When Wolfgang Vogelbein placed Pfiesteria into a container with small fish in his lab, he witnessed something startling: The so-called “cell from hell” literally ate the flesh off the fish.

“They just keep coming and coming,” said Vogelbein, a fish pathologist at the Virginia Institute of Marine Science, as a video showed hundreds of microscopic dots — Pfiesteria shumwayae cells — swarming to latch onto, and ultimately killing a larval fish.

More curious, though, was what Vogelbein and his colleagues didn’t see: evidence that the Pfiesteria cells used a toxin to stun or kill their prey.

“These observations don’t mean that toxic strains of Pfiesteria do not exist,” Vogelbein said. “It’s entirely possible that there are toxic strains. Our results show that there is an additional mechanism by which Pfiesteria kills fish. We would like to determine how strains used by others are really killing fish.”

The findings, recently published in the journal Nature, was one of three recent papers that challenge public perceptions about the organism that garnered national headlines when it was blamed for killing more than 30,000 fish and sickening more than three dozen people in Maryland’s Pocomoke River in 1997.

The papers not only raise questions about Pfiesteria’s toxicity, but also raise doubts about whether it is actually a bizarre organism with two dozen distinct life stages. Researchers in one of the new papers said they could document only seven or eight life stages. Another study failed to find any toxin-producing genes like those found in other types of algae known to harm fish.

Joann Burkholder, the North Carolina State University professor who discovered Pfiesteria, dismissed the new papers as part of a coordinated “press blitzkrieg attack” on the nearly 50 papers by herself and others describing Pfiesteria and its impacts. The new papers, she said, make “sweeping conclusions” based on research done largely with a single nontoxic, strain of Pfiesteria.

The papers have added fuel to what was already one of the ugliest disputes in environmental science. Researchers on both sides hurl highly unscientific slams such as “unprofessional,” “irresponsible,” and even accuse each other’s work as being “poor science.” Members of both camps have suggested that the others have acted to undercut grants, or set back the research, of their critics.

“It has been polarized for some time, and it’s not getting better, and in some ways it’s worse,” said Donald Boesch, president of the University of Maryland’s Center for Environmental Science. “It’s not the kind of professional conduct I condone, on either side.”

But he, and others, worry that the dispute has caused people to question the need for the nutrient control legislation that emerged.

Although nutrients had long been recognized as the Bay’s largest water quality threat — causing huge oxygen-depleted dead zones and devastating critical underwater grass beds — it was the Pfiesteria uproar that boosted legislation in Maryland to control runoff from farmland. “Even if there are not toxin producers, there are about 20 reasons why you want to reduce nutrient loadings,” Boesch said.

Adding more fuel to the dispute, the environmental group Waterkeeper Alliance in August accused the skeptics of not having the “appropriate expertise” to conduct Pfiesteria research. It filed a federal Freedom of Information Act request seeking volumes of data, even years of e-mail exchanges, about research by scientists who got funding from the millions of dollars that Congress has poured into the Pfiesteria issue since 1997.

Yet, despite the animosity, some outside experts say the Pfiesteria scientists may not be as far apart as it seems. “Just because there are differing hypotheses doesn’t mean they are necessarily mutually exclusive,” said Jonathan Kramer, director of the Maryland Sea Grant Program. “We haven’t worked out all the details yet. There is a lot going on out there, and it is very, very complicated.”

Less than 60 miles away from Vogelbein’s lab, other scientists scoff at any suggestion that Pfiesteria may not produce a toxin. “Ridiculous,” said Harold Marshall, an algae expert at Old Dominion University, whose lab was one of the first to collaborate with Burkholder and rear fish-killing Pfiesteria. “We are not saying that other things cannot kill fish,” he said. “But we are saying that Pfiesteria produces a toxin that does kill fish.”

That is the view that has dominated the published scientific literature for the last decade since Burkholder first reported the discovery of a “phantom dinoflagellate” — dinoflagellates are a type of swimming, single-cell plankton — that killed fish by the millions in the Albemarle-Pamlico estuary.

The bizarre organism, which Burkholder later named Pfiesteria piscicida — piscicida means “fish killer” — was often seen making rapid appearances that were associated with fish deaths, then disappearing.

Continued work by Burkholder and others suggested a highly unusual organism, fond of nutrient-laden water, which had two dozen or more life stages, including dormant cysts that remain on the bottom of the estuary, free-floating amoebae and swimming flagellates.

