Pearl River estuary (Sun et al. 2011). Moreover, available nutrients (N/P), pH, and salinity were considered as the important factors in in Fig. 1 The sampling sites in the Pearl River estuary situ bioremediation of PAHs (Lu et al. 2011). Therefore, further studies should be conducted to determine the rela- tionships between environmental factors and the diversity of Guangzhou Channel, the Shiziyang Channel, the Lingding the PAH-RHD gene in the Pearl River estuary. Bay, and the northern South China Sea, respectively. In Surface sediments across a five-site transect along the order to investigate the relationship between PAHs and the upstream to the downstream of the Pearl River estuary were PAH-RHD gene, a field background value of PAHs was examined. The bacterial composition was firstly detected conducted by gas chromatography/mass spectrometry based on denaturing gradient gel electrophoresis (DGGE) (GC/MS) (N6890/5975B, Agilent, USA) following the analysis of bacterial 16S rRNA gene (16S-DGGE). methods EPA 3550C and EPA 8270D. The total concentra- Secondly, the PAH-RHD gene was amplified to determine tions of 16 priority PAHs (∑PAHs) at sites A1, A8, A10, C1, the diversity and distribution of PAH-degrading bacteria. and C5 were 2.09, 1.11, 0.48, 0.32, and 0.17 mg/kg, respec- Finally, statistical analysis was employed to investigate the tively. The results further showed that naphthalene, phenan- main potential physicochemical contributors to the distribu- threne, pyrene, and benzo[a]pyrene were the dominant tion of the PAH-RHD[GN] gene. components. Previous studies also demonstrated that PAHs were higher at upstream, but lower at the downstream of the Pearl River estuary (Luo et al. 2008; Mai et al. 2001). Material and methods Sample collection Characteristics of the Pearl River estuary Surface sediment samples (0–10 cm) were collected by a static The Pearl River estuary, a subtropical estuary, is the second gravity corer in August, 2010. When the sediments were re- largest in China in terms of discharge volume. This estuary trieved, air-exposed outer surface of the sediments was can be roughly divided into four parts—Guangzhou Channel, discarded and the central cores were taken with a sterile stainless Shiziyang Channel, Lingding Bay, and the northern South steel knife. Then, the sediment samples were stored at −20 °C in China Sea. The Guangzhou Channel drains over the city of sterile collection bags. Meanwhile, the overlying water from the Guangzhou, the most urbanized and heavily populated district same sampling site was collected, filtered through 0.45 μm in the PRD region. The Shiziyang Channel receives inflows membranes, and preserved at −20 °C for biogeochemical from the Guangzhou Channel and the East river. The Lingding analysis. Physicochemical parameters of the overlying water Bay, a funnel-shaped estuary, receives about 1.7×1014 L/a were determined according to Wang et al. (2008). fresh water runoff and nearly 3×108 tons/a of suspended particles from four outlets, i.e., , Jiaomen, DNA extraction Hongqilimen, and Hengmen (Mai et al. 2001). In the present study, five sampling sites are selected (Fig. 1). Site A1, site Total community DNA (TC-DNA) of each sampling site A8, sites C1 and A10, and site C5 are located on the was extracted directly from 0.5 g sediment in triplicate using Appl Microbiol Biotechnol (2014) 98:875–884 877

TM the E.A.N.A. Soil DNA Kit (OMEGA, USA) following al. 2008). However, for the PAH-RHD[GP] gene, the first the manufacturer’s instructions. TC-DNA was visualized round of PCR was with primers Nid-for and Nid-rev1, while with 1 % agarose gel electrophoresis to assess its integrity. primers Nid-for and Nid-rev2 were then selected and used in All DNA preparations were stored at −20 °C prior to PCR the second round (Zhou et al. 2006). A high fidelity Ex Taq amplification. (Takara, Japan) was employed for amplification of the PAH- RHD gene. Positive PCR products were confirmed with elec- 16S-DGGE analyses of sediment bacterial community trophoresis on a 1.0 % agarose. Eventually, positive amplifi-

cation of the PAH-RHD[GN] gene was obtained from all the A nested PCR was used for amplification of bacterial 16S five samples, whereas that of the PAH-RHD[GP] gene was rRNA gene. Briefly, the first round of PCR was with primers only found at site A1. 27 F and 1492R (Suzuki and Giovannoni 1996), and the second round with primers 341 F (with a GC clamp attached Clone library analyses of the PAH-RHD gene to the 5′ end) and 907R (Muyzer and Ramsing 1995). At this stage, the PCR products obtained from the different triplicates The PCR products obtained from each sampling site in of each sampling site were individually pooled together before triplicate were pooled together and purified. Purified PCR loading on DGGE gel. products of the PAH-RHD gene were individually DGGE was performed using Dcode universal mutation inserted into the pMD18-T vector (Takara, Japan) for detection system as described in the manufacturer’s manual constructing clone libraries, and then positive clones (Bio-Rad, Hercules, CA, USA). Approximately 400 ng PCR were selected to sequence and analyze on an ABI3730 products were loaded onto 6 % (wt/vol) polyacrylamide (37. DNA Sequencer (USA) (Shanghai, China). The PAH-

