Genome Reconstruction and Gene Expression of “Candidatus

Aug 4, 2014 - Cyrielle Gasc , Céline Ribière , Nicolas Parisot , Réjane Beugnot , Clémence Defois , Corinne Petit-Biderre , Delphine Boucher , Eri...
63 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/est

Genome Reconstruction and Gene Expression of “Candidatus Accumulibacter phosphatis” Clade IB Performing Biological Phosphorus Removal Yanping Mao, Ke Yu, Yu Xia, Yuanqing Chao, and Tong Zhang* Environmental Biotechnology Laboratory, Department of Civil Engineering, The University of Hong Kong, Pokfulam, Hong Kong S Supporting Information *

ABSTRACT: We report the first integrated metatranscriptomic and metagenomic analysis of enhanced biological phosphorus removal (EBPR) sludge. A draft genome of Candidatus Accumulibacter spp. strain HKU-1, a member of Clade IB, was retrieved. It was estimated to be ∼90% complete and shared average nucleotide identities of 83% and 88% with the finished genome CAP IIA UW-1 and the draft genome CAP IA UW-2, respectively. Different from CAP IIA UW-1, the phosphotransferase (pap) in polyphosphate metabolism and V-ATPase in orthophosphate transport were absent from CAP IB HKU-1. Additionally, unlike CAP IA UW-2, CAP IB HKU-1 carried the genes for carbon fixation and nitrogen fixation. Despite these differences, the key genes required for acetate uptake, glycolysis and polyhydroxyalkanoate (PHA) synthesis were conserved in all these Accumulibacter genomes. The preliminary metatranscriptomic results revealed that the most significantly up-regulated genes of CAP IB HKU-1 from the anaerobic to the aerobic phase were responsible for assimilatory sulfate reduction, genetic information processing and phosphorus absorption, while the down-regulated genes were related to N2O reduction, PHA synthesis and acetyl-CoA formation. This study yielded another important Accumulibacter genome, revealed the functional difference within the Accumulibacter Type I, and uncovered the genetic responses to EBPR stimuli at a higher resolution.

■

INTRODUCTION Enhanced biological phosphorus removal (EBPR) has been applied for over 30 years in wastewater treatment plants (WWTPs) to remove phosphorus in order to prevent eutrophication. EBPR promotes the removal of phosphorus from wastewater by circulating sludge through famine anaerobic and feast aerobic phases. In the anaerobic phase, polyphosphateaccumulating organisms (PAOs) rapidly uptake short-chain volatile fatty acids (VFAs) to store as polyhydroxyalkanoate (PHA), coupled with the hydrolysis of intracellular polyphosphate (polyP) as the energy source and glycolysis of glycogen1 and/or the tricarboxylic acid (TCA) cycle2 as the reducing power provider, resulting in increased inorganic phosphorus (Pi) levels in the bulk liquid. In the aerobic phase, PHA is degraded to support PAO growth while intracellular polyP is biosynthesized using Pi from the bulk liquid.3 Sludge discharged at this point may contain phosphorus up to 0.05−0.15 mg P/mg VSS (volatile suspended solid),4 which is much higher than that in the nonEBPR activated sludge (AS) (0.02−0.03 mg P/mg VSS). This type of phosphorus-rich biosolid is more amendable than conventional chemical phosphate precipitates to the recovery of phosphorus, which is one of the major limited resources globally.5 EBPR is a compelling model system for scientific interest in microbial structure and biochemical functioning. Molecular techniques have revealed that Candidatus Accumulibacter phosphatis (henceforth referred to Accumulibacter), a genus in the subclass β-Proteobacteria, was one of the major PAOs.6 The © 2014 American Chemical Society

polyphosphate kinase gene (ppk1) was thought to encode the enzyme primarily catalyzing polyP synthesis and was used as a gene marker to separate the Accumulibacter lineage into two types (I and II) that could be further subdivided into five and seven clades, respectively.7−9 Metagenomics has enabled researchers to access the gene pools and functions of Accumulibacter even if these microbes have not been cultured. The research on reconstructing the Accumulibacter Clade IIA strain UW-1 (referred to CAP IIA UW-1) genome and its inferred metabolism3 was a milestone and laid the foundation for further studies in this field. Recently, another draft genome of Accumulibacter Clade IA strain UW-2 (referred to CAP IA UW-2) was retrieved, and significant functional differences were identified between these two genomes.10 However, the functional potentials of other clades in the Accumulibacter lineage remain unknown. Additional genomes and unprecedented functions of Accumulibacter from diverse environments will soon be uncovered. In addition to the functional potential revealed by metagenomics, EBPR gene expression dynamics has been studied using various approaches including reverse transcription quantitative PCR (RT-qPCR),11 metatranscriptomic array,12 metaproteomics13 and radiolabeled proteomics.14 However, the Received: Revised: Accepted: Published: 10363

May 31, 2014 August 1, 2014 August 4, 2014 August 4, 2014 dx.doi.org/10.1021/es502642b | Environ. Sci. Technol. 2014, 48, 10363−10371

Environmental Science & Technology

Article

Figure 1. Schematic of the study workflow. Each type of the molecular approachmetatranscriptomics, metagenomics, 16S rRNA gene pyrosequencing and ppk1 gene characterizationis shown, as are the subsequent, analysis tools and bioinformatic approaches that have been used to analyze the various data sets.

ammonium, acetate, glucose and yeast extract (Supporting Information (SI) Figure S1). The salinity of the influent was maintained at 1% by adding sea salt to simulate the saline sewage in Hong Kong due to the practice of seawater flushing16 (SI Table S1). The ratio of TOC (total organic carbon): N (nitrogen): P (phosphorus) was 100:13:15, which was similar to that used in previous studies.4,17 AS from a nitrifying lab-scale reactor, mixed with AS from Sha-Tin and Shek-Wu-Hui WWTPs in Hong Kong, was inoculated in the SBR. Total DNA Extraction. Total DNA was extracted from 1.5 mL of sludge collected at P0, P1 and P2 using FastDNA SPIN Kit for Soil (MP Biomedicals, Solon, OH). P0 and P1 sludge samples were collected at the end of the aerobic phase. Two technically duplicate DNA samples (DNA_1 and DNA_2) extracted from P2 sludge were used to evaluate the reproducibility of the metagenomic approach including DNA extraction, library construction and sequencing. 16S rRNA Gene Pyrosequencing. The V3 and V4 regions of 16S rRNA genes were amplified from the duplicate DNA extracts of each sludge sample and sent for pyrosequencing with the Roche 454 FLX Titanium platform at BGI (Shenzhen, China). To estimate the abundances of Accumulibacter, effective reads were aligned against the currently available Accumulibacter 16S rRNA genes using BLASTN (v. 2.2.27+)18 at the cutoffs of identity ≥97% and alignment length ≥400 bp. Polyphosphate Kinase Gene Characterization. The PCR assay was applied to amplify the ppk1 genes of the Accumulibacter lineage.7 Quantitative real-time PCR (qPCR) was employed to determine the proportions of ppk1 genes from different clades by using the published primer sets8 (SI Table S2).

