ÐÓ°ÉÔ­´´

THE REALISATION nearly six years ago that the disease AIDS was caused by a new virus HIV raised a deeper mystery – where did such a virus come from? The simple answer is still that no one knows. Nevertheless, some very strange hypotheses have been advanced. For example, a former Astronomer Royal postulated that such a virus could have come from outer space. Almost simultaneously, there was the conspiracy theory. To a politically cynical generation raised on Frederick Forsyth and John Le Carre, it seemed plausible that HIV had escaped from a germ warfare laboratory, prejudice dictating whether this was an act of Eastern or Western folly. In the current atmosphere of glasnost, this idea now seems embarrassingly silly.

More discomforting was the suggestion that HIV might have started life as a ‘harmless’ contaminant of a human vaccine which, once injected into people, changed its properties and produced a lethal disease. There are no grounds, epidemiological or biological, for believing that this has happened with HIV, but there are two good reasons for considering the possibility long enough to be able to discard it with an easy mind. The first is that such a phenomenon is not unheard of. Strains of virus have been known to jump species, with disastrous consequences. For example, the parvovirus, feline leukopenia virus in its natural host, the cat, is quite harmless. But the modified vaccine strain inoculated experimentally into dogs proved to be lethal to the new hosts. The second reason for considering HIV in the light of such a scenario is that the opportunity for such an incident existed, theoretically at least.

In the early days of polio vaccine production, the live virus vaccine was propagated in the laboratory on cultures of kidney cells taken from African green monkeys. These cells could conceivably have been contaminated with the simian immunodeficiency virus now known to be harboured by the African green monkey. This virus, which is called SIV_AGM was discovered only in 1985 by Professor Hayami in Tokyo. Although SIV_AGM causes no disease in the African green monkey, could it, as a vaccine contaminant, have become adapted to live in humans over the years, resulting in the new pandemic, AIDS? The answer is no – for two reasons. First, the epidemiology – the pattern of HIV infection and its timing – argues emphatically against such a possibility. Secondly, now that SIV_AGM has been cloned and sequenced, we know that it is not sufficiently close to HIV in its make-up to support the argument.

So, if such theories have no foundation, where has HIV come from? To answer this question scientists have been comparing the make-up of the HIVs with other known animal viruses to see if they are related and whether anything can be deduced about the evolution of the human immunodeficiency viruses.

Family resemblances

ÐÓ°ÉÔ­´´s compare viruses genetically rather in the way that parents and grandparents scan their offspring for family traits. In looking for similarities of character and personality among relations we look at the eyes, the nose, the hands and feet, the smile and mannerisms. ÐÓ°ÉÔ­´´s do the same with viruses. They look at the DNA sequence of the HIVs – that is, the order of the bases which make up the DNA and provide the information needed to produce proteins – and compare it with that of other viruses. In this way, researchers can estimate the overall similarity or homology, of two viruses.

But this approach can sometimes be misleading. Imagine that a baby resembles his father only by having the same shape of hands and perhaps a defect on one finger. That part of the child would constitute only a tiny, but telling, proportion of familial likeness, which might easily be missed if father and son were compared by making an overall comparison. If the son’s hand were compared with that of the father, however, one would look like a small replica of the other, so revealing a much higher degree of homology in this area. In much the same way, researchers compare viruses by looking at the genes in the viruses one at a time. To gain insight into HIV’s ancestry it makes sense to start by comparing it within the extended family of viruses that are similar in structure. Virologists have classified HIV as a ‘retrovirus’. Such viruses have no DNA initially and must convert their RNA to DNA by means of a special enzyme called reverse transcriptase. Retroviruses are further classified into subgroups on the basis of what they look like under the electron microscope and the diseases they cause. The viruses causing AIDS in people and animals, such as visna virus in sheep and equine infectious anaemia virus (EIAV), belong to the ‘slow’ or lentivirus group, so called because there is a long delay between infection and the onset of disease. These viruses have similarities within their pol gene (see Table 1). The fact that the pol gene has such ‘conserved’ sequences which remain stable from generation to generation indicates that the lentiviruses are all closely related to one another. But the pol gene of HIV-1 is not sufficiently similar to the others to suggest that the viruses have recently spread from one species to another. However, the 84 per cent homology within this gene between HIV-2 and SIV_MAC deserves closer scrutiny.

