Molecular techniques and genetic alterations in head and neck cancer

Abstract

It is well known that cellular DNA alterations can lead to the formation of cancer, and there has been much discovery in the pathways involved in the development of head and neck squamous cell carcinoma (HNSCC). With novel genome-wide molecular assays, our ability to detect these abnormalities has increased. We now have a better understanding of the molecular complexity of HNSCC, but there is still much research to be done. In this review, we discuss the well described genetic alterations and touch on the newer findings, as well as some of the future directions of head and neck cancer research.

Introduction

The genetic alterations associated with head and neck squamous cell carcinoma (HNSCC) are numerous and include a variety of different pathways. The accumulation and selection of these aberrant pathways can sometimes be due to random chance, but more commonly, they are due to a lifetime of environmental exposure, such as tobacco and alcohol. Thus, the opportunity for DNA damage is high, and there is often a multistep accumulation of genetic events that leads to the development of HNSCC.

The fundamental and simplified concept of the genetic basis behind cancer is the overexpression of oncogenes and/or the silencing of tumor suppressor genes. However, the field of molecular biology has gone well beyond simply looking for areas of deletion, mutation, or amplification. This review will cover many of the known genetic alterations in HNSCC and attempt to compartmentalize them based mostly on molecular techniques. It is important to note, however, that there are many different ways to assess a particular gene's importance in carcinogenesis. While much progress has been made in our understanding of the genetic basis of HNSCC, it is clear that we have only begun to understand the complex interactions between the gene products and pathways, the transcriptional regulation of these genes, and the post-translational modifications that occur. We will try to simplify what we do know while highlighting some of the newer, promising areas of research.

Cytogenetics and Loss of Heterozygosity

The cytogenetic analysis of cells has evolved from the gross visual analysis of chromosomes arrested in metaphase, to a detailed high-resolution map of the regions of chromosomal gain, loss, and translocation. Spectral karyotyping adds the use of chromophores specific to each chromosome to provide contrast, in addition to the normal banding pattern seen in metaphase. Comparative genomic hybridization is a technique whereby normal and tumor DNA is labeled and hybridized to normal metaphase chromosomes. The fluorescence pattern can then be analyzed for increased or decreased intensity. Fluorescent in-situ hybridization (FISH) employs sequence specific probes that are detected after hybridization, thus allowing for detection of particular genes of interest, as well as a direct visualization of copy number per cell.

These techniques have all been applied in HNSCC, and have shown rather consistent results. Some investigators have also elected to use the techniques in parallel in order to confirm their results. The common regions of chromosomal loss reported are at 1p, 3p, 4p, 5q, 8p, 10p, 11q, 13q and 18q, with gains at 1q, 3q, 5p, 7q, 8q, 9q, 11q, 12p, 14q, and 15q. ^1-5 A recent study looked at DNA ploidy using FISH and found that there was a significantly higher number of nondiploid cells in malignant lesions versus premalignant cells. ⁶

Single nucleotide polymorphisms (SNPs) are areas scattered around the genome that have altered DNA sequences that may not lead to an amino acid alteration, or altered sequences that do not seem to have any adverse effect in ‘normal’ individuals. These SNPs may be markers for disease predisposition, or may be used to genetically identify patients, as they tend to cluster with ethnic background. SNPs located in the DNA repair genes may be a marker for HNSCC development, and this has been suggested in many studies.^7-10 These polymorphisms have also been detected in enzymes thought to detoxify environmental toxins and certain patterns may confer higher or lower risks for HNSCC. ^11-15 The Pro/Pro genotype of the TP53 codon 72 polymorphism was found to be associated with an increased risk of oral cancer in non-smokers and associated with a worse prognosis. ¹⁶ SNPs can also be used as a surrogate representation of other particular genes of interest, providing information about loss or gain of that genomic region.

Microsatellites are tandem nucleotide repeats generally located within noncoding areas of the genome. They can have variable length and have been mapped to specific chromosomal regions, allowing for detection of adjacent genes of interest. Using simple PCR based techniques, one can identify if there is loss of genetic material, represented by complete deletion, or loss, of one allele (also known as loss of heterozygosity or LOH).

