Spillover

by David Quammen

  Group M was the most widespread and nefarious. The letter M stood for “main,” because that group accounted for most of the HIV infections worldwide. Without HIV-1 group M, there was no global pandemic, no millions of deaths. Group O was the second to be delineated, its initial standing for “outlier,” because it encompassed only a small number of viral isolates, mostly traceable to what seemed an outlier area relative to the hotspots of the pandemic: Gabon, Equatorial Guinea, and Cameroon, all in western Central Africa. By the time a third major group was discovered, in 1998, it seemed logical to label that one N, supposedly indicating “non-M/non-O” but also filling in the alphabetical sequence. (Years later, a fourth group would be identified and labeled P.) Group N was extremely rare; it had been found in just two people from Cameroon. The rarity of N and O put group M dramatically in relief. M was everywhere. Why had that particular lineage of virus, and not the other two (or three), spread so broadly and lethally around the planet?

  Parallel research on HIV-2, the less virulent virus, also found distinct groups but even more of them. Their labeling came from the beginning of the alphabet rather than the middle, and by the year 2000 seven groups of HIV-2 were known: A, B, C, D, E, F, and G. (An eighth group, turning up later, became H.) Again, most of them were extremely rare—each represented, in fact, by a viral sample taken from only one person. Groups A and B weren’t rare; they accounted for the majority of HIV-2 cases. Group A was more common than group B, especially in Guinea-Bissau and Europe. Group B was traceable mainly to countries on the eastern end of West Africa, such as Ghana and Côte d’Ivoire. Groups C through H, although tiny in total numbers, were significant in showing a range of diversity.

  As the new century began, AIDS researchers pondered this roster of different viral lineages: seven groups of HIV-2 and three groups of HIV-1. The seven groups of HIV-2, distinct as they were from one another, all resembled SIVsm, the virus endemic in sooty mangabeys. (So did the later addition, group H.) The three kinds of HIV-1 all resembled SIVcpz, from chimps. (The eventual fourth kind, group P, is most closely related to SIV from gorillas.) Now here’s the part that, as it percolates into your brain, should cause a shudder: Scientists think that each of those twelve groups (eight of HIV-2, four of HIV-1) reflects an independent instance of cross-species transmission. Twelve spillovers.

  In other words, HIV hasn’t happened to humanity just once. It has happened at least a dozen times—a dozen that we know of, and probably many more times in earlier history. Therefore it wasn’t a highly improbable event. It wasn’t a singular piece of vastly unlikely bad luck, striking humankind with devastating results—like a comet come knuckleballing across the infinitude of space to smack planet Earth and extinguish the dinosaurs. No. The arrival of HIV in human bloodstreams was, on the contrary, part of a small trend. Due to the nature of our interactions with African primates, it seems to occur pretty often.

  90

  Which raises a few large questions. If the spillover of SIV into humans has happened at least twelve times, why has the AIDS pandemic happened only once? And why did it happen when it did? Why didn’t it happen decades or centuries earlier? Those questions entangle themselves with three others, more concrete, less speculative, to which I’ve already alluded: When, where, and how did the AIDS pandemic begin?

  First let’s consider when. We know from Michael Gottlieb’s evidence that HIV had reached homosexual men in California by late 1980. We know from the case of Grethe Rask that it lurked in Zaire by 1977. We know that Gaëtan Dugas wasn’t really Patient Zero. But if those people and places don’t mark a real beginning point in time, what does? When did the fateful strain of virus, HIV-1 group M, enter the human population?

  Two lines of evidence call attention to 1959.

  In September of that year, a young print-shop worker in Manchester, England, died of what seemed to be immune-system failure. Because he spent a couple years in the Royal Navy before returning to his hometown and his job, this unfortunate man has been labeled “the Manchester sailor.” His health went into decline after his naval hitch, which he served mainly but not entirely in England. At least once he sailed as far as Gibraltar. Back in Manchester by November 1957, he wasted away, suffering some of the symptoms later associated with AIDS, including weight loss, fevers, a nagging cough, and opportunistic infections, including Pneumocystis jirovecii, but no underlying cause of death could be determined by the doctor who did the autopsy. That doctor preserved some small bits of kidney, bone marrow, spleen, and other tissues from the sailor—embedding them in paraffin, a routine method for fixing pathology samples—and reported the case in a medical journal. Thirty-one years later, in the era of AIDS, a virologist at the University of Manchester tested some of those archived samples and believed he found evidence that the sailor had been infected with HIV-1. If he was correct, then the Manchester sailor would be recognized retrospectively as the first case of AIDS ever documented in the medical literature.