In 1985, Danny Daniels at the New England Regional Primate Center isolated the first simian immunodeficiency virus (SIV_MAC) from a rhesus macaque. This was an exciting discovery, coming as it did the year after the discovery of HIV-1. One year later, a second AIDS (HIV-2) was discovered in Senegalese prostitutes. The second virus is restricted mainly to West Africa. Although the two viruses are similar in appearance, general structure and the disease they cause, they are sufficiently different to be considered separate viruses. For example, HIV-2 has an extra gene (the function of which is unclear) and the immune system appears to respond differently to infection by this virus. This is illustrated by the fact that antibodies made against HIV-1 are different from those made against HIV-2. (In a test-tube experiment, antibodies that can inhibit the infection of cells by one virus rarely inhibit the other as well.) The discovery of SIV_MAC alerted everyone to the possibility that the human AIDS virus might have originated in monkeys. This seemed a reasonable idea because SIV_MAC looked like HIV on electron microscope pictures, appeared to share biological properties with HIV and, most interesting of all, could induce an AIDS-like disease in macaques. Animals infected with SIVMAC wasted away and suffered from diarrhoea. Their immune systems were deficient, leading to opportunistic infections as well as lymphomas and encephalitis – all reminiscent of the human disease.

Several virus strains of SIV_MAC have now been cloned and the DNA sequences compared with HIV-1 and HIV-2. Interestingly, SIV_MAC appears to be more closely related to HIV-2 (there being a high overall sequence homology of 75 per cent) than to HIV-1 (where there is an overall 40 per cent homology). The homology is not restricted to the pol gene, but extends to other gene sequences as well (Table 2). There are further similarities. For example, like HIV-2, SIV_MAC has an extra gene. Furthermore, blood taken from a monkey infected with SIV_MAC may contain antibodies that can recognise and bind to some of the proteins of HIV-2 and vice versa (a reaction known as serological cross-reactivity).

It is tempting to speculate that SIV_MAC might be the forerunner of HIV-2, at least. There are, however, weaknesses in such an extrapolation. In the first place, although the two viruses are closely related, several of their genes are significantly different, which argues against a direct relationship. The extent of the divergence of the sequences (about 15 per cent in the gag and pol genes and about 30 per cent in the env gene) is higher than that found among the most dissimilar forms of HIV-1 (a maximum of 10 per cent in pol and between 15 and 20 per cent in env). We do not yet know how much these two groups of viruses vary genetically because not enough isolates of SIV_MAC or HIV-2 have been sequenced and their sequences compared.

Another problem with the theory that HIV evolved from the rhesus macaque virus is that SIV_MAC was isolated from a rhesus monkey kept captive in a laboratory, and very few captive macaques have been infected with this virus. SIV_MAC has never been found in wild macaques, which raises more unanswered questions: did the infected monkeys acquire the virus from some other species while in captivity or while in transit to a laboratory? Several different SIVs have now been isolated from unrelated species of monkey such as sooty mangabeys and mandrills. Their gene sequences and the characteristics of the diseases have been compared with HIV-1 and HIV-2 to reveal a complex picture. Interestingly, not all SIVs cause disease in their natural host (see Table 3). For example, SIV from the sooty mangabey causes simian AIDS in rhesus macaques, while SIV_AGM and SIV_MND appear to produce no symptoms whatever in the African green monkey and mandrill, respectively.

There is, however, a striking similarity between HIV-2 and SIV from the sooty mangabey (SIV_SMM). It is worth noting that the natural habitat of the sooty mangabey is the coastal forest belt of West Africa, stretching from Senegal to Ghana, the same region in which HIV-2 is endemic. There is not only a geographical link between the two viruses and a close molecular relationship but, just as for SIV_MAC and HIV-2, serological cross-reactivity as well.