This technique was the basis of the HNSCC progression model, indicating that there is a relatively common pattern of DNA allelic loss as one progresses from the premalignant to the malignant stage. ¹⁷ Presumably, a tumor suppressor gene (TSG) may be in the area of loss and thus would make the host more susceptible to dysfunction of this gene, leading to the development of cancer. LOH has some direct clinical implications in terms of prognosis and will be discussed further in another chapter.

Microarrays

Broadly, microarray technology involves the miniaturization of DNA sequence hybridization onto microscopic surfaces, which can then be read by a precise laser, able to detect and interpret the signal of these minute fluorophores. These microarrays can be used to detect DNA or RNA and have evolved to incorporate nearly the entire known human genome in a single experiment.

CGH or SNP arrays have the ability to finely map regions of chromosomal gain or loss, much more precisely than the conventional molecular techniques. CGH arrays have been used to profile HNSCC samples ^18,19, and one study identified a novel putative oncogene, LRP12 using this technique. ²⁰ SNP array has also been used in HNSCC and has drawn attention to loss of chromosome 6q. ²¹ CGH analysis has also been used to help predict differences in chemoradiosensitivity, as demonstrated by a recent publication where there was a clear distinction in genetic profile between the chemoradiosensitive and resistant patient groups²².

While CGH or SNP arrays focus on the DNA alterations, expression microarrays examine the mRNA component of samples. Early experiments focused on creating gene expression lists that differentiated tumor from normal ^23,24, or even from premalignant disease. ²⁵

More recent analyses have chosen to focus on particular subgroups and to determine what the expression differences are that contribute to the different behavior or characteristics between them. For example, one could develop a gene expression panel to differentiate those samples which presented with nodal metastasis. One group found a panel of 158 genes that differentiated metastatic versus nonmetastatic tumors²⁶, and another group found a 160 gene set²⁷. These gene products can then be further investigated to determine whether they have true functional roles in the development of metastasis. One study used microarrays focused on angiogenesis and metastatic pathways to look at predicting locoregional failure in patients undergoing primary chemoradiotherapy. The investigators found that MDM2 and erbB2 expression were independent negative predictors of locoregional disease-free survival.²⁸ Another similar study looked at expression arrays comparing HNSCC cell lines that exhibited relative radioresistance and radiosensitivity: 167 genes were detected, and of these, 25 were mapped as cancer related genes involved with growth, proliferation, apoptosis, and adhesion.²⁹ One recent study profiled 59 HNSCCs and found that there was a cluster of overexpression in hypoxia-regulated genes, which was an independent prognostic factor for recurrence-free survival.³⁰

There is no question that microarray technology has greatly improved our ability to identify meaningful targets and to condense years of work into one experiment. The analysis of such large amounts of data is an entire field of study unto itself, and in the future, integrating the data from DNA and mRNA arrays will be an especially powerful tool in identifying the key elements involved in HNSCC.

Tumor Suppressor Genes

The natural cell cycle includes programmed cell death, or apoptosis. Cells that are able to sidestep this process, or gain a selective growth advantage over the other cells have the potential to be come cancers. TSGs are those genes that normally function to maintain this growth homeostasis. They regulate the cell cycle, apoptosis, cell adhesion, and DNA repair. Aberrations such as deletions or mutations in these genes can lead to unchecked cell division and cancer formation.

TP53, a gene involved in apoptosis and cell cycle regulation, is one of the earliest TSGs discovered in cancer and found to be one of the most pervasive across all tumor types. ³¹ It was initially thought to be an oncogene, as there was aberrant overexpression of mutated forms of TP53, but it was subsequently shown that these forms were not functional. ³² Indeed, it was the silencing of the gene most often through various point mutations that led to tumor formation, and this is the case as well for premalignant disease ³³ and HNSCC. ³⁴ Complete knockout of the TP53 gene in a mouse model led to the development of metastatic squamous cell carcinoma in 100% of the animals, whereas the heterozygous mice developed tumors much later³⁵. While TP53 mutations are a rather complex set to study, a large-scale, recent analysis shows that disruptive TP53 mutations (those which lead to a large degree of alteration of the protein structure) were associated with reduced survival in HNSCC patients.³⁶ The TP53 story becomes even more intricate as the MDM class of oncogenes, specifically MDM2 and now potentially MDM4, have been shown to inhibit TP53 in the absence of TP53 mutation.³⁷ Overall, TP53 continues to be a major focus of study in HNSCC, particularly now that there is evidence that it has a role in predicting patient outcome and response to therapy. ^38-43

