Re: Could gene therapy activate evolutionarily suppressed viruses?

Almost anything is "possible". Almost anything "could happen". So the real question is not "could it happen?" but "what are the odds that it will or won't happen?".

The human genome contains roughly 3 billion bases of DNA. There are some number of copies of endogenous retroviruses in the human genome. Let's use the number 50 as a convenient estimate for now. Of those 50 endogenous retroviruses, a small percentage might be potentially functional. They might still have the gag, pol and env genes free of inactivating muations so that they could potentially encode an infectious virus. So, if there are 2 or 3 of these potentially activatable viruses in 3 billion bases of DNA, what are the odds that an element inserted into human DNA at random will insert just upstream of one of these endogenous viruses and activate it. The odds of inserting within 500 bases upstream of the viral genome are approximately one in 6 million. The odds that the insert will activate the virus are dependent on what is being inserted. I would estimate the overall odds of activating a virus to be about one in several billion. This seems to be backed up by the fact that millions of people have been infected with HIV-1, HIV-2, HTLV-I, HTLV-II, HCV and many other retroviruses which insert at random into the human genome, and as yet we are not aware of any activation of any endogenous retrovirus.

The second question is "What are the odds that a re-activated endogenous virus would be transmissible to other people and/or harmful to anyone?". There are a great many retroviruses known to humans, and surely many more that are as yet undiscovered. Plants, mammals, fish, insects and other life forms are known to carry endogenous retroviruses, mobile genetic elements possibly derived from viruses (or are the viruses derived from the mobile elements?), and fully transmissible retroviruses. Almost all viruses have a very small host range. Cat viruses do not infect humans. Dog viruses to not infect cats. Many of the viruses that humans have discovered, were discovered because they cause disease. Few humans are motivated to spend millions of dollars searching for and finding viruses that are totally harmless. Even with this bias toward finding harmful viruses, many harmless or nearly harmless viruses have been discovered. There are many theories about viruses evolving to be harmless or mearly harmless to their natural host. For example, it is believed that the feline immunodeficiency virus and all of the many known simmian immunodeficiency viruses are totally harmless to their natural hosts. It is only cross-species transfer that results in a virus which is harmful to the new host for several geneartions until it adapts to be harmless also to the new host. There are other theories about how the capture of a retrovirus into a host germ-line genome to become an endogenous retroviral element, result in essentially "vaccinating" the host against infection by this virus in its exogenous form. All of these factors tend to lessen the odds that a re-activated human endogenous retrovirus (such as HERV-K10, see GenBank entry with accession number M14123 at: http://www.ncbi.nlm.nih.gov/htbin-post/Entrez/query?uid=182227&form=6&db=n&Dopt= g for example) would be harmful to humans.

All human activities and technologies carry risks of harm as well as benefit. Driving automobiles puts humans at risk of crashes and gasoline fires, but prevents them from being kicked by horses for example. Nuclear power plants carry risks associated with radiation but prevent coal mine fires, global warming, acid rain, and other problems. For most advances in technology, it is not the physical risks and benefits that are evaluated, it is only the financial benefits. As long as coal-powered electricity is cheaper than wind or solar, it will be used regardless of the large and small risks. Humans seem to be increasingly able to look at the long and short term risks, and to factor them into the overall costs of a technology, for example by requiring coal-fired power plants to install expensive scrubbers to reduce acid rain. But we will never be able to calculate all risks, nor will we be willing to pay the costs of making every new technology 100% risk-free.

Humans currently are much more willing to accept "natural" risks, such as skin cancer from sunlight, than "man-made" risks such as cancer from man-made sources of radiation. We are also more tollerant of self-imposed risks, such as cancer from tobacco smoke than risks put upon us by others, such as cancer from some food aditive that we do not "choose" to eat. Risks from genetic engineering are both man-made and out of the hands of most individuals, so we weight the risks very highly. Most people also fear "high-tech" or "new" risks, more than old familiar risks. For example people have a greater fear of cancer from cell phones and microwave ovens than from skin cancer from the sun. In this regard, risks of genetic engineering are very highly feared, without a rational basis.

Developers of new technologies often discount irrational fear as if it was not important. They often do not realize that fear is fear, and whether it is rational or not makes no difference to the political and legal climate that will influence the use of their technology. People who develop new technologies of any sort should take lessons from such events as the discontinuation of the use of Alar to grow apples, and the failure in the USA of using nuclear energy to produce electricity. Even if a new technology is in fact very safe, it will fail if the public does not trust it, regardless if thier trust is rational or not.

The greater risk of the type of genetic engineering that could activate an endogenous retrovirus (this would be inserting new DNA at random into the human genome, rather than targetting the new DNA to a specific place) is that it will either activate a cancer-causing gene, or de-activate a gene needed for survival. The technology can be made much safer by targetting the inserted DNA to a specific place in the human genome. The success of human genetic engineering overall will be dependent on a general population that knows what DNA is, knows how genes work, and understands the difference between science and science fiction.

The odds of creating a new virus harmful to humans is probably about ten million fold less for any type of gene therapy, than the odds for bringing a pathogenic virus into humans by keeping a monkey as a household pet, or by paddling a canoe up a Brazilian stream. But we are willing to accept the latter type of risk much more than the former.

Brian Foley