Metatranscriptomic array analysis of “Candidatus Accumulibacter phosphatis”-enriched EBPR Sludge

Here we report a metatranscriptomic analysis of gene expression and regulation of “Candidatus Accumulibacter”-enriched lab-scale sludge during enhanced biological phosphorus removal (EBPR). Medium density oligonucleotide microarrays were generated with probes targeting most predicted genes hypothesized to be important for the EBPR phenotype. The objectives were to investigate which genes were expressed during EBPR and which genes were differentially expressed between the early stage of anaerobic and aerobic phases (defined as 15 min after acetate addition and 15 min after switching to aeration respectively). A medium density oligonucleotide Combimatrix microarray, CustomArrayTM 12K, (Combimatrix Corporation, Mukilteo, WA) was designed by Combimatrix to target many predicted genes identified during the metabolism reconstruction of "Candidatus Accumulibacter phosphatis" and the flanking community members, bacterial community composition markers (such as 16S rRNA) and viral genes from viral metagenomic analysis of the same bioreactor. The 12,544 spots on the array consisted of 10,644 spots for user-designed probes, 326 spots for manufacture-designed quality controls, 70 spots for negative controls and 1,504 empty spots. A total of 3,548 unique probes, each with triplicate spots were randomly distributed on the array. Among them, 2,108 probes were designed to target 1,300 predicted genes from the bacterial metagenome. The remaining 1,440 probes targeted contigs from the viral metagenome. Among the 1,300 bacterial genes, 404 genes were targeted by 3 different probes and the remaining 896 genes were targeted by only one probe. RNA samples were collected at the early stage of anaerobic and aerobic phases (15 min after acetate addition and switching to aeration respectively). Alexa Fluor 647-labeled aRNA from the anaerobic and aerobic samples were sequentially hybridized on the microarray to perform comparison of gene expression. Briefly, after the first hybridization and imaging, the hybridized aRNA on the array was stripped off under stringent conditions. The effectiveness of stripping was verified by imaging the array. After this, a second hybridization with a different sample was hybridized to this stripped microarray. The array manufacturer indicated that an array slide could be used with good reproducibility up to four times. The replicability of sequential uses was confirmed in our study. In this way, Alexa Fluor 647-labeled aRNA from the anaerobic and aerobic samples were sequentially hybridized on the same slide. Therefore, the comparison of gene expression profiles between anaerobic and aerobic samples was performed on two consecutive hybridizations on the same slide to minimize slide- and hybridization-dependent variation. A total of four hybridizations with alternating anaerobic and aerobic samples were applied on a single slide. Two microarray slides were used. Therefore, six pairs of such comparison on two separate slides were conducted (Hybridization 1 vs 2, 2 vs 3, 3 vs 4 on Slide 1, and Hybridization 5 vs 6, 6 vs 7 and 7 vs 8 on Slide 2).