immunologically related to a group of mouse mammary tumor virus (immunoperoxidase staining/paraffin sections/cancer antigen)

An antigen immunologically related to a group-specific antigen (gp52, a 52,000-dalton glycoprotein) of the mouse mammary tumor virus has been identified in paraffin sections of human breast cancers by means of the indirect im- munoperoxidase technique. The specificity of the reaction with antibody against mouse mammary tumor virus was examined by absorption of the IgG with the following: (a) purified gp52; (b) a number of virus preparations (mouse mammary tumor virus, Rauscher leukemia virus, simian sarcoma virus, baboon endogenous virus, and Mason-Pfizer monkey virus); (c) normal plasma, leukocytes, breast tissue, milk, actin, collagen, and hyaluronic acid, all of human origin; (d) sheep erythrocytes and mucin. Only mouse mammary tumor virus (from C3H or Paris RIII strains and grown in either murine or feline cells) and pu- rified gp52 eliminated the immunohistochemical reaction in the human breast tumors. Positive reactions were seen in 51 of 131 (39%) breast carci- nomas of various histologic types, a minimal estimate in view of the limited number of sections from each tumor that could be examined. Negative reactions were obtained in all 119 benign breast lesions (cystic disease, fibroadenoma, papilloma, gyne- comastia) and in all 18 normal breast tissues. With one excep- tion, 99 carcinomas from 13 organs other than breast and 8 cystosarcomas were all negative. Earlier investigations (1) have revealed key similarities between human breast cancer and the mouse mammary tumor experi- mental model. Human breast carcinomas contain particles with many of the characteristic biochemical and biophysical features of the RNA tumor viruses. These include the characteristic size and density and possession of outer membranes as well as cen- tral cores encapsulating an RNA-dependent DNA polymerase (reverse transcriptase) complexed to a 70S RNA molecule. Further, we (2-4) and others (5) have demonstrated a partial homology between RNA molecules found in human breast tumors and the RNA genome of the mouse mammary tumor virus (MMTV). It is at present futile to dwell on the etiologic implications of these findings, because no hard conclusions can be drawn until more definitive experiments become possible. However, whether they are the cause or the consequence of the human malignancy, the presence of these particles provides a novel opportunity to generate information of potentially practical importance for the diagnosis and the management of the dis- ease. Thus, if one or more of the particle proteins appear in the circulation, they could serve as systemic signals of the presence of the corresponding tumor. The feasibility of this approach was explored and established with the mouse mammary tumor