  But wait. Retesting of the same samples by a pair of scientists in New York, several years later, showed that the earlier HIV-positive result must have reflected a laboratory mistake. The bone marrow now tested negative. The kidney material again tested positive but in a way that rang alarms of doubt: HIV-1 evolves quickly, and the genetic sequence of virus from the kidney sample seemed far too modern. It looked more like a modern variant than like something that could have existed in 1959. That suggested contamination with some recent strain of the virus to account for the positive tests. Conclusion: The Manchester sailor may have died from immune-system failure but HIV probably wasn’t the cause. His case merely illustrates how tricky it can be to make a retrospective diagnosis of AIDS, even with the presence of what seems to be good evidence.

  Soon after that false lead from Manchester was debunked, another lead emerged in New York. By now it was 1998. A team of researchers including Tuofu Zhu, based at the Rockefeller University, obtained an archival specimen from Africa dating back to the same year as the sailor’s, 1959. This time it wasn’t tissues; it was a small tube of blood plasma, drawn from a Bantu man in what had been Léopoldville, capital of the Belgian Congo (nowadays Kinshasa, capital of DRC) and stored for decades in a freezer. The man’s name and his cause of death weren’t reported. His sample had been screened during an earlier study, in 1986, along with 1,212 other plasmas—some archival, others new—from various locations in Africa. This man’s was the only one that tested unambiguously positive for HIV. Tuofu Zhu and some colleagues probed further, working with what little remained of the original sample and using PCR to amplify fragments of the viral genome. Then they sequenced the fragments to assemble a genetic portrait of the Bantu man’s virus. In their paper, published in February 1998, they called the sequence ZR59, referencing Zaire (as the country had long been known) and the year 1959. Comparative analysis showed that ZR59 was quite similar to both subtype B and subtype D (finer divisions within the HIV-1 group M lineage) but fell about halfway between, which suggested that it must closely resemble their common ancestor. In other words, ZR59 was a glimpse back in time, a genuinely old form of HIV-1, not a recent contamination. ZR59 proved that HIV-1 had been present—simmering, evolving, diversifying—in the population of Léopoldville by 1959. In fact it proved more. Further analysis of ZR59 and other sequences, led by Bette Korber of the Los Alamos National Laboratory, yielded a calculation that HIV-1 group M might have entered the human population around 1931.

  For a decade, from the Zhu publication in 1998 until 2008, that landmark stood alone. ZR59 was the only known version of HIV-1 from a sample taken earlier than 1976. Then someone found another. This one became known as DRC60, and by now you can probably decode the label yourself: It came from the Democratic Republic of the Congo (same nation, latest name) and had been collected in 1960.

  DRC60 was a biopsy specimen, a piece of lymph node snipped from a living woman. Like the Manchester sailor’s bits of kidney and spleen, it had been locked away in a little pat of paraffin. Thu
s preserved, it needed no refrigeration, let alone freezing. It was as inert as a dead butterfly and less fragile. It could be stored and ignored on a dusty shelf—as it had been. After more than four decades, it emerged from a specimen cabinet at the University of Kinshasa and offered a new jolt of insight to AIDS researchers.

  91

  The University of Kinshasa sits on a hilltop near the edge of the city, reachable by an hour’s taxi ride through the broken streets, the smoggy sprawl, the snarled traffic of vans and busses and pushcarts, past the street-side vendors of funerary wreaths, the cell-phone-recharge kiosks, the fruit markets, the meat markets, the open-air hardware stores, the tire-repair shops and cement brokers, the piles of sand and gravel and garbage, the awesome decrepitude of a postcolonial metropolis shaped by eight decades of Belgian opportunism, three decades of dictatorial misrule and egregious theft, and then a decade of war, but filled with 10 million striving people, some of whom are dangerous thugs (as in all cities) and most of whom are amiable, hopeful, and friendly. The university campus, on its hill, loosely called “the mountain,” presents a relatively verdant and halcyon contrast to the city below. Students go there, climbing by foot from a crowded bus stop, to learn and to escape.

