UvA-DARE (Digital Academic Repository)
Molecular-cytogenetic characterization of head and neck cancer: Identification of novel prognostic factors and gene targets for therapy [double dissertation 2]
Wreesmann, V.B.; Singh, B.
Publication date 2004
Link to publication
Citation for published version (APA): Wreesmann, V. B., & Singh, B. (2004). Molecular-cytogenetic characterization of head and neck cancer: Identification of novel prognostic factors and gene targets for therapy [double dissertation 2].
General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).
Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.
UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl) Download date:02 Oct 2021 Molecularr Profiling in Head and Neck Cancer
Abstract t Squamouss cell carcinomas of the head and neck (HNSCC) constitute an anatomically het- erogeneouss group of neoplasms that share in common a causal association with tobacco and alco- holl exposure. Here we report the use of cDNA microarrays containing 17,840 clones to monitor genee expression changes associated with tumor progression in ten HNSCC patients. We identi- fiedd 366 over-expressed and 246 under-expressed genes when comparing invasive HNSCC tumorss to histologically normal surgical margins. In addition, we identified 55 over-expressed andd 71 under-expressed genes when comparing these invasive tumors to lymph node metastases fromm the same patients. A combination of these data sets identified genes that show a consistent patternn of expression during progression from normal to invasive tumor, and ultimately to metastaticc lymph node. We validated the value of this gene discovery approach by selecting moesin,, a member of the ERM family of cytoskeletal proteins, and one of the genes that consis- tentlyy increased during tumor progression, for subsequent immunohistochemical analysis using aa HNSCC tissue array.
eadd and neck squamous cell carcinoma (HNSCC) represents the fifth most commonn malignancy worldwide, representing a major international health Hproblem.. The five-year survival rate for this disease has improved only mar- ginallyy over the past decade; as a result, over 30,100 cases and 7,800 deaths occur annu- allyy in the United States.[7] There are few molecular markers that can be reliably used inn early detection or as indicators of prognosis. Several karyotypic or gross chromoso- mall aberrations have been investigated as markers of disease progression and/or out- come,, including gains of 3q, 8q, 9q, 20q, 7p, llql3, and 5p, losses of 3p, 9p, 21q, 5q, 13q, 18q,, and 8p.[2] Studiess involving cDNA microarrays have identified specific genes whose expres- sionn has changed in HNSCC samples compared to normal tissue. Villaret et al. [3], using arrayss containing 985 clones, examined sixteen HNSCC cases and identified nine over- expressedd genes. Utilizing laser capture microdissection of HNSCC cells to measure the expressionn of 588 known cancer-related genes, Leethanakul et al. [4] demonstrated increasedd expression of genes related to the wnt and notch signaling pathways, as well as aa decrease in expression of differentiation markers such as cytokeratins. Another expressionn profiling study using laser capture microdissection of normal and malignant orall epithelium identified about 600 differentially-expressed genes with oligonucleotide arrays.[5]] Microarray studies of HNSCC have also focused on molecular classification of HNSCC.. Hanna et al. [6] were able to identify 60 tumor-related genes from a cDNA microarrayy containing 1187 genes that could successfully predict the radiation response off tumor samples. Our preliminary study of molecular classification of HNSCC is the firstt use of microarray gene expression data to predict outcome in this disease.[7] Heree we report the use of cDNA microarrays containing 17,840 clones to monitor genee expression changes associated with tumor progression in ten HNSCC patients. We examinedd gene expression changes in invasive HNSCC tissue compared to normal sur- gicall margins, as well as gene expression in HNSCC invasive tumor tissue compared to
209 9 Chapterr 13 metastaticc lymph nodes from the same patient. We have utilized the statistical approachess outlined here to identify 366 over-expressed and 246 under-expressed genes whenn comparing invasive tumors to normal surgical margins. In addition, we have iden- tifiedd a number of genes whose expression is changed when comparing invasive tumor andd lymph node metastasis. The second aspect of our study of tumor progression involvess a combination of these data sets to identify genes that show a consistent pattern off expression during progression from normal mucosa to invasive tumor, and ultimate- lyy to metastatic lymph node. Our analysis to categorize genes based on their expression patternss have identified 140 genes which consistently increase in expression during pro- gressionn from normal tissue to invasive tumor, and subsequently to metastatic node (in att least 4 of the 9 cases studied). A similar list of 94 genes has been identified which decreasee in expression during tumor progression and metastasis. We validated the value off this gene discovery approach by selecting moesin, one of these candidate genes that consistentlyy increased during tumor progression, for subsequent immunohistochemical analysiss using a HNSCC tissue array. Our findings demonstrated that moesin protein expressionn correlated with the progression of this disease. Materialss and Methods Tissuee Samples and RNA Extraction Wee obtained Institutional Review Board approval and informed patient consent fromm each patient before obtaining invasive tumor tissue, matched normal surgical mar- gin,, and metastatic lymph node tissue at time of surgery for HNSCC at Memorial Sloan- Ketteringg Cancer Center. Only patients undergoing surgical treatment with curative intent,, having a history of smoking and no prior non-surgical treatment for head and neckk cancer were included in the study. In order to rule out gene expression alterations duee to stromal cell contamination, we confirmed that each tumor specimen used in our studyy contained greater than 70% cancer cells by analysis of corresponding hematoxylin andd eosin stained sections. Tissue samples were flashfrozen in liquid nitrogen and stored att -80°C prior to RNA extraction. Tissues were homogenized in Trizol reagent using a Brinkmannn Model PT 10/35 Tissue Homogenizer and total RNA extracted using Trizol reagentt following the protocol of the manufacturer (Invitrogen/Life Technologies, Carlsbad,, California). Total RNA was subsequently cleaned by column purification usingg an RNeasy Mini Total RNA purification kit (Qiagen, Valencia, California). T77 Linear Amplification and Hybridization to cDNA Microarrays AA set of 17,840 sequence-verified human IMAGE cDNA clones representing both knownn genes and expressed sequence tags (ESTs) were spotted onto poly-lysine coated microscopee slides using a custom robot designed and built at AECOM.fS] Prior to hybridization,, slides were pre-processed as described previously (http://microarraylk.aecom.yu.edu).. One round of linear amplification of tissue mRNA wass carried out using a modified T7 linear amplification protocol (9). Fluorescent label- ingg of probes was carried out by reverse transcription of amplified RNA with Superscript III reverse transcriptase (Invitrogen/Life Technologies) in a reaction containing 2.5 mM
210 0 Molecularr Profiling in Head and Neck Cancer dATP,, dCTP, dGTP, ImM dTTP, 10 Units of RNase Out (Invitrogen/Life Technologies),, 4 1 of Cy-labeled dUTP (Amersham-Pharmacia) in a final volume of 40 Reactions were carried out at 42°C for 2 hours. Fluorescently labeled cDNA probes weree combined and purified using a Microcon YM-50 spin column. Hybridization to cDNAA arrays was carried out overnight at 50°C in a buffer containing 30% formamide, 3XX SSC, 0.75% SDS and 100 ng of human Cot-1 DNA. Following hybridization, slides aree briefly washed with a solution of IX SSC, 0.1% SDS, then washed for 20 min at room temperaturee in 0.2X SSC, 0.1% SDS and 20 min at room temperature in 0.1X SSC (with- outt SDS). Slides are immediately dried as before and processed using the GenePix 4000AA microarray scanner (Axon Instruments, Foster City, California) and GenePix Pro 3.00 software for image processing and subsequent data analysis. Statisticall Analysis of Gene Expression Data Redd (Cy5) and green (Cy3) signal intensities for each element on the array were cal- culatedd using GenePix Pro 3.0. This software gives an integrated intensity per spot for eachh channel in addition to an integrated background count. In all subsequent analysis, wee used the mean background subtracted intensity for the two channels. For each spot, wee calculated the mean intensity over the spot in the two channels and from this sub- tractedd the median of the background intensity. We referred to this value simply as the intensityy in each channel and denoted them as: Ij and I2. Before using the spot intensi- tyy data for any analysis, it was crucial that the data from each microarray experiment be normalized.. We computed an intensity dependent normalization factor by first finding thee rank invariant subset of the spots (the spots that have equal or almost equal ranks in thee two channels).[70] Once this core of spots had been identified, a robust curve was fittedfitted using the lowess function from the R statistical package.[11] We then computed
_ thee intensity dependent normalization factor a(A) where A V i : was ^Q geo. metricc average intensity. We evaluated the variation in "chip to chip" comparisons by calculatingg the deviation for all array elements when a single experiment is replicated multiplee times. Our results showed that in four replicate experiments, the variance of thee log-transformed ratio was less than 0.40 for greater than 95% of the clones on our microarrayss (data not shown). This accuracy is well within cutoffs that have been cus- tomarilyy used in previous microarray studies. Ass part of our analysis, we were interested in genes that were consistently over- or underexpressedd throughout the patient tumor samples. One approach would be to sim- plyy sum the logs of the ratios for each gene, (iy=log(/7//2), and rank the genes which hadd the highest sums. This method was unsatisfactory for two reasons. First, noise lev- elss were often intensity dependent with noise increasing as the average intensity (A) decreased.. Therefore, a given ratio was less significant at lower than at higher intensi- ties.^]] In order to account for this effect, we computed an intensity dependent noise factor,, o(A), using a procedure similar to that used tocompute the intensity dependent normalization.. We used this factor to compute a weighted score for the log ratios simi-
211 1 Chapterr 13 larr to a statistical z-score. Genes could then be ranked by the sum of the noise-weighted logg ratios to order to identify those that were either over- or under-expressed in the sam- ples.. A second problem with summing the logs of the ratios was that a gene sometimes hadd a large expression ratio in one or two samples and changed only modestly (or not at all)) in the remaining ones. Since we were interested in genes that were consistently over- orr underexpressed, we transformed the weighted log ratio values as follows:
Thiss calculation bound the sum to lie between -N and N, where N was the number of tumorr samples analyzed. We used both the sum of the weighted log ratios and this "bounded-score"" to rank the genes and determine those that were either over- or under- expressedd in the tumor samples. Immunohistochemicall Analysis of Moesin Expression Threee different HNSCC microarrays were constructed for this study at Memorial Sloan-Ketteringg Cancer Center's Laboratory of Epithelial Cancer Biology. Five microm- eterr sections of normal and tumor tissues were embedded in paraffin and stained with hematoxylinn and eosin to identify viable, morphologically representative areas of the specimenn from which needle core samples could be taken. From each specimen, tripli- catee tissue cores with diameters of 0.6 mm were punched and arrayed onto a recipient paraffinn block using a precision instrument (Beecher Instruments, Silver Spring, MD).[/3]] Five micrometer sections of these tissue array blocks were cut and placed onto microscopee slides and used for immunohistochemical analysis. A total of 23 normal tonguee epithelium, 30 dysplastic epithelial lesions from HNSCC patients, 42 HNSCC andd 7 lymph node metastases were analyzed. Standardd immunoperoxidase procedures were used for immunohistochemical analy- siss of moesin expression. We utilized mouse monoclonal clone 38/87 at 1:50 (4 ug/ml) withh microwave pretreatment of the slides (Neomarkers, Fremont, CA); staining condi- tionss were optimized on sections from formalin-fixed, paraffin-embedded tissue controls ass specified by the manufacturer. Antibody reactivity was detected by using diaminobenzidinee as chromogen, and sections were counterstained with hematoxylin. Thee primary antibody was omitted for negative controls. All specimens (n=102) used forr analysis of the association between moesin expression and progression of HNSCC weree assessed using the non-parametric Wilcoxon-Mann-Whitney and Kruskall-Wallis tests.[14]] The consensus value of the three representative cores from each tumor sample arrayedd was used for statistical analyses. Expression values were displayed as median, 25%-75%% interquartile ranges, and mean values accompanied by 95% confidence inter- valss and range.