expression of some genes thought to be involved in EBPR metabolism was not (or not directly) detected using the above technologies due to their intrinsic limitation or immaturity, such as the limited number of qPCR primers and possible crosstalking among signals in the microarray. Moreover, such analyses were mostly conducted on Accumulibacter Type II-enriched mixed cultures, and very limited analyses involved Type I15 mostly due to the lack of the Accumulibacter Type I reference genome at that time. In those cultures with multiple clades of Accumulibacter and other microbes, uncovering the gene expression dynamics of a specific Accumulibacter clade was not possible. To achieve our objectives to retrieve novel Accumulibacter genomes and analyze the dynamics of community-wide transcription in response to the anaerobic and aerobic stimuli within the EBPR sludge, we conducted metatranscriptomic and metagenomic sequencing of a functionally stable and Accumulibacter-enriched community (Figure 1). The metatranscriptome helps reveal the gene expression levels in this complex microbial ecosystem by employing direct cDNA sequencing based on high throughput sequencing. Using an integrated metatranscriptomic and metagenomic approach, we generated a novel reference genome belonging to Accumulibacter Type I to improve our understanding of the intra- and intertype genomic differences and acquired its gene expression in the Pi-increasing anaerobic and Pi-decreasing aerobic stages.

■

EXPERIMENTAL SECTION EBPR SBR Operation. A 2 L cylinder sequencing batch reactor (SBR) was operated to remove excess phosphorus from synthetic wastewater containing phosphate (15 mg/L PO43−-P), 10364

dx.doi.org/10.1021/es502642b | Environ. Sci. Technol. 2014, 48, 10363−10371

Environmental Science & Technology

Article

Figure 2. Phylogenetic tree of ppk1 sequences. Sequences from P0, P1, and P2 sludge samples have been marked in green, blue, and orange, respectively. Numbers of the obtained sequences (based on >99% similarity of ppk1 nucleotide sequence) were in their brackets. Reference sequences are from NCBI database.

Total RNA Extraction, mRNA Enrichment and cDNA Synthesis. P2 sludge samples were collected for total RNA extraction at 1 h after adding acetate in the anaerobic phase and 1 h after switching to the aerobic phase of Pi accumulation and removal at the log phases (SI Figure S2). Total RNA was extracted by PowerSoil Total RNA Isolation Kit (MO-BIO Laboratories, Inc., CA). The RNA integrity numbers were 8.1 and 7.6, respectively, on a scale of 1−10 (10 indicating no degradation), showing the good quality of the extracted RNA. Messenger RNA (mRNA) was enriched by removing rRNA from 5 μg total RNA using the Ribo-Zero rRNA Removal Kit (MetaBacteria) (Epicenter Biotechnologies). The double-stranded cDNA was synthesized using the SuperScript Double-Stranded cDNA Synthesis Kit (Invitrogen, CA). DNA and cDNA Library Construction and Sequencing. DNA and cDNA library construction was performed following the manufacturer’s instruction.19 The extracted DNA samples were used for metagenomic sequencing using Illumina HiSeq 2000 with different strategies: (a) one lane of paired-end (2 × 100 bp) with a library of 200 bp insert size plus 3 Gb with a shotgun library of 800 bp insert size for the P2 sample; and (b) 6 Gb with a shotgun library of 800 bp insert size for the P1 sample. Metatranscriptomic libraries of cDNA samples were sequenced using Illumina HiSeq 2000 2 × 100 bp paired-end technology. “Omics” raw reads were removed when the ambiguous

nucleotides were more than 10% or more than 50% nucleotides had the quality scores of less than 20.20 Assembly, ORF Calling and Gene Annotation. The metagenomic reads were assembled using a de novo assembly algorithm integrated in the CLC Genomics Workbench version 6.0.2 (CLC Bio, Aarhus, Denmark) with a k-mer of 63.21 Open reading frames (ORFs) on contigs (≥1 kbp)21 were predicted using MetaGeneMark.22 Gene annotation was analyzed based on the BLASTP (v. 2.2.27+) results against the NCBI nr database and the KEGG database with parameters of -outfmt 6, -evalue 1e5, -max_target_seqs 10.23 Genome Binning for Dominant Accumulibacter. To bin the dominant Accumulibacter genome shared by P1 and P2 microbial communities, a combination of bidimensional coverage and tetranucleotide frequency patterns was applied as reported previously.21 Contig coverage was estimated by mapping the metagenomic reads from P1 and P2 to the assembled contigs using the CLC’s map reads to a reference algorithm with a minimum similarity of 90% over 95% of the read length.23 The 16S rRNA gene was reconstructed using EMIRGE with 80 iterations.24,25 To estimate the completeness of the genome bin, seven neighboring finished Rhodocyclaceae genomes including CAP IIA UW-1 were extracted from the IMG database and used as the pan genome. The Cluster of Orthologous Genes (COGs) shared 10365

dx.doi.org/10.1021/es502642b | Environ. Sci. Technol. 2014, 48, 10363−10371

Environmental Science & Technology

Article

among the pan genome, that is, those presenting in at least one copy in all of the seven neighbors, were collected as a set of common core COGs. The number of common core COGs observed in the genome bin showed the completeness of the draft genome.10,26 Another method to evaluate the completeness and potential contamination of the genome bin was using a set of 105 Hidden Markov Models of essential single copy genes (ESCGs) conserved in β-Proteobacteria.27,28 The ESCGs of the genome were obtained by searching against the ORFs using HMMER v3.1b1.29 Evaluation of Gene Expression Dynamics. Reads per kilobase per million mapped reads (RPKM) values for each mRNA sample (mRNA-RPKM) were adopted to analyze the gene expression levels. ORFs predicted from metagenomic contigs were used as the reference for mapping of the mRNA reads. The read mapping analysis in the CLC Genomics Workbench was performed with a minimum alignment length percentage of 0.8 and a similarity cutoff of 0.8.30 When analyzing the highly responsive genes related to EBPR stimuli, the genes with mRNA-RPKM values lower than 10 for both conditions or a length shorter than 100 aa were treated as low-quality coding sequences (CDSs) and were excluded from the analysis. Gene expression dynamics was measured by the fold change of mRNARPKM_Ae/mRNA-RPKM_An. Details of the methods are provided in the supplementary methods in SI. The metagenomic and metatranscriptomic sequencing data sets have been deposited in the NCBI Sequence Read Archive database under the accession number SRP038016.

