Virolution

What could cause viruses as terrifying as HIV-1 to change behaviour to this remarkable extent? This was a question guaranteed to fascinate me, so I made contact with Professor Essex, who was kind enough to agree to an interview with me.

By the spring of 1995, having digested the thinking of Terry Yates and Joshua Lederberg, and having interviewed many other leading virologists in the UK and Europe, from fields as diverse as botany, zoology and molecular biology, I was close to assembling what amounted to a jigsaw of understanding that came from adding the stories and views of these many experts into a single picture. By this stage I believed that our medical approach to the problem was inevitably skewed by our vocational, if entirely natural, concerns for humanity. Viruses had no such concerns. Thus, if I were to search for the evolutionary explanation for the emergence of plagues such as AIDS, I would have to abandon such vocational thinking, however deeply ingrained, and adopt a neutral stance that for me, as a doctor, felt curiously alien. I was obliged to ask myself this question: Is something very important going on in the world of plague viruses, something profound, which, if we could only grasp and define it, would give us a radically different perspective – a new level of understanding? So it was that, in February 1995, I made arrangements to return to America, where I planned to visit other experts at Harvard and Yale, and where I hoped to interview Essex among those based at Harvard. Unfortunately he was flying out of the country during the few weeks I would be in America, so I arranged a preliminary interview by phone two days before leaving England, to be followed up by a face-to-face meeting in the hustle and bustle of Washington National Airport.

Given my own earlier researches on the immune response to blood-borne viruses, I was very much looking forward to the prospect of talking to an investigator who shared my interest in immunity, and who had been a pioneer in linking animal and human retroviruses to the arrival of AIDS, and to the devastation of the immune system that gave rise to so many of its key symptoms and problems. During those heady years of fear and confusion, when AIDS was first challenging the world of medicine, he had also been one of the first to point to a retrovirus as the likely cause.

‘Why,’ I asked him, ‘had he become so involved with viruses?’

In fact, as he now explained to me, he had entered into a career in Veterinary Medicine because he was interested in viruses as a possible cause of chronic diseases, and cancer in particular. ‘At the time I was training, which was in the late 1960s, it was already clear that some forms of naturally occurring cancer were caused by viruses in animals, but it wasn’t yet clear in people, and there was debate about whether it would turn out to be true in people.’

We talked for a few more minutes about his earliest research, on diseases caused by retroviruses.

‘Two or three viruses were discovered in our laboratory during the time I was there, in either cats or monkeys, and it made a very solid case that long-term retroviruses – because they happened to be retroviruses – really could cause chronic diseases, such as leukaemias and certain other cancers. But then, as we were studying those parameters, we noticed that cats infected with these viruses developed a dramatic immunosuppression before they even developed cancer, and it even happened in the absence of developing cancer. Subsequently, it became clear that the immunosuppression was related to the strain of retrovirus … [This assumed a more societal importance when] it became clear that Gallo had identified the first human retrovirus.’

The virus to which Essex was referring is now known as the human T-cell leukaemia virus, or HTLV-1, which causes T-cell leukaemia, a cancer involving the white blood cells known as lymphocytes, which play a key role in the body’s fight against viruses, and indeed against any form of foreign invasion of the blood or tissues. The story actually goes back to 1976, when a Japanese researcher, Dr Kiyoshi Takatsuki, was studying a newly discovered form of leukaemia, in which the cancerous cells had nuclei so dramatically convoluted that they looked like the bunched up petals of flowers. Takatsuki noticed that almost all of the sufferers came from Kyushu, a large island to the southwest of Japan, so he travelled to Kyushu where he found that doctors in the local hospitals were treating many people with this bizarre leukaemia. At this time nobody knew the cause of this disease, but they decided they would call it “Adult T-cell Leukaemia”.

Then, in 1979, as Essex had just remarked, Robert Gallo and his collaborators at the US National Institutes of Health found the causative virus in the blood of a 28-year-old man from Alabama, who was suffering from a lymphoid cancer of his skin. Two years later, Japanese virologists Yorio Hinuma and Mitsuaki Yoshida isolated the same retrovirus from the Japanese leukaemia patients. This virus, which is now known as human T-cell leukaemia virus one, or HTLV-1, was the first human retrovirus to be isolated.

‘This was before AIDS was recognised?’ I asked Essex.

‘Before AIDS was recognised. It was probably about the time that, or shortly after, AIDS was already present in Americans but before it was recognised and appreciated as a clinical entity.’