The presence of a large number of fish triggers encysted cells to emerge and become toxic, secreting chemicals into the water that make the fish lethargic, according to Burkholder’s work.

The actual toxin produced by Pfiesteria has not been identified, although researchers contend progress is being made.

The chemicals cause the skin to slough off, often creating bloody sores. The Pfiesteria cells eat the sloughed skin, blood, and other materials that leak from the fish.

On a more ominous note, in the process of studying the organism, Burkholder and a colleague, Howard Glasgow, became ill, suffering headaches and short-term memory losses, apparently because of toxins that were released into the air. Studies by others have shown that rats injected with cultures from Pfiesteria suffer learning impairments.

But the presence of Pfiesteria by themselves will not necessarily kill fish. Burkholder, who has identified two species of Pfiesteria — P. piscicida and P. shumwayae — has found toxic, and nontoxic, strains of each.

Her conclusions have emerged after years of painstaking work learning to rear, and maintain, toxic forms of Pfiesteria in her laboratory. Anyone can grab a sample of water that has Pfiesteria. But actually getting it to show signs of toxicity in a laboratory is a different matter. It is, said Burkholder, “an especially cantankerous organism.”

Burkholder contends that one reason the new studies — almost all of which relied on the same strain of P. shumwayae — did not confirm her work is that it was improperly cultured in the laboratory. “What I believe those people had was actually a toxic strain of Pfiesteria which could not express toxicity because of woefully inappropriate culture conditions,” she said.

After years of research, Burkholder, Marshall and others recently published a complex, standardized bioassay procedure to detect and culture toxic Pfiesteria. They said the procedures, based on more than 2,000 bioassays conducted over 12 years, set the “gold standard” as the only reliable technique available to test for actively toxic strains of Pfiesteria, and call for repeatedly exposing fish in aquariums to a series of Pfiesteria cultures and consistently producing fish deaths.

In the last 18 months, VIMS scientist Jeff Shields has killed 4,500 fish in aquariums with P. shumwayae from North Carolina waters where fish kills had taken place. In the lab, fish began dying within four to six weeks, and haven’t stopped. “Our cultures kill fish daily now,” Shields said.

The rate of death goes in cycles related to Pfiesteria concentrations. As Pfiesteria numbers increase, fish mortality in the tanks grows — sometimes reaching 100 percent. The Pfiesteria concentrations decline, along with fish mortality, only to rise a few weeks later. With high numbers of Pfiesteria cells, death can take place in a matter of hours.

VIMS scientists, who contend the cells were raised following Burkholder’s protocols, wanted to know what was killing the fish, and how. But they were skeptical of the published “gold standard” tests set out by Burkholder and colleagues. Fish placed into aquariums carry such a host of contaminants that it’s impossible to say with certainty what kills the fish. Further, when fish begin dying and decomposing, they cause a proliferation of bacteria that may adversely affect the health of remaining fish.

So, they designed a series of “fractionation” experiments in which all of the components of water from an actively killing fish tank — bacteria, dinoflagellates and other materials — were separated using centrifuges and filters. Then, each constituent was exposed separately to larval fish.

The results showed that fish died only in samples where Pfiesteria came into contact with fish: assays that included dinoflagellates and fish, and assays in which fish were exposed to raw water from the fish tank, which would have included Pfiesteria cells. Assays that just included bacteria or water without cells proved harmless.

In another test, scientists placed fish larvae into tiny containers with Pfiesteria under two methods. In one case, the Pfiesteria were directly exposed to the fish, in another, the larval fish and Pfiesteria were in the same reservoir, but were separated by a porous membrane.

Theoretically, if the presence of the fish stimulated toxin production by Pfiesteria, the chemical should be able to pass through the membrane and poison the fish. Before the test, Vogelbein tested the membrane and showed that toxins produced by other kinds of algae could pass through.

Again, the results showed that Pfiesteria killed only when they were in physical contact with the larvae. And in those instances, the cells were not stunning the fish — they were eating them.Vogelbein reported signs of fish distress similar to those reported by Burkholder. Based on microscopic observations of direct attachment and feeding by the dinoflagellate on the skin, he attributed the reaction to the larvae being eaten alive.