212 2 Molecularr Profiling in Head and Neck Cancer
Resultss and Discussion Thee focus of our analysis was to identify genes that were consistently under- or over- expressedd during HNSCC tumor progression. Three types of analysis were conducted to minimizee effects of measurement and biologic variation on identifying consistent pat- ternss of gene expression with cDNA microarrays. First, Cy5/Cy3 ratios were determined forr each gene. Second, since noise levels in the dataset are intensity dependent, we uti- lizedd an intensitydependent noise factor to compute a weighted score for the log(Cy5/Cy33 ratio) of a given gene (Z-score, see Experimental Procedures). Third, becausee assessment of differential expression of a gene by summation of the logs of the ratioo measurements is susceptible to outlier values that occur often in microarray exper- iments,, we computed a "bounded-score" in which the values are bound between N and - N,, thereby reducing the effect of extremely large ratios on the sum (See Experimental Procedures).. This transformation avoids the inclusion of genes having a huge differen- tiall ratio in one or two samples and only modest changes in the remaining ones. Forr each patient in our study, gene expression patterns were compared as follows: histologicallyy normal adjacent mucosa was compared to invasive HNSCC tumor, and invasivee HNSCC tumor was then compared to metastatic lymph node. We used the Cy5/Cy33 ratio, the weighted log ratios and the transformed "bounded-score" to identify thosee genes that are consistently over- or under-expressed in these samples. Furthermore,, by combining the two datasets for each patient, we were able to partition geness into clusters based on their patterns of gene expression during tumor progression. Invasivee HNSCC vs Normal Adjacent Mucosa Wee compared gene expression in the invasive tumor tissue (Cy5) to the adjacent "normal"" surgical margin (Cy3) from the same patient. Of the 17,840 clones present on ourr human cDNA microarray, we identified 794 clones having a Cy5/Cy3 ratio greater thann 2.0 in at least 5 of the 10 patients, 705 clones in which the sum of the Z-score was greaterr than 11, and 883 clones in which the sum of the bounded-score was greater than 5.. We narrowed the list to 353 clones which were included in all three datasets. Similarly,, for the identification of underexpressed genes, we identified 612 clones hav- ingg a Cy5/Cy3 ratio less than 0.5 in at least five of the ten patients, 794 clones in which thee sum of the Z-score was less than -11, and 878 clones in which the sum of the bound- ed-scoree was less than -5. We narrowed this list to 246 clones which were included in all threee datasets. The cut-offs chosen for this analysis were largely arbitrary, in that we wantedd to limit the lists to a manageable number of genes. An abbreviated list of the dif- ferentiallyy expressed genes is given in Table 1. Complete lists of differentially expressed geness are available on our web site (http://microarray.info). Cellularr Adhesion Wee identified changes in the expression of several genes associated with cellular adhesion.. For example, the trans-membrane glycoprotein CD44 is well-established as an invasion-promotingg molecule, with the expression of variant exons by invasive and metastaticc oral SCCs.[/5] In fact, correlations have been made between the pattern of
213 3 Chapterr 13
Tablee IA. Partial list of genes over-expressed in invasive HNSCC compared to normal adja- centt mucosa. Included are the accession number for the clone printed on the array, gene description,, chromosomal location, number of patients in which the gene was over-expressed (R/GG ration >2), the median R/G ratio for all 10 patients, and the sums of the Z-score and bounded-score. .
Accessionn Gene Name Chromosomee Cy5/Cy3 Patients Z-score Bounded -score e ADHESION N H44784 4 Bullouss pemphigoid antigen (BPAG1) 6pl2 2 2.18 8 5 5 19.80 0 8.61 1 AA282906 6 CD444 antigen llpl3 3 3.56 6 8 8 22.72 2 8.08 8 AA194833 3 C!audin-11 (CLDN1) 3q28 8 4.61 1 9 9 35.52 2 8.14 4 R62612 2 Fibronectinn 1 (FN1) 2q34 4 3.3 3 9 9 22.20 0 6.57 7 T77595 5 Hexabrachionn (HXB) 9q33 3 5.08 8 8 8 28.24 4 8.20 0 AA148200 0 Integrinn linked kinase (ILK) llpl5 5 2.55 5 7 7 14.21 1 5.77 7 H44722 2 Integrin,, alpha 10 (ITGA10) lq21 1 2.01 1 5 5 13.53 3 6.23 3 AA424695 5 Integrin,, alpha 3 (ITGA3) 17q21 1 2.84 4 6 6 18.09 9 7.30 0 AA425451 1 Integrin,, alpha E (ITGAE) 17pl3 3 2.3 3 8 8 23.76 6 8.68 8 AA029934 4 Integrin,, alpha V (ITGAV) 2q31 1 2.18 8 7 7 13.45 5 5.88 8 AA598653 3 Osteoblastt specific factor 2 (OSF-2) 13ql3 3 3.02 2 5 5 21.87 7 5.31 1 AA775616 6 Secretedd phosphoprotein 1 (SPP1) 4q21 1 1.98 8 5 5 22.361 1 5.84 4 H38240 0 Thrombospondinn 2 (THBS2) 6q27 7 2.27 7 6 6 15.31 1 5.51 1 CELLL CYCLE AA4487555 CDC25B 20pl3 3 2.38 8 8 8 17.17 7 7.05 5 AA458994 4 Cyclinn A2 (CCNA2) 4q25 5 2.28 8 7 7 13.24 4 6.52 2 AA447662 2 Gemininn (GEM) 6p21 1 2.43 3 6 6 15.29 9 6.34 4 AA449336 6 Proteinn regulator of cytokinesis 1 (PRC1) 15q26 6 2.19 9 6 6 11.96 6 5.49 9 IMMUNE E AA670408 8 Beta-22 microglobulin(B2M) 15q21 1 3.22 2 8 8 19.47 7 7.70 0 AA701652 2 Granulysinn (GNLY) 2pl2 2 2.82 2 7 7 21.02 2 7.20 0 AA485353 3 Mac-22 binding protein 17q25 5 4.07 7 8 8 20.49 9 7.31 1 METABOLISM M AA6835788 Adenosine deaminase (ADA) 20ql2 2 3.08 8 7 7 18.92 2 6.42 2 AA0435511 Beta-1, 3-N-acetylglucosaminyltransferase 5 3q28 8 3.22 2 7 7 14.76 6 6.36 6 (B3GN-T5) ) AA4501899 Enolase 2 (gamma, neuronal) (NSE) 12pl3 3 2.09 9 5 5 23.73 3 6.33 3 AA4875822 Exostoses (multiple) 1 (EXT1) 8q24 4 4.37 7 8 8 26.38 8 6.96 6 R693077 Leucine aminopeptidase 4pl5 5 2.58 8 5 5 12.94 4 5.78 8 H700177 Mannosidase, alpha type II (MAN2A1) 5q21 1 3.24 4 7 7 12.11 1 5.84 4 R005955 N-acetylgalactosaminyltransferase 2 (GALNT2) lq41 1 2.64 4 5 5 15.44 4 6.48 8 AA4601155 Ornithine decraboxylase 1 (ODC1) 2p25 5 3.28 8 5 5 25.07 7 5.59 9 AA4463011 Paraoxonase 2 (PON2) 7q21 1 2.08 8 5 5 13.35 5 5.50 0 AA4018833 Sialidase 1 (NEU1) 6p21 1 2.25 5 6 6 15.69 9 7.86 6 AAOI12155 Spermidine/spermine Nl-acetyltransferase (SAT) Xp22 2 2.46 6 7 7 19.33 3 6.09 9 AA1361255 Spermine synthase (SMS) Xp22 2 2.36 6 5 5 12.98 8 5.37 7 PROLIFERATIONN AND GROWTH AA A1295 5 2 Forkhead box M1 12pl3 3 2.34 4 5 5 13.90 0 5.28 8 W427233 GROl oncogene 4q21 1 3.13 3 8 8 21.30 0 7.46 6 AA4545722 MCM2 3q21 1 2.14 4 5 5 13.51 1 5.41 1 AA6639955 MCM6 2q21 1 2.78 8 5 5 14.87 7 6.43 3 AA4502655 Proliferating cell nuclear antigen (PCNA) 20pl2 2 3.33 3 6 6 12.08 8 5.78 8 H547522 Replication factor C (RFC4) 3q27 7 2.9 9 8 8 16.51 1 6.81 1 AA6292622 Serine/threonine protein kinase (Polo-like) (PLK) 16ql2 2 2.12 2 5 5 13.72 2 6.31 1 PROTEINN MODIFICATION AA5989500 Cathepsin B (CTSB) 8p22 2 2.61 1 7 7 12.48 8 5.51 1 AA6440888 Cathespsin C (CTSC) llq!4 4 4.57 7 8 8 23.14 4 8.59 9
214 4 Molecularr Profiling in Head and Neck Cancer
Tablee 1 A. continued Accessionn Gene Name Chromosomee Cy5/Cy3 Patients Z-score Bounded -score e AA0460666 F-box and WD-40 domain protein 2 (FBXW2) 9q34 4 2.27 7 14.41 1 6.26 6 AA5986400 Midline 1 (MIDI) Xp22 2 2.16 6 13.71 1 5.25 5 R334822 P21 -activated kinase 2 (PAK2) 15 5 2.09 9 11.32 2 5.61 1 AA4814644 Peptidylprolyl isomerase B (cyclophin B) 15q21 1 2.59 9 13.32 2 7.16 6 AA2517700 Proteasome 26S subunit, ATPase, 2 7q22 2 2.47 7 11.47 7 5.65 5 AA9347622 Proteasome 26S subunit, non-ATPase, 11 17pl3 3 2.9 9 13.03 3 5.58 8 H513777 Proteasome 26S subunit, non-ATPase, 11 17pl3 3 2.7 7 11.58 8 5.43 3 R336088 Proteasome 26S subunit, non-ATPase, 5 9q33 3 2.38 8 14.04 4 6.91 1 H653955 Proteasome activator subunit 2 (PA28 beta) 14qll l 3.37 7 15.20 0 6.76 6 T687588 Proteasome subunit, beta type 1 6q27 7 2.69 9 12.56 6 5.52 2 AA8624344 Proteasome subunit, beta type, 9 6p21 1 3.26 6 21.34 4 8.33 3 R432700 Protein phosphatase 1, regulatory subunit 14C 6q24 4 5.66 6 32.44 4 5.94 4 AA4647299 Ubiquitin carrier protein (E2-EPF) 17pl2 2 2.6 6 14.81 1 7.10 0 AA2920744 Ubiquitin-conjugating enzyme (UBE2L6) llql2 2 3.53 3 20.16 6 7.58 8 SIGNALLING G AA011062 2 Bonee morphogenetic protein 2 (BMP-2) 20pl2 2.28 8 19.48 8 5.11 1 AA449773 3 CDC422 effector protein 4 (CEP4) 17q24 3.14 4 19.51 1 6.28 8 R42668 8 Chorionicc somatomammotropin hormone 1 (CSH1) 17q22 3.16 6 13.67 7 7.00 0 AA446994 4 Fibroblastt growth factor receptor 4, variant 1 (FGFR4)5q35 2.9 9 7 7 23.22 2 7.61 1 AA625981 1 FK506-bindingg protein 1 (FKBP-12) 20pl3 3.14 4 7 7 11.84 4 5.94 4 AA406420 0 Guaninee nucleotide binding protein, alpha 1 (GHANI1) 7q21 1 2.75 5 5 5 12.89 9 5.34 4 AA487912 2 Guaninee nucleotide binding protein, beta 1 (GNB1) lp36 2.11 1 5 5 16.58 8 6.51 1 AA436549 9 Nitricc oxide synthase interacting protein (NOSIP) 19ql3 2.35 5 6 6 11.04 4 5.18 8 AA425450 0 NMBB transmembrane glycoprotein (GPNMB) 7pl5 6.4 4 30.15 5 7.86 6 H48472 2 Non-kinasee Cdc42 effector protein (SPEC2) 5q 2.01 1 15.00 0 6.27 7 W72473 3 Phosphoinositide-3-kinase,, catalytic, alpha (PIK3CA)3q26 2.31 1 13.24 4 6.86 6 T87069 9 Prostaglandinn F2 receptor negative regulator (CD9P-l)lpl3 2.25 5 20.44 4 7.32 2 AA449333 3 RAB31,, member RAS oncogene family 18pll 3.67 7 22.34 4 6.85 5 AA057378 8 RAB32,, member RAS oncogene family 6q24 2.75 5 6 6 17.73 3 6.77 7 H63668 8 RAB7,, member RAS oncogene family 3q21 2.07 7 6 6 12.55 5 6.42 2 AAO17544 4 Regulatorr of G-protein signalling 1 (RGS1) lq31 2.95 5 7 7 20.32 2 6.74 4 AA284668 8 Urokinase-typee plasminogen activator (PLAU) 10q24 3.73 3 8 8 32.51 1 6.90 0 R83836 6 Yamaguchii sarcoma viral related oncogene homolog (LYN) 8ql3 3 3.03 3 6 6 30.52 2 8.60 0 STRESS S AA485036 6 Heatt shock protein 105kD beta (HSP105B) 13ql2 2 2.21 1 12.41 1 5.58 8 R71440 0 Heatt shock protein 47 (SERPINH2) llql3 3 3.74 4 23.34 4 9.51 1 AA598526 6 Hypoxia-induciblee factor 1, alpha subunit (HIF-1) 14q21 1 3.31 1 13.00 0 5.58 8 STRUCTURAL L AA4901722 Collagen, type 1, alpha-2 (COL1A2) 7q22 2 2.39 9 20.38 8 5.93 3 H875366 Collagen, type XVII, alpha 1 (COL17A1) 10q24 4 4.38 8 17.36 6 6.66 6 AA7060222 Keratin, type II cytoskeletal 1 (KRT1) 12qll l 3.12 2 30.74 4 6.63 3 AA0014322 Laminin, alpha 3 (LAMA3) 18qll l 2.59 9 16.25 5 7.30 0 AA6775344 Laminin, gamma 2 (LAMC2) lq25 5 5.1 1 34.40 0 9.62 2 AA1432011 Matrix metalloproteinase 1 (MMP1) llq21 1 6.12 2 34.30 0 8.87 7 W517944 Matrix metalloproteinase 3 (MMP3) llq22 2 2.37 7 19.21 1 6.91 1 R229777 Moesin Xqll l 3.79 9 21.55 5 9.15 5 AA0577966 Myosin IB 2ql2 2 4.3 3 26.83 3 8.13 3 AA1879777 Myosin X 5pl5 5 2.28 8 17.88 8 7.60 0 T699266 Myosin, heavy polypeptide 9 22ql3 3 2.65 5 14.94 4 6.64 4 R491444 Tubulin, alpha (TUBA1) 2q36 6 2.35 5 19.26 6 6.34 4 AA8654699 Tubulin, alpha (TUBA3) 12ql2 2 3.39 9 12.43 3 5.08 8 AAA 126760 Tubulin, gamma 2 (TUBG2) 17q21 1 2.09 9 11.61 1 5.75 5