Table 1. Estimated Abundance of Accumulibacter in the Whole Bacterial Community and Distribution of Different Accumulibacter Clades sample a

Accumulibacter (%) Clade IAb (%) Clade IIA (%) Clade IIB (%) Clade IICc (%) Clade IID (%) Other cladesd (%)

P0

P1

P2

3.4 0.4 25.0 1.1 17.5 0.4 55.7

18.3 1.7 1.3 0.0 13.7 0.0 83.3

4.0 0.9 1.5 0.1 2.4 0.0 95.2

a

Abundance of Accumulibacter in the bacterial community was estimated by the percentage of 16S rRNA pyrosequencing sequences assigned to Accumulibacter with cutoff of 97% identity and 400 bp alignment length. bThe percentage of each clade in the Accumulibacter lineage was obtained by dividing the ppk1 copy number by Accumulibacter 16S rRNA genes. The calculation had taken the following issues into account: ppk1 was the single copy gene in Accumulibacter and CAP IIA UW-1 genome had 2 copies of the rrn operon. cRelative abundance of Clade IIC was determined using the primer set targeting Acc-IIC ppk1 excluding OTU NS D3. dThe percentage of other clades was calculated by subtracting the total abundance of the above five Accumulibacter clades from the total Accumulibacter abundance.

technologies: duplicate DNA samples (P2-DNA-1, P2-DNA-2) for Illumina paired-end sequencing of two individual 200 bp libraries and one sample (P2-DNA-3) for Illumina paired-end sequencing of an 800 bp library. This effort produced a total of 300 million sequences from P2. Reproducibility between P2DNA-1 and P2-DNA-2 was satisfactory regarding the taxonomic binning and functional assignment, which was indicated by the correlation coefficients that were close to 1.000 (SI Figure S5). All metagenomic sequences from both P1 and P2 were assembled together using CLC’s de novo assembly algorithm and maximum k-mer of 63 bp,21 resulting in 80,141 contigs with a minimum length of 1 kbp and 51% of reads usage (Table 2). Accumulibacter Genome Binning. Using a combination of bidimensional coverage and tetranucleotide frequency,21,28 an Accumulibacter draft genome was extracted from P1 and P2 (Figure 3). This draft genome (defined as CAP IB HKU-1) had 3.6 Mbp in total with an average GC content of 64% (Table 3). To determine the completeness of this draft genome, a COGbased analysis was conducted by determining all COG functions that were shared among the seven neighboring Rhodocyclaceae genomes besides CAP IIA UW-1, generating 889 COG functions that were at least present once in all of the seven neighbors. This COG list was then compared with that of the CAP IB draft genome, and the results showed that 97 (11%) of the COG functions were missing from the CAP IB draft genome and the completeness was approximately 89%, consistent with the completeness calculated based on the ESCGs; 90 of the 105 ESCGs in β-Proteobacteria were carried by the CAP IB draft genome. Moreover, this draft genome had no genomic contamination from other populations that coexisted in the EBPR sludge, as indicated by the single copy of these 90 ESCGs. A ppk1 homologue identified in a long contig (contig_2292, 16,197 bp) shared 97% nucleotide identity with a previously identified Clade IB ppk1 gene (GenBank EU432912).9 An almost full-length of 16S rRNA gene (1,407 bp) was constructed using the EMIRGE approach24,25 and was clustered with the 16S rRNA genes of Accumulibacter Type I (SI Figure S6). The presence of the Clade IB ppk1 homologue and the reconstructed

■

RESULTS AND DISCUSSION Accumulibacter in the Microbial Community. Phosphorus removal in the EBPR SBR increased gradually from 22% (P0) to 98% (P1) after 50 days of enrichment (SI Figure S3). From P1 to P2, the SBR maintained consistent phosphorus removal of more than 80% for 4 months with a biomass phosphorus content of ∼0.07 mg P/mg VSS (SI Table S1). 454 pyrosequencing of 16S rRNA V3 V4 amplicons generated a total of 23,974 clean sequences. The Good’s coverage for each sludge sample was more than 92%, indicating good coverage of the major bacterial populations under this sequencing depth (SI Table S3). Among the top 30 operational taxonomic units (OTUs), the dominant Accumulibacter population (OTU_2) in P1 and P2 was 16.6% and 4.4%, respectively (SI Figure S4). The microbial diversity in Accumulibacter was not high, and the difference in abundances between the two Accumulibacter OTUs (OTU_2 and OTU_38) in P1 and P2 was significant. This preferred to exclude the genomic contamination from the dominant Accumulibacter genome bin based on the coveragedefined method. The ppk1 gene was used to partition the Accumulibacter lineage. A total of 50 partial ppk1 genes from each sludge sample (P0, P1, and P2) were sequenced after amplification. Accumulibacter in P1 and P2 were dominated by Clade IB, which was gradually enriched from P0 containing diverse ppk1 genes (Figure 2 and Table 1). Because currently available qPCR primer sets cannot target Clade IB ppk1 (SI Table S2), most of the ppk1 from P1 and P2 were assigned to other ppk1 clades. Metagenomic Sequencing and Assembly. The P1 biomass sample was collected for Illumina paired-end metagenomic sequencing using a library with an 800 bp insert size and generating 58 million effective sequences. The P2 sample was subject to metagenomic sequencing using a combination of 10366

dx.doi.org/10.1021/es502642b | Environ. Sci. Technol. 2014, 48, 10363−10371

Environmental Science & Technology

Article

Table 2. Assembled Contigs and Estimated Coverage of CAP IB HKU-1 P1

Table 3. Genome Characteristics, Key Metabolic Pathways and Gene Expression in CAP IB HKU-1

P2

characteristics

sample

DNA

DNA_1

DNA_2

DNA_3

library (bp) sequence number contig number (≥1 kbp)a maximum length (bp) N50 (bp) mapped readsb estimated coverage of CAP IB HKU-1c

800 58,420,806

200 135,741,636

200 135,659,488

800 28,542,368

genome characteristics

80,141 293,316

24,429,628 79

4,217 128,976,972 78

a

Contigs were assembled by total metagenomic sequences from both P1 and P2. bReads were mapped to contigs using CLC’s reference mapping algorithm, requiring 95% identity over 90% of the read length. cCoverage of CAP IB HKU-1 was roughly estimated based on Accumulibacter IB abundance (calculated according to 16S rRNA gene pyrosequencing and ppk1 gene clone library results) in the microbial communities and the genomic size of CAP IIA UW-1 (5.4 Mbp).

key metabolic pathways

gene expression

Figure 3. Extraction of the initial Accumulibacter clade IB genome bin from the metagenome using the coverage-defined method. Each circle represented an assembled contig with the size proportional to its length and colored by phylum. The box linked by 8 dots encloses contigs representing the initial CAP IB genome bin.