5:1 acrylamide/bisacrylamide) gels with a linear denaturing RHD[GN] gene sequences obtained from the five sam- gradient ranging from 40 to 60 % (100 % denaturant was pling sites were deposited in the GenBank database with 7 M urea and 40 % (wt/vol) deionized formamide) in 1× the following accession numbers: JX297626–JX297811,

TAE buffer. The electrophoresis was run at 60 °C and while the PAH-RHD[GP] gene sequences from sampling 100 V for 10 h. Gels were stained with ethidium bro- site A1 were submitted to GenBank with the following mide and photographed under a UV transilluminator accession numbers: KC183753–KC183772. (Alpha innotech, Japan). The aligned sequences of the PAH-RHD gene in each clone Dominant bands of DGGE profile were stabbed with sterile libraries were analyzed with the software Mothur (Schloss et pipette tip, resuspended in 40 μL of sterile ddH2O, and incu- al. 2009) to determine operational taxonomic units (OTUs) at bated at 4 °C overnight before they were re-amplified. The a 1 or 2 % dissimilarity cutoff. Simultaneously, Mothur was positions of the excised bands in the DGGE gel were con- used to estimate Chao1, Shannon index, Simpson index, and firmed with repeated DGGE. Bands showing the expected the coverage (Schloss et al. 2009). melting position were again amplified using 341 F (without GC clamps) and 907R primers. The PCR products were puri- Phylogenetic analyses of 16S rRNA and the PAH-RHD fied with the PCR purification kit (Takara, Japan) and subse- gene quently cloned into pMD18-T vector (Takara, Japan). Positive clones were submitted for sequencing using an ABI3730 The aligned PAH-RHD[GN] gene sequences from all five DNA Sequencer (USA) (Shanghai, China). All 24 sequences clone libraries were firstly analyzed with Mothur to deter- from the DGGE analysis were submitted to the GenBank mine OTUs at a 1 % dissimilarity cutoff, while OTUs about database (accession numbers: JX173762 to JX173785). the PAH-RHD[GP] gene at site A1 were determined at a 2 % Based on band intensity and position in the DGGE profile, dissimilarity cutoff. Close relatives and phylogenetic affili- Quantity One 4.6.2 (Bio-Rad, USA) was employed to analyze ations of the obtained sequences about 16S rRNA and the the community similarity by an unweighted pair group method PAH-RHD gene were determined from GenBank database arithmetic average (UPGMA) and to assess the phylum level by the BLASTN (http://blast.ncbi.nlm.nih.gov/Blast.cgi). composition of bacteria from the Pearl River estuary. Phylogenetic trees were constructed using MEGA 5 by the neighbor-joining algorithm and the Kimura 2-parameter dis- PCR amplification of the PAH-RHD gene tance estimation method with bootstrap analyses for 1,000 replicates (Tamura et al. 2011).

The PAH-RHD gene (PAH-RHD[GN] and PAH-RHD[GP])was amplified using a half-nested PCR. Briefly, for the PAH- UniFrac-based analyses of the PAH-RHD[GN] gene RHD[GN] gene, the first round of PCR was carried out with primers FRT5A and FRT3B. The half-nested round was In order to perform clone library comparisons, UniFrac sig- performed with primers FRT6A and FRT3B (Bordenave et nificant test and principal coordinate analyses were conducted 878 Appl Microbiol Biotechnol (2014) 98:875–884

Fig. 2 PCR-DGGE analysis of using the online software UniFrac (http://bmf2.colorado.edu/ 16S rRNA sequence fragments unifrac/index.psp) (Lozupone et al. 2006), which employs amplified from TC-DNA the genetic distances to evaluate the community similarity extracted from sediments of the based on the PAH-RHD gene sequence data. The Pearl River estuary (a) and [GN] UPGMA dendrogram environment clusters tree was projected in MEGA 5 constructed using DGGE (Tamura et al. 2011). profiles (b). The labels (A1, A8, A10, C1, and C5) Relationships between environmental factors and distributions represented corresponding sam- pling sites. Numbers on gels of PAH-RHD[GN] were the bands which were excised and sequenced Redundancy analysis (RDA) (CANOCO 4.5; Biometris, Wageningen, the Netherlands) was used to determine the correlations between bacterial community composition and environmental parameters (ter Braak and Smilauer 2002). Detrended correspondence analysis (DCA) was firstly done to decide between linear and unimodal response model for the bacterial species data (the 1 % distance cutoff). The length of the first DCA ordination axis was 1.727 for bacterial species data, so RDA was chosen to ordinate the bacterial community to the measured environmental parameters. A Monte Carlo permutation test (499 random permutations), including unrestricted permutation, was used to test the null hypothesis that bacterial profiles were unrelated to environmental variables (ter Braak and Smilauer 2002).

Results the temperature was lower compared to the other four Physicochemical parameters of the overlying water sampling sites.