The conclusion, Vogelbein said, is that scientists from VIMS and collaborating institutions in Florida and North Carolina saw results similar to those reported by Burkholder and her colleagues — including rapid fish death — but without a toxin. “Superficially, we are seeing the exact same thing that Burkholder and her colleagues have described,” he said. “However, we see an entirely different mechanism of fish killing. Has this mechanism been missed all these years, or are our strains really different?”

That can only be resolved, he said, if all Pfiesteria species and strains are retested side-by-side with the new methods.

Another test, reported in the Proceedings of the National Academy of Sciences and based on work with DNA primers, also raised questions about whether P. shumwayae produces a toxin. Most toxins produced by other harmful dinoflagellates — including all such toxins known to affect fish — are enzymes known as polyketides and non-ribosomal peptides.

Scientists from VIMS and the Rosenstiel School of Marine and Atmospheric Science developed special primers to look for those genes in P. shumwayae that killed fish in their experiments. But the primers did not find the genes, suggesting that if P. shumwayae produces a toxin, it is unlike any produced by other types of harmful algae.

“Although a set of the primers found part of a gene similar to a polyketide, it may be related to ordinary fatty acid biosynthesis, not toxin production,” said Kim Reece, a VIMS scientists who worked on developing the primers. “We did not find anything that was a common harmful algal exotoxin-type gene.”

In a third paper, a team of scientists led by Wayne Litaker of the National Oceanic and Atmospheric Administration’s Center for Coastal Fisheries and Habitat Research, conducted a detailed analysis of the Pfiesteria life cycle.

As part of the study, a special probe was developed that would bind to a unique DNA sequence found only in Pfiesteria. When the probe was combined with Pfiesteria cultures that had been exposed to fish or algae, it would enter the cells in each sample, but would only bind with those that carried the unique DNA sequence. The probe did not bind to any amoebae in the water samples, suggesting that they were actually contaminants introduced with the fish.

In the study’s second part, Pfiesteria cells were carefully separated from other organisms, then reared through their entire life cycle while being observed and photographed under high-resolution video microscopy. The scientists could confirm only seven to eight life stages, not the reported 24.

“When we had those single cell isolates, we never saw amoeba in them,” said Pat Tester, a NOAA scientist who participated in the study. “We really didn’t start out to do the life cycle work, but because we were finding things at odds with what were published, we thought we had better check this out so that we had a very firm footing for our other molecular work.”

A sign on the door of Andrew Gordon’s locked Pfiesteria laboratory at Old Dominion University warns it is a biosecure area that could pose a human health threat. Gordon has no doubt that the caution is for real: He is convinced that a Pfiesteria-related toxin is killing fish in the lab — and has been doing so for nearly two years.

In his bioassays, Pfiesteria from fish-killing cultures are placed in three, side-by-side, 2.5-gallon aquariums with fish. Meanwhile, three “control” aquariums are set up that are identical in every way — except for the Pfiesteria.

“We only see fish deaths in the tanks where we inoculate Pfiesteria,” he said. “It’s a pretty clear indication that the toxicity that we are seeing is related to Pfiesteria.”

Gordon was the lead researcher for a recent paper published in the journal Harmful Algae which, in the wake of the skeptical papers, received little media attention.

The paper reaffirmed much of what Burkholder’s research has shown: That both P. piscicida and P. shumwayae produce a toxin when reared in an aquarium with fish, but lose their toxicity when removed from the tanks and fed algal prey. Toxicity reappears when they are returned to tanks with fish.

But his work goes further. He has placed water from fish-killing tanks — with all Pfiesteria and bacteria filtered out — into new containers with fish. In those cases, water associated with Pfiesteria kills even though there are no cells to come into contact with fish.

“That’s a good, strong indication that we are looking at some kind of soluble toxin,” Gordon said. “We have clearly demonstrated that it is not just physical attacks by the dinoflagellates that are killing the fish.”

Still, the solution kills more slowly than when actual Pfiesteria cells are placed in the water. That suggests that predation in combination with the toxin may be important in hastening death, Gordon said.

But unlike the labs at VIMS, the ODU researchers do not see a clear relationship between the number of Pfiesteria cells and fish mortality. Sometimes high concentrations don’t kill fish; other times low concentrations do. “We believe there may be a suite of different toxins, and with different toxicities,” Gordon said.

