Consistent chromosomal anomalies in keratinocyte cell lines derived from untreated malignant lesions of the oral cavity

Spindle assembly checkpoint and centrosome abnormalities in oral cancer

Like many solid tumours, oral squamous cell carcinomas (OSCC) invariably exhibit chromosomal instability (CIN) leading to aneuploidy. The mechanisms responsible for CIN in OSCC, however, are largely unknown. This study examined the fidelity of the spindle checkpoint, together with the number, structure and function of centrosomes in a series of well-characterised aneuploid immortal OSCC-derived cell lines that harbour p53 and p16(INK4A) defects. The spindle checkpoints were fully functional in 2 of 7 cell lines and attenuated in the remaining 5 cell lines. Overexpression of the spindle checkpoint protein, Cdc20, was observed in 2 of the cell lines with attenuated checkpoints. Defects in centrosome number, size and localisation were detected in 5 of the cell lines. Clonal cell populations contained cells with both normal and abnormal numbers of centrosomes, suggesting that the control of centrosome number may be inherently unstable in OSCC-derived cell lines. Centrosomal abnormalities were then examined in tissue samples of oral epithelial dysplasias and carcinomas. Abnormal centrosomes were detected in all the tissues examined albeit in a low percentage of cells (<1% to >5%). The percentage of cells containing centrosome abnormalities was significantly higher in the carcinomas than in the dysplasias (p<0.02) and in the poorly differentiated SCCs relative to their moderately differentiated (p<0.04) and well-differentiated (p<0.01) counterparts. We suggest that the genetic alterations associated with the development of the immortal phenotype, together with spindle checkpoint and centrosome defects, are responsible, albeit in part, for the complex karyotypes observed in OSCC. The presence of centrosome abnormalities in oral dysplasias raises the possibility that such defects might contribute to malignant progression.

Organotypic co-cultures allow for immortalized human gingival keratinocytes to reconstitute a gingival epithelial phenotype in vitro

We report here that the organotypic co-culture (OCC) system allows for significant preservation of the tissue-specific phenotype of human gingival keratinocytes (IHGK) immortalized with the E6/E7 gene of the human papillomavirus type 16 (HPV16). The approach adopted is based on the OCC system facilitating spatially separated cell growth and cell-to-cell interactions via diffusible growth factors. Generally, IHGK reveal transcription of the HPV16 E6/E7 gene at rising passages. Fluorescence in situ hybridization performed for chromosomes 1, 8, 10, and 18 demonstrates that disomic fractions differ between the tested chromosomes but otherwise remain fairly constant. Monosomies of chromosome 18 are more prominent in late passages 81 and 83, while polysomies of chromosome 10 and 18 are detected in early passages 25 and 27. In comparison with corresponding monolayer cultures (MCs), IHGK in OCCs form stratified epithelia, proliferate, and express gingival-specific gene products in vitro. Moreover, mRNA gene transcription for growth factors interleukin 1beta, granulocyte-macrophage colony stimulating factor, fibroblast growth factor 7, and EGF in OCCs is different from that in MCs. When grafted onto nude mice, IHGK develop hyperplastic, differentiated surface epithelia devoid of malignant growth. We are not aware of any other OCC system comprising of IHGK, which allows for site-specific expression of gingival epithelial markers. This substantiates reconstitution of a gingival epithelial phenotype in vitro.

Comparative genomic hybridization analysis reveals 3q gain resulting in genetic alteration in 3q in advanced oral squamous cell carcinoma