  Professor Jean-Marie M. Kabongo is head of pathology in the university’s Department of Anatomic Pathology. He’s a small, natty man with a huge graying handlebar mustache and full muttonchops, making a forceful visual impression that’s vitiated by his gentle manner. When I met him in his office, on the second floor of a building that overlooks a grassy concourse shaded with acacias, he pleaded imperfect knowledge of DRC60 and the patient from whom that specimen came. An old case, after all, going back long before his time. Yes, a woman, he believed. His memory was vague but he could check the records. He began taking notes as I questioned him and suggested I come back in a couple days, when he might be better prepared with answers. But then I asked about the room where DRC60 had been stored, and he brightened. Oh, of course, he said, I can show you that.

  He fetched a key. He unlocked a blue door. Swinging it open, he welcomed me into a large sunlit laboratory with walls of white tile and two long, low tables down the middle. On one of the tables rested an old-fashioned folio ledger, with curling pages, like something from Chancery in the time of Dickens. On the far windowsill stood a row of beakers containing liquids in increments of color, beaker by beaker, from piss-yellow to vodka-clear. The yellowest, Professor Kabongo told me, was methanol. The clearest was xylol. We use these in preparing a tissue sample, he said. The point of such organic solvents is to extract the water; desiccation is prerequisite to fixing tissues for the long term. The methanol was darkened from processing many samples.

  He showed me a small orange plastic basket, with a hinged lid, about the size and shape of a matchbook. This is a “cassette,” Professor Kabongo explained. You take a lump of tissue from a lymph node or some other organ and enclose it in such a cassette; you soak the whole thing in the beaker of methanol; from the methanol, it goes through the intermediate baths in sequence; finally you dunk it in the xylol. Methanol draws out the water; xylol draws out the methanol, preparing your specimen for preservation in paraffin. And this device, Professor Kabongo said, indicating a large machine on one of the tables, delivers the paraffin. You take a leached tissue sample from its cassette, he explained, and, from that spigot, you dribble out a stream of warm, liquid paraffin. It cools on the sample like a pat of butter. Now you remove the cassette lid and label the base with an individual code—for instance, A90 or B71. That’s your archival specimen, he said. “A” means that it came from an autopsy. “B” indicates a biopsy. So the paraffin-caked bit of lymph node that yielded DRC60 would have been labeled B-something. Each coded specimen gets recorded in the big ledger. Then the specimens go into storage.

  Storage. Storage where? I asked.

  At the far end of the lab was another doorway, this one hung with a blue curtain. Professor Kabongo pushed the curtain aside and I followed him into a specimen pantry, narrow and tight, lined with shelves and cabinets along one side. The shelves and cabinets contained thousands of dusty paraffin blocks and old microscope slides. The paraffin blocks were in stacks and cartons, some of the cartons dated, some not. It appeared to be organized chaos. A wooden stool awaited use by any curious, tireless soul wishing to rummage through the samples. Although I didn’t plan to rummage, my tour had suddenly come to its crescendo. Here? Yes, just here, said the professor. This is where DRC60 sat for decades. He could have added, with local pride: before becoming a Rosetta stone in the study of AIDS.

  92

  From the pantry behind the blue curtain, that sample and hundreds of others had traveled a circuitous route, to Belgium and then the United States, ending up in the laboratory of a young biologist at the University of Arizona. Michael Worobey is a Canadian, originally from British Columbia, whose specialty is molecular phylogenetics. After his undergraduate work he went to Oxford on a Rhodes scholarship, which ordinarily means two years of mildly strenuous academic work plus lots of tea, sherry, tennis on grass, and genteel anglophilia before the “scholar” returns to professional school or a career. Worobey put Oxford to more serious use, staying on, finishing a doctorate and then a postdoc fellowship in evolutionary biology at the molecular level. From there he returned to North America in 2003, accepting an assistant professorship at Arizona and building himself a BSL-3 lab for work on the genomes of dangerous viruses. Several years later, it was Worobey who detected evidence of HIV in a certain Congolese biopsy specimen from 1960.

