eddy@boulder.Colorado.EDU (Sean Eddy) (06/21/87)
[the following article is transferred from sci.med.aids for possible open discussion. The first article in the series quoted a West German newspaper article that claimed HIV - the AIDS virus - was created in maximum-containment biowarfare facilities at Fort Detrick, Maryland. Subsequent discussion has centered on the dangers of virological research and the possibility of a human pathogen being created by accident in a research lab engaged in virological research.] Eddy@boulder.colorado.EDU writes: (quotes marked with '|') |- eddy@boulder.colorado.EDU !{hao,nbires}!boulder!eddy |In article <1411@sigi.Colorado.EDU> tmb@PREP.AI.MIT.EDU writes: |>PS: please reconsider your statement that 'HIV is extremely unlikely |>to be the result of genetic engineering'... | |Sorry, but I can't buy it. First of all, the technology did not exist |in 1976, when Craig says the first case was recognized in the U.S. |HTLV-I had not yet even been discovered. No sequence data was |available for visna or HTLV-I. Your idea [*] that random mutagenesis |of visna virus could result in the accidental production of HIV |fights astronomical odds; HIV is so different from visna, the odds |of making the right 'mutations' without achieving a lethal hit in |the virus genome approach infinity. Your comments are factually correct but irrelevant. (As an aside, I never claimed [*] and agree with you that [*] is an extremely unlikely event. The sequence homology between visna and HIV is probably the result of a common ancestor rather than a direct derivation). During the evolution of HIV's ancestors, two important steps must have taken place: (1) creation of a human viral pathogen with novel properties and (2) transfer of this virus into human beings. Both of these events are unlikely, but they did occur, as the existence of HIV demonstrates. Your argument is that any work with the expressed goal of carrying out these steps in a laboratory is doomed to failure, and you are right. My argument is that there are certain techniques that greatly increase the likelihood of either event. Any work involving recombinant DNA research with mammalian viruses and subsequent packaging and culturing of these viruses greatly increases the likelihood of (1). The reasons for this are technical: o DNA/RNA preparations are always contaminated to a considerable degree with sequences derived from the host genome; recombinant DNA experiments therefore generate a small but significant fraction of 'clones' with unexpected and unpredictable constitution o mammalian genomes contain a large number of inactive, altered, or dormant integrated viruses that could recombine with the virus under study or use it as a helper virus o mammalian genomes contain a large number of genes that could become integrated into a viral genome and enhance the pathogenicity of the virus, alter its effects, or alter its host range (proto-oncogenes being the most well-known example; the recent discoveries about the causes of neuro-pathogenicity of HIV suggest that HIV has indeed acquired some mammalian intercellular signal protein) o cell-culture techniques alter the normal selective and evolutionary mechanisms at work in whole animals; viruses can propagate and cross for many more generations than they could in vivo. Replication defective viruses can rely on the presence of helper viruses to traverse large distances in sequence space None of these events are, of course, events that could not also occur in nature. However, the point is that certain kinds of work greatly increase the likelihood of generating pathogens with novel properties and may also greatly speed up the evolution of such pathogens. Techniques that greatly increase the likelihood of (2) are sloppy mammalian tissue culture techniques, handling of infected laboratory animals, deliberate transfer of viruses, tissues, and animal products to human beings (e.g. cross-species transplantation, vaccines). (Some non-research route of transfer of animal viruses to humans are consumption of raw meat (the transmission of kuru, a slow brain disease, is an intra-species example), and bites from animals). Concretely, in the case of HIV, this means that work on biological warfare could be 'responsible' for the existence of HIV by greatly increasing the likelihood that HIV came into existence and greatly increasing the likelihood that HIV was transmitted to humans. The ancestors of the HIV need not have been directly involved in the work, and the goal of the work need not have been to produce a virus with HIV like properties. All that counts is that work on biological warfare involves generation and culturing of viable recombinant viruses and deliberate transfer of such viruses into mammals closely related to humans and human cell lines (if not humans). Now, all these considerations do not mean that 'recombinant DNA work', per se, is dangerous. However, it is important to remember that the only justification for current safety procedures in molecular biology is that, so far, no significant accidents seem to have occurred. This means, on the other hand, that we must constantly reevaluate procedures and reassess risks. Furthermore, we do not exactly know how an 'accident' would manifest itself. It was only a century after organic chemistry had begun, for example, that people began to realise that many organic compounds are potent carcinogens. Personally, I do not feel at risk working with invertebrates and invertebrate cell lines. Spilling a Schneider line tissue culture is a nuisance because I have to redo the experiment, but any virus that might be in there is very unlikely to infect mammals, and I do not culture viruses anyhow. Any kind of molecular biology involving organisms or pathogens of organisms that distant from humans is unlikely indeed to result in the creation of a human pathogen (I don't think there are known pathogens that are dangerous to both an invertebrate species and humans). Working with mammalian tissue culture and mammalian tissues probably presents a much higher risk, and working with mammalian viruses (regardless of whether they are human pathogens or not) is even more dangerous. As I argued above, the risk is not necessarily the virus itself (which the researcher presumably understands well), but rather the increased chance that a pathogen with novel properties arises by recombination, and that it survives. Finally, the measures that people take to deal with these risks, containment, decontamination, &c., are probably not what is really important. Spilling a mammalian tissue culture that is believed to be 'virus free' on yourself is probably much more dangerous than flushing a moderately infectious virus down the drain (I am not suggesting to do the latter, however). What counts much more is the attitude of the individual researcher, his comprehension of the subject, and his admittance to himself of the ignorance of the field about many issues. Altogether, yes, recombinant DNA is a very powerful, very beneficial technology, and there is no reason to give it up. However, given how little we know about how viruses come into existence and evolve, what their relationship to their hosts is, and how they function, there are good reasons to be cautious, in particular in areas involving recombinant DNA work with viruses and reintroduction into animals. Work on biological warfare (it cannot even be called 'research') is the ultimate in lack of caution, and the ultimate in taking unjustified and pointless risks. It does not matter how HIV arose, accidentally in biological weapons work, accidentally in the development of some vaccine using recombinant DNA techniques, or in some monkey in the middle of Africa. We should all the same abandon biological weapons research since the risks don't outweigh the benefits; we should always remind ourselves as researchers in basic and medical research of the possible dangers and not-understood areas of our fields; and we should stop eating half-cooked monkey brains. Thomas. PS: again, please forward if you feel it warranted, and feel free to comment. I had the feeling you either did not read my previous letter very carefully, or it was too obscure. [ I don't know how to stop my .sig, sorry ] - Sean Eddy - MCD Biology; U. of Colorado at Boulder; Boulder CO 80309 - eddy@boulder.colorado.EDU !{hao,nbires}!boulder!eddy - - "Are you with the police?" "No, ma'am. We're musicians."
