Identification, Purification and Characterization of Proteins with Cytotoxic/Antitumor Activity from Chondrosia reniformis

The majority of medicines come from natural resources, and in particular a great number of bioactive molecules (more than 10,000) have been isolated until now from the marine environment, with hundreds of new compounds still being discovered every year [1]. The richest sources of sea natural products are invertebrates with Porifera being the most prolific phylum. Sponge-derived bioactive compounds show the most varied activities: antifungal, anti-HIV, antiviral, antibacterial, but the predominant action is cytotoxic/antitumor with more than 800 compounds so far identified in this category. Sporadic observations on the behavior of the marine sponge Chondrosia reniformis pointed out a possible production of cytotoxic compounds in stress conditions able to kill neighboring organisms. The aim of this work was to purify and possibly characterize the compound causing said activity by means of a step by step purification followed by in vitro cytotoxicity analyses (MTT test) on a series of human tumor cell lines (leukemia, breast cancer, pulmonary carcinoma) as well as on healthy primary cells (fibroblast and blood mononuclear cells). The MTT tests showed a predominant cytotoxicity on tumor cells compared to healthy cells, thus pointing out a possible antitumor activity. Furthermore, preliminary experiments surprisingly showed that most cytotoxic/antitumor activity was confined to the protein fraction of the crude hydrophilic extract. Thus, different methods of protein fractionation were used to isolate the bioactive protein for a final high definition mass spectrometry (HD-MS) characterization. Namely, the purification steps were: a 10 kDa-cut-off dialysis, a 30% ammonium sulphate protein fractionation and a HPLC separation by gel filtration. After the latter step, a particular low-MW HPLC peak retained the antitumor activity, thus this fraction was separated in a 2D gel electrophoresis showing the presence of three protein bands. The bands from the gel were then trypsin-digested and the derived peptides analyzed in HD-MS obtaining the sequence of a series of peptides. Finally, through an in silico analysis comparing the peptide sequences to the C. reniformis transcriptome previously obtained in our lab, we were able to identify three possible protein candidates for the cytotoxic/antitumor activity: a protein with a trefoil factor domain and two unknown proteins showing no homology to known sequences. Actually, homologues of the first protein from higher Metazoa already show in literature an involvement in stress conditions, especially in epithelial injuries. Finally, we plan for the future to obtain the three proteins in recombinant form for further characterization studies. Research supported by EU (FP7 grant agreement n: 266033 SPonge Enzyme End Cell for Innovative AppLication-SPECIAL).