The 9p21 locus has been a topic of study in HNSCC, as there is a high incidence of LOH and cytogenetic abnormality at this site. The CDKN2a locus codes for both p16 and p14(ARF), which are putative TSGs that regulate the cell cycle and stabilize TP53 through MDM2 respectively. Despite the uniformly low expression of these TSGs in HNSCC, there has been a lack of concomitant genetic alterations, such as mutation or LOH. The explanation is most likely that this locus is under epigenetic control, such as promoter hypermethylation, which serves to silence the gene.^44,45

There have been numerous other putative TSGs involved in HNSCC. More recently, the deleted in colorectal cancer (DCC) gene located at 18q21 has been under study for its involvement in HNSCC as a conditional TSG which requires binding to netrin-1 for its growth effects. ^46,47 High resolution CGH array analysis of HNSCC also helped to detect loss of the FAT gene at 4q35. FAT is thought to play a role in cell-cell adhesion within the cadherin family. ⁴⁸ 8-oxoguanine DNA glycosylase 1 (OGG1) is a DNA repair enzyme whose reduction in function has been shown to be a risk factor in lung cancer and head and neck cancer. Interestingly, OGG function returned to baseline after successful treatment of the patients' HNSCC⁴⁹.

Our concept of tumorigenesis has not changed in that the role of TSGs continues to be studied in carcinogenesis. It is likely that there are a myriad of TSGs that are silenced in HNSCC, but the pattern of gene silencing may be quite different between patients. Identifying the relevant genes is an important aspect to our understanding of the development of cancer, and in many cases seems to have prognostic as well as therapeutic implications.

Oncogenes

The identification of oncogenes has been somewhat more limited due to technical aspects of molecular biology – it is easier to identify a mutation or loss of DNA, than it is to identify the meaningful overexpression of a gene. It makes intuitive sense that the aberrant overexpression of those genes that are useful in promoting the growth, survival, and spread of cells could lead to the development of a cancerous cell, but the degree and timing of overexpression needed is less evident.

One of the more studied oncogenes in HNSCC is the epidermal growth factor receptor (EGFR). It has been shown to be overexpressed in a majority of HNSCC tumors⁵⁰ and has been linked to improved survival. ⁵¹ There are many therapeutic implications of EGFR, as it works through the tyrosine kinase cascade. This gene will be further discussed in Chapter 19.

The ΔNp63-α gene is a homologue to TP53 and has been shown to be overexpressed in HNSCC with oncogenic effects. ⁵² Detailed mechanistic studies have suggested that it leads to the accumulation of beta-catenin, a transcription factor that may play a role in cell adhesion and in the Wnt signaling pathway. A more recent study showed that ΔNp63-α inhibited p73, a TSG related to TP53, which is also involved in the bcl-2 mediated apoptotic pathway. ⁵³

Matrix metalloproteinases (MMPs) are a family of genes thought to be involved in cell adhesion, proliferation, and migration. MMP7 gene expression was associated with a risk of early HNSCC.⁵⁴ In an early study, MMP2 was found to be related to lymph node metastasis ^55,56, and a subsequent study confirmed that MMP9 and other MMPs were related to infiltrative growth and lymph node involvement. ^56,57 In an oral SCC cell lines, MMP3 appeared to be necessary for anchorage-independent growth. ⁵⁸

Often times, the cytogenetic analyses will focus researchers to particular amplified regions of the cancer genome to identify putative oncogenes. For example, 11q13 amplification has consistently been shown to be present in HNSCC, which has sparked many investigations of the possible targets located at this locus. Recent CGH-array based analysis isolated a 1.7Mb region with 13 possible genes. The investigators finally settled on the Fas-associated death domain as an important gene of interest, that may also be linked to taxol sensitivity.⁵⁹ Other possible oncogenes involved in HSNCC include c-Jun NH2-terminal kinases (JNK), which are involved in T-cell differentiation and apoptosis.⁶⁰ Integrated arrays have looked at combining CGH array with expression to match those areas of relative gain to increased expression. Some novel targets on chromosome 7p14 have been identified, including NT5C3, ANLN, and INHBA.⁶¹

As our molecular techniques advance into more detailed mapping of the genome, our ability to identify oncogenes and TSGs has become more refined as well. As with TSGs, the range of oncogenes being identified is quite broad, and the often interrelated mechanisms by which they work are becoming more complex.

