Clionapyrrolidine A--a metabolite from the encrusting and excavating sponge Cliona tenuis that kills coral tissue upon contact.

The Caribbean encrusting and excavating sponge Cliona tenuis successfully competes for space with reef corals by undermining, killing, and displacing live coral tissue at rates of up to 20 cm per year. The crude extract from this sponge, along with the more polar partitions, kills coral tissue and lowers the photosynthetic potential of coral zooxanthellae. We used a bioassay-guided fractionation of the extract to identify the compound(s) responsible. The crude extract, the aqueous partition, and compound 1, herein named clionapyrrolidine A [(-)-(5S)-2-imino-1-methylpyrrolidine-5-carboxylic acid], when incorporated into gels at close to natural volumetric concentrations, killed coral tissue when brought into forced contact with live coral for periods of 1-4 days. This is the first report of a pure chemical produced by a sponge that kills coral tissue upon direct contact. The results are consistent with the localized coral death that occurs when C. tenuis-colonized coral fragments are thrown forcibly against live coral during storms. However, healed C. tenuis fragments placed directly onto live coral were killed readily by coral defenses, and fragments placed in close proximity to coral did not have any effect on the adjacent coral tissue. Solutions of clionapyrrolidine A in sea water were only slightly toxic against live coral. Hence, the coral death naturally brought about by C. tenuis when undermining live coral does not occur through external release of allelochemicals; below-polyp mechanisms must be explored further. N-acetylhomoagmatine (2), originally isolated from Cliona celata from the Northeastern Atlantic, was also assayed for comparison purposes because of its structural similarity to siphonodictidine, a toxic compound produced by a coral excavating sponge of the genus Aka. The lack of activity of N-acetylhomoagmatine at close to natural concentrations seems to indicate that the guanidine moiety, which is also present in siphonodictidine, is not a sufficiently strong structural motif for activity against corals.