215 5 Chapterr 13
Tablee 1 A. continued Accessionn Gene Name Chromosome e Cy5/Cy3 3Patient s sZ-scor e eBounde d d -score e TRANSCRIPTION N AA7753555 DNA HELICASE II (XRCC5) 2q35 5 2.17 7 5 5 13.03 3 5.71 1 H20908 8 DNAA repair helicase (ERCC3) 2q21 1 2.34 4 7 7 13.65 5 7.55 5 AA504348 8 DNAA Topoisomerase II alpha (T0P2A) 17q21 1 2.82 2 8 8 15.12 2 5.73 3 AA481769 9 DNA/chromatinn binding protein (PLU-1) lq32 2 2.03 3 5 5 12.12 2 6.35 5 H14359 9 ETSS domain transcription factor (ELF4) Xq26 6 3.07 7 8 8 17.57 7 5.95 5 AA844141 1 ETSS oncogene family (ELK1) Xpll l 2.09 9 5 5 14.56 6 5.60 0 AA455929 9 KETT gene (p53 related) 3q27 7 5.27 7 8 8 25.40 0 6.94 4 AA011232 2 Methyl-CpGG binding domain protein 4 (MBD4))) 3q21 2.02 2 5 5 13.13 3 6.85 5 H23021 1 Retinoblastoma-bindingg protein 8 (RBBP8) 18qll l 2.44 4 7 7 19.35 5 5.03 3 AA425238 8 Runt-relatedd transcription factor 1 (AML1C) 21q22 2 2.29 9 5 5 14.48 8 6.45 5 AA486367 7 STAT1 1 2q32 2 4.1 1 8 8 23.69 9 7.70 0 H40921 1 ZIC22 protein (ZIC2) 13q32 2 2.49 9 7 7 22.99 9 8.14 4 N64741 1 Zincc finger protein SLUG 8qll l 1.92 2 5 5 11.09 9 5.03 3 TRANSLATIONN AND RNA PROCESSING AA6764711 Euk. translation initiation factor (eIF3), subuni1t99 9 7p22 3.13 3 7 7 15.31 1 7.56 6 R63918 8 Neuronatinn alpha (NNAT) 20qll l 2.19 9 6 6 12.69 9 6.40 0 AA599178 8 Ribosomall protein L27a (RPL27A) llpl5 5 3.24 4 8 8 16.76 6 5.17 7 H06113 3 Ribosomall protein L3 (MRPL3) 2.04 4 5 5 15.01 1 7.16 6 T65211 1 SFRSS protein kinase 2 (SRPK2) 7q22 2 2.33 3 5 5 13.19 9 6.73 3 AA779221 1 snRNP-specificc protein (U5-116KD) 17q21 1 2.08 8 6 6 11.30 0 5.94 4 AA496787 7 Splicingg factor, arginine/serine-rich 4 (SFRS4) lp33 3 2.08 8 5 5 12.06 6 5.57 7 TRANSPORT T AA6782800 Beta-3A-adaptin (AP3BI) 5pl4 4 1.93 3 5 5 13.89 9 5.67 7 R76281 1 Copperr uptake protein (CTR2) 9q31 1 2.08 8 5 5 14.69 9 6.04 4 AA630776 6 Delta-adaptinn (AP3D1) 19pl3 3 3.65 5 8 8 26.15 5 9.67 7 H58873 3 Glucosee transporter, type 1 (GLUT1) lp35 5 3.33 3 8 8 17.40 0 6.88 8 AA676460 0 Karyopherinn alpha 2 (KPNA2) 17q23 3 3.74 4 6 6 11.36 6 5.67 7 H29407 7 LIV-11 (Breast cancer, estrogen regulated) 18ql2 2 3.92 2 8 8 25.02 2 8.58 8
Tablee IB. Partial list of genes under-expressed in invasive HNSCC compared to normal adja- centt mucosa. Included are the accession number for the clone printed on the array, gene description,, chromosomal location, number of patients in which the gene was under-expressed (R/GG ratio <0.5), the median R/G ratio for all 10 patients, and the sums of the Z-score and bounded-score. .
Accession n Genee Name Chromosomee Cy5/Cy3 Patients Z-score Bounded -score e ADHESION N AA480851 1 Claudinn 10 (CLDN10) 13q31 1 0.39 9 6 6 -27.61 1 -6.02 2 AA709271 1 Neurall cell adhesion molecule 2 (NCAM2) 21q21 1 0.33 3 5 5 -18.39 9 -7.43 3 H94471 1 Occludinn (OCLN) 5ql3 3 0.45 5 5 5 -18.15 5 -8.08 8 CELLL CYCLE T81764 4 Celll division cycle 27 (CDC27) 17ql2 2 0.40 0 5 5 -15.57 7 -6.82 2 AA454080 0 Inhibitorr of DNA binding 4 (ID4) 6p22 2 0.43 3 5 5 -18.71 1 -6.04 4 W94630 0 Minichromosomee maintenance deficient (MCM5) 22ql3 3 0.58 8 5 5 -12.85 5 -6.20 0 IMMUNE E AA227594 4 Mai,, T-cell differentiation protein 2ql3 3 0.25 5 6 6 -45.34 4 -6.22 2 AA127741 1 RU2S S 6p22 2 0.43 3 6 6 -16.91 1 -6.74 4 AA868278 8 Testiss specific protein 1 (TPX1) 6p21 1 0.41 1 6 6 -17.73 3 -6.24 4 METABOLISM M W443400 Aconitase 2, mitochondrial (AC02) 22ql3 3 0.48 8 5 5 -12.68 8 -6.26 6 N934288 Alcohol dehydrogenase, beta polypeptide (ADH2) 4q21 1 0.15 5 7 7 -39.19 9 -7.27 7
216 6 Molecularr Profiling in Head and Neck Cancer
Tablee IB continued Accessionn Gene Name Chromosomee Cy5/Cy3 Patients Z-scoree Bounded -score e AA8448188 Amylase, alpha 2A; pancreatic (AMY2A) lp21 0.51 1 5 5 -14.78 8 -15.17 7 AA6994699 Carbonic anhydrase VA, mitochondrial (CAVA) 16q24 0.50 0 5 5 -14.03 3 -6.58 8 W848688 Fatty acid omega-hydroxylase (CYP4A11) 1 p 0.49 9 6 6 -13.68 8 -6.63 3 R440055 Glutamate decarboxylase 2 (GAD2) lOpll 0.53 3 5 5 -12.77 7 -6.25 5 AA6824233 Monoadmine oxidase B (MAOB) Xpl 1 0.51 1 5 5 -14.42 2 -5.87 7 H933188 Niphedipine oxidase (CYP3A) 7q21 0.46 6 6 6 -15.69 9 -6.81 1 H685099 UDP glycosyltransferase 2, BIO 4ql3 0.42 2 6 6 -14.82 2 -7.13 3 PROTEINN MODIFICATION H913899 Lipoprotein-associated coagulation inhibitor (LACl)2q31 0.54 4 5 5 -17.30 0 -5.98 8 H040288 Serine proteinase 22 (P11) 12ql3 3 0.35 5 7 7 -32.35 5 -5.92 2 SIGNALLING G N66871 1 Glucocorticoidd receptor alpha, variant 3' UTR 5q q 0.57 7 5 5 -12.72 2 -5.45 5 AA878391 1 Glypican-55 (GPC5) 13q32 2 0.51 1 5 5 -12.62 2 -6.07 7 N63500 0 Intersectinn 2 (ITSN2) 2p25.1 1 0.46 6 5 5 -14.89 9 -6.84 4 AA443157 7 Phosphodiesterasee 4D interacting protein (PDE4DIP)lql2 0.41 1 5 5 -14.57 7 -5.32 2 H03040 0 Pregnancy-specificc beta-1 glycoprotein (PSG) 19ql3 3 0.49 9 5 5 -17.19 9 -7.55 5 R55786 6 Proteinn kinase A anchor protein 6 (AKAP6) 14ql2 2 0.54 4 5 5 -15.08 8 -6.45 5 R89715 5 Proteinn kinase C, gamma (PRKCG) 19ql3 3 0.55 5 5 5 -16.34 4 -6.76 6 STRUCTURAL L AA6291899 Keratin 4 (KRT4) 12ql3 3 0.24 4 6 6 -23.87 7 -6.31 1 W600577 Keratin 13, type I, cytoskeletal (KRT13) 17q21 1 0.45 5 6 6 -33.57 7 -7.38 8 TRANSCRIPTION N R96155 5 Defectivee mariner transposon (HSMAR2) 0.43 3 6 6 -24.14 4 -6.36 6 T97599 9 Deltexx homolog 1 (DTX1) 12q24 0.57 7 5 5 -15.25 5 -7.36 6 AA057425 5 Myeloid/lymphoidd or mixed-lineage leukemia (MLLT2) 4q21 1 0.43 3 6 6 -22.10 0 -5.65 5 AA037352 2 Pairedd box gene 1 (PAX 1) 20p 11 0.35 5 5 5 -22.68 8 -6.41 1 R70541 1 Transcriptionn factor (Myc/Max/Mad family member)7qll 0.40 0 5 5 -16.00 0 -5.93 3 AA449189 9 Zincc finger protein (ZFP95) 7q22 0.34 4 6 6 -27.41 1 -6.31 1 T91080 0 Zincc finger protein Ikaros (ZNFN1A1) 17p 13 0.47 7 5 5 -11.79 9 -5.37 7 TRANSPORT T N59214 4 ATP-bindingg cassette, sub-family G, member 2 (ABCG2) 4q22 2 0.54 4 5 5 -13.84 4 -7.14 4 H79534 4 Hemoglobin,, epsilon 1 (HBE1) llpl5 0.36 6 6 6 -18.72 2 -5.17 7 N49856 6 Neurotransmitterr transporter (BTG1) 12pl3 0.47 7 6 6 -11.08 8 -5.12 2 N62948 8 Organicc anion transporting polypeptide (OATP) 12pl2 0.47 7 6 6 -12.77 7 -5.78 8 AA975384 4 Potassiumm voltage-gated channel member 5 (KCNA5)12pl3 0.55 5 5 5 -13.42 2 -6.03 3 W90588 8 Synaptogyrinn 1 (SYNGR1) 22ql3 0.49 9 5 5 -14.47 7 -6.66 6
CD444 variant exons and clinicopathological variables, including metastasis and sur- vival.. [7 6] In addition to the fibronectin subunit FN1, a gene previously shown to be over-expressedd in papillary thyroid carcinoma [17], we observed an increase in the expressionn of the ECM glycoprotein tenascin (TN). Strong expression of TN in HNSCC wass confirmed previously.[75] Interestingly, TN expression in laryngeal SCC correlated withh the expression of fibronectin, CD44 and others.[19] We also observed an increase inn the expression of Bullous perphigoid antigen 1 (BPAG1), an antigen previously detectedd in extracts of human SCC cells as part of hemidesmosomal structures promot- ingg the adhesion of epithelial cells to the underlying basement membrane. [20] This changee may be part of an alteration of the hemidesmosome during squamous cell car- cinogenesis.. Enhanced expression of another family member, BPAG2, had been
217 7 Chapterr 13 observedd in pre-cancerous and cancerous tissues, including invasive SCC.[27] Our group off adhesion-related genes also included members of the integrin family and associated proteins.. In oral SCC, expression of integrins is highly variable, but some patterns, such ass the losses of pi integrins, a6|34, av(35, and the expression of av(36, have emerged.[22] Wee observed the over-expression of four integrin a subunits in invasive SCC (a10, a3, aE andd av). Although not previously identified in oral SCC, flow cytometric analysis of a colonn cancer-derived cell line showed that a3 (CD49c) was involved in an increase in metastaticc activity.[25] The increase in the ECM protein osteopontin (OPN) confirmed detectionn of the protein in pre-malignant and malignant lesions arising from oral epithe- lium.^]] Although its function in carcinogenesis remains unclear, it had been demon- stratedd that tumor-derived OPN was able to inhibit macrophage function and enhance survivall of SCC metastases.[25] Wee also observed a decrease in the expression of some adhesion-related genes in invasivee oral SCCs, including claudin 10 and occludin (Table IB). Evidence had previ- ouslyy demonstrated a loss of some claudin family members in HNSCC.[26] Most inter- estingg is the loss of expression of occludin (OCN), one of the major structure-determin- ingg transmembrane proteins located at cellular tight junctions. Using an in vitro invasion assayy with oral SCC cell lines, Shibata et al [27] showed that inhibition of invasion by malotilatee involved development of cell-cell adhesions and a dose-dependent increase in zonulaa occludin. The loss of occludin had also been observed in poorly differentiated gastrointestinall adenocarcinomas.^] Proliferationn and Cell Cycle Regulation Thee emergence of genes related to cellular proliferation in our study was not sur- prising.. Expression of genes such as proliferating cell nuclear antigen (PCNA) had pre- viouslyy been shown to correlate with the prognosis of patients with oral SCC.[29] Expressionn of the proangiogenic chemokine growth regulated oncogene 1 (GROl) was alsoo increased in invasive SCC. Previous observations confirmed that increased expres- sionn of this gene in SCC was associated with metastatic tumor progression, and was reg- ulatedd by the activation of nuclear factor kappa B (NF-kappaB).[50] The overexpression off the Forkhead box transcription factor FoxMl was intriguing as it had previously been shownn to upregulate expression of cyclin Bl.[37] Cyclin Bl overexpression predicted poorr prognosis in several cancers, including oral SCC.[32] With an increase in cellular proliferationn and tumor growth, we expected to observe an increase in cell cycle-related genee expression. We observed over-expression of CDC25B, a phosphatase identified as a potentiall human oncogene.[J3] Strong expression had been associated with deep inva- sionn in gastric carcinoma and SCC of the esophagus, and was a significant predictor of poorr prognosis in colorectal carcinoma.[34, 35, 36] In HNSCC, both CDC25A and CDC25BB were overexpressed in a large percentage of invasive tumors.[37] Also increased wass PLK, a novel Polo-like serine/threonine kinase identified as a marker for cellular proliferation.. [38] It had been established that constitutive expression of this gene result- edd in malignant transformation of mammalian cells.[39] Furthermore, PLK phosphory-