eddy@boulder.Colorado.EDU (Sean Eddy) (06/22/87)
In article <1418@sigi.Colorado.EDU> tmb@PREP.AI.MIT.EDU writes: > During the evolution of HIV's ancestors, two important >steps must have taken place: (1) creation of a human viral >pathogen with novel properties and (2) transfer of this virus >into human beings. > > Both of these events are unlikely, but they did occur, as >the existence of HIV demonstrates. You seem to have a problem imagining that viruses can evolve on their own to become human pathogens, without the help of Fort Detrick. The very fact that viral pathogens exist is proof that new human pathogens can indeed arise without the intervention of science. In recent history, there are, in fact, several cases of highly virulent viruses that have appeared out of nowhere. Rocio virus appeared in 1975 to cause an encephalitis epidemic in Brazil. Marburg virus suddenly appeared in 1967, killing several people in Germany. O'nyong nyong virus (what a name!) spread as in epidemic proportions through East Africa in 1959 -- and virtually disappeared three years later. > Any work involving recombinant DNA research with mammalian >viruses and subsequent packaging and culturing of these >viruses greatly increases the likelihood of (1). The reasons >for this are technical: > > o DNA/RNA preparations are always contaminated to a > considerable degree with sequences derived from the > host genome; recombinant DNA experiments therefore > generate a small but significant fraction of 'clones' > with unexpected and unpredictable constitution HIV has no sequences that seem to be derived from the host genome. > o mammalian genomes contain a large number of > inactive, altered, or dormant integrated viruses > that could recombine with the virus under study > or use it as a helper virus HIV is not an endogenous human retrovirus. No homologous sequences have been found in the human genome by hybridization studies. > o mammalian genomes contain a large number of genes > that could become integrated into a viral genome > and enhance the pathogenicity of the virus, alter > its effects, or alter its host range (proto-oncogenes > being the most well-known example; the recent > discoveries about the causes of neuro-pathogenicity > of HIV suggest that HIV has indeed acquired some > mammalian intercellular signal protein) The HIV genome does not have room for the 'intracellular signal protein' you suggest. It is more likely that the 'neurovirulence' is coded by natural variants of the HIV CD4-specific antireceptor. (There is some recent evidence that only some, not all, strains of HIV can infect cells in the CNS.) >None of these events are, of course, events that could not >also occur in nature. However, the point is that certain >kinds of work greatly increase the likelihood of generating >pathogens with novel properties and may also greatly speed up >the evolution of such pathogens. No, I disagree. Considering Nature is trying out all these things every second of the day, in every living thing in the planet, I think there is little a researcher can do to speed up the course of evolution by accident. And we've covered the 'speed up evolution deliberately' argument for HIV already. > Personally, I do not feel at risk working with >invertebrates and invertebrate cell lines. Spilling a >Schneider line tissue culture is a nuisance because I have to >redo the experiment, but any virus that might be in there is >very unlikely to infect mammals, and I do not culture viruses >anyhow. Any kind of molecular biology involving organisms or >pathogens of organisms that distant from humans is unlikely >indeed to result in the creation of a human pathogen (I don't >think there are known pathogens that are dangerous to both an >invertebrate species and humans). Speaking of knowing the risks one is dealing with -- what kind of invertebrate cells are you working with? Drosophila (Sindbis virus can infect Drosophila cells; 1 lab-related fatality)? Mosquito, tick, or mite (many laboratory fatalities from arboviruses)? While these accidents all occurred in virus labs, not just from handling cell lines, don't be so sure invertebrates don't carry human pathogens. > It does not matter how HIV arose, accidentally in >biological weapons work, accidentally in the development of >some vaccine using recombinant DNA techniques, or in some >monkey in the middle of Africa. We should all the same >abandon biological weapons research since the risks don't >outweigh the benefits; we should always remind ourselves as >researchers in basic and medical research of the possible >dangers and not-understood areas of our fields; and we should >stop eating half-cooked monkey brains. > > Thomas. In fact, it matters a hell of a lot how HIV arose. Scientifically, because the origins of the virus could tell us a lot about the mechanisms of its pathogenicity, and could indicate the proper model systems to study to learn more. Socially, it matters a hell of a lot too. Wild rumors about the origins of HIV in a virus research lab, or as a result of the smallpox vaccine, or God knows what else people will think of, cannot in any way further the progress of AIDS research. There is plenty of distrust of science in this field ("they're not telling us everything") already. - Sean Eddy - MCD Biology; U. of Colorado at Boulder; Boulder CO 80309 - eddy@boulder.colorado.EDU !{hao,nbires}!boulder!eddy - - "You see, we're on a mission from God." - -Elwood Blues