The basic physicochemical parameters of the overlying Bacterial DGGE profiles water were summarized in Table 1. The variation of these parameters at the different sampling sites was ob- Bacterial composition of the five sampling sites was determined vious. Compared to other sampling sites, higher concen- by 16S-DGGE (Fig. 2a). UPGMA based on comparison of trations of NH4-N, PO4-P, NO3-N, and NO2-N were DGGE patterns indicated that the highest community similarity detected at sampling site A1. But the concentration of was recorded between sites A10 and C5 (78 %). However,

SiO3-Si at sampling site C1 was the highest. Nevertheless, bacterial community at site A1 showed the lowest similarity salinity, pH, and water depth at sampling site C5 were higher with site C5 (53 %) (Fig. 2b). The phylogenetic analyses of the due to its being adjacent to the South China Sea, but sequenced bands demonstrated that the majority of sequences

Table 1 Physicochemical properties of the overlying water in the Pearl River estuary

Sites NH4-N SiO3-Si PO4-P (mg/L) NO2-N Temperature (°C) Water depth (m) pH Salinity (psu) NO3-N (μmol/L) (μmol/L) (μmol/L) (μmol/L)

A1 281.61 55.02 0.11 28.65 30.4 5 6.38 0.19 289.78 A8 80.15 18.07 0.01 19.43 30.0 4.5 7.66 1.45 116.17 A10 30.48 80.61 0.02 13.29 30.0 6 8.02 12.13 111.79 C1 7.91 106.27 0.06 16.40 31.2 5 7.17 2.65 99.56 C5 2.97 29.42 0.01 3.51 28.87 15 8.36 22.59 4.34

NH4-N ammonium, SiO3-Si silicate, PO4-P phosphate, NO2-N nitrite, NO3-N nitrate Appl Microbiol Biotechnol (2014) 98:875–884 879 in this study were affiliated with , Molecular and phylogenetic diversity of the PAH-RHD Acidobacteria, Gammaproteobacteria, Deltaproteobacteria, gene Actinobacteria, and uncultured bacteria (Fig. 3). According to the band intensity and position in the DGGE profile, phylum Molecular diversity of the PAH-RHD[GN] gene was com- level composition using fingerprinting methods showed that pared at a 1 or 2 % distance cutoff (Table 2). Overall, a total Gammaproteobacteria were the most dominant class of 30 OTUs were observed at a 1 % distance cutoff, whereas (22.14 %) and Deltaproteobacteria can account for 20.31 % only four OTUs at a 2 % distance cutoff. However, the of the bacterial community. In general, Gram-negative bacteria coverage of each clone library can reach to 85.7–100 %, were dominant in sediments from the Pearl River estuary indicating that clone libraries covered adequately for the

(79.26 % of the total) compared with Gram-positive bacteria PAH-RHD[GN] gene (Table 2). OTU numbers and Shannon (10.15%)andunculturedbacteria(10.59%)(Fig.4). index of site A1 were relatively higher than the other sites,

Fig. 3 Phylogenetic tree of 16S 56 Oceanimonas baumannii GB6 (NR025027) rRNA genes about selected Alcanivorax balearicus MACL04 (NR043109) DGGE bands. Bootstrap values 52 100 2 (JX173763) represent 1,000 replicates and Uncultured gamma proteobacterium (GU061264) Uncultured gamma proteobacterium (DQ394995) only values above 50 % are 63 1 (JX173762) Gammaproteobacterium shown. The scale bar represents Methylonatrum kenyense AMT 1 (NR043963) the expected number of changes 13 (JX173774) per nucleotide position 18 (JX173779) 90 12 (JX173773) (2 % nucleotide substitution). 91 66 11 (JX173772) GenBank accession numbers 100 Uncultured bacterium band 8(HQ010589) are in parentheses Uncultured Sinobacteraceae bacterium E4 (HQ003539) Uncultured gamma proteobacterium (JQ726868) 6 (JX173767) 99 Uncultured beta proteobacterium A3-104 (JN371560) Betaproteobacteria 52 23 (JX173784) 100 Uncultured as1-98 (GU257563) 95 10 (JX173771) 100 purpureus (NR044679) 100 Rhodocyclus purpureus (FN995214) 80 17 (JX173778) 100 Burkholderiales bacterium JOSHI_001 (JF916701) 100 Comamonas aquatica LMG 2370 (NR042131) 100 4 (JX173765) 95 Uncultured Comamonadaceae bacterium (EU641895) 24 (JX173785) 57 Novosphingobium sp. FW-6 (JQ349046) Erythromicrobium sp. S23436 (D84627) 72 100 Alphaproteobacteria 7 (JX173768) 83 Rhodobium gokarnense JA173 (NR042475) 93 Uncultured Rhodobacteraceae bacterium F9P4 (HQ242304) 9 (JX173770) 100 Uncultured bacterium D8W_135 (HM057730) Uncultured bacteria 98 Uncultured delta proteobacterium TRAN-113 (JF344529) 53 5 (JX173766) 14 (JX173775) Deltaproteobacteria 100 Uncultured delta proteobacterium MS-A223 (FJ949275) 8 (JX173769) 100 Uncultured Desulfuromonadales bacterium X-128 (HQ132471) Uncultured Acidobacteria bacterium (DQ676329) 100 3 (JX173764) 19 (JX173780) 100 Uncultured Acidobacteria bacterium RII-TR055 (JQ580288) Acidobacteria 100 15 (JX173776) 58 Uncultured Acidobacteria bacterium (HM991563) Nitrospira calida (HM485589) 86 20 (JX173781) 100 Uncultured Nitrospirae bacterium (JQ579759) Nitrospirae 99 Nitrospira moscoviensis SBR1024 (AF155153) Actinobaculum massiliense (NR041909) 57 100 100 Actinobaculum massiliense 301 (FJ390134) Arthrobacter sp. CN2a (GQ332343) 100 21 (JX173782) Uncultured actinobacterium HAHS13.051 (HQ396929) Actinobacteria 95 89 100 22 (JX173783) Uncultured actinobacterium JBS_8n505 (EU702794) 81 100 16 (JX173777) Acidimicrobium ferrooxidans DSM 10331 (NR074390)

