A metagenome of a full-scale microbial community carrying out enhanced biological phosphorus removal

Abstract

Enhanced biological phosphorus removal (EBPR) is widely used for removal of phosphorus from wastewater. In this study, a metagenome (18.2 Gb) was generated using Illumina sequencing from a full-scale EBPR plant to study the community structure and genetic potential. Quantitative fluorescence in situ hybridization (qFISH) was applied as an independent method to evaluate the community structure. The results were in qualitative agreement, but a DNA extraction bias against gram positive bacteria using standard extraction protocols was identified, which would not have been identified without the use of qFISH. The genetic potential for community function showed enrichment of genes involved in phosphate metabolism and biofilm formation, reflecting the selective pressure of the EBPR process. Most contigs in the assembled metagenome had low similarity to genes from currently sequenced genomes, underlining the need for more reference genomes of key EBPR species. Only the genome of 'Candidatus Accumulibacter', a genus of phosphorus-removing organisms, was closely enough related to the species present in the metagenome to allow for detailed investigations. Accumulibacter accounted for only 4.8% of all bacteria by qFISH, but the depth of sequencing enabled detailed insight into their microdiversity in the full-scale plant. Only 15% of the reads matching Accumulibacter had a high similarity (>95%) to the sequenced Accumulibacter clade IIA strain UW-1 genome, indicating the presence of some microdiversity. The differences in gene complement between the Accumulibacter clades were limited to genes for extracellular polymeric substances and phage-related genes, suggesting a selective pressure from phages on the Accumulibacter diversity.

Figure 1

Comparison of the metagenome and qFISH results for Aalborg East EBPR plant. (a) Detailed qFISH results. ‘% of all bacteria' is the specific probe compared with the EUBmix probes. The probe-defined groups in the top left panel have been grouped according to functional class and colored according to their phylogenetic classification. (b) qFISH results summed to phylum level to facilitate comparison with the metagenome data. (c) Overview of the metagenome binning based on MEGANs lowest common ancestor approach. ORFs were assigned based on a 10% bitscore difference to other BLAST hits; only nodes with over 200 ORFs assigned are shown. The numbers after descriptions denotes, respectively, the number of ORFs assigned to the particular node and the sum of ORFs assigned below the particular node. AOB, ammonia-oxidizing bacteria; GAO, glycogen-accumulating organism; NOB, nitrite-oxidizing bacteria; PAO, polyphosphate-accumulating organism.

Figure 2

Analysis of the 10 species with the highest number of ORFs assigned. ORFs were assigned based on the best BLASTP hit. (a) The number of ORFs assigned to a particular species as a function of percent identity of the OFRs compared with the species. (b) The number of ORFs assigned, the percent of the coding genome observed and percent of genes observed.

Figure 3

SEED subsystem comparison of the functional potential of the Aalborg East EBPR metagenome with a non-EBPR WWTP metagenome (Sanapareddy et al., 2009) and an average of 26 microbial metagenomes from six distinct biomes (Dinsdale et al., 2008). The data for the 26 microbial metagenomes is shown in a box plot with median, 25th percentile, 75th percentile, minimum and maximum values depicted.

Figure 4

Graphical overview of the mapping of the short metagenome reads to the genome of Accumulibacter clade IIA strain UW-1 as a function of percent identity of each read. rRNA and tRNA genes have been masked. In the back of the figure, the summed coverage for a given genome position is shown. To the right of each figure, an insert presents the overall read identity compared with the genome. (a) Overview of the full genome. Each bar represents an average read coverage of 10 000 bases. (b) Zoom on the first 50 000 bases. Each bar represents an average read coverage of 200 bases. Selected core and hypothetical genes have been marked on the figure to highlight the difference in coverage of functional annotated genes versus hypothetical proteins.

Figure 5

Circular visualization of gene clusters unique to Accumulibacter clade IIA strain UW-1. Ring two represents the full Accumulibacter genome and in ring one and three selected areas of the genome have been expanded. Below each ring, the coverage profile is shown (yellow fill). The coverage profiles were calculated on the basis of the reference mapping and binned in either 10-kb (ring two) or 200-bp fragments (ring one and three). The coverage profile maximum was set at 10 to facilitate visualization (average coverage). Small marks represent 10 kb of sequence in all rings.