We had arrived at what appeared to be a key moment in the early investigation of AIDS and I encouraged him to talk about it further.

‘There was a workshop in Seattle. It was contacted by our National Cancer Institute, in collaboration with the Japanese, to see how important such human retroviruses might be. I was invited to co-chair it … I went there with Gallo and about four or five people from his lab, together with six or seven Japanese investigators working on these viruses, such as Hinuma, Miyoshi, Yoshida and Takatsuki – the people who defined the leukaemia. It became clear during the discussion that there were lots of things that really needed to be done, and they just weren’t happening. The first thing I suggested doing – and was encouraged to do myself – was to question whether or not some of those viruses, like the human T-cell leukaemia virus, might be immunosuppressive, the way retroviruses of cats were, because they infected the same T-lymphocytes.’

‘Which immune cell in particular is infected by the HTLV-1 virus?’

‘The T4 lymphocyte – exactly the same cell as AIDS. It’s not absolutely clear that it’s through the same receptor, but it’s definitely the same cell. You find all the same transmission strategies, such as sex, and mother to infant, although HIV is a little more effectively transmitted cell-free by blood.’

Even today, HTLV-1 infection remains an important source of disease in Japan, America, in the Aborigines of Australia, Peru, Colombia, Ecuador, Africa and the Caribbean – and the Inuit of Northern Canada. As Essex explained, it is transmitted in the same way as AIDS, although the pattern of transmission varies from country to country. In Japan it is mainly transmitted from mother to child through breast-feeding, while in America, Australia and South America it is mainly through sexual intercourse and through contaminated needles, syringes and blood products. The disease pattern also resembles AIDS, with most of the deaths resulting from immunosuppression. A small percentage of sufferers get progressive nerve damage, and in the very long term it can cause cancers such as leukaemia and lymphoma. A closely related virus, HTLV-2, infects intravenous drug users in America and the Caribbean, which is also associated with nerve damage.

I listened attentively as Essex moved deeper into the heart of the preliminary investigation of the AIDS pandemic.

‘So we were studying HTLV from the standpoint of immune suppression and then, with Gallo, we put forward the hypothesis that retroviruses should be considered as a possible cause of AIDS. It was at that time too, let’s say ’82 – after the clinical syndrome of human AIDS had been announced, and before any human viruses had been claimed for discovery – that the head of the New England Primate Centre, a guy named Ron Hunt, called me. He said that he had seen immunosuppression similar to human AIDS, and to the immunosuppression we had described in cats, in his monkey colony at the Harvard-associated facility about 50km away from the medical campus. He asked me if I would come out and talk to them and make some suggestions about how they might find the cause. I went out there and had discussions with Ron, and with Norman Letvin and two or three others, and I made the suggestion that we look at blood and tissue samples. We found that there was a virus in the animals that had developed lymphoma, and in the ones that were housed with those that developed lymphoma, even though they might not have lymphoma. We showed that it was a retrovirus on serology and by electron microscopy.’

I should explain that the afflicted monkeys were not African monkeys – they were Asian macaques. I asked him if the antibodies they were finding in the macaques in the monkey colony suggested that they were infected not with the human virus, HTLV-1, but with a related retrovirus.

‘Right. It showed that a lot of the monkeys that had lymphoma – and some of the ones that did not have lymphoma but were in the same facility, and were immunosuppressed – had a virus that was cross-reactive with, and morphologically similar to, HTLV-1 in Japanese people. Then, when [HIV-1] the actual cause of AIDS was discovered,

we already had samples of the causative virus [of the monkey immunosuppression] in our laboratory, and we then asked ourselves whether or not the sick monkeys had a virus exactly like [HIV-1] – and whether or not it clustered with the development of immunodeficiency.

‘I collaborated with Ron Desrosiers and Norm Letvin, with the work in my own laboratory coordinated by Phyllis Kanki, who was a doctoral student of mine at that time. We found that the monkeys that had the AIDS-like immunosuppression, and some of the ones with lymphoma too, were infected with a new virus, which we initially called STLV-III. Later, of course, it was called SIV.’

SIV is the simian immunodeficiency virus, and its discovery would play a major role in our understanding of the origins of AIDS. But there was an additional, important extrapolation that came from its study. At this time nobody knew where AIDS had come from, geographically or virologically.