0.02 880 Appl Microbiol Biotechnol (2014) 98:875–884

Gammaproteobacteria gene fragments closely related to the nahAc gene from Betaproteobacteria Pseudomonas sp. 9816–4 (sequence identity within the Deltaproteobacteria – Actinobacteria cluster 99.2 99.5 %). The sequence identity between cluster Acidobacteria A and cluster B was about 80.5 %. With regard to the PAH- Alphaproteobacteria RHD[GP] gene, 20 sequenced clones from site A1 showed Nitrospirae high sequence similarity (97–99 %) to the nidA gene from uncultured bacteria Mycobacterium species (Fig. 6). Gram-positive bacteria Gram-negative bacteria UniFrac-based analyses of the PAH-RHD[GN] gene 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 percentage of phylum-level composition An all-environment UniFrac significance test indicated a Fig. 4 Phylum level composition of bacteria from the Pearl River significant difference (P≤0.001) among all the PAH- estuary based on 16S-DGGE analysis RHD[GN] assemblages in the current study. The weighted UniFrac PCA analysis indicated that sites A1 and C1 sepa- yet the two values showed that bacterial community of the rated from other sites with larger discrepancy, while the northern South China Sea (site C5) had the lowest diversity other three sites grouped together with higher support (Table 2). The Chao 1 demonstrated that the richness at sites values, while site C5 was separated from the other two sites. A1 and A8 was greater than sites A10, C1, and C5. However, The separation was explained along the first principal coor- the low Simpson index was observed at the Pearl River dinate (P1) by 97.28 % of the total community structure estuary (Table 2). Although the PAH-RHD[GP] gene was variation among all samples (Fig. 7). amplified only at site A1, the richness of the PAH-RHD[GP] gene at site A1 was relatively high. The Shannon index can Relationships between environmental factors reach to 1.85 when the dissimilarity cutoff was set at 2 % and distributions of the PAH-RHD[GN] gene (Table 2). Phylogenetic analyses showed that 186 bacterial se- RDA was performed to discern possible linkages between quences of the PAH-RHD[GN] gene recovered in this study environmental factors and distributions of the PAH-RHD[GN] were distributed into two clusters (Fig. 5). A large propor- gene (Fig. 8). Each environmental variable in the RDA biplot tion of sequences (171 out of 186 clones) were grouped in was represented by an arrow; the length of the individual cluster A. This cluster contained PAH-RHD[GN] gene frag- arrow indicated how much variance was explained by that ments closely related to the nagAc gene described for variable. Eigenvalues (indicating strength of the model) for Ralstonia sp. U2 (sequence identity within the cluster 98. the first two multivariate axes were 0.407 and 0.312, respec-

7–99.7 %); however, cluster B contained PAH-RHD[GN] tively. The sum of all canonical eigenvalues was 1.000. The

Table 2 Diversity and richness indices of PAH-RHD gene se- Sites Cutoff Clone numbers OTUs Shannon Simpson Chao Coverage (%) quence from the surface sedi- ments in the Pearl River estuary PAH-RHD gene from Gram-negative bacteria (PAH-RHD[GN]) A1 0.01 40 12 2.17 0.12 17.0 87.5 0.02 40 3 0.86 0.48 3.0 100 A8 0.01 35 9 1.79 0.18 19.0 85.7 0.02 35 2 0.57 0.61 2.0 100 A10 0.01 35 9 1.58 0.28 14.0 86.1 0.02 35 2 0.35 0.80 2.0 100 C1 0.01 40 8 1.72 0.21 8.3 95.0 0.02 40 2 0.59 0.59 2.0 100 C5 0.01 36 6 1.31 0.30 9.0 91.7 0.02 36 1 0 1 1.0 100 Total 0.01 186 30 2.61 0.10 47.5 91.9 a Positive PCR products of the 0.02 186 4 1.05 0.43 4.0 100 PAH-RHD[GP] gene were only a found at sampling site A1, PAH-RHD gene from Gram-positive bacteria (PAH-RHD[GP]) while no PCR product was A1 0.01 20 13 2.42 0.06 25 55.0 found at sampling sites A8, 0.02 20 8 1.85 0.14 14 80.0 A10, C1, and C5 Appl Microbiol Biotechnol (2014) 98:875–884 881

