Oral Carcinogenesis: Studies With Human Gingival Keratin

癌变是一个多步骤的过程，涉及细胞原癌基因的激活和肿瘤抑制基因的失活。上皮细胞经历不同阶段的表型和基因型改变，逐渐获得恶性生长特征。尽管口腔癌的发病率相对较高，但人们对导致疾病发生和发展的分子事件知之甚少。以前我们报道了正常的人口腔角质形成细胞（NHOK）可以在高Ca2+的培养下诱导终末分化。与NHOK相比，hpv16e6 /E7永生人口腔角化细胞（IHOK）形成角化膜的能力显著降低，并与转谷氨酰胺酶1的诱导减少相关。此外，IHOK细胞抵抗tgf - β的生长抑制，可能是由于hpv16e7使Rb通路失活。与NHOK和IHOK相比，一些HNSCC系完全不能诱导转谷氨酰胺酶1和确证包膜形成。我们研究的目的是了解HNSCC癌变过程中发生的遗传和分子变化。为此，我们开展了头颈部鳞状细胞癌（HNSCC）系的基因表达研究，以将基因表达谱的变化与表型改变联系起来。

Carcinogenesis is a multi-step process involving the activation of cellular proto-oncogenes and inactivation of tumor suppressor genes. Epithelial cells undergo different stages of phenotypic and genotypic alterations and gradually acquire malignant growth characteristics. In spite of the relatively high frequency of oral cancers, little is known about the molecular events that lead to disease initiation and progression. Previously we reported that normal human oral keratinocytes (NHOK) can be induced to undergo terminal differentiation by high Ca2+ in culture. The capacity of the HPV 16 E6/E7 immortalized human oral keratinocytes (IHOK) to form cornified envelopes was significantly reduced compared to NHOK and correlated with the reduced induction of transglutaminase 1. Furthermore, IHOK cells were resistant to growth inhibition by TGF-beta, probably due to inactivation of the Rb pathway by HPV16 E7. In contrast to NHOK and IHOK, several HNSCC lines were totally incapable of induction of transglutaminase 1 and cornified envelope formation. The goal of our study is to understand the genetic and molecular changes that occur during HNSCC carcinogenesis. To this end we carried out gene expression studies of head and neck squamous cell carcinoma (HNSCC) lines to relate the changes in the gene expression profiles with phenotypic alterations. Global gene expression profiles of human head and neck squamous carcinoma cell lines : We have conducted microarray experiments using a NCI-Onco chip with 9700 cDNA clones for a systematic analyses of gene expression profiles in 25 HNSCC cell lines and IHOK. We used NHOK as a reference in this study to characterize gene expression changes associated with HNSCC carcinogenesis. Genes primarily involved in cell cycle regulation, oncogenesis, cell proliferation, differentiation, apoptosis and cell adhesion were widely altered in HNSCC cells. Genes consistently upregulated in most HNSCC cells included growth factors (EGF receptor pathway substrate 8, IGF binding proteins3 and -6), DNA binding proteins and transcription factors (topoisomerase II alpha, PCNA, replication protein 3), cell cycle regulatory genes (CDC20, CDC25B, serine/threonine kinase 6), oncogenes (DEK oncogene, ECT2, and stathmin1), TTK protein kinase, cytokines IL8 and TNFSF10, antiapoptotic gene survivin, and stress related gene DHFR. The most remarkable feature of the gene expression profiles of HNSCC was the coordinated pattern of downregulation of a wide range of genes involved in terminal differentiation (transglutaminase 1 and -3, SPRR1A and -B, SPRR3, various cytokeratins 1,5,6A, 7, 13, 14, 15, 16, and 17) and cell adhesion (annexins 1,2 and 3, cadherins 1 and 13, laminins alpha 3, beta 3 and gamma 2, integrins alpha 5 and beta 1). This finding is consistent with the previous observation of the loss of competency to form cornified cell envelopes in several HNSCC cell lines. These findings suggest that the loss of the cellular differentiation machinery represents a critical feature of HNSCC. Furthermore, hierarchical clustering analysis, as well as principal component analysis, revealed two distinct subtypes of gene expression patterns among the 26 cell lines, reflecting a degree of heterogeneity in HNSCC. By applying Significance Analysis of Microarrays, 136 genes were selected for being distinctively expressed between the two groups. Genes differentially expressed in the two subgroups include cell proliferation-related genes, IGFBP6, EGFR, and VEGFC; tumor suppression and apoptosis related genes such as Tp53, Tp63 as well as cell cycle regulators such as CCND1 and CCND2 (cyclins D1 and D2) suggesting that the two subgroups might have undergone different pathways of carcinogenesis. Group I, that seems to represent more malignant properties include Hep2, KB, HN8, HN22, HN13, HN19, UMSCC-22B and UMSCC-74B. Group II consists of 18 cell lines including IHOK, HN4, HN6, HN12, HN1117, HN26, HN30, HN31, PCII-15B, and UMSCC-11A, -11B, -14A, -14B, -14C, -17A, -17B, and -22A. It is interesting to note that the segregation of 11 HN series cell lines from this study is in total agreement with the subgrouping from the previous microarray analysis (6700 cDNA chip) done with a human cancer RNA pool as a reference. Cells in group I showed lower expression of proliferation genes and cell adhesion genes than group II cells, but stronger expression of anti-apoptotic genes. These findings suggest that group I cells represent a later stage of tumorigenesis than group II cells. This idea is also line with the finding that IHOK clustered with group II. We have searched for potential genes that may be responsible for the across the board loss of terminal differentiation in HNSCC cells. We have found that the expression of a putative tumor suppressor gene, KLF4, and of TGF beta receptors 1 and -2 was significantly reduced in HNSCC cells. The role of these genes and a potential of the TGF beta pathway signaling in the regulation of the zinc-finger transcription factor, KLF4, in oral carcinogenesis is under study.