We analyzed DNA sequence copy number aberrations (DSCNAs) in 17 primary oral squamous cell carcinomas (OSCCs) by comparative genomic hybridization. DSCNAs were detected frequently at 3q25-qter (7/17), Xp21 (5/17), and Xq12-q23 and 8q23-q24 (4/17), and losses were detected frequently at 13q21-q22 (5/17), 3p21-pter, 4p15-pter and 17p13 (4/17), and 8p22-pter and 9p21-pter (3/17). Four tumors showed amplifications of seven loci: 3q11-qter, 3q13, 3q26, 7q21-q22, 8q23-qter, 9p22-pter, and 12p11. The total number of DSCNAs was significantly greater in stage III and stage IV tumors than in stage I and stage II tumors (P=.008). Furthermore, 3q gain was detected preferentially in stage III and stage IV tumors (6/8) rather than in stage I and stage II tumors (1/9, P=.013). In our study, all tumors with gain of 3q also contained one or more loss(es) in common regions. On the other hand, all tumors with gain of 9p did not contain 3q gains. These observations indicate that gain of 3q and accumulation of DSCNAs are strongly associated with tumor progression in OSCC. Furthermore, 3q gain and loss of one or more additional loci in common aberration regions appears to be a group of DSCNs associated with dominant genetic pathways of leading to advanced OSCCs.

Chromosomal alterations in squamous cell carcinomas of the head and neck: window to the biology of disease

BACKGROUND

Cytogenetic alterations underlie the development of squamous cell carcinomas of the head and neck (SCCHN). Because many of the molecular genetic changes in SCCHN result from chromosomal alterations, a complete perspective on the genetic changes in tumors requires a basic introduction to cytogenetics. This review presents a brief description of the latest cytogenetic techniques and a description of chromosomal alterations in SCCHN, their molecular correlates, and clinical implications.

RESULTS

The most frequent cytogenetic alterations in SCCHN are gains of 3q, 8q, 9q, 20q, 7p, 11q13, and 5p and losses of 3p, 9p, 21q, 5q, 13q, 18q, and 8p. The karyotypes often provide an explanation for the mechanism by which the molecular genetic alterations arose. For example, the coordinate gains and losses involving whole arms of chromosomes 3, 5, 7, 8, and 9 often result from isochromosome formation. In addition, apparent allelic imbalances may not represent loss of heterozygosity but gene amplification. These results suggest that cytogenetic analysis is valuable for placing the molecular genetic findings in perspective at the cellular level.

CONCLUSIONS

Cytogenetic endpoints may be useful tools for dissecting clinical differences in tumor behavior and response to therapy. Numerous studies are underway to examine the biology of and genetic alterations in SCCHN that will lead to additional markers for use as rapid, noninvasive screening methods for individuals at high risk for primary or recurrent SCCHN. Our goal is to minimize morbidity and mortality from SCCHN by identifying useful predictors of disease and recurrence risk and response to therapy to implement earlier detection and more effective prevention and/or treatment strategies.

The structure of the human LIM protein ACT gene and its expression in tumor cell lines

We describe the human ACT genomic and cDNA sequence which like its murine counterpart contains the defining secondary structure of the FHL (Four-and-a-Half LIM-domain) LIM-protein family. The coding region of the human ACT gene spans five exons. This distribution is very similar to the FHL1 gene and includes the arrangement of split codons across exon boundaries suggesting that these genes share a common ancestor. The human ACT gene was not detected by Northern analysis in the adult testis although this is the only known site of expression found with its murine counterpart. However, the human ACT gene was found to be expressed in a panel of human tumor cell lines derived from squamous cell carcinomas, melanomas, and leukemias. Interestingly, FHL1, FHL2, and FHL3 were also found to be expressed in some of these cell lines and the results suggest an important role for FHLs in tumor biology.

P53 and cyclin D1 staining patterns of malignant and premalignant oral lesions in age-dependent populations

OBJECTIVE

Recent epidemiologic studies have identified a trend of increasing cancer incidence in younger patients. The purpose of this study was to determine whether this might be reflected by different molecular mechanisms for tumor development.

STUDY DESIGN

Dysplastic and malignant oral lesions from age-distinct patient populations were immunohistochemically analyzed for expression of p53 and cyclin D1. Chi-square analysis was used to determine statistical significance.

RESULTS

Eighty-two percent of "older" and 75% of "younger" carcinomas stained positively with p53; 63% of carcinomas in the older population and 55% of carcinomas in the younger population showed cyclin D1 positivity. Dysplasias showed similar cyclin D1 staining in both groups. Interestingly, 100% of "younger" dysplasias stained positively for p53, whereas 35.3% of "older" dysplastic lesions showed immunoreactivity. Staining of carcinomas was not statistically significant, whereas p53 staining of dysplasias proved highly significant (P < .025).