CAP IB HKU-1

num of contig

824

num of bases (mbp) mean length of contigs (bp) average coverage of the draft genome in P1 average coverage of the draft genome in P2 num of essential single copy genes completeness (%) GC content (%) total ORFs average ORF size (bp) function assigned num of CDS 16S rRNA gene ppk1 gene coverage of ppk1 gene in P1 coverage of ppk1 gene in P2 ANI with CAP IIA UW-1 shared genes with CAP IIA UW-1

3.6 4,340 77

90 ∼90 64 3,720 884 3,705 3,621 reconstructeda Gene_380 76 59 82.7 903

acetate uptake

+

glucose uptake glycolysis (EMP) glycolysis (ED) TCA cycle carbon fixation PHA synthesis polyphosphate metabolism nitrate reduction nitrogen fixation assimilatory sulfate reduction

+ + − + + + + + + +

mRNA_RPKM_An mRNA_RPKM_Ae genes only expressed in anaerobic phase genes only expressed in aerobic phase genes up-regulated in aerobic phase genes down-regulated in aerobic phase

34,061 39,872 91

60

181 444 381

a

A16S rRNA gene of 1,407 bp was reconstructed using EMIRGE method.24,25

16S rRNA gene helped assign the draft genome to a proper phylogenetic unit. A total of 1,986 genes were shared between the CAP IA UW-2 and CAP IB HKU-1 genomes based on the nucleotide sequence determined using reciprocal best-hit analysis.10,31 These genes shared an average nucleotide identity (ANI) of 88% (SI Figure S7), indicating that HKU-1 was a species that was different from UW-1 or UW-2 based on the species cutoff of 95%−96%.32 The ANI between CAP IB HKU-1 and CAP IIA UW-1 was 83%, which was relatively higher than that of 78% between CAP IA UW-2 and CAP IIA UW-1. Metatranscriptomic Analysis. After filtering out the rRNAlike sequences (approximately 30% of the raw reads) from the metatranscriptomic data set using a BLASTN search against the SILVA SSU and LSU database with an e-value cutoff of 1e-20,33 the mRNA reads in the remaining part were mapped to the metagenomic ORFs to analyze the gene expression under

anaerobic and aerobic conditions.30 The results showed that ∼30% of the mRNA reads mapped to these ORFs. Genes in CAP IB HKU-1 were expressed with a total mRNARPKM of 34,061 under the anaerobic condition and 39,872 under the aerobic condition, indicating similar overall gene expression levels with an average fold change of 1.2. It was found that 3,621 ORFs, 97% of the 3,720 total ORFs in CAP IB, were transcribed (at least one mRNA mapped to the ORF) under one set of conditions and were treated as CDSs30 with average mRNA_RPKM_An of 9.4 and average mRNA_RPKM_Ae of 11.0. The gene was identified as significantly up-regulated if the fold change of aerobic mRNA-RPKM_Ae and anaerobic mRNARPKM_An was larger than 5-fold and moderately up-regulated if the fold change ranged from 2−5-fold. In contrast, the gene was 10367

dx.doi.org/10.1021/es502642b | Environ. Sci. Technol. 2014, 48, 10363−10371

Environmental Science & Technology

Article

Figure 4. Metabolic model of CAP IB HKU-1 inferred from its draft genome. Specific gene expression dynamics between anaerobic and aerobic samples were indicated by colors. Blue, expressed, but no significant expression variation between the two samples (fold change = 0.5−2.0); green, expressed more in the aerobic sample (fold change ≥2.0); orange, expressed more in the anaerobic sample (fold change ≤0.5). Dash line means the key gene were absent from the CAP IB HKU-1 draft genome in the pathway. If a gene had more than one copy, the most significant expression difference among the copies was showed here.

formation of PHA, conversion to pyruvate and entering the TCA cycle.3 We detected robust expression of acetyl-CoA synthetase (acs) activating acetate to AcCoA both in the anaerobic and aerobic samples with mRNA-RPKM values of 40.4 and 33.7, respectively (Figure 4 and SI Table S4). Similar to the microarray results,12 key genes were implicated in PHB formation, i.e., AcCoA acetyl transferase (phaA), acetoacetyl-CoA reductase (phaB), and polyhydroxyalkanoate synthase (phaC), and were detected at high expression levels under anaerobic conditions, consistent with the anaerobically higher expression levels of aceEF, which catalyze the transformation of pyruvate to AcCoA. Moreover, the robust expression of a cluster of pyruvate ferredoxin oxidoreductase genes (porABG) demonstrated that AcCoA participated in pyruvate metabolism. Here we only compared the transcription patterns at two stages when the acetate became depleted in the SBR; therefore, the expression profile of acetate inducing genes of acs, CS, and porABG11 did not show a significant difference in the two phases. Moreover, unlike CAP IA UW-2,10 CAP IB HKU-1 carried the key genes for carbon fixation (SI Table S4). PHA Synthesis and Glycogen Degradation. The conversion of acetate to PHA requires reducing power generated via glycogen degradation and/or the TCA cycle. In addition, the source of this reducing power is one of the most debatable aspects of EBPR metabolism.1,2,34 In this CAP IB HKU-1 draft genome, a complete TCA cycle was identified, which was the same as in the other Accumulibacter genomes. Moreover, the observed expression of methylmalonyl-CoA mutase (MCM), fumarate reductase subunit C ( f rdC) and fumarate reductase flavoprotein subunit (FR) demonstrated the alternative scenario of anaerobic TCA cycle in CAP IB HKU-1. The anaerobic TCA cycle is very important for EBPR metabolism by alleviating

