Medicine October 2000

The Hunt for the Origin of AIDS

The notion that AIDS arose from a polio vaccine made with contaminated chimpanzee cells—the thesis of the best-selling book The River—is far from the only theory about how the epidemic started, and it is hotly disputed. The quest for the source of the epidemic is intensifying, as researchers scour the jungle for clues and try to "walk back" the disease genetically with the help of the world's most powerful computers
The Conclusions of "Nirvana"

MEANWHILE, dozens of journalists gathered in San Francisco, for the 2000 retrovirus conference. This year they swarmed around Bette Korber, a scientist at the Los Alamos National Laboratory. Korber had just presented new data that, like Beatrice Hahn's, challenged the OPV theory. The frenzied media reaction created the false impression that she had solved the origin mystery and disproved the OPV theory. It also overshadowed another origin talk presented that day, one that deserved at least equal attention.

The little-noted presentation came from François Simon, the chief researcher at the Franceville lab, who also works in Cameroon. Simon announced the discovery of the three SIV-infected chimps in Cameroon and two new SIVcpzs that had been isolated from them. He and his colleagues had also found strains of HIV-1 in two Cameroonians which closely matched the simian viruses. This was the first hard evidence that an SIVcpz had recently made a leap into human beings and established an HIV-1 infection. It significantly bolstered the natural-transfer argument.

Although the data that Korber had compiled, in collaboration with the biologist Carla Kuiken, Hahn, and others, were more theoretical, they offered a tidier news story. Korber maintains a database that catalogues HIV isolates and viral fragments that have been genetically sequenced. Using one of the world's fastest computers, "Nirvana" (an array of up to 2,048 processors that operate in concert), Korber and her colleagues analyzed the virus's evolutionary history. They focused on the main group of HIV-1, which has eleven strains, or subtypes.

By plugging in known mutation rates of HIV and comparing the genetic makeup of these subtypes, the researchers calculated that an ancestor common to the subtypes dated to 1930, plus or minus twenty years. As a control, the group showed that the program's "molecular clock" could accurately date the 1959 sample from Léopoldville and also the virus's subtype E, which is known to have surfaced in Thailand around 1986.

Korber is a soft-spoken researcher who shies away from dramatic pronouncements and appends caveats to almost everything she says. "Our results don't disprove [the OPV] hypothesis but make it unlikely," she said, pointing out that the CHAT trials did not begin until 1957. Hooper had received advance word of Korber's results and had already challenged them at a press conference about The River. He noted, rightly, that no one knows when this common ancestor jumped from chimpanzees to human beings. He then suggested that the various subtypes could have evolved in chimps, and speculated that the OPV manufacturers had pooled kidney cells from ten Lindi chimps infected with distinct SIVcpzs. This, he said, could account for the diverse viruses in human beings. Korber replied with further understatement: "This seems to be quite an unlikely scenario," she said.

The press persuaded Korber to hold an impromptu conference. She grimaced her way through it, explaining at great length and in carefully measured terms precisely why the OPV hypothesis underwhelmed her. "I wouldn't go so far as to say it could not have happened," she said in reply to a question from Lawrence K. Altman, the New York Times medical reporter, who last November wrote a favorable column about The River that greatly increased the amount of attention paid to the book. Korber went on to say that the OPV thesis simply clashed with the genetic diversity of the main HIV group. If the OPV theory were true, she explained, each of the known HIV-1 subtypes would have had to contaminate the vaccine, enter a human being in a viable form, establish an infection, and then spread exponentially; each "founder" would have had to start a distinct epidemic. "I find the numbers unlikely because it's difficult for zoonosis to take," Korber said. (Hooper and others have also posited that a single chimp had the prototype for every HIV subtype, but Korber says that's even more of a stretch.)