CONCLUSIONS

p53 immunoreactivity is detectable at an earlier stage of carcinogenesis in younger patients than in the traditional risk population for oral cancer.

Frequent allelic loss and homozygous deletion in chromosome band 8p23 in oral cancer

Frequent loss of heterozygosity on chromosome 8p in a variety of human malignancies, including head and neck cancers, has suggested the presence of a tumor suppressor gene (or genes) associated with the pathogenesis of these cancers. To test the role of genetic alterations at 8p23 in oral carcinogenesis, we studied 51 squamous cell carcinomas of the head and neck and 29 oral squamous cell carcinoma cell lines for allelic loss using 7 microsatellite markers spanning approximately 5 cM of chromosome band 8p23. Twenty-three of 51 tumors (45%) and 23 of 29 cell lines (79%) showed allelic loss at 1 or more loci. Three cell lines showed homozygous deletion of loci within a 3 cM region defined by the markers D8S1781 and D8S262. Our results suggest that a tumor suppressor gene (or genes) is located in 8p23 and is associated with the development and/or progression of oral carcinomas.

Deletion mapping defines three discrete areas of allelic imbalance on chromosome arm 8p in oral and oropharyngeal squamous cell carcinomas

Deletions on chromosome arm 8p, as defined by allelic imbalance, are a frequent event in many different types of malignant tumors, including those of the head and neck. These regions are thought to harbor tumor suppressor genes. In order to define a high-density deletion map of this chromosomal arm in oral and oropharyngeal squamous cell carcinomas, we have tested for allelic imbalance in 35 such tumors with 22 short tandem-repeat polymorphisms. Overall, 21 (60%) of the 35 tumors showed allelic imbalance at one or more loci on chromosome arm 8p. Interstitial deletions defined three discrete areas of deletion: at 8p23, 8p22, and 8p12-p21. Tumors of TNM stages II-IV showed a significantly higher frequency of allelic imbalance on 8p than did TNM stage I tumors. Our data suggest that there are least three tumor suppressor loci on chromosome arm 8p that may be implicated in oral carcinogenesis. Furthermore, inactivation of such genes may be associated with high-grade tumors.

Assessment of chromosomal gains and losses in oral squamous cell carcinoma by comparative genomic hybridisation

Cytogenetic studies have demonstrated that oral squamous cell carcinomas (OSCCs) are usually characterised by complex karyotypes with many marker chromosomes. We analysed the genetic changes of six OSCC cell cultures by comparative genomic hybridisation (CGH). The CGH technique provides information on chromosomal gains and losses of the whole tumour genome in a single experiment and can therefore identify regions that harbour putative tumour suppressor genes (in the case of loss of chromosomal material) or oncogenes (in the case of gain or amplification of chromosomal material). Recurrent losses were detected at chromosome arms Xp and 3p (four cases). Gains consistently occurred at chromosome arms 8q and 9q (four cases) and at 1q, 3q, 5p, 7p, and 9p (three cases). The same six tumour cultures have previously been analysed by classical karyotyping. An important discrepancy between the two techniques was the number of losses detected: 55 with karyotyping versus 26 with CGH. On the basis of the cytogenetic complexity of these tumours and on FISH experiments that confirmed the CGH results, we conclude that genetic changes, particularly losses, can be more reliably detected by CGH analysis.

Chromosomal changes observed in immortalized human keratinocytes transformed by ionizing radiation