considered to be significantly down-regulated if the fold change was lower than 1/5 and moderately down-regulated if the fold change ranged from 1/5−1/2-fold. As a result, 80% of the genes carried by CAP IB HKU-1 were expressed without obvious differences, 10% of the genes were up-regulated and 3% were down-regulated in response to these stimuli. Highly Responsive Genes Related to EBPR Stimuli. To address the important genes related to EBPR stimuli in CAP IB HKU-1, we closely evaluated the CDSs that showed expression changes between the anaerobic and aerobic phases (SI Figure S8). Of the 169 total CDSs showing a ≥ 2-fold change in expression level, we found that 14 CDSs (≥5-fold) were highly responsive to the condition change. A total of 27 of the KEGGbased functional annotations of the moderately up-regulated CDSs were related to genetic information processing, such as RNA polymerase responsible for transcription, ribosomal protein in translation and enzymes in RNA degradation, indicating that gene regulation was more active in the aerobic phase during which cell growth mainly occurred. Notably, a gene cluster (gene_715, 716, 717, 719, and 3513) encoding enzymes for assimilatory sulfate reduction and sharing a similar gene organization with CAP IIA UW-1 and CAP IA UW-2 was significantly up-regulated under the aerobic condition (≥8-fold), indicating more rapid growth of this bacterium in the aerobic phase. The highly responsive genes are likely to be important for the adaptation of CAP IB HKU-1 to the EBPR stimuli. Additionally, the gene expression dynamics relative to core EBPR metabolic pathways were analyzed as follows. Carbon Metabolism. Acetate was the dominant carbon source fed into the SBR in this study. Metabolic reconstruction of CAP IIA UW-1 proposed multiple possible routes for AcCoA (acetyl-CoA) to participate in carbon metabolism such as the 10368

dx.doi.org/10.1021/es502642b | Environ. Sci. Technol. 2014, 48, 10363−10371

Environmental Science & Technology

Article

draft genome. F-ATPase was robustly transcribed with a maximum mRNA_RPKM_Ae of 239.2 in the aerobic phase, coinciding with the higher intensive F-ATPase signals compared to that of the V-ATPase detected on the microarray12 and only FATPase was detected with metaproteomics.15 Nitrogen Metabolism. Previous studies indicated that Accumulibacter of different clades had different nitrate reducing abilities.39 The respiratory nitrate reductase gene (nar), which initiated a prokaryotic denitrification pathway,40 could not be identified in either CAP IIA UW-13 or CAP IA UW-2 genomes,10 whereas all of the subunits (narGHIJ) were found in the Tetrasphaera species.41 Similar to CAP IA UW-2 and IIA UW-1, no nar gene was identified on CAP IB HKU-1. Given that both CAP IA UW-2 and CAP IB HKU-1 are not complete, it is still too early to conclude the absence of nar genes within Accumulibacter of Type I, but the potential presence of nar genes is quite unlikely based on current findings. CAP IIA UW-1 and CAP IA UW-II contained a napC/nirT homologue that had been reported to perform nitrate reduction if enough proton-motive force for energy production is provided.10,42 Although this homologue was not found in CAP IB HKU-1, periplasmic nitrate reductase (napA) was identified (Figure 4) and most likely performed a significant role in nitrate reduction as a result of the 110 times expression level over nasA (the assimilatory nitrate reductase catalytic subunit) in the anaerobic sample based on mRNA-RPKM_An values. In the dissimilative nitrate reducing CAP IB HKU-1 pathway, the nitrous-oxide reductase (nosZ) that catalyzes the reversible conversion of N2O to nitrogen was moderately up-regulated in the anaerobic phase (Figure 4), which was consistent with the microarray12 and RT-qPCR assay results.11 Similar to CAP IIA UW-1, CAP IB HKU-1 carried the genes required for nitrogen fixation. Subunits of nitrogenase molybdenum nif D and nif K were identified in CAP IB with a similar organization in CAP IIA UW-1 (SI Figure S9). This indicated that CAP IB HKU-1 might survive in a nitrogendeficient environment although the expression levels of these gens were low (mRNA-RPKM < 1.0) in this study, most likely due to sufficient ammonia as a nitrogen source for CAP IB growth in the influent (14 mg/L NH4+-N). Advantages and Limitations of “Omics” Approaches in Studying EBPR Processes. The metagenome has enabled researchers to determine the microbial functional pool even before the microbes are cultured. EBPR research, including the compelling model system dominated by uncultured microbes, has benefited from this technique. Several genomes of important PAOs3,10,41 and their competitor glycogen accumulating organisms (GAOs) 28,43 have been retrieved. Here we reconstructed another Accumulibacter genome CAP IB HKU1 from metagenomic sequences of an EBPR SBR. The genome was estimated to be ∼90% complete and carried a ppk1 gene clustered in Clade IB. The functional annotation of CAP IB HKU-1 showed that metabolic genes that are essential for acetate uptake, glycolysis (EMP), the TCA cycle and PHA synthesis were conserved in all available Accumulibacter genomes although there were some differences between CAP IA UW-2 and CAP IB HKU-1, such as the key genes for carbon fixation and nitrogen fixation. Combining metatranscriptomics with draft genomes binned from the metagenome of the same microbial community enables the profiling of expression dynamics at the individual genome level. Here, we compared the gene expression changes from Piincreasing anaerobic to Pi-decreasing aerobic conditions. The