Near the end of the question-and-answer session a reporter remarked that Stanley Plotkin was in the room. Plotkin is currently a consultant to the French vaccine manufacturer Aventis Pasteur; he collaborated closely with Koprowski in the making and testing of CHAT, subsequently developed several vaccines himself, and co-authored the classic textbook on the subject. "I wasn't planning to participate in this conference, which is about Dr. Korber's work," Plotkin said. "I would reiterate, first of all, that no chimpanzee tissues were used in any oral polio vaccine. As far as her work is concerned, clearly it supports the idea that the introduction of HIV from chimpanzees occurred well before the vaccination campaign in the Belgian Congo."

Plotkin had recently helped to find the labs to test the remaining CHAT in Wistar's freezers. After the press conference I asked him what his reaction would be if the vaccine tested positive for chimp DNA—or, even more damning, for SIVcpz. "As a scientist, I have to accept scientific data," Plotkin said. "From what I know, I consider that unlikely." Plotkin told me recently that after the publication of The River he got in touch with all the principal scientists and technicians who had worked on CHAT at Wistar and in the labs in Belgium and the Congo. To a person, they said that they had not worked with chimpanzee cells. Plotkin planned to detail their statements, which he had received in writing, at the Royal Society meeting.

Both support for and criticism of Korber's data and her 1930 date (since revised to 1931, with a range of 1915 to 1941) have come from unlikely quarters. Jim Moore, a primatologist at the University of California at San Diego, more or less stumbled into the debate over her theory, having published an origin thesis of his own in AIDS Research and Human Retroviruses on the eve of the conference. Moore (no relation to the AIDS researcher John Moore) and his co-workers combed the literature on the Belgian Congo and pre-World War II French Equatorial Africa—the region that includes the Republic of the Congo, Gabon, and the Central African Republic. They concluded that the human epidemic probably started between 1890 and 1930.

Moore began by documenting how the colonial powers' use of forced labor to build railroads and other infrastructure led many African villagers to flee to the forest, where they would have depended more on bush meat than they had previously. Labor camps separated families and encouraged promiscuity, and many villagers moved to newly established cities (Kinshasa's population, he notes, grew tenfold from 1905 to 1940). And wars and civil unrest, of course, have been constants in most African countries for much of this time.

In addition to these social forces, which HIV naturally exploits, Moore discovered evidence of shoddy medical practices that could have greatly amplified the spread of the virus. As many as 35,000 people living in French Equatorial Africa from 1893 to 1912 may have been vaccinated against smallpox directly from another person's arm: doctors would withdraw liquid from a pox blister on someone who had recently been vaccinated and inject it into someone else (this method was adopted to save the trouble of transporting large quantities of vaccine to remote locales). A 1917-1919 campaign against sleeping sickness used six syringes for 90,000 people.

I met with Moore earlier this year in his office, whose walls display detailed maps of Tanzania, where he has done fieldwork. He immediately made it clear that he is not hostile to the OPV theory. He greatly respects his late colleague William Hamilton, and is also impressed by Hooper. "Ed has spent ten years of his life pursuing a hypothesis," Moore says, adding that he likes the OPV hypothesis because it is plausible and can to some extent be tested. "I'm betting when it's tested, it will be falsified. But it's not obviously wrong. How are you going to falsify the scratched-hunter hypothesis?" he asks. "It's almost impossible."

However, Moore is skeptical about Hooper's emphasis on geographic links between early cases and CHAT vaccination sites. "I want to see an epidemiologist look at where they were taking samples," he says. "It sounds, as Hooper puts it, like it's remarkably coincident, but until it's looked at carefully, quantitatively, there are grounds for suspicion. I'm just guessing, but it strikes me as plausible that people would be vaccinated at places where there are clinics or where doctors go regularly—and where the clinics are and the doctors go is where you'd have samples."