218 8 Molecularr Profiling in Head and Neck Cancer latess cyclin Bl at the cytoplasmic retention signal, thereby regulating its entry into the nucleuss during prophase.[40] PLK transcript levels were elevated in most HNSCC tumors,, and this increase in expression correlated with both nodal stage and metasta- sis.^/] ] Metabolicc Pathways Thee increase in genes related to various metabolic pathways underscored the obser- vationn that tumor cells alter the synthesis and degradation of a variety of cellular com- ponents.. For example, we observed an increase in the expression of genes such adeno- sinee deaminase (ADA), leucine aminopeptidase and the glycolytic enzyme enolase 2, all geness previously identified in HNSCC. In fact, serum levels of leucine aminopeptidase andd enolase 2 were related to the extent of lymph node spread in HNSCC and poor prog- nosiss in NSCLC.[42, 43] Of interest were the emergences of genes related to glycosyla- tion.. We know that oncogenic transformation in HNSCC is accompanied by many alter- ationss in glycosylation at the surface of tumor cells.[44] Specific carbohydrate structures participatee both in cell-cell and cell-matrix interactions, and can play a major role in tumorr metastasis.[45] We observed an increase in the expression of alpha mannosidase II,, a key enzyme in N-glycan processing and a target for anticancer therapies. Inhibition off this enzyme by the alkaloid swainsonine reduced incidence of metastasis of melanoma cells,, and had been selected for clinical testing based on its anticancer activity. [46] Anotherr area of interest was the increase in the expression of genes related to polyaminee metabolism, specifically ornithine decarboxylase (ODC1), spermidine/sper- minee Nl-acetyl transferase (SAT) and spermine synthase (SMS). Weiss et al [47] estab- lishedd that squamous cell tumorigenesis was accompanied by an increase in polyamine biosynthesis,, and elevated levels of ODC, the rate limiting enzyme in that pathway. We havee confirmed the increase expression of ODC in our study, as well as the increased expressionn of spermine synthase (SMS), another enzyme in the pathway of polyamine biosynthesis.. These observations agreed with studies showing that depletion of polyaminee levels in HNSCC cells by difluormethylornithine (DMFO) inhibited prolif- erationn and increased apoptosis in these cells, an effect that was reversed by the addition off putrescine or spermidine.[45] Similarly, intracellular concentrations of polyamines andd ODC activity were increased in colorectal cancer tissue and in premalignant polyps.[49]] We were surprised to find an increase in the expression of spermidine/sper- minee Nl-acetyl transferase (SAT), a rate-limiting enzyme of the polyamine degradation pathway.. However, the simultaneous increase in activity of SSAT and ODC has been reportedd previously in bladder carcinogenesis. [50] The role of these seemingly opposing activitiess in SCC tumorigenesis remains to be understood. Post-translationall Modification and Protein Degradation Changess in protein modification and degradation pathways are known to be part of thee tumorigenic process. Cathepsins B and C both increased in expression in invasive orall SCC. Cathepsin B had been identified as a serum tumor marker for HNSCC, and wass shown to be significantly associated with advanced tumor stage and poor histologi-
219 9 Chapterr 13 call grade in oral cancer. [57] There was an increase in the expression of many proteaso- mall components, a reflection of the involvement of proteasomes in the invasive behav- iorr of oral SCC.[52] It had been reported that proteasome inhibitors, such as PS-341, inhibitedd growth and angiogenesis of SCC. [53] In fact, proteasome inhibition leading to p27Kipp accumulation was a predictor of malignant behavior in oral SCC.[54] The decreasee in expression of lipoprotein-associated coagulation inhibitor (LACI) had not beenn previously reported in oral SCC, but was intriguing for several reasons. Immunohistochemicall analysis of laryngeal SCC sections demonstrated the presence of LACII in connective tissue stroma adjacent to the tumor, and normal squamous epithe- liall cells. [55] Furthermore, downregulation of LACI expression using an antisense tran- scriptt in the human lung cancer cell line A549 increased the invasive potential of these cells.[56] ] Signall Transduction Comparisonss between invasive HNSCC tumors and corresponding "normal" tissue revealedd many changes in the expression of genes related to signal transduction path- ways.. In the interest of brevity, we highlighted only some of these for mention here, acknowledgingg that there were many genes of interest within this group. The catalytic subunitt of phosphatidylinositol-3-kinase (PIK3CA) was elevated in 8 of the 10 patients studied.. The association of this enzyme with the malignant behavior of cells has been welll established.[57] In the case of HNSCC, signaling through the EGFR/PI3K pathway cann lead to radioresistance of these cells.[58] Also increased was the urokinase-type plas- minogenn activator (PLAU), a protein implicated in invasion and metastasis of oral SCC, andd believed to be valuable in evaluating the clinical course of patients with this dis- ease.. [59] Also increased was the fibroblast growth factor receptor 4 (FGFR4) gene, amplificationn of which had been observed in human breast and gynecological can- cers.. [60] Specific allelic variation had been associated with lymph node metastasis and advancedd tumor stage, leading to the possibility that specific allelic variation could pre- disposee these patients to accelerated disease progression.[6/] Enhanced expressions of FGFF and alternate receptors had been previously identified as increased in HNSCC, and itt had been postulated that these contributed to tumor growth and metastasis via the mediationn of angiogenesis. [62] Structurall Proteins AA variety of genes encoding structural proteins increased in expression in invasive tumors,, including keratin 1 and collagens type I and type XVII. The degree of kera- tinizationn appears to be a valuable parameter in identifying oral SCC patients at risk for developingg regional metastases to the neck.[63] The increase in keratin 1 expression was alsoo accompanied by a loss of cytokeratins 4 and 13 (CK4 and CK13). Loss of both pro- teinss had been previously observed in tongue SCC, with the loss of CK13 being related too the invasive and metastatic ability of the cancer cell line.[64, 65] We also identified an increasee in expression of two subunits (alpha 3 and gamma 2) of laminin 5, one of the principall components of anchoring filaments and keratinocyte adhesion. Interestingly,
220 0 Molecularr Profiling in Head and Neck Cancer thesee specific subunits of laminin 5 were previously reported in the cytoplasm of bud- dingg carcinoma cells at the invasive front of oral SCC.[66] The increased expression of myosinn contractile proteins (MY09, MYO10 and MYOIB) was of interest in that it had beenn previously reported that malignant cells, including oral SCC, contained an increasedd number of contractile proteins.[67] These proteins may be involved in the activee migration of tumor cells, and may therefore play a role in the invasion of sur- roundingg tissue.[68] Transcription n Alterationss in gene expression are largely attributable to changes in the levels of transcriptionn factors and associated regulatory molecules that affect the expression of downstreamm "effector" genes. Our study revealed that in invasive HNSCC, there was an increasee in the expression of the zinc-finger transcription factor SLUG, which is known too induce growth-factor induced epithelial/mesenchymal transition during development viaa desmosome dissociation.[69] In breast carcinoma cells, expression of SLUG strong- lyy correlated with loss of E-cadherin transcripts, an adhesion molecule important in car- cinomaa progression.[70] Finally, the DNA-binding protein PLU-1 was also overex- pressedd in invasive HNSCC tumors. This gene was consistently overexpressed in the invasivee and in situ components of invasive breast cancer.[71] Overall,, our data demonstrated that these tumors have gene expression alterations thatt parallel those found in other kinds of tumors. Furthermore, we hypothesize that specificc combinations of these alterations are likely to be responsible for the specific propertiess of these tumors, including clinical staging, resistance to chemotherapeutic strategies,, and patient outcome. Invasivee HNSCC vs Lymph Node Metastasis Inn the second dataset, we compared gene expression in the metastatic lymph node tissuee (Cy5) to the original invasive tumor (Cy3) from the same patient. Due to the fact thatt one of the RNA samples (metastatic lymph node tissue from patient MSK143) was nott analyzable, we obtained comparison data for only 9 of the 10 patients in our study. Off the 17,840 clones present on our cDNA microarray, we identified 139 clones having aa Cy5/Cy3 ratio greater than 2.0 in at least 4 of the 9 patients, 617 clones in which the summ of the Z-score was greater than 11, and 716 clones in which the sum of the bound- ed-scoree was greater than 5. We narrowed the list to 55 clones which were included in alll three datasets. Similarly, for the identification of under-expressed genes, we identi- fiedd 214 clones having a Cy5/Cy3 ratio less than 0.5 in at least 4 of the 9 patients, 522 cloness in which the sum of the Z-score was less than -11, and 378 clones in which the summ of the bounded-score was less than -5. We narrowed the list to 71 clones which were includedd in all three datasets. Abbreviatedd lists of these differentially expressed genes are given in Table 2. In com- parisonn of gene expression patterns between invasive HNSCC tumor tissue and corre- spondingg lymph node metastasis, we see much fewer changes in the expression of genes thann observed when comparing invasive tumors to corresponding adjacent mucosa. One
221 1 Chapterr 13
Tablee 2A. Partial list of genes over-expressed in metastatic lymph node tissue compared to invasivee HNSCC. Included are the accession number for the clone printed on the array, gene description,, chromosomal location, number of patients in which the gene was over-expressed (R/GG ratio >2), the median R/G ratio for all 9 pateints, and the sums of the Z-score and bound- ed-score. .