Fig. 5 Neighbor-joining tree A1(3), A8(2), A10-29, C5(3) based on the PAH-RHD[GN] gene A10-28 sequences. Bootstrap values A10-8 represent 1,000 replicates and A8-12, A8-31, A10-10, C1-20, C1-40, C5-9 only values above 50 % are A1-39 shown. The scale bar represents A8-26 A1-10, A8(3), A10(6),C1(2), C5-10 the expected number of changes A1(6), A8(3), A10(3), C1(2), C5(6) per nucleotide position (5 % A1-5 nucleotide substitution). Numbers A8-19 in parenthesis after the letters A10-9 (A1, A8, A10, C1, and C5) A1(6), A8(4), A10(3), C1(7), C5(7) correspond to the analyzed clone C5-32, C5-7, C5-5 numbers phnAc Alcaligenes faecalis (AB024945) nagAc Ralstonia sp. U2 (AF036940) 78 A10-6, A8-13 C1-22 Cluster A A1-32 C1-14, C1-13 A8-14 A1(7), A8(3), A10(5), C1-25, C5(3) 9354 A1(5), A8(9), A10(8), C1(9), C5(8) A8-2 A10-20 51 A10-33 A1-13, A1-28 84 A1-27 A1(2), A8(3), A10(2), C1(3), C5(4) 100 rhd gene Uncultured bacterium (FN597603) phnAc Delftia acidovorans Eh2-1 (AY367788) 85 DntAc Burkholderia cepacia (AF169302) phnAc Comamonas testosteroni (AY568278) nahAc Pseudomonas stutzeri (AF039533) ndoC1 Pseudomonas putida (AF004284) 57 87 A1-7, C1(9) 100 C1-6 A1-16, C1-26 51 pahd Uncultured Pseudomonas sp. (AM743136) Cluster B nahAc Pseudomonas sp. 9816-4 (U49496) nahAc Pseudomonas chlororaphis SY-02 (HM623873) A1-4, A1-1 nahAc Staphylococcus sp. KW-07 (HQ380297) phnAc Burkholderia sp. RP007 (AF061751)

0.05 first canonical axis was positively correlated with temperature ∑PAHs and pH (Fig. 8). In detail, the composition of the and SiO3-Si; the second canonical axis was positively corre- PAH-RHD[GN] gene at site A1 was positively correlated with lated with NH4-N and PO4-P but negatively correlated with NH4-N and NO3-N, while that at sites A8 and A10 had

Fig. 6 Neighbor-joining tree rhd gene Mycobacterium gilvum Spyr1 (CP002385) based on the PAH-RHD[GP] nidA Bacterium py129 (HM049717) gene sequences from the 85 A1-19, A1-8 sampling site A1. Bootstrap A1-20, A1-18, A1-14, A1-5, A1-2 values represent 1,000 68 nidA Mycobacterium tuberculosis SE12 (GU586859) nidA Mycobacterium replicates and only values sp. MCS (AY330102) rhd gene Mycobacterium sp. HH2 (DQ192654) above 50 % are shown. The 93 83 nidA Mycobacterium sp. py136 (HM049718) scale bar represents the A1-3 expected number of changes 99 A1-15 per nucleotide position A1-16 (5 % nucleotide substitution) nidA3 Mycobacterium sp. py147 (HM049740) 97 nidA3 Mycobacterium sp. py145 (HM049738) rdoA5 Mycobacterium gilvum (EU872102) 100 A1-17, A1-7, A1-12, A1-10, A1-9, A1-1 A1-13, A1-6 98 A1-11, A1-4 rhd gene Mycobacterium rhodesiae NBB3 (CP003169) 50

0.05 882 Appl Microbiol Biotechnol (2014) 98:875–884

influenced negatively on that gene composition. However, the

effect of environmental factors on the PAH-RHD[GN] gene at site C5 was opposite to that at site C1 (Fig. 8).

Discussion

Bacterial community composition in the sediments of the Pearl River estuary

Because the DGGE technique was frequently used as an effec- tive and popular tool to identify the dominant bacterial com- munity in the environment (Muyzer and Ramsing 1995;Feng et al. 2009;Sunetal.2011;Liuetal.2012), the phylogenetic diversity of bacterial community was determined by 16S- DGGE in surface sediments from the Pearl River estuary. The sequence analyses showed that Gammaproteobacteria were the predominant class (22.14 %) in sediment samples (Fig. 3). This observation was similar to previous studies. Ravenschlag et al. (2001)reportedthatGammaproteobacteria accounted for up to Fig. 7 The weighted UniFrac PCA analysis of the bacterial PAH- 10.5 % of the total cell counts and 20 % of prokaryotic rRNA in RHD[GN] gene sequences from the Pearl River estuary (individual sampling sites as circles) the Smeerenburgfjorden sediments. Feng et al. (2009)also found that this class was dominant in the sediments of coastal positive correlation with pH and ∑PAHs but negative corre- area from the East China Sea. Furthermore, 79 % of bacteria in lation with PO4-P. The concentration of SiO3-Si influenced the sediments of the Pearl River estuary were Gram-negative positively on the gene composition at site C1, but water depth bacteria. This result was also consistent with previous findings that Gram-negative bacteria were the most significant clade present in surface marine sediments, comprising over 50 % of PO4 -P site A1 the bacterial biomass (Bowman et al. 2003;Fengetal.2009; NH4 -N Liu et al. 2012). Therefore, Gram-negative bacteria, especially NO3 -N Gammaproteobacteria, were dominant in the sediments of the

NO2 -N Pearl River estuary.

