Gerwyn (he who must not be named) has turned up an early paper on virus which appeared when tissue samples were 'passaged' through nude mice. Type C virions include retroviruses; some sizes can hardly be anything else. I've selected some sentences from the abstract which practically jumped off the page at me.
> In another instance type C virus particles were seen replicating in the chondroblastic human cells of a xenografted osteosarcoma. The type C virus produced in the human cells failed to transform NIH/3T3 cells, the C-127 rat cell line, or mink cells. Nucleic acid hybridization studies in which a human endogenous retroviral probe and a xenotropic murine leukemia virus envelope probe were used suggested that the retrovirus present in the human osteosarcoma cells is related to murine leukemia viruses. Intracisternal A-particles (IAP) were also detected in the human osteosarcoma cells. Their presence in the human cells was demonstrated by simultaneous visualization of IAP and human HLA determinants at the cell surface
Several points are worth emphasizing:
1) the particles were definitely associated with human cells;
2) the virus failed to infect offered mouse, rat or mink cells;
3) DNA hybridization showed it resembled X-MLV;
4) human cells showed both IAP and HLA epitopes on their surfaces;
5) the strain of nude mice, NIH/3T3 is the same one in which sequences called pre-XMRV1 and pre-XMRV2 were found.
There is even a relation to prostate cancer, where XMRV was originally discovered. Osteosarcomas are primary cancers of bone, but when prostate cancer metastasizes it frequently spreads to bone. The PC-3 cell line was derived from such tissue.
More germane to the current controversy, this took place at least 9 years prior to the alleged recombination event claimed to create XMRV. Were there two such unique events? How many more should we assume? Is this a more parsimonious assumption than one assuming a common human source?
(A side issue here is that the human cells were producing IAPs, expressed on their surfaces. This says tests for contamination by mouse DNA relying on IAP sequences may also exclude infected human cells, a point I had previously made w.r.t. the Shin/Singh paper.)
If this was, in fact, the origin of XMRV, and we reject the idea that such a highly-specific event could happen twice, we have evidence the resulting virus could escape detection at a later date, (a key point in current debate.) If this early event was not the origin, we must ask where a virus capable of infecting human cells most likely originated in mice carrying human xenografts. My answer would be that it was passed from humans with the disease under study.
What distresses me most here is that deniers (of a human retrovirus) often seem to be intelligent, energetic and knowledgeable, but impervious to perception of circularity in their own reasoning. One aspect has to do with the possibility that the virus in nude mice, where the virus in the 22Rv1 cell line was first detected, actually came from the human xenograft, yet remained undetected until testosterone used to stimulate tumor growth activated latent provirus which had previously escaped detection.
The current definition of XMRV used by deniers says it is identical to the retrovirus contaminating the 22Rv1 cell line derived from human xenografts in the strain of nude mice above. The original XMRV virus was described as 94% homologous to X-MLV (mentioned in the reference above,) and 95% homologous to a HERV soon after discovery. This would leave only a tiny gap for a distinct virus. Defining anything in this gap as a contaminant neatly plugs the hole. This argument was possible the day the original paper in Science was published, without any laboratory work.
The problem with defining the virus as a contaminant is that the hypothesis of a previously undetected human retrovirus associated with prostate cancer would also produce contamination of cell lines derived from human prostate cancer xenografts in mice made prior to assays to detect that virus. All this furor does nothing to distinguish competing hypotheses.
(To escape a flame war, let me say I'm aware the assumption of an undetected retrovirus in humans may also lead to circular reasoning. I am trying hard to find an argument which allows people to come out of debate by a different door than the one through which they entered.)
The recombination-origin hypothesis simply takes the circularity into a larger circle. Let me explain.
There are good reasons to believe the endogenous pre-XMRV sequences are of recent origin, in evolutionary terms. They have only been found in one strain of nude laboratory mice. They have only been seen as two fragments. The retrovirus which inserted them would have left a larger set of overlapping sequences allowing its full genome to be reconstructed if it had been around very long. The nude mice in question have not been around too long, they would not survive in the wild.
The peculiar characteristic of nude mice is a defective immune system. They have no thymus. They do not produce either CD4+ or CD8+ T cells, making them exceptionally vulnerable to viral infection. The purpose is to allow xenografts of human tissue. They are isolated from most sources of infection, except humans. Humans control every aspect of their environment. If you had to select a strain of mice most vulnerable to viral infection from humans this would be it. If they were any more vulnerable, they might be kept in bubbles.
One other point, we don't find any complete copies of the virus inserting the pre-XMRV sequences in the NIH/3T3 strain. Had nude mice carrying the gene shown signs of active viral infection they would have been culled.
The contamination/recombination argument is not falsifiable. In seeking to attack a new finding, deniers have gone outside of science entirely. Every possibility has been neatly covered, even for viruses not yet discovered.
Anyone want to place a bet on never finding a previously undiscovered human retrovirus resembling a mouse virus?
Two Unique Events?
Blog entry posted by anciendaze, Jun 25, 2011.