This then begs the obvious question: has HIV-2 evolved from SIV_SMM? ÐÓ°ÉÔ­´´s would feel happier about supporting such an idea if the one SIV_SMM virus which has been cloned, sequenced and compared to HIV-2 had been isolated from a wild monkey living in its natural habitat under normal conditions. It was not. As for SIV_MAC, the mangabey virus was isolated from a monkey which had lived for more than 20 years in captivity in two American primate centres. The trail has, thus, led us almost to the same point reached with SIV_MAC – with one important difference – sooty mangabeys in the wild are naturally infected with SIV_SMM. Because particles of SIV_SMM look like those of SIV_MAC, it is possible that captive macaques acquired their SIV_MAC originally from sooty mangabeys. Is it also the true ancestor of HIV-2? Or was it the other way around? Did the monkey infect man or did man infect the monkey? Whatever the truth of the matter SIV_SMM, SIV_MAC and HIV-2 could have in some way evolved either from one another or from a close common ancestor.

What about HIV-1? Until very recently, the molecular data available from studying the simian viruses suggested that HIV-1 was no more closely related to them than it was to HIV-2. Indeed, until last year, no candidate ancestor of HIV-1 had emerged from any of the primate viruses. Then scientists working at the Centre International de Recherches Medicales (CIRM) in Gabon reported two cases of wild-born chimpanzees which were positive for antibodies to HIV-1. These animals were never experimentally exposed to HIV-1, nor had they ever been inoculated with human blood products.

A retrovirus called SIV_cp2 was isolated from one of the chimpanzees and this virus has now been sequenced by Simon Wain-Hobson’s group at the Pasteur Institute in Paris. Although the sequence is very close to that of HIV-1, it is also sufficiently divergent from the human virus to be considered a separate and distinct lentivirus. In other words, it does not appear to be just another form of HIV-1. Is then SIV_cp2 the forerunner of the human AIDS virus? There are a number of reasons why such speculations would be premature.

In the first place, a similarity in the sequence of lentiviruses isolated from different species does not necessarily argue in favour of cross-species contamination. It is quite possible that SIV_cp2, for example, evolved in chimpanzees in parallel with the evolution of HIV-1 in humans. The human and nonhuman primate lentiviruses may always have been present in their respective hosts and remain undetected for any number of reasons – the SIVs rarely produce obvious signs of disease; the HIVs could have been restricted to small or remote populations. Latter-day lifestyles would, of course, contribute much to the effective spread of the human virus – fast international travel, sexual liberty, reuse of syringes and needles clinically in Africa and by drug-abusers in the West, to mention a few possible contributory factors.

There is, however, a more obvious reason why scientists must be cautious about reading too much into the origin of HIV based on one HIV-1-like virus isolated from one wild-caught chimpanzee. Sadly, hundreds of chimpanzees have been used in AIDS research, particularly in the US, since HIV-1 was found to replicate, without causing symptoms of disease, in these primates. Not one of these captive animals, which were screened for HIV antibodies before they were infected experimentally, has ever been found to carry a virus similar to HIV-1. Indeed, out of the 50 chimpanzees born in the wild in different areas of Gabon which were tested for HIV antibodies at CIRM, only two were found to be positive, and the virus (SIV_cp2) was isolated from only one.

Furthermore, routine screening of captive primates in Britain has failed to show evidence of HIV/SIV infection in any chimpanzee to date. All of which raises questions about SIV_cp2. Is this indeed a naturally infected chimpanzee? How prevalent is infection in the wild? If other isolates exist, how closely related are they to one another, and to HIV-1? Only answers to these questions will tell us whether SIV_cp2 is relevant to the origin of the human AIDS virus.

But is it possible that monkeys could have infected humans with an HIV-like virus? The answer is yes, first, because transmission might have been possible via a number of routes such as monkey bites, scratches, eating monkey meat or taking ritual preparations obtained from primates. Secondly, mathematical calculations on the sequence divergence existing between different strains of HIV-1, HIV-2 and SIV_MAC suggest that such differences could have evolved in no more than 40 years. Hence it is not impossible that HIV evolved to its present ability to cause disease and spread into people within the past few decades. The earliest retrospectively identified cases of AIDS were in West Africa in the 1960s.

In conclusion, no one can yet be certain of the origin of HIV. But in endeavouring to find a solution to this question, scientists will undoubtedly find out more about lentiviruses in general and may even discover new members of this group of viruses in both animals and people. Preventing AIDS in the future is the real challenge, but an understanding of the origins of HIV and why not all SIVs cause disease in their natural hosts may help us to control HIV in the years to come.

Dr Myra McClure is a virologist at the Institute of Cancer Research in London.