Diversity of the PAH-RHD gene in the sediments of the Pearl River estuary Temperature The diversity index of the PAH-RHD gene was relatively site C5 site C1 [GN] low in the sediments of the Pearl River estuary (Shannon index Depth SiO3 -Si from 0 to 0.86) when the distance cutoff was set at 2 % (Table 2), whereas Shannon index can reach to 1.74 at Berre lagoon microbial mat (Bordenave et al. 2008)andto0.96at Salinity King George Island (Jurelevicius et al. 2012). The following site A8 reasons can account for the observed low diversity of the PAH- site A10 RHD[GN] gene. First, the selection of primers targeting the PAHs PAH-RHD gene had great effects on the bacterial diversity. The design of degenerate primers is a compromise between pH amplification of evolutionarily divergent and potentially novel genes on one hand and specificity on the other (Bordenave et

al. 2008). For example, the primers Rieske_f/Rieske_rcanfind -0.4 1.0 novel dioxygenase, resulting in 15 % of the clone library with low matches (<50 % identity), but the primers were highly Fig. 8 RDA of the relationships between the bacterial PAH-RHD[GN] gene sequences and environmental factors, with environmental factors degenerate and the amplicon size was too short (only 78 bp) as arrows and individual sampling sites as stars (NiChadhainetal.2006). After comprehensive consideration Appl Microbiol Biotechnol (2014) 98:875–884 883 of primer selection—coverage, specificity, and PCR product showed that bacterial communities at sites A8, A10, and C5 length—a half-nested PCR was chosen to make sure of the grouped together but separated with site A1 and site C1 specificity and stability of the clone library analysis. Second, (Fig. 7). Interestingly, the distribution of sampling sites in the PCR condition was also important. A high fidelity Taq the UniFrac PCA analysis was consistent with the results of enzyme, Ex Taq (Takara, Japan), was used in the current study. RDA (Fig. 8). To further explore the relationships between This enzyme with proof reading activity can reduce the muta- environmental factors and the distribution of the PAH- tions induced by the PCR process (Zhou et al. 2006). RHD[GN] gene, RDA was conducted. Previous studies be- Therefore, the occurrence possibility of novel dioxygenase will lieved that environmental factors (temperature, water depth, also reduce. Third, the presence of dioxygenase genes was well salinity, and nutrient concentrations) can significantly influ- correlated with the presence of certain genera in one region ence the bacterial community in the natural environment (Kumar and Khanna 2010). Most sequences of the PAH- (Zhang et al. 2009;Sunetal.2011; Cao et al. 2012).

RHD[GN] gene (171 out of 186 clones) belonging to cluster Because site A1, in Guangzhou channel, received a large A had the obvious nucleotide identity with the nagAc gene of number of inorganic and organic nutrients from the city of Ralstonia sp. U2 (Fig. 5). Fourth, and most important, richness Guangzhou, bacterial community at site A1 showed a strong of the PAH-RHD[GN] gene showed a positive relationship with correlation with NH4-N, whereas, site C5, located in the the contamination level of PAH in the environment (Gomes et outer estuary, was significantly influenced by the invasion al. 2007; Bordenave et al. 2008;Floccoetal.2009; Ding et al. of seawater from the open ocean, therefore salinity and 2010; Jurelevicius et al. 2012). In the current study, richness of water depth were the key factors shaping bacterial commu- the PAH-RHD[GN] gene and PAH contamination level were nity at site C5. The bacterial community at the Shiziyang highest at site A1, but lowest at site C5 (Table 2). As for the Channel and the Lingding Bay were, however, controlled by

PAH-RHD[GP] gene, positive PCR amplicon was found only at SiO3-Si, pH, and ∑PAHs. site A1, indicating a similar relationship between the gene In conclusion, this was the first report to investigate the richness and environmental PAH (Table 2). Previous studies diversity and distribution of the PAH-RHD gene along a have also demonstrated this relationship (Gomes et al. 2007; PAH contamination gradient from the Guangzhou Channel Bordenave et al. 2008;Floccoetal.2009;Dingetal.2010; to the northern South China Sea. The richness of the PAH- Jurelevicius et al. 2012). Even no PCR amplicon of the RHD gene was higher in the upstream of this region, while

PAH-RHD gene was obtained when TC-DNA from lower at the downstream. NH4-N, ∑PAHs, pH, SiO3-Si, and noncontaminated soils or groundwater was used (Ding et water depth were the key factors to affect the distribution of al. 2010). Thus, amplification efficiency depends on the the PAH-RHD[GN] gene in the Pearl River estuary. The targeted gene abundance in the natural environment. detection of functional genotypes and the bacterial isolates Positive amplification occurs more often when using carrying the PAH-RDH gene in the Pearl River estuary still DNA from samples with high targeted gene abundance. needs investigation.