Accessionn Gene Name omosom eem Cy5/Cy3 Patients sZ-scor e eBounde d d -score e ADHESION N AA A132090 CD533 glycoprotein antigen (CD53) lpl3 3 3.71 1 6 6 34.31 1 5.65 5 AA055862 2 Glycoproteinn A33 (transmembrane) (GPA33) lq23 3 2.30 0 5 5 21.46 6 6.12 2 N64384 4 Integrin,, alpha X (ITGAX) 16pll l 2.06 6 5 5 24.65 5 7.21 1 R21535 5 Intercellularr adhesion molecule 2 (ICAM-2) 17q23 3 2.45 5 5 5 17.54 4 5.95 5 W73144 4 L-plastinn (LCP1) 13ql4 4 2.55 5 6 6 32.66 6 6.15 5 IMMUNE E AA476285 5 CD44 antigen (p55) 12pl3 3 2.02 2 5 5 22.87 7 6.67 7 W74668 8 Glycophorinn C (GPC) 2ql4 4 2.01 1 5 5 21.81 1 5.98 8 AA701652 2 Granulysinn (GNLY) 2p!2 2 1.46 6 4 4 19.10 0 6.13 3 AA634028 8 HLA-DPA1 1 6p21 1 1.97 7 4 4 27.57 7 6.80 0 AI004331 1 HLA-DQB1 1 6p21 1 1.83 3 4 4 22.00 0 5.72 2 N67007 7 Lectin-likee NK cell receptor (LLT1) 12pl3 3 2.04 4 5 5 17.79 9 5.89 9 AA454784 4 Lymphocytee chemoattractant factor (LCF) 15q26 6 2.64 4 5 5 22.09 9 6.54 4 T83159 9 Lymphocyte-specificc protein 1 (LSP1) llplS S 1.98 8 4 4 21.50 0 5.96 6 AA131406 6 Monokinee induced by gamma interferon (MIG) 4q21 1 1.97 7 4 4 12.57 7 5.40 0 METABOLISM M N957611 Fucosidase, alpha-L-1 (FUCA1) lp34 4 2.71 1 6 6 27.83 3 6.70 0 AA455 5197 Glutathione peroxidase 4 (GPX4) 19pl3 3 1.53 3 4 4 17.09 9 5.68 8 AA2787599 Proteoglycan 1, secretory granule (PRG1) 10q22 2 2.15 5 5 5 24.92 2 6.55 5 H612433 Uncoupling protein 2 (UCP2) llql3 3 2.98 8 5 5 22.77 7 6.08 8 PROTEINN MODIFICATION N639433 Lysozyme (LYZ) 12ql3 3 2.45 5 5 5 20.39 9 5.34 4 R972266 PI-kinase-related kinase (SMG-1) 16pl3 3 2.03 3 5 5 17.46 6 6.79 9 SIGNALLING G R982955 Cat eye syndrome chromosome region, candidate 1 22qll l 1.76 6 4 4 14.59 9 5.86 6 (( CECR1) AA6764533 CD37 antigen 19pl3 3 2.39 9 5 5 23.29 9 5.09 9 N707655 Dedicator of cyto-kinesis 2 (DOCK2) 5q35 5 2.36 6 6 6 24.90 0 5.60 0 AA4245755 Hematopoietic Lineage Cell Specific Protein 3ql3 3 3.31 1 6 6 31.94 4 5.49 9 (HCLS1) ) AA0212099 NOD2 16ql2 2 2.67 7 5 5 19.96 6 5.70 0 H571800 Phospholipase C, gamme 2 (PLCG2) 16q24 4 2.11 1 5 5 20.82 2 5.06 6 AA4791022 Protein kinase C, beta 1 (PRKCB1) 16pll l 1.85 5 4 4 17.05 5 6.51 1 AA2366177 Guanine nucleotide exchange factor 6 Xq26 6 1.78 8 4 4 14.06 6 5.13 3 (ARHGEF6) ) W385711 Ras homolog gene family, member H (ARHH) 4pl3 3 2.88 8 5 5 26.01 1 5.66 6 AA4537744 Regulator of G-protein signalling 16 (RGS16) lq25 5 2.01 1 5 5 19.85 5 6.05 5 W960999 Retinoid X receptor, gamma (RXRG) lq22 2 1.96 6 4 4 18.42 2 7.52 2 AA4876344 Rho-GDP-dissociation inhibitor Ly-DGI beta 12pl2 2 2.15 5 5 5 29.07 7 6.41 1 (ARHGDIB) ) R838366 Yamaguchi sarcoma viral related oncogene 8ql3 3 2.00 0 26.62 2 6.64 4 homologg (LYN) STRUCTURAL L R097777 Breast cancer associated protein (BRAP1) 12ql3 3 3.35 5 5 5 22.48 8 5.05 5 R785300 Coactosin-like protein (CLP) 17pll l 1.68 8 4 4 27.16 6 5.84 4 AA6641799 Keratin 18 (KRT18) 12ql3 3 2.11 1 5 5 21.04 4 6.22 2
222 2 Molecularr Profiling in Head and Neck Cancer
Tablee 2B continued Accessionn Gene Name Chromosome Cy5/Cy3 Patients Z-score Bounded -score e TRANSCRIPTION N N321466 Avian reticuloendotheliosis viral oncogene 2pl3 3 1.68 8 4 4 23.15 5 6.43 3 homologg (REL) AA2348977 MADS box transcription enhancer factor 2C 5ql4 4 1.66 6 4 4 20.24 4 6.39 9 (MEF2C) ) AA4902677 Pleckstrin (PLEK) 2pl4 4 2.66 6 5 5 27.16 6 6.52 2 AA909033 Pleckstrin homology domain binding protein 2qll.2 2 2.33 3 5 5 18.57 7 5.29 9 (PSCDBP) ) AA0555855 Proto-oncogene BCD1 10pl5 5 2.52 2 6 6 25.67 7 5.19 9 H593655 Upstream stimulatory factor (USF1) lq22 2 1.95 5 4 4 20.38 8 5.33 3 TRANSLATION/RNAA PROCESSING AA4050000 Ribonuclease 6 precursor (RNASE6PL) 6q27 7 1.55 5 4 4 20.20 0 6.79 9 AA7015455 Ribonuclease K6 (RNASEK6) 14qll l 2.26 6 5 5 20.36 6 6.70 0 UNKNOWN N AA5212322 HSPC022 22 1.74 4 27.85 5.51
off the most interesting findings when comparing these tissues was the increase in expressionn of L-plastin, a member of the plastin family of actin-binding proteins. This genee had not been previously identified in HNSCC, although a previous study examin- ingg colorectal cancer progression with microarrays demonstrated an increase in L-plas- tinn expression when comparing gene expression profiles of colon cancer cell lines derivedd from invasive and metastatic tumors from a single patient. [72] In the same study,, immunological analysis of L-plastin in 58 colorectal cancer specimens revealed a statisticallyy significant correlation between expression of the protein and the progres- sionn of cancer staging, suggesting that L-plastin could be a useful metastatic marker. We alsoo observed a decrease in the expression of adhesion-related proteins, including junc- tionn plakoglobin (JUP), connexin 26, annexin Al, and desmosomal molecules desmo- plakinn and desmocollin-2 (Table 2B). Losses of desmoplakin and plakoglobin had been associatedd with distant metastasis in oral SCC, and the expression of plakoglobin had beenn shown to have predictive value for nodal metastasis in this disease.[73] One meta- bolicc gene of interest here was alpha-L-fucosidase. Although not previously identified in orall SCC, activity of this enzyme assayed on endometrial, cervical and ovarian tissue had shownn elevated activity in malignant tissue.[74] In fact, it had been suggested that the increasedd activity of fucosidase in tumors may be a mechanism by which cancer cells subvertt the process of macrophage activation. Outt of the group of genes whose expression was decreased in metastatic lymph nodes wass DOC1, a putative oral cancer suppressor that negatively regulates cyclin dependent kinasee 2 (CDK2) activity. Previous studies demonstrated that decreased DOC1 expres- sionn significantly correlated with cervical lymph node metastasis and the 10-year sur- vivall status of patients with surgically resected oral SCC. [75] Also included in this group off genes are the apoptosis-related protein APR-3, the human P2Y1 receptor, the apopto- sis-relatedd Napl gene (NCKAP1), and a novel BNIP-2 like gene (BNIP-S). The loss of expressionn of the cysteine proteinase inhibitors cystatins A and B in metastatic lymph
223 3 Chapterr 13
Tablee 2B. Partial list of genes under-expressed in metastatic lymph node tissue compared to invasivee HNSCC. Included are the accession number for the clone printed on the array, gene description,, chromosomal location, number of patients in which the gene was under-expressed (R/GG ration <0.5), the median R.G ratio for all 9 patients, and the sums of the Z-score and bounded-score. .
Accessionn Gene Name Chotnosoi m eem Cy5/Cy3 Patients sZ-scor e Boundee d d -score e ADHESION N AA4533044 AF-6 (Fushion partner for ALL-1) 6q27 7 0.49 9 5 5 -20.62 2 -6.63 3 H630777 Annexin Al (ANXA1) 9ql2 2 0.39 9 6 6 -27.46 6 -7.36 6 AA4906888 Connexin 26 (GJB2) 13qll l 0.13 3 7 7 -42.24 4 -7.78 8 AA0746777 Desmocollin-2 (DSC2) 18ql2 2 0.39 9 6 6 -29.88 8 -5.05 5 R334566 Desmoplakin I (DSP) 6p24 4 0.35 5 6 6 -37.29 9 -7.25 5 R064177 Junction plakoglobin (JUP) 17q21 1 0.60 0 4 4 -25.26 6 -5.15 5 R639711 LBP protein 32 (LBP-32) 2p25 5 0.53 3 4 4 -26.50 0 -5.88 8 AA0407033 Profilin 2 (actin monomer-binding protein) 3q25 5 0.48 8 5 5 -13.63 3 -5.05 5 (PFN2) ) CELLL CYCLE R786077 Deleted in oral cancer-1 (DOC1) 12q24 4 0.60 0 -15.322 -5.10 IMMUNE E R624600 UL16-binding protein 2 (ULBL2) 6q25 5 0.55 5 -16.33 3 -5.12 2 METABOLISM M AA4273988 Acetylserotonin O-methyltransferase-like protein Xp22 2 0.55 5 4 4 -14.95 5 -5.17 7 AA1716133 Carbonic anhydrase XII (CA12) 15q22 2 0.60 0 4 4 -18.47 7 -5.21 1 R315622 CDP-diacylglycerol synthase (CDS1) 4q21 1 0.49 9 5 5 -23.56 6 -6.63 3 AA4482077 Deoxycytidylate deaminase (DCTD) 4q35 5 0.54 4 4 4 -18.77 7 -8.08 8 T601111 Fatty acid binding protein 5 (FABP5) 8q21 1 0.33 3 5 5 -37.93 3 -6.71 1 H868122 Heparan sulfate 3-0-sulfotransferase 1 (HS3ST1) 4pl6 6 0.41 1 5 5 -15.60 0 -5.09 9 AA0110966 Monoamine oxidase A (MAOA) Xqll l 0.41 1 5 5 -20.99 9 -5.75 5 PROLIFERATION N W427233 GROl oncogene 4q21 1 0.40 0 5 5 -24.72 2 -7.38 8 PROTEINN MODIFICATION W722077 Cystatin A (CSTA) 3q21 1 0.23 3 8 8 -43.98 8 -8.99 9 H229199 Cystatin B (CSTB) 21q22 2 0.41 1 4 4 -33.70 0 -7.28 8 R546644 Leukocyte Elastase Inhibitor (SERPINB1) 6p25 5 0.48 8 5 5 -20.91 1 -5.23 3 SIGNALLING G AA4858988 Apoptosis related protein 2p33 3 0.64 4 4 4 -13.46 6 -5.51 1 W487133 Avian erythroblastic leukemia viral oncogene homolog 7pl2 2 0.61 1 4 4 -20.58 8 -6.56 6 (EGFR) ) AA0267555 BNIP-2 homolog (BNIP-S) lq21 1 0.55 5 4 4 -17.42 2 -5.18 8 T985599 Decorin (DCN) 12ql3 3 0.52 2 4 4 -15.39 9 -5.67 7 N902466 Ephrin receptor (EPHA1) 7q32 2 0.60 0 4 4 -14.74 4 -6.06 6 R704622 Erythroblastic leukemia viral oncogene homolog 2 17q21 1 0.59 9 4 4 -19.54 4 -7.32 2 (C-ERB-B2) ) R249699 GABA A receptor, beta 1 (GABRB1) 4pl2 2 0.42 2 6 6 -26.75 5 -7.06 6 AA4545977 Golgi phosphoprotein 2 (GP73)) 9p24 4 0.67 7 4 4 -18.52 2 -5.75 5 H537033 Growth factor receptor-bound protein 7 (GRP7) 17ql2 2 0.60 0 4 4 -13.15 5 -5.57 7 AA9367577 Heparin-binding growth factor binding protein 4pl6 6 0.59 9 4 4 -33.32 2 -7.12 2 (HBP17) ) AA4537833 Mai, T-cell differentiation protein 2 (MAL2) 8q23 3 0.48 8 5 5 -28.20 0 -5.56 6 AA0991055 NCK-associated protein 1 (NAP1) 2q32 2 0.64 4 4 4 -13.30 0 -5.16 6 AA1213666 P2Y1 3q25 5 0.51 1 4 4 -25.59 9 -6.96 6 AA4278877 Prostaglandin F2 receptor negative regulator lp!3 3 0.59 9 4 4 -17.28 8 -5.69 9 (SMAP-6) ) AA5986766 Reticulocalbin 2, EF-hand calcium binding domain 15q23 3 0.64 4 -16.099 -7.04
224 4 Molecularr Profiling in Head and Neck Cancer
Tablee 2B continued Accession n Genee Name Chromosome Cy5/Cy3 3 Patients s Z-score eBounde d d -score e (RCN2) ) AA598508 8 Retinoicc acid-binding protein 2 (RBP6J lq21 1 0.31 1 5 5 -37.00 0 -6.20 0 W93592 2 Wingless-typee MMTV integration site family, 5A 3p21 1 0.54 4 4 4 -24.28 8 -6.57 7 (WNT5A) ) STRESSS INDUCIBLE PROTEINS AA504943 3 Crystallin,, alpha B (CRYAB) llq22 2 0.60 0 4 4 -32.88 8 -7.39 9 AA478479 9 Heatt shock 70-related protein (APG-1) 4q28 8 0.39 9 5 5 -22.87 7 -7.42 2 STRUCTURAL L W60057 7 Keratinn 13(KRT13) 17q21 1 0.35 5 6 6 -46.66 6 -5.74 4 H44051 1 Keratinn 14 (KRT14) 17ql2 2 0.47 7 5 5 -27.86 6 -5.58 8 AA878048 8 Keratinn 15 (KRT15) 17q21 1 0.53 3 4 4 -29.07 7 -7.05 5 AA150532 2 Keratinn 6 isoform K6a (KRT6A) 12ql2 2 0.47 7 5 5 -31.70 0 -5.45 5 W94063 3 Smalll proline rich protein 4 (SPRR4) lq21 1 0.58 8 4 4 -22.76 6 -6.75 5 TRANSCRIPTION N AA455929 9 KETT gene (TP63) 3q27 7 0.46 6 5 5 -24.44 4 -5.96 6 H45711 1 Zincc finger transcription factor GKLF (KLF4) 9q31 1 0.48 8 4 4 -30.77 7 -6.58 8 TRANSPORT T AA778392 2 BENE E 2ql3 3 0.72 2 4 4 -16.34 4 -5.38 8 AA732931 1 Syntaxinn 4A (STX4A) 16pll l 0.69 9 4 4 -16.45 5 -6.07 7 AA127685 5 Transmembranee 9 superfamily member 1 14qll l 0.54 4 4 4 -16.09 9 -5.99 9 (TM9SF1) )