Human epithelial cancer cells were induced by concerted action of DNA tumor virus and X-ray radiation. Treatment of nontumorigenic early passage AD12-SV40 immortalized epithelial cells (RHEK-1) at passage 23 with radiation, resulted in further changes in their growth properties. One day old cultures of these RHEK-1 cells were irradiated with graded doses of X-rays (0, 2, 4, 6, and 8 Gy i.e. RHEK-1, RHEK-1/200R, RHEK-1/400R, RHEK-1/600R, and RHEK-1/800R). Morphologic alterations, the ability to grow in soft agar, and to form rapidly-growing squamous cell carcinomas in nude mice were concomitantly acquired properties of the radiation transformed cell lines RHEK-1/200R and RHEK-1/ 400R. On the basis of commonality in having addition of some extra material in chromosome 11 in the region between q14/q22 in all tumorigenic cell lines RHEK-1/200R and RHEK-1/400R, and deletion of the same region in nontumorigenic irradiated cell lines-RHEK-1/600R and RHEK-1/800R, it is deduced this region may have some important oncogene/s or other gene/s that play an important role in tumorigenesis. When compared to squamous cell carcinoma data, the duplication observed in the present study is also observed in 28 to 38% of head and neck and also in 25% of cases of untreated malignant lesions of oral squamous cell carcinoma. Thus, this study shows the correlation between in vitro induced squamous cell carcinoma to in vivo tumors.

Cell, tissue and organ culture as in vitro models to study the biology of squamous cell carcinomas of the head and neck

In vitro models are currently being used to study head and neck squamous cell carcinoma (HNSCC). Several hundred HNSCC cell lines have been established by various investigators and used to study a broad spectrum of questions related to head and neck cancer. The head and neck model with respect to multistage carcinogenesis is now complete. Several techniques exist for the culture of normal epithelial cells from the upper aerodigestive tract (UADT). The biology of these UADT cells (oral cavity, oropharynx, hypopharynx and larynx) is being studied. Successful culture of premalignant lesions (dysplastic mucosa, leukoplakia, erythroplakia) has resulted in establishment of a limited number of premalignant cell lines and cell cultures. HPV infection of normal oral epithelial cells for immortalization (approximately premalignant cells) coupled with transformation with carcinogens (malignant cells) has established an experimental model for progression. Two in vivo models for oral carcinogenesis, the 7,12 dimethylbenz(a)anthracene-induced hamster cheek pouch model and the 4-nitroquinoline-N-oxide rat oral model, have been established in culture. Thus, multistage carcinogenesis models have been established from both human tissues and animal models and include cultures of normal, premalignant and malignant cells. Culture techniques for growing dissociated primary tumor cells for short term experimental analysis are being used. The culture of normal or tumor tissue as organ/explant cultures allows for the maintenance of normal cell-cell and cell-matrix interaction, but limits experimentation since these cultures cannot be propagated. Several three dimensional model systems are being used to obtain this histological complexity but allow for experimentation. The ability to culture normal, premalignant and malignant cells coupled with the use of a variety of culture techniques, should allow for the continued growth and experimentation in head and neck cancer research.

Centromeric breakage as a major cause of cytogenetic abnormalities in oral squamous cell carcinoma

Cytogenetic analysis of short-term explant tumor cultures derived from 11 human oral squamous cell carcinomas (nine from primary tumors and two from nude mouse xenograft cultures) revealed clonal chromosomal aberrations with multiple numerical and structural changes in all tumors. Recurrent breakpoints were located at chromosomal bands 1p13 (five tumors), 11q13 (four tumors), 3q27-29 (three tumors), and 12q13 (three tumors). Four tumors had a homogeneously staining region at band 11q13. Consistent chromosomal losses included 3p, 9p13-pter, and 18q22-qter, each occurring in eight tumors. Gain of material was observed for chromosome arms 3q, 5p, 7p, and 8q. As many as 134 of a total of 218 chromosomal breakpoints (61%) occurred in centromeric regions, often resulting in isochromosomes and unbalanced whole-arm translocations. Using fluorescence in situ hybridization with chromosome-specific centromeric alphoid repeat probes, two whole-arm translocations, der(Xq;11q) and a der(3q;11q), each from a different tumor, were shown to contain juxtaposed centromeric sequences of both participating chromosomes, strongly suggesting that the breakpoints were within the centromeres. We propose that centromeric breakage is an important mechanism for the generation of genetic imbalance in the development of oral squamous cell carcinoma.