The obtained PAH-RHD[GN] gene sequences were related to nagAc gene from Ralstonia sp. U2 or nahAc gene from Acknowledgments This work was supported by the Knowledge Pseudomonas sp. 9816–4 (Fig. 5). Both nagAc gene and Innovation Programs of the Chinese Academy of Sciences (KSCX2- EW-G-12C and No. KSCX2-SW-132), the National Natural Science nahAc gene were considered as the key catabolic gene to Foundation of China (No. 41076070 and No. 41176101), and the key mineralize naphthalene (Gibson et al. 1995;Zhouetal. projects in the National Science & Technology Pillar Program in the Eleventh Five-year Plan Period (No. 2009BADB2B0606 and No. 2001). However, PAH-RHD[GP] gene sequences from site A1 had the highest nucleotide identity with the nidA gene of 2012BAC07B0402). Mycobacterium species (Fig. 6). The nidA gene of Mycobacterium spp. was considered as molecular biomarkers for high molecular weight PAHs degradation (e.g., pyrene and References fluoranthene) (DeBruyn et al. 2007;Pengetal.2010). Therefore, the higher richness of the PAH-RHD gene at site Bordenave S, Goni-urriza M, Vilette C, Blanchard S, Caumette P, A1 indicated that PAH bioremediation potential at the up- Duran R (2008) Diversity of ring-hydroxylating dioxygenases in stream may be much greater than the downstream. pristine and oil contaminated microbial mats at genomic and transcriptomic levels. Environ Microbiol 10:3201–3211 Bowman JP, McCammon SA, Gibson JAE, Robertson L, Nichols PD Environmental factors to shape the distribution (2003) Prokaryotic metabolic activity and community structure in of the PAH-RHD[GN] gene Antarctic continental shelf sediments. Appl Environ Microbiol 69:2448–2462 The present study showed that bacterial communities with Cao H, Hong Y, Li M, Gu JD (2012) Community shift of ammonia- oxidizing bacteria along an anthropogenic pollution gradient from the PAH-RHD[GN] gene in the Pearl River estuary had the Pearl River Delta to the South China Sea. Appl Microbiol geographical difference. The weight UniFrac PCA analysis Biotechnol 94:247–259 884 Appl Microbiol Biotechnol (2014) 98:875–884