FADH2 accumulation when external electron acceptors are lacking in the anaerobic phase.35 The transcription level along this pathway did not show significant down-regulation from the anaerobic to the aerobic sample because it would decrease only when the Pi level became very low at the end of the aerobic phase.11 Another alternative reducing power for PHA formation is based on anaerobic degradation of glycogen to pyruvate and further to AcCoA.1 Similar to the CAP IIA UW-1 and CAP IA UW-2 genomes, the CAP IB HKU-1 genome bin conserved an Embden-Meyerhof-Parnas (EMP) glycolysis pattern and all key genes of the Entner-Doudoroff (ED) pathway appeared to be absent. This result showed that EMP was a prevailing glycolysis pattern for Accumulibacter, consistent with that of the reported metagenomic,3 metaproteomic13,14 and metatranscriptomic array12 studies. Polyphosphate Metabolism. The most intensively studied genes in bacterial polyP metabolism are ppk1 and exopolyphosphatase (ppx), which catalyze the biosynthesis and hydrolysis of polyP, respectively. The ppx gene was identified at the downstream of the ppk1 gene, and its transcription depended on the ppk1 promoter, indicating a polyP operon.36,37 A polyP operon was identified in contig_2292 of CAP IB HKU-1 and shared approximately the same organization as that in CAP IIA UW-1 (SI Figure S9). Previous metaproteomic15 and microarray12 studies rarely detected the expression of ppx likely due to technical limits. Here, the expression of ppx was clearly detected with mRNA-RPKM values of 7.2 and 4.6 in the anaerobic and aerobic samples, respectively. Both ppk1 and ppx genes showed little expression difference under the two conditions with a fold change of less than 2. However, the phosphotransferase (pap) that catalyzes the reversible conversion of AMP to ADP at the expense of polyP was absent from the CAP IB HKU-1 draft genome or the other assembled contigs. Whether pap is encoded in the unread part of the CAP IB genome requires further validation. Pi Transport. Pi transport is a critical step for EBPR metabolism and directly affects the performance of phosphorus removal. Two active Pi transport systems, low-affinity Pi transport (pit) and phosphate-specific transport (pst), were identified in the CAP IIA UW-1 genome.3 Pit is formed by a single protein and functions as periplasmic Pi-binding proteins, mediating Pi-translocation and energizing Pi transport collaboratively, whereas pst is a complex system composed of four proteins, pstS, pstC, pstA, and pstB.38 Possibly due to the limited probes targeting phosphate/sulfate permease in the pit system and low detection sensitivity, respectively, microarray12 and metaproteomics13 detected no or very low expression of pit. Here, CAP IB HKU-1 pit was moderately up-regulated in the aerobic sample indicated by the 2.8-fold change. The pstC, pstA, and pstB sequences were identified in a CAP IB HKU-1 contig, whereas pstS was missing from the metagenomic assembly or the raw reads. The three pst (C, A, and B) proteins were found to be moderately up-regulated under the aerobic conditions with a maximum fold change of 2.1. The induction of pit and pst most likely occurred when cells were starved of Pi; thus it is not surprising that these genes were up-regulated in the initial aerobic stage after Pi was almost depleted from the intracellular storage and accumulated in bulk liquid in the anaerobic stage. Pi was thought to combine with ADP to form ATP after entering the cells through the F-type H+-transporting ATPase (FATPase) and/or the V-type H+-transporting ATPase (VATPase).3 V-ATPase was missing from the CAP IB HKU-1 10369

dx.doi.org/10.1021/es502642b | Environ. Sci. Technol. 2014, 48, 10363−10371

Environmental Science & Technology

■

ACKNOWLEDGMENTS The study was funded by Hong Kong General Research Fund (HKU7190/12E). Y. Mao, K. Yu, Y. Xia, and Y. Chao thank HKU for the Postgraduate Scholarship. We gratefully acknowledge Jing Guo and Huanzi Zhong from BGI (Shenzhen, China) for their help on metagenomic and metatranscriptomic sequencing.

most significantly up-regulated genes were responsible for assimilatory sulfate reduction (cysIHDN), genetic information processing (ribosomal protein) and Pi absorption (pst and pit), whereas the down-regulated genes were related to N2O reduction (nosZ), PHA synthesis (phaC), and transformation from pyruvate to acetyl-CoA formation (aceEF). However, the metatranscriptomic experiment employing only two sludge samples without any replication or internal controls limits the confidence of the interpretations for fold-changes in gene expression. Since both RNA samples were extracted when acetate was almost depleted, the gene responses to famine and feast conditions, such as those for carbon metabolism, the TCA cycle and glycolysis, were not thoroughly determined. The integrated “omics” approach has some advantages over RT-qPCR and microarray in revealing gene dynamics responsive to stimuli. This approach may uncover expression changes of unknown or ignored genes such as the hypothetical proteins, whereas the latter two methods can only be used to analyze the regulation of well-defined genes using limited primers/probes. Moreover, this combined method is useful for determining the interaction between coexisting microbes, that is, the surrounding genomes of CAP IB are retrieved to determine their cooperation or competition in the EBPR process. However, the “omics” approaches are still costly and time-consuming at this time being. Both RT-qPCR and microarray have been shown to have quite low detection limits and can quantify the gene expression levels within a short time. Therefore, combining “omics” with RTqPCR and/or microarray will be more powerful for exploring the microbial community and their functional dynamics.44 In addition to metagenomics and metatranscriptomics, metaproteomics has been well developed and is being increasingly used in further microbial functional studies45,46 although it shows some limits in the resolution of protein identification and characterization. In summary, although the application of the integrated “omics” approach in retrieving the Accumulibacter genome and analyzing its gene responses to EBPR stimuli is still in an emerging stage, this study generated another Accumulibacter reference genome, CAP IB HKU-1, for studies on inter- and intra- physiological differences among the Accumulibacter types, and the understanding of its gene regulation patterns at the aerobic/anaerobic stages will assist further proteomic exploration. Reporting of this data is significant because of the importance of Accumulibacter in wastewater treatment and the limited data on these uncultured strains. The potential functional difference among the Accumulibacter genomes may contribute to derive new methods for isolating this PAO and trouble-shooting in EBPR plants.