‘As soon as we realised there were viruses related to HIV and HTLV in monkeys, it seemed likely these viruses must be coming from Africa, and perhaps the common link with the human AIDS virus would be African. A year or two earlier, Belgian and Dutch researchers published work on the clinical recognition of AIDS in African people. So we said, “Gee, maybe we should look at people in Africa who are high risk for this sort of infection – like female prostitutes and male patients attending STD clinics, and perhaps infectious disease patients – and see if they have a virus that fits somewhere within the spectrum between the monkey viruses and the human viruses.” So we looked at blood samples from these high-risk people and found some cross-reactive antibodies and subsequently we also found actual virus.’

Which virus was he now speaking about?

‘People were clearly infected with a virus very closely related to the monkey virus, in fact virtually indistinguishable from it. And this new virus was clearly related to HIV-1 – but it was also clearly distinguishable from HIV-1.’

Like HIV-1, this second human retrovirus would subsequently be isolated by Luc Montagnier at the Pasteur Institute, and identified as the second human immunodeficiency virus, or HIV-2. But what now interested me was the very close evolutionary link between HIV-2 and the virus Essex’s group had earlier discovered, the simian immunodeficiency virus, SIV.

‘What has been shown since then is that the monkeys in West Africa have a range of SIV viruses. Some of these viruses, from monkeys in exactly the endemic area we were studying, and from mangabeys in particular, have a simian immunodeficiency virus that is indistinguishable from the HIV-2 in people in that area. Yet there are HIV-2 viruses infecting people two or three countries away, like Ghana, that are distinguishable from the HIV-2 infecting people in Senegal, which is 80km away – even though the viruses in people and monkeys in each country are not distinguishable from each other.’

I sat back to reflect on what Essex was telling me. The simian immunodeficiency virus and the human immunodeficiency virus, HIV-2, are actually one and the same virus. In his words: ‘It’s just that you call it one thing if it’s in people and another thing if it’s in mangabey monkeys.’ But there was a further, crucial, implication of what he had discovered. We believe that HIV-1, the main virus of AIDS, was transferred to people from a specific group of chimpanzees. We also know that, in chimpanzees, HIV-1 grows freely and reproduces in their internal organs and tissues, but it causes no evidence of disease. And like HIV-1 in chimpanzees, SIV produces no evidence of disease in mangabey monkeys, even though the virus also multiples freely in the monkeys’ tissues. Yet it seems altogether likely that, on first contact between the viruses and these animal hosts, the viral behaviour is likely to have been very aggressive. If we need any confirmation, we only need recall what happened when the SIV-carrying African mangabeys were housed in the same facility as the Asian macaques at the facility near Harvard. No more is it surprising that when chimpanzees, carrying an SIV virus closely related to what we now recognise as HIV-1, came into contact with people, the forerunner of HIV-1 hopped species to cause the aggressively fatal AIDS in humans.

I posed some relevant questions:

‘Let us say that a particular virus has been infecting an animal for a very long time and the animal and virus have reached the stage where they are coexisting without the virus causing serious disease in the animal. Now say another species of animal – a similar species – comes into contact with the host. It seems likely that the virus will cross species in a very vicious manner – it may prove to be highly lethal. Is it possible that what we are seeing here is an evolutionary mechanism? I also ask myself this: What if this is a symbiotic pattern of evolution, a symbiotic relationship between virus and host? In these circumstances, what might the host animal be getting out of it? And what occurs to me is that one of the things it could be getting out of it is the advantage that if a rival, for food or whatever, comes into its ecological niche, the virus jumps species and wipes out the rival.’

‘Yes, I think that’s a very logical hypothesis. You know the system that most shaped my own thoughts on that and made me write some of the things I did, such as the Scientific American article in which I compound the monkey virus behaviour in the different species with Frank Fenner’s discussion of the myxomatosis epidemic in Australia. And the bottom line of that is that when Europeans brought captive rabbits into Australia for the first time, the rabbits escaped into the wild. And because there were no foxes or natural enemies to control the rabbit populations, they multiplied in numbers and started destroying the crops. So the people there decided they needed to kill off the rabbits. They brought in a myxomatosis virus that those rabbits had not seen before. The myxomatosis virus they brought in killed right away – because it spread very well – some 99.8% of the rabbits. But then two things happened. Number one – within four years, the resistant minority grew so you had a different population of disease-resistant rabbits. Now, even if you brought in a virulent strain it didn’t kill them. And number two – the myxomatosis virus that remained [as a persistent infection in the rabbits] was less virulent, so I think there is crystal-clear evidence that both the host and the virus attenuated themselves for optimal survival in that situation. Now, were you to bring in new rabbits, the new rabbits would be disadvantaged. The surviving rabbits still live with the virus but they are now resistant, so that they can then be totally healthy and function normally while retaining a myxomatosis virus that is still virulent enough to prove a threat to any rival rabbits coming in.’