Donald Burke, formerly the head of the U.S. military's AIDS research program and currently the director of the Center for Immunization Research at Johns Hopkins University, questions the validity of every model that has estimated the epidemic's time of origin, because none has yet allowed for a process called recombination—essentially, the splicing of genes from two sources. A person may become infected with two distinct subtypes of HIV, and they may recombine and form a new strain. Burke believes that HIV may well have relied on recombination to create a strain that could readily spread from person to person. The urbanization of Africa and the imposition by colonialists of a common language, Burke argues, brought together groups—and HIV subtypes—that had previously had little contact with one another, providing opportunities for the development of recombinants and for the spread of viable strains. It is as if someone poured all the various viral strains into one container and then shook it, Burke says. And because recombinant strains form branches of odd lengths in unusual places on a phylogenetic tree, current modeling programs cannot factor in recombination with any accuracy.

Gerald Myers, a scientist at Los Alamos who was once Korber's boss, firmly believes in the usefulness of computer modeling, and has high praise for the process that has allowed Korber to estimate the beginnings of the epidemic; however, he, too, has strong reservations about Korber's conclusions, for reasons that further complicate matters. "The burden is on proponents of natural-transfer theory to tell us what happened between the 1930s and the 1970s," Myers says. Myers's own group approached the question from the opposite direction, working forward from a number of given dates to the present day. "We can model this as a believable process if it started as late as the late 1950s," he says. "We cannot model it, so far, if we start earlier."

Myers is a polymath who speaks slowly and deliberately. He pointed out a pattern that puzzles him: the various HIV subtypes, when placed on the virus's evolutionary tree, are equally distant from one another, indicating that they branched from a common ancestor at roughly the same time. Standard Darwinian evolution displays a more uneven pattern, he explains. "We expect a lot of variance in the HIV tree, and there's very, very little—surprisingly little. So you have to say at first glance at the mathematics that something extraordinary has happened. It could be needles. It could be that somebody blew a whistle and funny things started happening, like in an Anthony Burgess novel."

Myers and his colleagues have analyzed the evolutionary tree formed by subtypes of feline immunodeficiency virus (FIV), a distant relative of HIV that causes an AIDS-like disease in cats. "A lot of things humans did happened to cats," Myers says. "They've suffered urbanization. Cats moved to the cities, they changed their sex habits, and who knows what." And the FIV-subtype tree looks nothing like the one for HIV. "It's a very messy tree by comparison," he says.

HIV1 family tree Myers clearly recognizes that all modeling is based on assumptions and therefore has limitations; he stresses that at best his work and Korber's will help to guide other scientists in their thinking about which experiments to conduct. "The problem will come back to haunt us if we don't try to get to the bottom of this," he says. "None of us wants the polio-vaccine story to hold up. It's a horrible thought. But I think we need to go ahead and face it squarely."

The Hunt Continues

FRANÇOIS D'Horpock, a pathologist in Abidjan, Côte d'Ivoire, led me into a dark and dusty, long-ignored room in his laboratory at the University of Treichville. Scattered about were more than 2,000 tissue samples from autopsies dating as far back as 1967; each was numbered and stored in a chunk of paraffin the size of an ice cube. Metal cabinets along the walls held thousands of numbered slides containing blood smears and the like. Giant ledgers contained the case history of each sample.

Before going to Gabon, I met with AIDS researchers in Côte d'Ivoire, Kenya, Uganda, Rwanda, and South Africa. I sought out D'Horpock and other pathologists to see if anyone had old tissue or blood samples that researchers might test for evidence of HIV. Such samples could shore up what is perhaps the single weakest aspect of origin-of-AIDS studies generally: researchers have identified very few cases in Africa prior to 1980, and have even fewer samples of old HIV isolates.

Shortly after the epidemic surfaced, several research groups did screen old medical records and archives of stored blood; this is what led to the discovery of the "plausible" AIDS cases and the positive blood samples that Hooper documents. But, as he details in The River, other stored samples exist that no one has checked. My hunch is that Albert Schweitzer's old hospital in Lambaréné, Gabon, which was founded in 1913, might contain samples of particular relevance: the hospital is located in a region where some of the SIVcpz-infected chimpanzees were found.