■

■

REFERENCES

(1) Mino, T.; van Loosdrecht, M. C. M.; Heijnen, J. J. Microbiology and biochemistry of the enhanced biological phosphate removal process. Water Res. 1998, 32 (11), 3193−3207. (2) Lemos, P. C.; Serafim, L. S.; Santos, M. M.; Reis, M. A. M.; Santos, H. Metabolic pathway for propionate utilization by phosphorusaccumulating organisms in activated sludge: 13C labeling and in vivo nuclear magnetic resonance. Appl. Environ. Microbiol. 2003, 69 (1), 241−251. (3) Martín, H. G.; Ivanova, N.; Kunin, V.; Warnecke, F.; Barry, K. W.; McHardy, A. C.; Yeates, C.; He, S.; Salamov, A. A.; Szeto, E.; Dalin, E.; Putnam, N. H.; Shapiro, H. J.; Pangilinan, J. L.; Rigoutsos, I.; Kyrpides, N. C.; Blackall, L. L.; McMahon, K. D.; Hugenholtz, P. Metagenomic analysis of two enhanced biological phosphorus removal (EBPR) sludge communities. Nat. Biotechnol. 2006, 24 (10), 1263−1269. (4) Bond, P. L.; Erhart, R.; Wagner, M.; Keller, J.; Blackall, L. L. Identification of some of the major groups of bacteria in efficient and nonefficient biological phosphorus removal activated sludge systems. Appl. Environ. Microbiol. 1999, 65 (9), 4077−4084. (5) Yuan, Z.; Pratt, S.; Batstone, D. J. Phosphorus recovery from wastewater through microbial processes. Curr. Opin. Biotechnol. 2012, 23 (6), 878−883. (6) Hesselmann, R. P. X.; Werlen, C.; Hahn, D.; van der Meer, J. R.; Zehnder, A. J. B. Enrichment, phylogenetic analysis and detection of a bacterium that performs enhanced biological phosphate removal in activated sludge. Syst. Appl. Microbiol. 1999, 22 (3), 454−465. (7) McMahon, K. D.; Yilmaz, S.; He, S.; Gall, D. L.; Jenkins, D.; Keasling, J. D. Polyphosphate kinase genes from full-scale activated sludge plants. Appl. Microbiol. Biotechnol. 2007, 77 (1), 167−173. (8) He, S.; Gall, D. L.; McMahon, K. D. “Candidatus Accumulibacter” population structure in enhanced biological phosphorus removal sludges as revealed by polyphosphate kinase genes. Appl. Environ. Microbiol. 2007, 73 (18), 5865−5874. (9) Peterson, S. B.; Warnecke, F.; Madejska, J.; McMahon, K. D.; Hugenholtz, P. Environmental distribution and population biology of Candidatus Accumulibacter, a primary agent of biological phosphorus removal. Environ. Microbiol. 2008, 10 (10), 2692−2703. (10) Flowers, J. J.; He, S.; Malfatti, S.; del Rio, T. G.; Tringe, S. G.; Hugenholtz, P.; McMahon, K. D. Comparative genomics of two ‘Candidatus Accumulibacter’ clades performing biological phosphorus removal. ISME J. 2013, 7 (12), 2301−2314. (11) He, S.; McMahon, K. D. ‘Candidatus Accumulibacter’ gene expression in response to dynamic EBPR conditions. ISME J. 2011, 5 (2), 329−340. (12) He, S.; Kunin, V.; Haynes, M.; Martin, H. G.; Ivanova, N.; Rohwer, F.; Hugenholtz, P.; McMahon, K. D. Metatranscriptomic array analysis of ‘Candidatus Accumulibacter phosphatis’-enriched enhanced biological phosphorus removal sludge. Environ. Microbiol. 2010, 12 (5), 1205−1217. (13) Wilmes, P.; Wexler, M.; Bond, P. L. Metaproteomics provides functional insight into activated sludge wastewater treatment. PLoS One 2008, 3 (3), e1778. (14) Wexler, M.; Richardson, D. J.; Bond, P. L. Radiolabelled proteomics to determine differential functioning of Accumulibacter during the anaerobic and aerobic phases of a bioreactor operating for enhanced biological phosphorus removal. Environ. Microbiol. 2009, 11 (12), 3029−3044. (15) Wilmes, P.; Andersson, A. F.; Lefsrud, M. G.; Wexler, M.; Shah, M.; Zhang, B.; Hettich, R. L.; Bond, P. L.; VerBerkmoes, N. C.; Banfield, J. F. Community proteogenomics highlights microbial strain-variant

ASSOCIATED CONTENT

S Supporting Information *

This file contains Tables S1−S4, Figures S1−S9 and supplementary methods. This material is available free of charge via the Internet at http://pubs.acs.org/.

■

Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest. 10370

dx.doi.org/10.1021/es502642b | Environ. Sci. Technol. 2014, 48, 10363−10371

Environmental Science & Technology

Article

relationship to whole-genome sequence similarities. Int. J. Syst. Evol. Microbiol. 2007, 57 (Pt 1), 81−91. (32) Richter, M.; Rosselló-Móra, R. Shifting the genomic gold standard for the prokaryotic species definition. Proc. Natl. Acad. Sci. U. S. A. 2009, 106 (45), 19126−19131. (33) Mackelprang, R.; Waldrop, M. P.; DeAngelis, K. M.; David, M. M.; Chavarria, K. L.; Blazewicz, S. J.; Rubin, E. M.; Jansson, J. K. Metagenomic analysis of a permafrost microbial community reveals a rapid response to thaw. Nature 2011, 480 (7377), 368−371. (34) Comeau, Y.; Hall, K. J.; Hancock, R. E. W.; Oldham, W. K. Biochemical model for enhanced biological phosphorus removal. Water Res. 1986, 20 (12), 1511−1521. (35) Oehmen, A.; Lemos, P.; Carvalho, G.; Yuan, Z.; Keller, J.; Blackall, L.; Reis, M. Advances in enhanced biological phosphorus removal: From micro to macro scale. Water Res. 2007, 41 (11), 2271−2300. (36) Akiyama, M.; Crooke, E.; Kornberg, A. An exopolyphosphatase of Escherichia coli. J. Biol. Chem. 1993, 268 (1), 633−639. (37) Lee, S. J.; Lee, Y. S.; Lee, Y. C.; Choi, Y. L. Molecular characterization of polyphosphate (PolyP) operon from Serratia marcescens. J. Basic Microbiol. 2006, 46 (2), 108−115. (38) Aguena, M.; Yagil, E.; Spira, B. Transcriptional analysis of the pst operon of Escherichia coli. Mol. Genet. Genomics 2002, 268 (4), 518−524. (39) Flowers, J. J.; He, S.; Yilmaz, S.; Noguera, D. R.; McMahon, K. D. Denitrification capabilities of two biological phosphorus removal sludges dominated by different ‘Candidatus Accumulibacter’ clades. Environ. Microbiol. Rep. 2009, 1 (6), 583−588. (40) Tavares, P.; Pereira, A. S.; Moura, J. J.; Moura, I. Metalloenzymes of the denitrification pathway. J. Inorg. Biochem. 2006, 100 (12), 2087− 2100. (41) Kristiansen, R.; Nguyen, H. T.; Saunders, A. M.; Nielsen, J. L.; Wimmer, R.; Le, V. Q.; McIlroy, S. J.; Petrovski, S.; Seviour, R. J.; Calteau, A.; Nielsen, K. L.; Nielsen, P. H. A metabolic model for members of the genus Tetrasphaera involved in enhanced biological phosphorus removal. ISME J. 2013, 7 (3), 543−554. (42) Gonzalez, P. J.; Correia, C.; Moura, I.; Brondino, C. D.; Moura, J. J. Bacterial nitrate reductases: Molecular and biological aspects of nitrate reduction. J. Inorg. Biochem. 2006, 100 (5−6), 1015−1023. (43) Nobu, M. K.; Tamaki, H.; Kubota, K.; Liu, W. T. Metagenomic characterization of ‘Candidatus Defluviicoccus tetraformis strain TFO71,’ a tetrad-forming organism, predominant in an anaerobicaerobic membrane bioreactor with deteriorated biological phosphorus removal. Environ. Microbiol. 2014, DOI: 10.1111/1462-2920.12383. (44) Parkhill, J. What has high-throughput sequencing ever done for us? Nature Rev. Microbiol. 2013, 11 (10), 664−665. (45) Maron, P.-A.; Ranjard, L.; Mougel, C.; Lemanceau, P. Metaproteomics: A new approach for studying functional microbial ecology. Microbial. Ecol. 2007, 53, 486−493. (46) Kleiner, M.; Wentrup, C.; Lott, C.; Teeling, H.; Wetzel, S.; Young, J.; Chang, Y. J.; Shah, M.; VerBerkmoes, N. C.; Zarzycki, J.; Fuchs, G.; Markert, S.; Hempel, K.; Voigt, B.; Becher, D.; Liebeke, M.; Lalk, M.; Albrecht, D.; Hecker, M.; Schweder, T.; Dubilier, N. Metaproteomics of a gutless marine worm and its symbiotic microbial community reveal unusual pathways for carbon and energy use. Proc. Natl. Acad. Sci. U. S. A. 2012, 109 (19), E1173−1182.