nodess were consistent with previous findings that uncontrolled proteolysis in the absencee of these inhibitors played a role in tumor metastasis.[76] Also decreased was the decorinn gene, which coded for a small leucine-rich proteoglycan known as a biological ligandd for the EGF receptor.[77] It had been demonstrated that decorin was capable of suppressingg growth of A431 squamous cell carcinoma cells.[78] Furthermore, immuno- histochemicall staining had demonstrated a loss of decorin in tumor stroma at the inva- sivee site of highly invasive oral squamous cell carcinomas. [79] Tumorr Progression in HNSCC Inn addition to the comparison of gene expression between different tissue samples fromm the same patient, we were also interested in identifying genes that showed a con- sistentt pattern of expression across each patient's three tissue samples. For example, a genee which consistently decreased its expression in experiments comparing "normal" adjacentt mucosa to invasive tumor tissue, and decreased expression again when com- paringg invasive tumor to metastatic lymph node tissue, may represent a tumor suppres- sorr gene or a potential target for therapeutic intervention. Our approach was two-fold. Wee initially combined the two datasets for each patient such that for each of the 17,840 cloness on the array, we had two measurements of differential gene expression, one com- paringg invasive tumor to adjacent normal mucosa, and one comparing metastatic lymph nodee to invasive tumor. Second, we partitioned our entire clone set according to their patternn of expression across the 9 patients for which a complete dataset was collected (Figuree 1). Using this algorithm, a clone that increased in expression in tumor tissue rel- ativee to normal tissue, and then increased again in metastatic node tissue relative to the originall invasive tumor was placed in category 1. Conversely, a clone which consecu-
225 5 Chapterr 13 tivelyy decreased in expression was placed in category 9. Clones that were unchanged, or forr which data was discarded due to low signal, were placed in category 5 by default. Initiall partitioning was based on the category in which a given clone most often appeared.. We initially identified 347 clones appearing in category 1 and 303 clones in categoryy 9. When these datasets were refinedd by insisting that a given clone appearr in a single category at least four Metastatic c times,, we identified 144 genes that consis- Normal l Invasive e Lymph h tentlyy increased during tumor progres- Mucosa a HNSCC C Node e 347 7 sion.. Among the category 9 dataset, 95 >* >* (1) ) geness consistently decreased in expres- 2,560 0 1,142 2(2 ) ) sionn (at least 4 of 9 patients). It is impor- 1,071 1(3 ) ) tantt to point out that these are genes / / 1,470 0(4 ) ) whichh may not have appeared in our ini- / / tiall queries due to the application of strin- 17,840 0 12,877 7 10,595 5(5 ) ) gentt criteria for gene selection, but appear w w 812 2(6 ) ) heree due to a consistent trend in expres- \ \ sionn pattern during progression from nor- ^* ^* 881 1(7 ) ) mall tissue to invasive tumor, and ulti- 2,403 3 1,219 9(8 ) ) matelyy to metastatic disease. ^ ^ 303 3(9 ) ) Alll genes included in these two datasetss are listed in Tables 3A and 3B. Amongg the category 1 genes identified Figuree 1. Partitioning of 17840 cDNA human cDNA clones wass moesin, a member of the ERM family intoo nine distinct categories (parentheses) based on expres- off proteins that function as membrane- sionn data for 9 HNSCC patients. The set of 17,840 human cDNAA clones were initially partitioned into 3 groups cytoskeletall linkers and play a role in the dependingg on whether expression of the gene of interest regulationn of cell adhesion and cortical increased,, decreased, or remained relatively unchanged morphogenesis.. [80] Moesin had not been basedd on measurement of the bounded-score when compar- ingg normal adjacent mucosa to invasive HNSCC tumor tis- previouslyy identified in HNSCC. suee from the same patient. Each of these 3 groups was then However,, immunohistochemical analysis subdividedd into 3 additional categories based on the same off skin tumors revealed that the staining criteriaa when comparing the invasive HNSCC tissue to metastaticc lymph node tissue from the same patient. Each patternn for moesin varied among different clonee is assigned to a category to which it was most fre- typess of skin tumors, with invasive squa- quentlyy assigned over the course of the 9 patients examined. Cloness that were unchanged or for which data was discarded mouss cell carcinoma showing an intense duee to low signal were placed in category 5 by default. andd heterologous staining of the cyto- Categoriess 1 and 9 are highlighted by thick boxes. plasmm and cell membrane.[S/] Overexpressionn of another family member, ezrin, had been identified as being associat- edd with metastasis in a murine model of osteosarcoma.[82] High ezrin expression had alsoo been shown to be involved in the process of invasion by endometrial cancer cells.[83]] Stanniocalcin-1 (STC1), a polypeptide hormone believed to be a regulator of minerall homeostasis, also showed a category 1 pattern of expression in 4 of the 9 patients
226 6 Molecularr Profiling in Head and Neck Cancer
Tablee 3A. Partial list of category 1 genes. Included are the accession number for the clone printedd on the array, gene description, chromosomal location, number of patients for which thee gene expression pattern was category 1, and the number of patients for which the gene expressionn pattern was either category 1 or 2.
Accessionn Gene Name C omosomee Categor yy 1 Category y 11 or 2 ADHESION N W214822 Basement membrane-induced gene (ICB-1) lp33 3 5 5 6 6 H459766 Catenin, alpha 2 (CTNNA2) 2pl2 2 4 4 4 4 AA1320900 CD53 glycoprotein lpl3 3 5 5 5 5 R626122 Fibronectin 1 C-terminat domains (FN1) 2q34 4 5 5 5 5 T775955 Hexabrachion (HXB) 9q33 3 4 4 4 4 N643844 Integrin, alpha X (ITGAX) 16pll l 5 5 5 5 AA7756166 Osteopontin (SPP1) 4q21 1 4 4 4 4 H382400 Thrombospondin 2 (THBS2) 6q27 7 5 5 6 6 CELLL CYCLE AA4870311 Solid tumor-associated protein 1 (STAG1) 20ql3 3 IMMUNE E H292955 Butyrophilin (BTF2) 6p22 2 4 4 5 5 AA4785855 Butyrophilin (BTF3) 6p22 2 5 5 6 6 AA7016522 Granulysin (GNLY) 2pl2 2 4 4 7 7 T693044 TAP binding protein (TAPBP) 6p21 1 4 4 6 6 METABOLISM M AA5644744 Apolipoprotein C II (APOC2) 19ql3 3 4 4 5 5 H515744 Arachidonate 5-lipoxygenase (ALOX5) lOqll l 6 6 6 6 AA8572I22 Galactose-1-phosphate uridyl transferase (GALT) 9pl3 3 4 4 5 5 AA2911633 Glutaredoxin 5ql4 4 5 5 7 7 AA593800 Isocitrate dehydrogenase 3 (NAD + ) gamma (IDH3G) Xq28 8 4 4 5 5 R693077 Leucine aminopeptidase (LOC51056) 4pl5 5 4 4 5 5 AA252244 Methionine adenosyltransferase II, beta (MAT2B) 5q34 4 4 4 4 4 AA2787599 Proteoglycan 1 (PRG1) 10q22 2 4 4 6 6 AA0112155 Spermidine/spermine Nl-acetyltransferase (SAT) Xp22 2 4 4 6 6 PROLIFERATION N AA1577877 Kinetochore associated 1 (KNTC1) 12q24 4 5 5 8 8 H061133 Ribosomal protein L3 (MRPL3) 5 5 8 8 PROTEINN MODIFICATION T950522 Caspase 1 (CASP1) llq23 3 4 4 6 6 AA6440888 Cathepsin C (CTSC) llql4 4 5 5 5 5 AA6640044 Lysosomal pepstatin insensitive protease (CLN2) llplS S 4 4 7 7 H703733 Nedd4 binding protein 2 (N4BP2) 4pl4 4 4 4 5 5 AA2920744 Ubiquitin-conjugating enzyme E2L 6 (UBE2L6) llql2 2 4 4 7 7 SIGNALLING G R330311 Adaptor-related protein complex 3, sigma 2 (AP3S2) 15q25 5 4 4 4 4 AA4853711 Bone marrow stromal cell antigen 2 (BST2) 19pl3 3 4 4 7 7 W811966 CDC42 effector protein 2 (CEP2) llql3 3 4 4 6 6 AA2849544 Colony stimulating factor 1 receptor (CSF1R) 5q33 3 5 5 6 6 AA0864755 Cullin 5 (VACM-1) llq22 2 4 4 5 5 AA6303744 Dual specificity phosphatase 6 (DUSP6) 12q22 2 4 4 7 7 AA4469944 Fibroblast growth factor receptor 4 variant 1 (FGFR4) 5q35 5 5 5 5 5 H619355 Nuclear orphan receptor LXR-alpha (NR1H3) llqll l 5 5 6 6 H918266 Oncogene related to SRC, FGR, YES (FYN) 6q21 1 4 4 4 4 R439566 Pleckstrin homology, Sec7 and coiled/coil domains 22ql2 2 4 4 4 4 (PSCI) ) AA0175444 Regulator of G-protein signalling 1 (RGS1) lq31 1 5 5 5 5 H485011 SCN circadian oscillatory protein (SCOP) 18q21 1 5 5 6 6
227 7 Chapterr 13
Tablee 3A continued Accessionn Gene Name Chromosomee Category 1 Category y 11 or 2
AA0853188 Stanniocalcin 1 (STC1) 8p21 AA4957244 Tumor endothelial marker 6 (TEM6) 7pl2 R838366 Yamaguchi sarcoma viral related oncogene homolog 8ql3 (LYN) ) STRESSS INDUCIBLE PROTEINS R714400 Serine (or cysteine) proteinase inhibitor (SERPINH2) llql3 STRUCTURAL L R785300 Coactosin-like protein (CLP) 17pll l 5 5 6 6 AA6775344 Laminin, gamma 2 (LAMC2) lq25 5 4 4 6 6 AA47624ÜÜ Lysyl hydroxylase (PLOD) lp36 6 4 4 7 7 R229777 Moesin (MSN) Xqll l 5 5 9 9 AA4645788 Rho/Rac guanine nucleotide exchange factor Iq21 1 5 5 5 5 (ARHGEF2) ) TRANSCRIPTION N N321466 Avian reticuloendotheliosis viral oncogene homolog (REL)2pl 3 3 5 5 5 5 N704633 B-cell translocation gene 1 (BTG1) 12q22 2 4 4 6 6 AA9539755 I kappa B epsilon (NFKBIE) 6p24 4 5 5 6 6 AA0112322 Methyl-CpG binding domain protein 4 (MBD4) 3q21 1 4 4 5 5 AA4474811 Sp-100 nuclear autoantigen 2q36 6 5 5 6 6 AA7019144 Sterol regulatory element binding protein-2 (SREBF2) 22ql3 3 4 4 5 5 N924433 Upstream binding transcription factor (UBTF) 17q21 1 4 4 4 4 W901288 X-box binding protein 1 (XBP1) 22ql2 2 4 4 5 5 TRANSPORT T AA6782800 Adaptor-related protein complex 3, beta 1 (AP3B1) 5pl4 4 4 4 5 5 AA6307766 Adaptor-related protein complex 3, delta 1 (AP3D1) 19pl3 3 5 5 8 8 H386500 Glucose transport-like 5 (GLUT5) lp36 6 6 6 7 7 R639188 Neuronatin alpha (NNAT) 20qll l 4 4 8 8 N678166 Nucleoporin (NUP43) 6q24 4 4 4 5 5
Tablee 3B. Partial list of category 9 genes. Included are the accession number for the clone printedd on the array, gene description, chromosomal location, number of patients for which thee gene expression pattern was category 9, and the number of patients for which the gene expressionn pattern was either category 8 or 9.