Although positive evidence of HIV in Abidjan from, say, 1967 would rewrite the history of the epidemic, D'Horpock's samples are not old enough to reveal much about the OPV hypothesis. Furthermore, he says, they were treated with formaldehyde, a process that may have degraded the viral DNA. Still, no one has even approached him to find out whether the samples would be worth testing, though there are large American- and French-funded AIDS-research efforts on the campus. "Research takes time and is expensive," D'Horpock says with a shrug.

Even if a stored sample does provide evidence of an HIV infection from 1956 or earlier, it will have to yield highly detailed information, or Hooper will simply point out that natural transfer sometimes occurs. If the sample contains antibodies but not the virus itself, researchers will not be able to do a phylogenetic analysis. Should someone isolate an archival HIV, it will have to belong to the main group of HIV-1s in order to bear on the OPV theory. And the genetic sequence will have to fit into the evolutionary tree at the right point in time.

Still, there are compelling reasons to try to settle the issue. Kevin De Cock, a leading AIDS epidemiologist at the U.S. Centers for Disease Control and Prevention, who started a large AIDS-research project in Abidjan in 1988 and now works in Kenya, used to question the relevance of origin studies. "Does any of this matter?" De Cock says. "I've changed my mind. You can't have fifty million infected, sixteen million of whom have died, and say it doesn't matter how this came about. Conditions that expose humans to these simian viruses have probably increased. You can't say from a public-health perspective that those are of no importance."

A better understanding of the origin of AIDS could have an impact on efforts to combat other diseases as well. Omu Anzala, an AIDS-vaccine researcher at the University of Nairobi, says that ripples from The River have "caused many problems" in Kenya. "When people read The River and they are not very scientific, the arguments are pretty convincing," he says. Claiming that the vaccine contains HIV, some Kenyan clergymen have recently discouraged their countrymen from taking part in the current campaign to eradicate polio (which uses a thoroughly tested, contaminant-free OPV). "In the last six or seven months we've been trying to vaccinate as many people as we can," Anzala says. "But certain segments of society have been saying, 'Who knows whether the vaccine isn't contaminated?'"

PARTWAY down the packed-mud road that leads from Franceville to Okondja, Telfer stopped his Land Cruiser so that we could speak with two hunters. One man had a shotgun slung over his shoulder; the other had a machete and a tall wooden cane. Both appeared to be at least in their fifties. Through an interpreter who works with Telfer, I asked the men their ages. Neither had any idea. The interpreter asked them a question of his own: "Have you ever heard of World War Two?" Again they both lifted their shoulders and shook their heads no.

Telfer was making the four-hour round-trip drive to Okondja to take a blood sample from a pet monkey he had found on an earlier expedition. Farther along the route he spotted a dead monkey for sale, splayed on top of an upside-down empty fuel can. Telfer does not want to pay for animals and thereby indirectly encourage trade in bush meat, and so, with his interpreter's help, he delicately negotiated with a woman living in a nearby shack. She agreed to let him take samples from the monkey, a member of Cercopithicus cephus—a species in which no one had yet found an SIV. After putting on latex gloves, he plucked some hair from the animal; he would later do a DNA test to determine where the monkey fits on the cephus family tree. Telfer wrapped a strip of rubber around the monkey and hung it from a hand-held scale: 6.6 pounds. He opened the monkey's mouth and clipped off a bloody piece of tongue. He would have liked to open the belly and take the spleen—a likelier place for SIV to hide—but the village men who by then had clustered around were becoming agitated, and the interpreter discreetly herded us back to the car. A half hour or so later we came upon another recently killed cephus in front of a roadside shack. After similar negotiations with this monkey's owner, Telfer took a tissue sample from the animal's rectum—a location where HIV is frequently found in people.