DeBruyn JM, Chewning CS, Sayler GS (2007) Comparative quantita- sediments from the Pearl River Delta and the northern South tive prevalence of Mycobacteria and functionally abundant nidA, China Sea. Arch Environ Contam Toxicol 55:11–20 nahAc, and nagAc dioxygenase genes in coal tar contaminated Mai BX, Fu JM, Zhang G, Lin Z, Min YS, Sheng GY, Wang XM sediments. Environ Sci Technol 41:5426–5432 (2001) Polycyclic aromatic hydrocarbons in sediments from the Ding GC, Heuer H, Zuehlke S, Spiteller M, Pronk GJ, Heister K, Pearl river and estuary, China: spatial and temporal distribution Koegel-Knabner I, Smalla K (2010) Soil type-dependent re- and sources. Appl Geochem 16:1429–1445 sponses to phenanthrene as revealed by determining the diversity Meynet P, Hale SE, Davenport RJ, Cornelissen G, Breedveld GD, and abundance of polycyclic aromatic hydrocarbon ring- Werner D (2012) Effect of activated carbon amendment on bac- hydroxylating dioxygenase genes by using a novel PCR detection terial community structure and functions in a PAH impacted urban system. Appl Environ Microbiol 76:4765–4771 soil. Environ Sci Technol 46:5057–5066 Doyle E, Muckian L, Hickey AM, Clipson N (2008) Microbial PAH Muyzer G, Ramsing NB (1995) Molecular methods to study the degradation. Adv Appl Microbiol 65:27–66 organization of microbial communities. Water Sci Technol 32:1–9 Feng BW, Li XR, Wang JH, Hu ZY,Meng H, Xiang LY,Quan ZX (2009) Ni Chadhain SM, Norman RS, Pesce KV, Kukor JJ, Zystra GJ (2006) Bacterial diversity of water and sediment in the Changjiang estuary Microbial dioxygenase gene population shifts during polycyclic and coastal area of the East China Sea. FEMS Microbiol Ecol aromatic hydrocarbon biodegradation. Appl Environ Microbiol 70:236–248 72:4078–4087 Flocco CG, Gomes NCM, Mac Cormack W, Smalla K (2009) Occurrence Peng JJ, Cai C, Qiao M, Li H, Zhu YG (2010) Dynamic changes in and diversity of naphthalene dioxygenase genes in soil microbial functional gene copy numbers and microbial communities during communities from the Maritime Antarctic. Environ Microbiol degradation of pyrene in soils. Environ Pollut 158:2872–2879 11:700–714 Pohlman JW, Coffin RB, Mitchell CS, Montgomery MT, Spargo BJ, Gibson DT, Resnick SM, Lee K, Brand JM, Torok DS, Wackett LP, Steele JK, Boyd TJ (2002) Transport, deposition and biodegrada- Schocken MJ, Haigler BE (1995) Desaturation, dioxygenation, and tion of particle bound polycyclic aromatic hydrocarbons in a tidal monooxygenation reactions catalyzed by naphthlene dioxygenase basin of an industrial watershed. Environ Monit Assess 75:155– from Psedumonas sp. strain 9816–4. J Bacteriol 177:2615–2621 167 Gomes NCM, Borges LR, Paranhos R, Pinto FN, Kroegerrecklenfort E, Ravenschlag K, Sahm K, Amann R (2001) Quantitative molecular Mendonca-Hagler LCS, Smalla K (2007) Diversity of ndo genes in analysis of the microbial community in marine Arctic sediments mangrove sediments exposed to different sources of polycyclic aro- (Svalbard). Appl Environ Microbiol 67:387–395 matic hydrocarbon pollution. Appl Environ Microbiol 73:7392–7399 Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M, Hollister EB, Guan YF, Samuel Sojinu OS, Li SM, Zeng EY (2009) Fate of Lesniewski RA, Oakley BB, Parks DH, Robinson CJ, Sahl JW, Stres polybrominated diphenyl ethers in the environment of the Pearl B, Thallinger GG, Van Horn DJ, Weber CF (2009) Introducing River Estuary, South China. Environ Pollut 157:2166–2172 mothur: open-source, platform-independent, community-supported Hong YG, Yin B, Zheng TL (2011) Diversity and abundance of software for describing and comparing microbial communities. anammox bacterial community in the deep-ocean surface sediment Appl Environ Microbiol 75:7537–7541 from equatorial Pacific. Appl Microbiol Biotechnol 89:1233–1241 Sun FL, Wang YS, Wu ML, Wang YT, Li QP (2011) Spatial hetero- Horner-Devine MC, Lage M, Hughes JB, Bohannan BJM (2004) A geneity of bacterial community structure in the sediments of the taxa–area relationship for bacteria. Nature 432:750–753 Pearl River estuary. Biologia 66:574–584 Iwai S, Johnson TA, Chai B, Hashsham SA, Tiedje JM (2011) Com- Suzuki MT, Giovannoni SJ (1996) Bias caused by template annealing parison of the specificities and efficacies of primers for aromatic in the amplification of mixtures of 16S rRNA genes by PCR. dioxygenase gene analysis of environmental samples. Appl Envi- Appl Environ Microbiol 62:625–630 ron Microbiol 77:3551–3557 Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S (2011) Jurelevicius D, Alvarez VM, Peixoto R, Rosado AS, Seldin L (2012) MEGA5: molecular evolutionary genetics analysis using maxi- Bacterial polycyclic aromatic hydrocarbon ring-hydroxylating mum likelihood, evolutionary distance, and maximum parsimony dioxygenases (PAH-RHD) encoding genes in different soils from methods. Mol Biol Evol 28:2731–2739 King George Bay, Antarctic Peninsula. Appl Soil Ecol 55:1–9 ter Braak CJF, Smilauer P (2002) CANOCO reference manual and Kumar M, Khanna S (2010) Diversity of 16S rRNA and dioxygenase CanoDraw for Windows user’s guide: software for canonical com- genes detected in coal-tar-contaminated site undergoing active munity ordination (version 4.5). Microcomputer Power, Ithaca bioremediation. J Appl Microbiol 108:1252–1262 Wang YS, Lou ZP, Sun CC, Sun S (2008) Ecological environment Liu G, Yu L, Li J, Liu X, Zhang G (2011) PAHs in soils and estimated changes in Daya Bay, China, from 1982 to 2004. Mar Pollut Bull air soil exchange in the Pearl River Delta, South China. Environ 56:1871–1879 Monit Assess 173:861–870 Zhang Y, Dong J, Yang B, Ling J, Wang Y, Zhang S (2009) Bacterial Liu Z, Huang S, Sun G, Xu Z, Xu M (2012) Phylogenetic diversity, community structure of mangrove sediments in relation to environ- composition and distribution of bacterioplankton community in mental variables accessed by 16S rRNA gene-denaturing gradient the Dongjiang River, China. FEMS Microbiol Ecol 80:30–44 gel electrophoresis fingerprinting. Sci Mar 73:487–498 Lozupone C, Hamady M, Knight R (2006) UniFrac—an online tool for Zhou HW, Guo CL, Wong YS, Tam NFY (2006) Genetic diversity of comparing microbial community diversity in a phylogenetic con- dioxygenase genes in polycyclic aromatic hydrocarbon-degrading text. BMC Bioinform 7:371–384 bacteria isolated from mangrove sediments. FEMS Microbiol Lett Lu XY, Zhang T, Fang HHP (2011) Bacteria-mediated PAH degradation 262:148–157 in soil and sediment. Appl Microbiol Biotechnol 89:1357–1371 Zhou NY, Fuenmayor SL, Williams PA (2001) nag genes of Ralstonia LuoXJ,ChenSJ,MaiBX,ShengGY,FuJM,ZengEY(2008) (formerly Pseudomonas) sp. strain U2 encoding enzymes for Distribution, source apportionment, and transport of PAHs in gentisate catabolism. J Bacteriol 183:700–708