Effects of transforming growth factor beta-1 on growth-regulatory genes in tumour-derived human oral keratinocytes

This study examined the effect of transforming growth factor beta-1 (TGF-beta 1) on c-myc, RB1, junB and p53 expression together with pRb phosphorylation, in carcinoma-derived and normal human oral keratinocytes with a range of inhibitory responses to this ligand. Amplification of c-myc was observed in eight of eight tumour-derived cell lines and resulted in corresponding mRNA expression. The down-regulation of c-myc expression by TGF-beta 1 predominantly reflected growth inhibition by TGF-beta 1, but in two of eight tumour-derived cell lines which were partially responsive to TGF-beta 1 c-myc expression was unaltered by this ligand. While RB1 mRNA levels were unaltered by TGF-beta 1, the ligand caused the accumulation of the underphosphorylated form of the Rb protein in all cells irrespective of TGF-beta 1-induced growth arrest. junB expression was up-regulated by TGF-beta 1 in cells with a range of growth inhibitory responses. All cells contained mutant p53. TGF-beta 1 did not affect p53 mRNA expression in both tumour-derived and normal keratinocytes and there was no alteration in p53 protein levels in keratinocytes expressing stable p53 protein following TGF-beta 1 treatment. The data indicate that TGF-beta-induced growth control can exist independently of the presence of mutant p53 and the control of Rb phosphorylation and c-myc down-regulation. It may be that TGF-beta growth inhibition occurs via multiple mechanisms and that the loss of one pathway during tumour progression does not necessarily result in the abrogation of TGF-beta-induced growth control.

TGF-beta receptor regulation mediates the response to exogenous ligand but is independent of the degree of cellular differentiation in human oral keratinocytes

This study examined the expression of TGF-beta cell-surface receptors, the response to exogenous TGF-beta 1 and the autocrine production of TGF-beta in normal and squamous cell carcinoma-derived human oral keratinocytes with variable degrees of cellular differentiation. TGF-beta receptor expression, the response to exogenous ligand and the autocrine production of TGF-beta appeared unrelated to cellular differentiation. Cells expressed variable proportions of type-I, -II and -III TGF-beta receptors. The expression of type-III receptors correlated inversely with the expression of type-I receptors, but there was no relationship between type-II and either type-I or type-III TGF-beta receptors. Normal cells and the majority (7 of 8) of tumour-derived keratinocytes were inhibited by exogenous TGF-beta 1 and the degree of inhibition correlated with the expression of type-I, but not type-II or type-III, TGF-beta receptors. One tumour-derived cell line was refractory to exogenous TGF-beta 1 although it expressed all 3 receptor types. Endogenous TGF-beta was produced by both normal and tumour-derived keratinocytes and correlated inversely to the expression of type-I, but not type-II, TGF-beta receptors. Further, cells that produced more autocrine TGF-beta had a diminished response to exogenous TGF-beta 1. The data indicate a complex interaction between the expression of TGF-beta cell-surface receptors, endogenous ligand production and the cellular response to exogenous TGF-beta 1.

Epidermal growth factor and transforming growth factor alpha characteristics of human oral carcinoma cell lines

This study examined the expression of epidermal growth factor (EGF) cell-surface receptors, the response to exogenous ligand and the autocrine production of transforming growth factor alpha (TGF-alpha) in normal and carcinoma-derived human oral keratinocytes. One of eight malignant cell lines overexpressed EGF receptors, while the remainder expressed receptor numbers similar to normal cells. Exogenous EGF stimulated incorporation of tritiated thymidine in a dose-dependent manner. In keratinocytes expressing normal numbers of EGF receptors, the cellular response to exogenous EGF correlated positively with total EGF receptor number. SCC-derived keratinocytes produced more TGF-alpha than normal cells. There was no statistical correlation between the autocrine production of TGF-alpha, EGF cell-surface receptor expression and cellular response to exogenous EGF. While the growth-stimulatory effects of exogenous TGF-alpha were inhibited by the addition of a neutralising antibody, the presence of this antibody in conditioned medium failed to produce a similar decrease in growth. The results indicate that overexpression of EGF receptors is not an invariable characteristic of human oral squamous carcinoma-derived cell lines. Further, the contribution of TGF-alpha to the growth of normal and carcinoma-derived human oral keratinocytes in vitro may be less significant than previously documented.