When we got to Okondja and located the pet, a cephus monkey living in a makeshift cage alongside a house, the owner was not at home. We headed to the marketplace for lunch, where I had a stew made of porcupine (which is not a protected species). The locals took obvious delight in watching us eat there, which pleased Telfer: an important part of his job, he says, is building trust. Over lunch we met another man who has a pet monkey, which Telfer hopes to sample in the future.

We headed back to the caged pet cephus. Its owner had returned, and he agreed to let Telfer take a sample. The first step was to anesthetize the animal with a shot of ketamine. Telfer donned a long leather glove, lifted the tin top off the cage, and tried to grab the monkey. The monkey, of course, wanted no part of this, and madly dodged Telfer's every lunge. Tension rose, amplified by the crowd that had gathered. At last Telfer succeeded in anesthetizing the animal. He placed it on the ground, drew blood from the femoral vein, measured its weight and size, and marked it with a tattoo gun.

As we prepared to leave, a neighbor came up to tell us that she, too, had a pet monkey. We walked a few houses over and saw the animal, a baby cephus, but Telfer decided that it was too small to draw blood from, so we headed back to Franceville. Telfer was pleased with the day's tally: tongue from one fresh kill, rectal tissue from another, blood from a pet, and one adult and one baby pet for future testing.

In June, Preston Marx told me that the pet cephus whose blood we had sampled showed evidence of SIV infection. I sent Telfer an e-mail about it; the subject line on his reply read, "interesting? maybe." As often happens in this kind of work, the preliminary data turned out to be as confusing as they are intriguing: the monkey's blood showed SIV antibodies in one test but not in another. However, researchers extracted strands that appear to be pieces of an SIV, and analysis is proceeding.

Further information about the origin of AIDS will soon emerge from the many research efforts under way. Paul Telfer and his colleagues plan to go to the village where Za-Za came from and take blood samples from human beings there. Telfer is also preparing to collect fecal samples from chimps in the wild. Beatrice Hahn has begun to agree with primatologists who question whether chimp subspecies truly exist. She recently analyzed fecal and urine samples from chimps in Uganda and Côte d'Ivoire and blood from Pan paniscus in the Congo. All tested negative for SIVs, which has led her to wonder whether any chimps outside west central Africa are infected. Martine Peeters, a French researcher who has done pioneering work on the chimp isolates from Gabon, has been testing primate meat from a marketplace in Cameroon. Donald Burke has begun working in Cameroon and hopes to start testing primates in the wild there. At the international AIDS conference held last July in Durban, South Africa, Anne-Mieke Vandamme, a Belgian researcher, presented data that support Bette Korber's 1931 date by means of an entirely different methodology. Even more intriguing, Vandamme's work dates the emergence of SIVcpz and HIV-1 from a common ancestor to about 1700. The analysis of the CHAT samples may suggest directions for new investigations. And the Royal Society meeting will undoubtedly fuel new research as well.

But as Telfer so trenchantly observed, standing quietly in the jungle and looking at nothing in particular, 90 percent of the game is waiting and not seeing much—until the day we see something spectacular.

Presented by

Jon Cohen is a correspondent with Science and the author of Almost Chimpanzee (Times Books/Henry Holt & Company, 2010), from which his October 2010 Atlantic story was adapted.

Join the Discussion

After you comment, click Post. If you’re not already logged in you will be asked to log in or register with Disqus.

Please note that The Atlantic's account system is separate from our commenting system. To log in or register with The Atlantic, use the Sign In button at the top of every page.

blog comments powered by Disqus


Cryotherapy's Dubious Appeal

James Hamblin tries a questionable medical treatment.


Confessions of Moms Around the World

In Europe, mothers get maternity leave, discounted daycare, and flexible working hours.


How Do Trees Know When It's Spring?

The science behind beautiful seasonal blooming

More in Technology

More back issues, Sept 1995 to present.

Just In