Mitochondrial Mutations

Mitochondria are certainly mysterious elements within the cellular structure, as they are self-contained organelles with a separate genome and even possess a unique genetic code. While they are known to play a role in cellular respiration and the generation of ATP, their role in carcinogenesis has only recently come to light.

There is an accumulation of mitochondrial mutations in HSNCC, as evidenced by the increased mitochondrial DNA content seen as one progresses from premalignant to malignant head and neck lesions. ⁶² One study found that point mutations in the ATP subunit 6 may have an antiapoptotic effect in cancer cell lines. ⁶³ A more direct mechanistic link between cancer formation and mitochondria involve the accumulation of succinate (part of the tricarboxylic acid cycle), which inhibits the degradation if hypoxia-inducible factor(HIF)-1alpha, an angiogenic factor activated by tumor hypoxia.⁶⁴ Direct sequence analysis of HNSCC samples showed that mitochondrial mutation was a frequent event and also confirmed the accumulation of HIF-1alpha in cell lines. ⁶⁵

The role of mitochondrial mutations in carcinogenesis will continue to evolve, and they are an attractive target as markers of disease, as there is an abundance of mitochondrial DNA present in every cell. Furthermore, there may be an expanded array of possible therapeutic interventions given their unique role in cellular function.

Other Genetic Alterations

There are many other newer avenues of research that have emerged, and their impact in HNSCC remains to be seen.

Proteomics involves the high-throughput analysis of proteins within samples. Such assays have been used in HNSCC samples and have shown promising results. ^66-69 The patterns detected and the particular proteins identified can be used to characterize the tumors, as well as provide insights into the mechanisms involved in carcinogenesis. There is much more that needs to be done in terms of the bioinformatic analysis of these assays and in the proper identification of the protein peaks. Once there is a better understanding of the data produced by proteomic studies, it opens the door to a broader array of serum analyses whereby the diagnosis, response to therapy, and recurrence might be detected by a simple blood test.

The cancer stem cell hypothesis proposes that there are subsets of cancer cells which have the ability to self-renew, and it is these cells that are the main progenitor cells that fuel carcinogenesis. A recent study sorted HNSCC cells based on the cell surface marker CD44 found that the CD44(+) population had the ability to give rise to new tumors in vivo, and could be passaged and differentiate in vitro. ⁷⁰ While this experiment has not yet been replicated in HNSCC, it is an intriguing concept in that most experimental molecular biology assumes that there is a clonal population of tumor cells, all sharing the same genetic alterations. The cancer stem cell hypothesis would argue that eliminating the subpopulation of progenitor cells would be much more efficient in treating tumors. Thus, there are widespread clinical and research implications, making it a very exciting area of study.

MicroRNA (mir) molecules are small ∼22 nucleotide, non-coding RNA molecules that have been shown to regulate post-transcriptional gene expression. ⁷¹ These mirs are thought to be involved in a host of cellular processes including differentiation, apoptosis, and proliferation. MicroRNA arrays are now being performed on HNSCC, but the interpretation of the results is still at its early phase.⁷² The interactions of the mirs can be difficult to predict, and each one may have several putative targets. However, there is no question that they play a large role in the regulation of gene transcription, and they are also a powerful experimental tool in studying the effects of gene silencing in vitro or in vivo.

Conclusion

The field of molecular biology has been growing exponentially, with the global expansion of our understanding of cancer biology in general, and the introduction of newer molecular techniques. Our understanding of the fundamental mechanisms behind HNSCC has improved, but this has also led to a greater appreciation for the complexity of this disease process. The lab research led to some clinical applications as discussed in a subsequent chapter, but there is much more that needs to be understood before we are able to make a significant impact on the treatment of patients.