Accessionn Gene Name Chromosomee Category 9 Category 88 or 9 ADHESION N H630777 Annexin Al (ANXA1) 9ql2 AA0540733 Carcinoembryonic antigen-related cell adhesion molecule 6 19ql3 (CEACAM6) ) H947455 Contactin associated protein-like 2 (CASPR2) 7q35 H944711 Occludin (OCLN) 5ql3 CELLL CYCLE AA0465233 Centrin (CETN3) 5ql4 IMMUNE E AA2275944 Mai, T-cell differentiation protein (MAL) 2ql3 T824144 RAB2, member RAS oncogene family 8qll METABOLISM M AA4538599 Alcohol dehydrogenase 5 chi subunit (ADH5) 4q21 H034366 Beta-1, 3-N-acetylglucosaminyltransferase 3 19pl3 (B3GNT3) ) AA8570355 Biliverdin reductase B (BLVRB) 19ql3
228 8 Molecularr Profiling in Head and Neck Cancer
Tablee 3B continued Accessionn Gene Name CIromosom ee 1Categor yy 9 Category 88 or 9 AA6780655 Biphosphoglycerate (2,3-) mutase (BPGM) 7q31 1 4 4 4 4 T493555 Choline phosphotransferase 1 (CPT1) 12q q 4 4 6 6 AA6641800 Glutathione peroxidase 3 (GPX3) 5q23 3 4 4 7 7 H868122 Heparan sulfate 3-O-sulfotransferase 1 (HS3ST1) 4pl6 6 4 4 4 4 AA0194822 Mitochondrial creatine kinase (CKMT1) 15ql5 5 4 4 5 5 PROTEINN MODIFICATION W722077 Cystatin A (CSTA) 3q21 1 5 5 5 5 H229199 Cystatin B (CSTB) 21q22 2 4 4 4 4 R546644 Leukocyte elastase inhibitor (SERPINB1) 6p25 5 4 4 4 4 AA4904977 Ubiquitin-like 3 (UBL3) 13ql2 2 4 4 5 5 SIGNALLING G R937822 Bax-interacting factor 1 (BIF-1) lp22 2 4 4 5 5 AA0267555 BNIP-2 similar (BNIP-S) lq21 1 4 4 5 5 W654611 Dual specificity phosphatase 5 (DUSP5) 10q25 5 4 4 6 6 H844811 Ephrin receptor (EPHA2) lp36 6 4 4 4 4 R800411 Formyl peptide receptor-like 1 (FPRL1) 19ql3 3 4 4 7 7 T993033 Guanine nucleotide binding protein (GNA15) 19pl3 3 4 4 4 4 AA4873700 Myosin regulatory light chain (MLC-B) 18pll l 4 4 4 4 R739099 Pregnancy specific beta-1-glycoprotein 11 (PSG11) 19ql3 3 4 4 6 6 W844455 Ras-related associated with diabetes (RRAD) 16q22 2 5 5 6 6 AA1876811 Stromal membrane-associate protein (SMAP1) 6ql2 2 4 4 4 4 AA4166855 UNC13 9pl2 2 6 6 7 7 STRESSS INDUCIBLE PROTEINS AA1868044 EROl-like protein (EROIL) 14q22 2 4 4 4 4 R951288 Oxidative-stress responsive protein 1 (OSR1) 3p22 2 4 4 5 5 H574944 Protein kinase Hll 12q24 4 5 5 6 6 STRUCTURAL L AA4614733 Actin-binding Z-disk protein Nebulette (NEBL) 10pl2 2 4 4 4 4 AA6295422 Brush-i (WASF3) 13ql2 2 5 5 7 7 AA4475999 Cytoplasmic linker associated protein 2 (CLASP2) 3p22 2 4 4 5 5 W600577 Keratin 13 (KRT13) 17q21 1 6 6 7 7 AA8780488 Keratin 15 (KRT15) 17q21 1 4 4 4 4 AA6291899 Keratin 4 (KRT4) 12ql3 3 5 5 6 6 TRANSCRIPTION N AA489555 DEAD/H box polypeptide 17 (DDX17) 22ql3 3 5 5 6 6 H228266 LIM domain only 7 (LM07) 13q21 1 4 4 5 5 H865588 MAX dimerization protein (MAD) 2pl3 3 4 4 4 4 AA7058866 MAX-interacting protein 1 (MXI1) 10q24 4 4 4 5 5 AA2842322 RNA polymerase II elongation factor (ELL2) 5q21 1 4 4 4 4 R541933 Trichorrhinoplhalangeal syndrome 1 (TRPS1) 8q24 4 4 4 4 4 AA0018355 Zinc finger protein 262 (ZNF262) lp32 2 5 5 7 7 N307044 Zinc finger protein ZFP26 llplS S 4 4 4 4 AA4796933 Zinc finger protein 212 (ZNF212) 7q36 6 5 5 6 6 TRANSPORT T AA0456411 ATP-binding cassette, sub-family B, member 10 (ABCB10CB10)) ) lq42 2 4 4 AA778392 2 BENEE protein 2ql3 3 6 6 6 6 AA165679 9 Kinesinn family member IC (KIF1C) 17pl3 3 4 4 4 4 T50082 2 Mitochondriall solute carrier (HT-015) 8pl2 2 4 4 5 5 N73680 0 Solutee carrier family 11, member 2 (SLC11A2) 12ql3 3 4 4 5 5 W90588 8 Synaptogyrinn 1 (SYNGR1) 22ql3 3 4 4 7 7 AA732931 1 Syntaxinn 4A (STX4A) 16pll l 4 4 4 4 AA757170 0 Taxx interaction protein 33 (VELI1) 12q21 1 4 4 7 7 UNKNOWN N W204388 AuxotrophiAuxotrophinnn (AXOT(AXOT)) 2q24
229 9 Chapterr 13
Tablee 3B continued Accessionn Gene Name Chromosome Category 9 Category 88 or 9 AA0046388 H19, imprinted maternally expressed llplS 4 4 untranslatedd mRNA AA0453088 Insulin induced protein 2 (LOC51141) 2q21 4 7 AA4485699 Sushi-repeat-containing protein (SRPX) Xp21 4 6 studied.. Enhanced expression of STC1 had been observed in tissues of hepatocellular carcinomaa and colorectal cancer compared to corresponding cancer-free tissue, suggest- ingg that it might be a useful molecular marker for detection of tumor cells.[84] Expressionn of STC1 had also been observed in the hypoxic regions of human breast and colonn cancer cells, a result of interest in that it was known that the presence of hypoxic regionss within solid tumors was associated with a more aggressive tumor phenotype and poorerr prognosis.[85] Also of interest in category 1 was the facilitative glucose trans- porterr GLUT5. GLUT5 exhibited a category 1 expression pattern (increasing expres- sion)) in 6 of the 9 patients for which data were collected. A similar increasing pattern of GLUT55 expression in metastasis had been observed in invasive lung cancers and their liverr metastases, and glucose uptake was known to be associated with growth of malig- nantt head and neck tumors.[86] Amongg the genes identified in category 9 (decreasing in expression) were four poten- tiall tumor suppressor genes: Brush-1, Mxil, oxidative stress-response (OSR1) gene and occludinn (see earlier discussion). Brush-1 was part of a multi-protein complex linking signall transduction to actin, thereby regulating cellular shape and motility. Furthermore,, loss of Brush-1 expression in both breast cancer cell lines and invasive tumorss identified it as a potential tumor suppressor.[87] Mxil belongs to the Mad fam- ilyy of transcription factor proteins, which function as potent antagonists of Myc onco- proteins.^]] It had been shown that Mxil can suppress prostate tumor cell proliferation inin vitro.[89] Immunohistochemicall Analysis of Moesin Expression in HNSCC Tissue Wee examined the expression of moesin in tissue arrays containing normal epitheli- um,, as well as corresponding squamous cell carcinoma of the tongue and lymph nodes. Whilee normal squamous epithelium of the tongue showed negative or basal membrane stainingg for moesin, advanced squamous cell carcinoma showed a strong positive stain- ingg in the membrane and cytoplasm of most tumor cells (Figure 2). These results agreed withh the moesin staining patterns observed in invasive squamous cell carcinoma of the skin.[81]] Interestingly, tumor cells invading lymph nodes also displayed a positive moesinn staining. Both membrane and cytoplasmic expression of moesin increased when comparingg normal epithelium (median percentage of positive cells: 10.0%) to dysplastic epitheliumm (median percentage of positive cells: 20.0%) to tumor samples (median per- centagee of positive cells: 81.5%) (Table 4). Each of these comparisons reached a statisti- callyy significant association (Mann-Whitney, /)<0.O0O5). However, moesin expression wass not increased when comparing tumor samples and lymph node metastases (median
230 0 Molecularr Profiling in Head and Neck Cancer
percentagee of positive cells: 90.0%, Mann- A.. Normal Mucosa Whitney,, p=0.817). All lymph nodes invaded byy tumor cells displaying positive moesin stain- ingg showed overexpression of this protein in the HNSCCC tumor. Overall, moesin expression was foundd to be significantly associated with HNSCCC progression (Kruskall-Wallis, p
231 1 Chapterr 13
Tablee 4. Percentage of cells displaying positive staining for moesin in a HNSCC tissue array. Alll specimens (n = 102) used for analysis of the association between moesin expression and pro- gressionn of HNSCC were assessed using the non-parametric Wilcoxon-Mann-Whitney and Kruskall-Walliss tests. Tissue' ' Mediann (%) 25%-75% %Interquartil e e Meann (%) 95%% CI Rang* * rangee (%) off the Mean Normall epithelium 10.0 0 10.0 0 7.2 2 4.3-10.1 1 0-30 0 Dysplasticc epithelium 20.0 0 13.5 5 27.7 7 19.7-35.8 8 0-100 0 Primaryy tumor 81.5 5 28.7 7 70.9 9 60.2-81.5 5 0-100 0 Nodall metastasis 90.0 0 100.0 0 66.4 4 24.2-108.7 7 0-100 0 *AA total of 23 normal tongue epithelium, 30 dysplastic epithelial lesions from HNSCC patients, 42 HNSCC and 7 lymphh node metastases were assessed. The consensus value of 3 representative cores from each tumor sample arrayedd was used for statistical analyses.
References s 1.. Greenlee RT, Hill-Harmon MB, Murray T, Thun M. Cancer statistics, 2001. CA Cancerr J Clin 2001; 51:15-36. 2.. Gollin SM. Chromosomal alterations in squamous cell carcinomas of the head and neck.. Head Neck 2001; 23:238-253. 3.. Villaret DB, Wang T, Dillon D, Xu J, Sivam D, Cheever MA, Reed SG. Identificationn of genes overexpressed in head and neck squamous cell carcinoma usingg a combination of complementary DNA subtraction and microarray analysis. Laryngoscopee 2000; 110:374-381. 4.. Leethanakul C, Patel V, Gillespie J, Pallente M, Ensley JF, Kootongkaew S, Liotta LA,, Emmert-Buck M, Gutkind JS. Distinct pattern of expression of differentiation andd growthrelated genes in squamous cell carcinoma of the head and neck revealed byy the use of laser capture microdissection and cDNA arrays. Oncogene 2000; 19:3220-3224. . 5.. Alevizos I, Mahadevappa M, Zhang X, Ohyama H, Kohno Y, Posner M, Gallagher GT,, Varvares M, Cohen D, Kim D, Kent R, Donoff RB, Todd R, Yung CM, Warringtonn JA Wong DT. Oral cancer in vivo gene expression profiling assisted byy laser capture microdissection and microarray analysis. Oncogene 2001; 20:6196- 6204. . 6.. Hanna E, Shrieve DC, Ratanatharathorn V, Xia X, Breau R, Suen J, Li S. A novel alternativee approach for prediction of radiation response of squamous cell carcino- maa of head and neck. Cancer Res 2001; 61:2376-2380. 7.. Belbin TJ, Singh B, Barber I, Socci ND, Wenig B, Smith R, Prystowsky MB, Childs,, G. Molecular classification of head and neck squamous cell carcinoma usingg cDNA microarrays. Cancer Res 2002; 62:1184-1190. 8.. Cheung VG, Morley M, Aguilar F, Massimi A, Kucherlapati R, Childs G. Making andd reading microarrays. Nature Genet (Supplement) 1999; 21:15-19. 9.. Wang E, Miller LD, Ohnmacht GA, Liu ET, Marincola FM. High-fidelity mRNA amplificationn for gene profiling. Nature Biotech 2000; 18:457-459.
