August 12, 2002
Volume 80, Number 32
CENEAR 80 32 p. 9
ISSN 0009-2347


Putative toxin's role in fish kills is questioned by two studies


A toxin produced by Pfiesteria dinoflagellates has been thought to be the culprit behind dramatic East Coast fish kills and sickness in people who've had contact with infested waters. But two new studies cast doubt on whether such a toxin actually exists.

One study, aimed at detecting toxins produced by P. shumwayae
strain CCMP 2089, finds no evidence for such compounds. The other finds that the strain kills fish by feeding on them, not by releasing toxins.

In the first study, a team led by Robert E. Gawley, a chemistry professor at the University of Miami, finds that seawater in which CCMP 2089 has been actively killing fish loses its lethal activity after centrifugation. Even when cells in the water are lysed before centrifugation, the supernate has no effect on fish. And when infested seawater is lyophilized and extracted with dichloromethane and methanol, none of the extracts is active (Proc. Natl. Acad. Sci. USA, published online Aug. 5,

The team also searched for genes involved in polyketide synthesis, because fish toxins produced by other dinoflagellates are known to be polyketides. The researchers found no such genes. Nor did they find genes encoding certain peptide toxins.

Meanwhile, biologist Wolfgang K. Vogelbein and others at Virginia Institute of Marine Science (VIMS), Gloucester Point, have demonstrated that fish do not die if they are separated from CCMP 2089 by a membrane through which toxins, if present, could pass. Using various techniques, they show that CCMP 2089 kills fish by attaching to their skin and sucking out nutrients (Nature, published online Aug. 7,

The two studies challenge current views about Pfiesteria activity.

SKEPTIC Gawley says that, although Pfiesteria indeed kills fish, it does not do so by producing toxins.
For example, the feeding observed by Vogelbein causes damage to fish skin that is different from lesions found in sickened menhaden, which are often assumed to be due to Pfiesteria toxic activity. But Vogelbein says that previous work by him and others has shown that "
Pfiesteria-characteristic" lesions of menhaden are due to a fungus-like agent.

In addition, the studies call into question the assumption that the ill effects of Pfiesteria-infested waters on human health are due to a Pfiesteria toxin. Those effects may be due to other organisms, Gawley says. For example, at a fish kill in Delaware in 2000, his coauthor Daniel G. Baden found an organism that produces brevetoxin, which is a fish toxin and a neurotoxin, he points out.

Joann M. Burkholder, the director of the Center for Applied Aquatic Ecology at North Carolina State University, Raleigh, claims that her group has isolated a neurotoxin from Pfiesteria cultures. Once the patent is in place, she says, a paper will be published. The existence of this purified toxin--and previous demonstrations by her group and others that Pfiesteria can kill fish without physical contact--refutes the conclusions of Gawley and Vogelbein, she says.

Their studies, she says, used strains of Pfiesteria that she considers benign and are consistent with previous findings for benign strains. She adds that toxic strains are difficult to maintain, suggesting that the inability to detect a toxin may reflect the researchers' "lack of training" in handling toxic strains.

The Gawley and Vogelbein teams assert that their CCMP 2089 cultures are indeed toxic: They cause stress, disease, and death in fish bioassays.

"In light of the two studies, it will be incumbent on the Burkholder group, when they produce this toxin that they say they have isolated, to show that Pfiesteria made it," Gawley says. "If it's a polyketide, for example, I'm going to be very skeptical that it came from Pfiesteria."


Chemical & Engineering News
Copyright © 2002 American Chemical Society