Yet much remains unknown about the exact conditions that trigger toxic reactions. Even when Gordon puts the exact same Pfiesteria cultures in three identical tanks, they may have dramatically different mortalities. Sometimes, one tank kills fish, and the other two do not.

That means it may be possible that other factors, such as the presence of certain bacteria, may be needed to trigger or enhance toxic effects, according to the ODU researchers. “The problem now is figuring out the exact conditions that control it,” said Marshall, who works closely with Gordon.

But, Marshall is worried that “irresponsible” challenges about Pfiesteria’s toxicity will de-emphasize Pfiesteria research and — even worse — signal that there are no human health impacts. “People in Andy [Gordon’s] lab wear respirators,” he said. “There is a physical danger there.”

Can different strains of a single-cell organism actually be as different as described in the conflicting papers? It’s possible, some say.
Wayne Coats, a dinoflagellate specialist with the Smithsonian Environmental Research Center in Maryland, said dinoflagellates in general are notoriously difficult to work with. While he praised the Litaker paper — which failed to find the amoeboid stages of Pfiesteria — as a “nice piece of work,” he cautioned that “some strains of Pfiesteria could produce life history stages that were not observed.”

Likewise, he said, the notion that some Pfiesteria strains kill by eating fish, and others by a toxin “is not mutually exclusive.”
“Science is investigation,” Coats said. “You may come up with what appear to be contradictory results at times, but when you look at them carefully, they are not always contradictory.”

Other aspects of the story may yet unfold. Allen Place, of the University of Maryland’s Center of Marine Biotechnology, believes toxicity attributed to Pfiesteria in the wild may actually stem from another dinoflagellate, Karlodinium micrum, which is often found in association with Pfiesteria.

Karlodinium has been linked to fish kills in the region, recently described in the journal Harmful Algae, and is related to the algae species which cause the toxic red tides along the Gulf Coast. Place said it releases a toxin in self defense which can stun fish if they are nearby.

Allen and colleagues have had no luck identifying any toxin stemming from Pfiesteria, although they have seen it attack fish as reported by Vogelbein and his colleagues. Allen hypothesizes Pfiesteria found at fish kills may actually be preying upon fish stunned by Karlodinium. “To me,” Place said, “Pfiesteria is more of a scavenger.”

The issue of whether Pfiesteria acts through toxins, predation — or both — will not be resolved until there is the scientific equivalent of an Old West shootout: subjecting the different strains of Pfiesteria used in different laboratories to the exact same tests.

That’s not likely to happen soon. The cultures used by VIMS scientists and their colleagues are available to other researchers — Gordon and Burkholder both believe they can coax toxicity out of them. But the strains Burkholder has raised are not available for the VIMS researchers to apply their techniques.

“This controversy will not be resolved until those cultures are made available,” Vogelbein said. “It’s that simple.”

Burkholder’s refusal to share has been criticized by a number of scientists in the field. The journal Science recently called the refusal of scientists to share materials used in a published experiment as “tasteless.”

Burkholder and her supporters defend the decision to not freely release the cultures for a variety of reasons. It took years to perfect the conditions under which the cells would produce toxins, they say, and it is expensive to produce. Without enough grant support, their ability to produce large amounts is limited. In one instance, Burkholder said she gave Pfiesteria cultures to a skeptic who then threw it out.

She also said handing out cultures would compete with the research under way by herself and her graduate students.

Burkholder said she had made material available to 29 researchers at 21 laboratories, most of whom she works closely with and who had come to her lab to get extensive training on how to successfully — and safely — rear and maintain toxic cultures. Without that training, she said, simply handing over cultures is pointless.

“This is not just a culture issue,” she said. “This is an expertise issue. There is an apprenticeship sort of thing required here even though the protocols are published because it’s a complicated set of procedures. It has taken us almost 10 years to learn how to do it.”

Marshall, who has worked closely with Burkholder and collaborated on several papers, said he has never had problems obtaining her cultures. Burkholder said the VIMS research team never made a formal request for cultures until Aug. 5, the same day their Nature paper was published.

VIMS scientists say numerous requests have been made, both by themselves and other interested researchers, both in writing, and at public forums. Place also said he had made requests.

While the debate over cultures goes on, Burkholder and colleagues are formally challenging the recently published papers and their “sweeping, falsely based conclusions.” Vogelbein welcomed the challenge, and said he and his colleagues are responding to the complaint.

In the meantime, scientists on both sides say more papers are forthcoming.