The purposes of this study were to characterize the complexation of toxic metals with algal toxins and to determine the effects of arsenic and copper on the growth of Karenia brevis under specific experimental conditions. Microcystins, pahayokolides, brevetoxins and okadaic acid were used as representatives of algal toxins while arsenic, copper, cadmium, cobalt, iron, manganese and mercury were selected as typical toxic metals (including metalloids here) in the aquatic environment. The stabilities of the toxin-metal complexes were determined using equilibrium dialysis and/or centrifugal ultrafiltration technique. A direct exposure of arsenic and copper to the K. brevis was carried out to determine the effects of these metals to the growth of the algal cell. The results indicated that Cu2+, Hg2+, Co2+, Cd2+ and Fe2+ were capable of complexing with the algal toxins. Moreover, the exposure experiments demonstrated that the high concentration of arsenic and copper could affect the growth of the K. brevis.

Pahayokolides A-D are cytotoxic cyclic polypeptides produced by the freshwater cyanobacterium Lyngbya sp. strain 15-2 that possess an unusual β-amino acid, 3-amino-2,5,7,8-tetrahydroxy-10-methylundecanoic acid (Athmu). The absolute configuration of pahayokolides A-D was determined using advanced Marfey’s method. It was also confirmed that a pendant N-acetyl-N-methyl leucine moiety in pahayokolide A was absent in pahayokolides B and pahayokolides C-D were conformers of pahayokolide A. Feeding experiments indicated that the biosynthesis of the Athmu sidechain arises from leucine or α-ketoisovalerate, however could not be further extended by three rounds of condensation with malonate units. Putative four peptide and one unique polyketide synthetases in Lyngbya sp. strain 15-2 were identified by using a PCR method and degenerate primers derived from conserved core sequences of known NRPSs and PKSs. Identification of one unique KS domain conflicted with the logic rule that the long side chain of Athmu was assembled by three rounds of ketide extensions if PKSs were involved. A gene cluster (pah) encoding a peptide synthetase putatively producing pahayokolide was cloned, partially sequenced and characterized. Seven modules of the non-ribosomal peptide synthetase (NRPS) were identified. Ten additional opening reading frames (ORFs) were found, responsible for peptide resistance, transport and degradation. Although the predicted substrate specificities of NRPS agreed with the structure of pahayokolide A partially, the disagreement could be explained. However, no PKS gene was found in the pah gene cluster.