protein expression within activated sludge performing enhanced biological phosphorus removal. ISME J. 2008, 2 (8), 853−864. (16) Zhang, T.; Shao, M. F.; Ye, L. 454 pyrosequencing reveals bacterial diversity of activated sludge from 14 sewage treatment plants. ISME J. 2012, 6 (6), 1137−1147. (17) Liu, Y.; Zhang, T.; Fang, H. H. P. Microbial community analysis and performance of a phosphate-removing activated sludge. Bioresour. Technol. 2005, 96 (11), 1205−1214. (18) Altschul, S. F.; Gish, W.; Miller, W.; Myers, E. W.; Lipman, D. J. Basic local alignment search tool. J. Mol. Biol. 1990, 215 (3), 403−410. (19) Yu, K.; Zhang, T. Metagenomic and metatranscriptomic analysis of microbial community structure and gene expression of activated sludge. PLoS One 2012, 7 (5), 1−13. (20) Qin, J.; Li, R.; Raes, J.; Arumugam, M.; Burgdorf, K. S.; Manichanh, C.; Nielsen, T.; Pons, N.; Levenez, F.; Yamada, T.; Mende, D. R.; Li, J.; Xu, J.; Li, S.; Li, D.; Cao, J.; Wang, B.; Liang, H.; Zheng, H.; Xie, Y.; Tap, J.; Lepage, P.; Bertalan, M.; Batto, J. M.; Hansen, T.; Le Paslier, D.; Linneberg, A.; Nielsen, H. B.; Pelletier, E.; Renault, P.; Sicheritz-Ponten, T.; Turner, K.; Zhu, H.; Yu, C.; Li, S.; Jian, M.; Zhou, Y.; Li, Y.; Zhang, X.; Li, S.; Qin, N.; Yang, H.; Wang, J.; Brunak, S.; Dore, J.; Guarner, F.; Kristiansen, K.; Pedersen, O.; Parkhill, J.; Weissenbach, J.; Meta, H. I. T. C.; Bork, P.; Ehrlich, S. D.; Wang, J. A human gut microbial gene catalogue established by metagenomic sequencing. Nature 2010, 464 (7285), 59−65. (21) Albertsen, M.; Hugenholtz, P.; Skarshewski, A.; Nielsen, K. L.; Tyson, G. W.; Nielsen, P. H. Genome sequences of rare, uncultured bacteria obtained by differential coverage binning of multiple metagenomes. Nat. Biotechnol. 2013, 31 (6), 533−538. (22) Zhu, W.; Lomsadze, A.; Borodovsky, M. Ab initio gene identification in metagenomic sequences. Nucleic Acids Res. 2010, 38 (12), e132. (23) Albertsen, M.; Hansen, L. B. S.; Saunders, A. M.; Nielsen, P. H.; Nielsen, K. L. A metagenome of a full-scale microbial community carrying out enhanced biological phosphorus removal. ISME J. 2012, 6 (6), 1094−1106. (24) Miller, C. S.; Baker, B. J.; Thomas, B. C.; Singer, S. W.; Banfield, J. F. EMIRGE: Reconstruction of full-length ribosomal genes from microbial community short read sequencing data. Genome biol. 2011, 12 (5), R44. (25) Miller, C. S.; Handley, K. M.; Wrighton, K. C.; Frischkorn, K. R.; Thomas, B. C.; Banfield, J. F. Short-read assembly of full-length 16S amplicons reveals bacterial diversity in subsurface sediments. PLoS One 2013, 8 (2), e56018. (26) Hess, M.; Sczyrba, A.; Egan, R.; Kim, T. W.; Chokhawala, H.; Schroth, G.; Luo, S.; Clark, D. S.; Chen, F.; Zhang, T.; Mackie, R. I.; Pennacchio, L. A.; Tringe, S. G.; Visel, A.; Woyke, T.; Wang, Z.; Rubin, E. M. Metagenomic discovery of biomass-degrading genes and genomes from cow rumen-supporting information. Science 2011, 331 (6016), 463−467. (27) Dupont, C. L.; Rusch, D. B.; Yooseph, S.; Lombardo, M. J.; Richter, R. A.; Valas, R.; Novotny, M.; Yee-Greenbaum, J.; Selengut, J. D.; Haft, D. H.; Halpern, A. L.; Lasken, R. S.; Nealson, K.; Friedman, R.; Venter, J. C. Genomic insights to SAR86, an abundant and uncultivated marine bacterial lineage. ISME J. 2012, 6 (6), 1186−1199. (28) McIlroy, S. J.; Albertsen, M.; Andresen, E. K.; Saunders, A. M.; Kristiansen, R.; Stokholm-Bjerregaard, M.; Nielsen, K. L.; Nielsen, P. H. ‘Candidatus Competibacter’-lineage genomes retrieved from metagenomes reveal functional metabolic diversity. ISME J. 2013, 8 (3), 613− 614. (29) Finn, R. D.; Clements, J.; Eddy, S. R. HMMER web server: Interactive sequence similarity searching. Nucleic Acids Res. 2011, 39, W29−37. (30) Ishii, S.; Suzuki, S.; Norden-Krichmar, T. M.; Tenney, A.; Chain, P. S.; Scholz, M. B.; Nealson, K. H.; Bretschger, O. A novel metatranscriptomic approach to identify gene expression dynamics during extracellular electron transfer. Nat. Commun. 2013, 4, 1601. (31) Goris, J.; Konstantinidis, K. T.; Klappenbach, J. A.; Coenye, T.; Vandamme, P.; Tiedje, J. M. DNA-DNA hybridization values and their

■

NOTE ADDED AFTER ASAP PUBLICATION The Supporting Information file has been revised from the original posting of August 13, 2014. The correct version was posted August 14, 2014.

10371

dx.doi.org/10.1021/es502642b | Environ. Sci. Technol. 2014, 48, 10363−10371