medpundit |
||
|
Thursday, January 16, 2003I've been meaning to post something myself, but I wanted to look into the experimental details to see if, as you suggest, they might explain why this group seems have such bad luck compared to others. One thing that probably does not play a role is the nearness of the SCID gene to a potential oncogene. The viral vector integrates at random, not necessarily anywhere near the gene that causes the disease. However, if the virus (which carries the wildtype SCID gene, to cure the disease) lands near an oncogene, it can turn up expression of that oncogene and lead to leukemia. But this should work just as well in non-SCID cells, and with any virus, not only one used for gene therapy. In fact, researchers have recently had a field day using retroviral mutagenesis in mice (here and here.) It seems that that’s probably what happened. The researchers have a letter to the editor in today’s New England Journal of Medicine that they wrote before the second patient developed leukemia: One proviral integration site was found, located on the short arm of chromosome 11 within the LMO-2 locus, as determined with the use of linear-amplification mediated polymerase-chain-reaction analysis. This proviral integration within the LMO-2 locus was associated with aberrant expression of the LMO-2 transcript in the monoclonal T-cell population. Aberrant expression of LMO-2 has been reported in acute lymphoblastic leukemia arising from T cells with / receptors, usually with the chromosomal translocation t(11;14). Tests for replication-competent retrovirus were repeatedly negative in our patient's lymphocytes. After the child developed leukemia, they discovered a different translocation which may be responsible for the disease. Here’s the original paper describing the therapy. They mention in it that there were three different insertion sites for the gene in the patients treated, but don’t say where those sites were. They also used a “defective Moloney murine leukemia virus” as the vector for gene insertion. Could that have anything to do with the leukemia? I would have to defer to Charlie on that one. UPDATE: Charles Murtaugh says the retrovirus probably has nothing to do with it: Probably the fact that it was a leukemia virus isn't relevant -- I'm not sure about this particular human virus, but in mice, leukemia viruses cause the disease mainly by virtue of the genes they themselves carry, i.e. viral oncogenes. But as the researchers point out, if the genome of even an innocuous retrovirus integrates near a so-called "cellular oncogene," it can cause that gene to be transcriptionally upregulated, and drive uncontrolled proliferation. Cellular oncogenes are generally genes that promote normal replication, and as such they are under tight regulatory control to prevent abberant growth. If you pop a virus in there, this regulation can be disrupted. One unfortunate property of blood cells, actually, is that it is very easy to turn them cancerous, compared to solid tissue cells. I strongly doubt that gene therapy directed to, say, the liver would ever give rise to a spontaneous tumor. FINAL UPDATE: Charles Murtaugh emailed again to say the Moloney leukemia virus does cause cancer. (We should note that the research team used a defective version of it, that is one that's not supposed to be able to reproduce.) But, how defective are those samples used in gene therapy? Are they 100% defective, or are there a small percentage of active virus particles in the mix? For more on the Moloney virus click here.) posted by Sydney on 1/16/2003 08:08:00 AM 0 comments 0 Comments: |
|