I pressed him a little further: ‘I’m aware also that certain herpes viruses appear to be particularly venomous when they cross species in monkeys. Is it your opinion that we are seeing the same thing?’

‘Oh, absolutely! I think that some of the most dramatic examples in primates are viruses like Herpesvirus simiri and anteles. They have co-evolved in one species of monkey, like spider or squirrel monkey, and when you put them into contact with any other species of monkey they are highly, highly lethal, but in the resident co-evolving species they do nothing.’

‘Would you agree that, here again, we’re looking at an example of an evolutionary role?’

‘My guess is you could even find evidence that the monkeys that are most susceptible occupy the same ecological niche and are eating the same food, as opposed to some of the ones, even if they cover the same territory, that eat a different food and fit a different living niche, and that are not quite as susceptible [to the virus].’

When two or more partners enter into a mutualistic symbiosis, each partner will contribute an innate ability, or trait, that the other partner lacks. It is obvious what a host contributes to a virus-host partnership, since it offers the virus shelter and the use of the host’s own genetic machinery to make more copies of itself. Without the host, the virus would not survive. But although it might appear less obvious, there is in fact a key ability that the virus possesses – in evolutionary terms, a “trait” – that the host does not. This is the innate potential for lethal aggression. In the example of Elysia chlorotica, we witnessed how the retroviruses that are long-established partners within its nucleus and tissues, may end the slug’s life cycle with what appears to be a ferocious demonstration of aggression. In fact, following my researches into viral behaviour, in Virus X, I first put forward the evolutionary concept of “aggressive symbiosis” as an important mechanism – I have never claimed it to be an exclusive one – in a number of situations in nature, and in particular in the interaction between plague viruses and their hosts. But coining the mechanism was merely the first step in the hypothesis I was attempting to formulate. I now had to figure out how such a mechanism might work, in evolutionary terms, as part of the evolving partnerships of viruses and their hosts in nature.

I began by making a couple of reasonable assumptions. Up to this time, virtually all evolutionary research within the discipline of virology had been Darwinian in concept. I was familiar with its conclusions, which were central to medical virology, and, by and large, I agreed with them. I also took the view that Darwinism and symbiogenesis were not mutually exclusive. There was overwhelming evidence that both mechanisms operated in nature. This suggested that each virus-host relationship needed to be examined in its own right: but it also needed to be examined through the binocular vision of both evolutionary mechanisms, and not merely through one. Sometimes the dominant mechanism would fit the Darwinian paradigm, such as the operation of selection at selfish individual, or selfish gene, level. Sometimes the dominant mechanism would more closely fit the symbiotic paradigm, with selection operating to an important degree at the level of the partnership of virus and host. Indeed, I saw no reason why, in certain situations, both paradigms would not apply, with a dynamic that might start with dominance at selfish level, but might evolve to end up with a dominant effect at partnership level. This would fit with the original thinking of de Bary. Moreover, it would also fit with the mathematical derivations of two Oxford-based Professors, Anderson and May, who, in the early 1990s, had spent a lot of time examining virus-host dynamics, including co-evolution – a Darwinian concept that came very close to the symbiotic concept of partnership.

Over the ensuing years, I continued to work on the dynamic of emerging plague viruses, and I discovered that aggressive symbiosis worked through a series of very specific steps. It began when the virus invaded a new, or virgin, host. The interaction could result in a variety of different behaviours, depending on whether the virus came from a closely related species, and was thus pre-evolved in its infectious strategies, or whether it came from a more distantly related species, when its infectious strategies would not be so efficiently pre-evolved. The Sin Nombre hantavirus came from a rodent and, though it killed a high percentage of the people it infected, it could not efficiently transmit between people. Several recently notorious viruses Marburg, Ebola, Lassa fever and the South American haemorrhagic fever viruses, such as Machupo and Junin, did exactly the same. Lassa, Machupo and Junin all came from rodents and were not sufficiently transmissible between people. While the hosts of Marburg and Ebola were still uncertain, their failure to transmit efficiently between people suggested a distantly related host. In such cases, the genetic differences been former and new host ensured that the evolutionary dynamic ended there. But where a virus came from a closely related species, such as the rabbit myxomatosis virus, which was symbiotic with the Brazilian wood rabbit, or, as with HTLV and HIV, when it hopped species from monkeys and chimpanzees to the closely related humans, the genetic similarities paved the way for a new evolutionary dynamic. Since the tissue and immune barriers of the original hosts were very similar to those of the new host, these viruses would possess pre-evolved strategies that would work pretty much in the new host as they did in the old. Moreover, all of these viruses had a very important characteristic in common. Once they entered an individual, or species, they never went away, not in terms of the individual, and not in terms of the entire affected population, or even the species. The biological term for such a relationship is “persistence” and the viruses are said to be “persistent viruses”. The very nature of such a long-term, and inevitably intimate, relationship has major implications for the virus-host evolutionary dynamics.

