Head and neck squamous cell carcinomas (HNSCC) are characterized by a marked propensity for local invasion and spread to cervical lymph nodes, with distant metastases developing in 30-40% of cases. HPV-16 is an important risk factor for HNSCC. How HPV enhances susceptibility to HNSCC is not fully understood, but seems to involve cofactors. In this study, we examined the effect of the cooperation between HPV-16 and the tyrosine kinase receptor ErbB-2 on E-cadherin/catenin complex patterns and neoplastic transformation of human normal oral epithelial (NOE) cells. We report that overexpression of ErbB-2 or E6/E7 alone does not affect E-cadherin/catenin complex patterns nor does it induce cell transformation of NOE cells. In contrast, coexpression of E6/E7 and ErbB-2 downregulates E-cadherin and catenin expression. This is accompanied by cytoplasmic localization of E-cadherin, as well as nuclear translocation of a, b, and c-catenins. Furthermore, we demonstrate that E6/E7 cooperate with overexpressed ErbB-2 to induce tumor formation in nude mice and to upregulate cyclin D1 and c-myc expression. Our data suggest that E6/E7 cooperate with ErbB-2 in head and neck carcinogenesis, at least in part, via the conversion of b-catenin from a cell adhesion to a nuclear function, that is, to act as a potential transcriptional regulator. This conversion leads to the upregulation of cyclin D1, c-myc and other oncoproteins necessary for alteration of the E-cadherin/catenin complex and cell transformation of NOE cells.

Dysregulation of cell cycle is important in oncogenesis. We analyzed the inactivation of the INK4 family CKI/CDK/RB pathway by gene promoter hypermethylation in leukemogenesis. The methylation-specific polymerase chain reaction (MSP) with primers for methylated (M-MSP) and unmethylated (U-MSP) alleles of the p15, p16, p18, and RB genes was used to study five leukemic cell lines, 50 acute myeloid leukemia (AML) and 25 acute lymphoblastic leukemia (ALL) samples. None of the leukemic cell lines showed p18 and RB methylation. p15 was methylated in Raji, while p16 was methylated in U937 and Raji. In NB4 and Jurkat, both alleles of p15 and p16 appeared to be deleted. At diagnosis, p15 methylation occurred in 29 (58%) AML patients, and 10 (40.0%) ALL patients. p16 methylation occurred in two (4%) AML and two (8%) ALL patients. Only one each of AML and ALL patients had concurrent p15 and p16 methylation. None of the patients had methylation of p18 or RB. In AML, p15 methylation was associated with M2 subtype ( p=0.018). Patients with and without p15 methylation had similar complete remission (CR) rates and projected 5-year overall survival (OS) or disease-free survival (DFS). Therefore, methylation inactivation of the INK4/CDK/RB pathway in leukemia involved primarily p15 and occasionally p16, but not p18 or RB. In AML, p15 gene methylation was associated with the M2 subtype, but was not prognostic for CR, OS, or DFS.

SOCS1 and SHP1 are negative regulators of the Jak/STAT signalling pathway that is implicated in leukaemogenesis. We studied if aberrant methylation of SOCS1 and SHP1 might be involved in the pathogenesis and prognostication of acute leukaemias by methylation-specific polymerase chain reaction (MSP). At diagnosis, methylation of SHP1 occurred more frequently in acute myeloid leukaemia (AML) (n=26, 52%) than acute lymphoblastic leukaemia (ALL) (n=6, 24%) (p=0.02). Methylation of SOCS1 was absent in both AML and ALL patients. SHP1 methylation was not associated with specific clinicopathologic features and had no prognostic impact on AML patients. Frequent methylation of SHP1, but not SOCS1, may be important in the pathogenesis, but not prognosis, of acute leukaemias.

Summary. The gene CEBPA, encoding the transcription factor C/EBPa, is crucial for granulocyte differentiation. We investigated the frequency of aberrant CEBPA promotor methylation with the methylation-specific polymerase chain reaction in 70 patients with acute myeloid leukaemia (AML). Two patients, both with M2 morphology, were found to have methylated CEBPA. In one of them, the fusion gene AML1/ETO, reported to cause transcription repression of CEBPA, was also present, suggesting that more than one mechanism might collaborate to suppress CEBPA gene expression. Aberrant CEBPA methylation is infrequent in AML, but may occur preferentially in the M2 phenotype.