IMMUNOHISTOCHEMICAL DETECTION OF THE CYCLIN-DEPENDENT KINASE-INHIBITOR-2 MULTIPLE TUMOR-SUPPRESSOR GENE-1 (CDKN2 MTS1) PRODUCT P16(INK4A) IN ARCHIVAL HUMAN SOLID TUMORS - CORRELATION WITH RETINOBLASTOMA PROTEIN EXPRESSION/

The retinoblastoma (RB) and cyclin-dependent kinase inhibitor 2/multip le tumor suppressor gene 1 (CDKN2/MTS1) tumor suppressor genes play im portant roles in the regulation of the cell cycle, The protein product s of these two genes, pRB and p16(INK4A) (''p16''), respectively, inhi bit progression from G(1) to S phase, Moreover, p16 has been shown to exert its function through inhibition of CDK4-mediated phosphorylation of pRB, Both genes have been found to be mutated or deleted in a wide range of primary human tumors and tumor cell lines, However, the pres ence of CDKN2/MTS1 containing nonneoplastic elements in every tumor sp ecimen may contribute to the apparent lower deletion detection rate in resected neoplasms compared to cell lines, We have developed an immun ohistochemical assay that allows us to assess p16 expression in formal in-fixed, paraffin-embedded tissues, ris controls, we used paraffin-em bedded pellets of cell lines with wed-defined p16 status (four positiv e and four negative lines), as well as routinely processed nude mouse xenografts of two p16-positive cell Lines, p16-negative cells were cha racterized by the absence of nuclear staining, whereas cytoplasmic sta ining was variable, In neoplastic and normal tissues, the level of p16 generally appeared to be low, We tested 75 random human malignancies from 4 different anatomic sites for p16 expression and correlated the findings with the immunohistochemical presence or absence of pRB. Twen ty % of tumors selectively lacked pRB, while 37% of neoplasms had unde tectable p16, In 43% of all carcinomas, both pRB and p16 could be dete cted, Significant differences existed in the expression of both tumor suppressor genes between carcinomas from different sites, Breast cance rs had the highest rate of p16 negativity (13 of 20), Our data show th at: (a) immunohistochemistry may be a suitable modality to screen for RE and CDKN2/MTS1 abnormalities in paraffin-embedded tissues; (b) unde tectable levels of p16 expression occur at a relatively high frequency ; (c) p16 and pRB expression in common human malignancies is not mutua lly exclusive; (d) loss of function of both tumor suppressor genes app ears to be a distinctly uncommon phenomenon; and (e) different types o f carcinomas have variable rates of disturbance in the p16/pRB pathway .