Marine micro-organisms are important components of various biogeochemical cycles, complex food webs and ecological niches. Metagenomic sequencing can provide rapid profile of metabolic activities within the sponge and resident microbes. However, the study of metatranscriptomes from sponges using high throughput sequencing technology has only recently begun. Through this study we isolated, characterized and compared metatranscriptome profiles of Axinella corrugata host and sponge-specific microbial communities using 454 pyrosequencing technology. Four cDNA libraries (two eukaryotic and two prokaryotic) were generated from Axinella corrugata sponge samples collected in December 2009 and May 2010, and were characterized to a) reveal which metabolic genes were actively expressed and b) reveal possible interactions between the sponge and its microbial symbionts. The techniques used for isolation of mRNA and cDNA normalization also helped in optimization of whole-transcriptome amplification. More than 130,000 ESTs were generated for the two seasonal sponge samples and the metagenomic data sets were analyzed using bioinformatics tool, MG-RAST. Several stress-related transcripts were found which can increase our understanding of sensitivity of the sponge to changes in physical parameters in nature. The involvement of the sponge and its microbial consortia is depicted through actively expressed nitrogen and sulfur metabolism genes. Novel genes involved in several functional pathways may be discovered upon further studying hypothetical genes found across all four metagenomic data sets. Metatranscriptomic data sheds light on the functional role of microbes within the sponges and the extent of their involvement in sponge metabolism. 16S rRNA analysis was also carried out using genomic DNA of the same samples, to better elucidate the bacterial taxa abundance in the sponge. This study provides a profile of active mRNA trancripts in Axinella corrugata which include eukaryotic as well as prokaryotic sequences. The data analysis of this research provides new information at the cross-disciplinary interface between molecular biology and computational science.