  Worobey amplified fragments of the viral genome, pieced the fragments together, recognized them as an early version of HIV-1, and named the sequence DRC60. Comparing his sequence with ZR59, the other earliest known strain, he reached a dramatic conclusion: that the AIDS virus has been present in humans for decades longer than anyone thought. The pandemic may have gotten its start with a spillover as early as 1908.

  To appreciate Worobey’s discovery and how it splashed down amid previous ideas, you’ll need to know a little context. That context involved a heated dispute over just how HIV-1 entered the human population. The prevailing notion as of the early 1990s, based on what had been learned about HIV-2 and the sooty mangabey, among other factors, was that HIV-1 also came from an African primate, and that it had probably gotten into humans by way of two separate instances (for groups M and O, the ones then recognized) of butchering bushmeat. This became known as the cut-hunter hypothesis. In each instance, a man or a woman had presumably butchered the carcass of an SIV-positive primate and suffered exposure through an open wound—maybe a cut on the hand, or a scratch on the arm, or a raw spot on any skin surface that got smeared with the animal’s blood. A wound on the back might have sufficed, if the carcass were draped over shoulders for carrying home. A wound in the mouth, if some of the meat were consumed raw. All that mattered was blood-to-blood contact. The cut-hunter hypothesis was speculative but plausible. It was parsimonious, requiring few complications and no unlikelihoods. It fit the known facts, though the known facts were fragmentary. And then in 1992 a contrary theory arose.

  This one was heterodox and highly controversial: that HIV-1 first got into humans by way of a contaminated polio vaccine tested on a million unsuspecting Africans. The vaccine itself, by this theory, had been an unintended delivery system for AIDS. Someone, according to the theory, had monumentally goofed. Someone was culpable. Scientific hubris had overridden caution, with catastrophic results. The scariest thing about the polio-vaccine theory was that it also seemed plausible.

  Viruses are subtle, as you’ve seen. They get in where they shouldn’t. Laboratory contaminations occur. Even viral or bacterial contamination of a vaccine at the production level—it has happened. Back in 1861, a group of Italian children vaccinated against smallpox, with material direct from a “vaccinal sore,” came down with syphilis. Smallpox vaccine administered to kids in Camden, New Jersey, at the start of the twentieth century, seems to have
been contaminated with tetanus bacillus, resulting in the death of nine vaccinated children from tetanus. Around the same time, a batch of diphtheria antitoxin prepared in St. Louis, using blood serum from a horse, also turned out to carry tetanus, which killed another seven children. Producers then began filtering vaccines, an effective precaution against bacterial contamination; but viruses passed through the filters. Formaldehyde was sometimes added to inactivate a target virus, and that supposedly killed unwanted viruses too, but the supposition wasn’t always correct. As late as midcentury, some of the early batches of the Salk polio vaccine were contaminated with a virus known as SV40, endemic in rhesus macaques. SV40 in vaccine became a hot issue, several years later, when suspicions arose that this virus causes cancer.

  Whether vaccine contamination happened with HIV-1, and far more consequentially, is another matter. That the vaccine in question had been given to Africans was not in dispute. Between 1957 and 1960, a Polish-born American researcher named Hilary Koprowski—a lesser-known competitor in the same vaccine-development race that engaged Salk and Sabin—arranged for his candidate vaccine to be widely administered in areas of the eastern Belgian Congo and adjacent colonial holdings. These were parts of what would eventually be DRC, Rwanda, and Burundi. Koprowski himself visited Stanleyville, in 1957, and made contacts who later oversaw the trials. Children and adults lined up trustingly, in places like the Ruzizi Valley north of Lake Tanganyika, to receive oral doses of liquid vaccine from a tablespoon or a squirting pipette. Spritz, you’re good. Next! The numbers are uncertain. By one account, roughly seventy-five thousand kids were vaccinated just in Léopoldville. The heterodox theory argued two additional points about this enterprise: First, that Koprowski’s vaccine was produced by growing the virus in chimpanzee kidney cells (rather than in monkey kidney cells, the standard technique); second, that at least some batches of that vaccine were produced from chimpanzee kidneys drawn from animals infected with SIVcpz.