For a virus to enter into such a persistent relationship with a host, it is obvious that the host must be able to survive the long-term presence of the virus. In some cases, this may not necessarily lead to any major manifestations of disease. It is possible, I would even venture likely, that many viruses in nature enter into benign partnerships without the manifestations of what we recognise as a plague – but such interactions, by their very nature, are likely to pass unnoticed. But with myxomatosis, as with HIV, the virgin host, whether Australian rabbit or global human population, cannot live in a benign harmony with the newly arrived virus. A variable proportion, 99.8% of the Australian rabbits, and perhaps as many as 98% of the human species, cannot survive the initial contact with the myxomatosis virus and HIV-1 respectively. Here the first step of aggressive symbiosis kicks in. The invading virus kills all those who cannot live with its presence. I labelled this brutal dynamic “plague culling”. This is what we saw with myxomatosis, and I’m afraid it is what we would likely have seen with AIDS had it arrived among our human, or pre-human, ancestors when they inhabited the geographically contiguous area of the hinterlands of the African rainforest. We can anticipate that plague culling would reduce the new host population to a rudiment, selected by the lethality of the virus, and genetically distinct from the majority of its former population, or species. Although the survivors might be sickly, or have their lives somewhat shortened, the key implication, from the evolutionary perspective, is that they are capable of living with the persisting presence of the virus. The second step of aggressive symbiosis involves long-term co-evolution of virus and host – with the potential of mutualistic partnership.

So went my theory. Some people might disagree with my conclusions. But I could also point out to such sceptics that it was capable of being confronted. If I was right and aggressive symbiosis, evolving to mutualism, was commonplace in nature, it could be confirmed or refuted by looking for the pathognomonic signal of such a partnership: natural selection operating at the level of the partnership.

One of the most useful probes to come from the applications of molecular genetics to evolutionary biology has been our ability to follow the changes in DNA sequences over the vastness of evolutionary time, so we can distinguish genetic sequences that have been conserved by natural selection from those that have not. Viruses, as we have seen, evolve at fantastic speeds. This means that, in the examination of established virus-host partnerships, if we detect highly conserved viral sequences, this would be suggestive of selection working at the level of the partnership. If we could then take this a step further and demonstrate that those same conserved viral sequences were contributing to host survival – or, in symbiotic terms, to survival at holobiontic level – this would provide conclusive evidence for the evolutionary paradigm of virus-host symbiosis.

In fact, when we look for evidence of selection at partnership level in nature, it proves not unduly difficult to find it.

Many of my readers, whether biologists or non-biologists, will be familiar with the cruel life cycle of the parasitic wasps, where approximately 25,000 species of insects have entered into aggressive symbiotic partnerships with approximately 20,000 species of polydnaviruses. The partnership has become so intimate that many of the viruses have entered the germ line of the wasps, to emerge, as fully formed viruses, when the wasp is laying its eggs. Whether the viruses live around the wasp ovaries, or whether they emerge from the wasp genome at the time of egg-laying, they are inevitably injected into the caterpillar prey along with the wasp’s eggs. In normal circumstances the wasp’s eggs would not survive – they would be detected and destroyed by the immune system of the caterpillar. But here viral aggression comes into play, paralysing the immune system of the caterpillar, and then taking over key aspects of its internal chemistry to convert it into a brood chamber for the emerging wasp larvae. The full complexity of the symbiosis has proved to be a source of wonder to biologists, with viruses compelling the caterpillar to produce sugars to feed the larvae, and even going so far as to disrupt the caterpillar’s hormonal system, thereby preventing its natural metamorphosis into a butterfly or moth.

There can be little doubt that here we see selection operating at the level of the partnership, with viral genes and behaviour greatly enhancing the survival potential of the wasp. And when biologists, such as Provost and Whitfield, investigated the wasp-virus partnership, they found that it dated back to a single symbiotic union, approximately 74 million years ago, during the age of the dinosaurs.