232 2 Molecularr Profiling in Head and Neck Cancer
10.. Tseng GC, Oh MR, Rohlin L, Liao JC, Wong WH. Issues in cDNA microarray analysis:: quality filtering, channel normalization, models of variations and assess- mentt of gene effects. Nucleic Acids Res 2001; 29:2549-2557. 11.. Dudoit S. Statistical methods for identifying differentially expressed genes in repli- catedd cDNA microarray experiments. U.C. Berkeley Department of Statistics Technicall Report #578. http://www.stat.berkeley.edU/tech-reports/578.ps.Z, 2000. 12.. Hughes TR, Marton MJ, Jones AR, Roberts CJ, Stoughton R, Armour CD, Bennett HA,, Coffey E, Dai H, He YD, Kidd MJ, King AM, Meyer MR, Slade D, Lum PY, Stepaniantss SB, Shoemaker DD, Gachotte D, Chakraburtty K, Simon J, Bard M, Friend,, SH. Functional discovery via a compendium of expression profiles. Cell 2000;; 102:109-129 (plus supplemental material). 13.. Hoos A, Urist MJ, Stojadinovic, A, Mastorides S, Dudas ME, Leung DHY, Kuo D, Brennann MF, Lewis JJ, Cordon-Cardo C. Validation of tissue microarrays for immunohistochemicall profiling of cancer specimens using the example of human fibroblasticc tumors. Am J Pathol 2001; 158:1245-1251. 14.. Tudor G, Koch GG. Review of nonparametric methods for the analysis of crossover studies.. Stat Methods Med Res 1994; 3:345-381. 15.. Oliveira DT, Sherriff M, Odell EW. Expression of CD44 variant exons by primary andd metastatic oral squamous carcinomas. J Oral Pathol Med 1998; 27:303-307. 16.. Rudzki Z, Jothy S. CD44 and the adhesion of neoplastic cells. Mol Pathol 1997; 50:57-71. . 17.. Huang Y, Prasad M, Lemon WJ, Hampel H, Wright FA, Kornacker K, LiVolsi V, Frankell W, Kloos RT, Eng C, Pellegata NS, de la Chapelle A. Gene expression in papillaryy thyroid carcinoma reveals highly consistent profiles. Proc Natl Acad Sci 2001;98:15044-15049. . 18.. Herold-Mende C, Andl T, Laemmler F, Reisser C, Eichhorn S. Expression and localizationn profile of tenascin in squamous cell carcinomas of the head and neck. HNOO 1999; 47:723-729. 19.. Goussia AC, Ioachim EE, Peschos D, Assimakopoulos DA, Skevas A, Agnantis NJ. Expressionn of the extracellular matrix protein tenascin in laryngeal epithelial lesions:: correlation with fibronectin CD44 cathepsin D and proliferation indices. Virchowss Arch 2000; 436:579-584. 20.. Ebihara T, Hashimoto T, Kudoh J, Gamou S, Shimizu N, Nishikawa T. Detection off the 170-kDa bullous pemphigoid antigen by immunoprecipitation. J Invest Dermatoll 1993, 100:176-179. 21.. Yamada T, Endo R, Tsukagoshi K, Fujita S, Honda K, Kinoshita M, Hasebe T, Hirohashii S. Aberrant expression of a hemidesmosomal protein bullous pem- phigoidd antigen 2 in human squamous cell carcinoma. Lab Invest 1996; 75:589- 600. . 22.. Thomas GJ, Jones J, Speight PM. Integrins and oral cancer. Oral Oncol 1997; 33:381-388. . 23.. Okazaki K, Nakayama Y, Shibao K, Hirata K, Sako T, Nagata N, Kuroda Y, Itoh
233 3 Chapterr 13
H.. Establishment of a human colon cancer cell line (PMF-kol4) displaying highly metastaticc activity. Int J Oncol 2000; 17:39-45. 24.. Devoll RE, Li W, Woods KV, Pinero GJ, Butler WT, Farach-Carson MC, Happonenn RP. Osteopontin (OPN) distribution in premalignant and malignant lesionss of oral epithelium and expression in cell lines derived from squamous cell carcinomaa of the oral cavity. J Oral Pathol Med 1999; 28:97-101. 25.. Crawford HC, Matrisian LM, Liaw L. Distinct roles of osteopontin in host defense activityy and tumor survival during squamous cell carcinoma progression in vivo. Cancerr Res 1998; 58:5206-5215. 26.. Al Moustafa AE, Alaoui-Jamali MA, Batist G, Hernandez-Perez M, Serruya C, Alpertt L, Black MJ, Sladek R, Foulkes WD. Identification of genes associated with headd and neck carcinogenesis by cDNA microarray comparison between matched primaryy normal epithelial and squamous carcinoma cells. Oncogene 2002; 21:2634- 2340. . 27.. Shibata T, Nagayasu H, Hamada J, Konaka S, Hosokawa M, Kawano T, Kitajo H, Arisuee M. Inhibitory effects of malotilate on in vitro invasion of lung endothelial celll monolayer by human oral squamous cell carcinoma cells. Tumour Biol 2000; 21:299-308. . 28.. Kimura Y, Shiozaki H, Hirao M, Maeno Y, Doki Y, Inoue M, Monden T, Ando- Akatsukaa Y, Furuse M, Tsukita S, Monden M. Expression of occludin tight-junc- tion-associatedd protein in human digestive tract. Am J Pathol 1997; 151:45-54. 29.. Storkel S, Reichert T, Reiffen KA, Wagner W. EGFR and PCNA expression in oral squamouss cell carcinomas-a valuable tool in estimating the patient's prognosis. Eurr J Cancer B Oral Oncol 1993; 29B:273-277. 30.. Loukinova E, Chen Z, Van Waes C, Dong G. Expression of proangiogenic chemokinee Gro 1 in low and high metastatic variants of Pam murine squamous cell carcinomaa is differentially regulated by IL-lalpha EGF and TGF-betal through NF-kappaBB dependent and independent mechanisms. Int J Cancer 2001; 94:637- 644. . 31.. Leung TW, Lin SS, Tsang AC, Tong CS, Ching JC, Leung WY, Gimlich R, Wong GG,, Yao KM. Over-expression of FoxMl stimulates cyclin Bl expression. FEBS Lettt 2001; 507:59-66. 32.. Hassan KA, El-Naggar AK, Soda JC, Liu D, Hong WK, Mao L. Clinical signifi- cancee of cyclin Bl protein expression in squamous cell carcinoma of the tongue. Clinn Cancer Res 2001; 7:2458-2462. 33.. Nilsson I, Hoffmann I. Cell cycle regulation by the Cdc25 phosphatase family. Prog Celll Cycle Res 2000; 4:107-114. 34.. Kudo Y, Yasui W, Ue T, Yamamoto S, Yokozaki H, Nikai H, Tahara E. Overexpressionn of cyclin-dependent kinase-activating CDC25B phosphatase in humann gastric carcinomas. Jpn J Cancer Res 1997; 88:947-952. 35.. Nishioka K, Doki Y, Shiozaki H, Yamamoto H, Tamura S, Yasuda T, Fujiwara Y, Yanoo M, Miyata H, Kishi K, Nakagawa H, Shamma A, Monden M. Clinical signif-
234 4 Molecularr Profiling in Head and Neck Cancer
icancee of CDC25A and CDC25B expression in squamous cell carcinomas of the oesophagus.. Br J Cancer 2001; 85:412-421. 36.. Takemasa I, Yamamoto H, Sekimoto M, Ohue M, Noura S, Miyake Y, Matsumoto T,, Aihara T, Tomita N, Tamaki Y, Sakita I, Kikkawa N, Matsuura N, Shiozaki H, Mondenn M. Overexpression of CDC25B phosphatase as a novel marker of poor prognosiss of human colorectal carcinoma. Cancer Res 2000; 60:3043-3050. 37.. Gasparotto D, Maestro R, Piccinin S, Vukosavljevic T, Barzan L, Sulfaro S, Boiocchii M. Overexpression of CDC25A and CDC25B in head and neck cancers. Cancerr Res 1997; 57:2366-2368. 38.. Yuan J, Horlin A, Hock B, Stutte HJ, Rubsamen-Waigmann H, Strebhardt K. Polo-likee kinase a novel marker for cellular proliferation. Am J Pathol 1997; 150:1165-1172. . 39.. Smith MR, Wilson ML, Hamanaka R, Chase D, Kung H, Longo DL, Ferris DK. Malignantt transformation of mammalian cells initiated by constitutive expression off the polo-like kinase. Biochem Biophys Res Commun 1997; 234:397-405. 40.. Yuan J, Eckerdt F, Bereiter-Hahn J, Kurunci-Csacsko E, Kaufmann M, Strebhardt K:: Cooperative phosphorylation including the activity of polo-like kinase 1 regu- latess the subcellular localization of cyclin Bl. Oncogene 2002; 21:8282-8292. 41.. Knecht R, Elez R, Oechler M, Solbach C, von Ilberg C, Strebhardt K. Prognostic significancee of polo-like kinase (PLK) expression in squamous cell carcinomas of thee head and neck. Cancer Res 1999; 59:2794-2797. 42.. Garg LN, Yadav SP, Lal H. Serum leucine aminopeptidase in head and neck can- cer.. J Laryngol Otol 1994; 108:660-662. 43.. Hatzakis KD, Froudarakis ME, Bouros D, Tzanakis N, Karkavitsas N, Siafakas NM.. Prognostic value of serum tumor markers in patients with lung cancer. Respirationn 2002; 69:25-29. 44.. Bergler W, Riedel F, Schwartz-Albiez R, Gross HJ, Hormann K. A new histobio- chemicall method to analyze sialylation on cell-surface glycoproteins of head and neckk squamous-cell carcinomas. Eur Arch Otorhinolaryngol 1997; 254:437-441. 45.. Goss PE, Baker MA, Carver JP, Dennis JW. Inhibitors of carbohydrate processing: aa new class of anticancer agents. Clin Cancer Res 1995; 1:935-944. 46.. Dennis JW, Laferte S. Tumor cell surface carbohydrate and the metastatic pheno- type.. Cancer Metastasis Rev 1987; 5:185-204. 47.. Weiss RL, Calhoun KH, Ahmed AE, Stanley D. Ornithine decarboxylase activity inn tumor and normal tissue of head and neck cancer patients. Laryngoscope 1992; 102:855-857. . 48.. Bock JM, Pickart MA, Pink JJ, Harari PM. Modulation of tumor cell proliferation andd apoptosis by polyamine depletion in cells of head and neck squamous cell car- cinomas.. RadiatRes 1999; 152:604-610. 49.. Wallace HM, Caslake R. Polyamines and colon cancer. Eur J Gastroenterol Hepatol 2001;; 13:1033-1039. 50.. Matsui-Yuasa I, Otani S, Yano Y, Takada N, Shibata MA, Fukushima S. 235 5 Chapterr 13
Spermidine/sperminee Nl-acetyltransferase a new biochemical marker for epithelial proliferationn in rat bladder. Jpn J Cancer Res 1992; 83:1037-1040. 51.. Vigneswaran N, Zhao W, Dassanayake A, Muller S, Miller DM, Zacharias W. Variablee expression of cathepsin B and D correlates with highly invasive and metastaticc phenotype of oral cancer. Hum Pathol 2000; 31:931-937. 52.. Ikebe T, Takeuchi H, Jimi E, Beppu M, Shinohara M, Shirasuna K. Involvement off proteasomes in migration and matrix metalloproteinase-9 production of oral squamouss cell carcinoma. Int J Cancer 1998; 77: 578-585. 53.. Sunwoo JB, Chen Z, Dong G, Yeh N, Crowl Bancroft C, Sausville E, Adams J, Elliottt P, Van Waes C. Novel proteasome inhibitor PS-341 inhibits activation of nuclearr factor-kappa B cell survival tumor growth and angiogenesis in squamous celll carcinoma. Clin Cancer Res 2001; 7:1419-1428. 54.. Kudo Y, Takata T, Yasui W, Ogawa I, Miyauchi M, Takekoshi T, Tahara E, Nikai H.. Reduced expression of cyclin-dependent kinase inhibitor p27Kipl is an indica- torr of malignant behavior in oral squamous cell carcinoma. Cancer 1998; 83:2447- 2455. . 55.. Wojtukiewicz MZ, Zacharski LR, Rucinska M, Zimnoch L, Jaromin J, Rozanska- Kudelskaa M, Kisiel W, Kudryk BJ. Expression of tissue factor and tissue factor pathwayy inhibitor in situ in laryngeal carcinoma. Thromb Haemost 1999; 82:1659- 1662. . 56.. Lakka SS, Konduri SD, Mohanam S, Nicolson GL, Rao JS. In vitro modulation of humann lung cancer cell line invasiveness by antisense cDN A of tissue factor path- wayy inhibitor-2. Clin Exp Metastasis 2000; 18:239-244. 57.. Berrie CP. Phosphoinositide 3-kinase inhibition in cancer treatment. Expert Opin Investigg Drugs 2001; 10:1085-1098. 58.. Gupta AK, McKenna WG, Weber CN, Feldman MD, Goldsmith JD, Mick R, Machtayy M, Rosenthal DI, Bakanauskas VJ, Cerniglia GJ, Bernhard EJ, Weber RS, Muschell RJ. Local recurrence in head and neck cancer: relationship to radiation resistancee and signal transduction. Clin Cancer Res 2002; 8:885-892. 59.. Nozaki S, Endo Y, Kawashiri S, Nakagawa K, Yamamoto E, Yonemura Y, Sasaki T.. Immunohistochemical localization of a urokinase-type plasminogen activator systemm in squamous cell carcinoma of the oral cavity: association with mode of invasionn and lymph node metastasis. Oral Oncol 1998; 34: 58-62. 60.. Jaakkola S, Salmikangas P, Nylund S, Partanen J, Armstrong E, Pyrhonen S, Lehtovirtaa P, Nevanlinna H. Amplification of fgfr4 gene in human breast and gynecologicall cancers. Int J Cancer 1993; 54:378-382. 61.. Bange J, Prechtl D, Cheburkin Y, Specht K, Harbeck N, Schmitt M, Knyazeva T, Mullerr S, Gartner S, Sures I, Wang H, Imyanitov E, Haring HU, Knayzev P, Iacobellii S, Hofler H, Ullrich A. Cancer progression and tumor cell motility are associatedd with the FGFR4 Arg(388) allele. Cancer Res 2002; 62:840-847. 62.. Dellacono FR, Spiro J, Eisma R, Kreutzer D. Expression of basic fibroblast growth
236 6 Molecularr Profiling in Head and Neck Cancer
factorr and its receptors by head and neck squamous carcinoma tumor and vascular endotheliall cells. Am J Surg 1997; 174:540-544. 63.. Lopes MA, Nikitakis NG, Reynolds MA, Ord RA, Sauk J Jr. Biomarkers predictive off lymph node metastases in oral squamous cell carcinoma. J Oral Maxillofac Surg 2002;60:142-147. . 64.. Vaidya MM, Borges AM, Pradhan SA, Bhisey AN. Cytokeratin expression in squa- mouss cell carcinomas of the tongue and alveolar mucosa. Eur J Cancer B Oral Oncoll 1996; 32B:333-336. 65.. Morifuji M, Taniguchi S, Sakai H, Nakabeppu Y, Ohishi M. Differential expres- sionn of cytokeratin after orthotopic implantation of newly established human tonguee cancer cell lines of defined metastatic ability. Am J Pathol 2000; 156:1317- 1326. . 66.. Kosmehl H, Berndt A, Strassburger S, Borsi L, Rousselle P, Mandel U, Hyckel P, Zardii L, Katenkamp D. Distribution of laminin and fibronectin isoforms in oral mucosaa and oral squamous cell carcinoma. Br J Cancer 1999; 81:1071-1079. 67.. Gabbiani G, Csank-Brassert J, Schneeberger JC, Kapanci Y, Trenchev P, Holborow EJ.. Contractile proteins in human cancer cells. Immunofluorescent and electron microscopicc study. Am J Pathol 1976; 83:457-474. 68.. Kuffer R, Fiore-Donno G, Lopez-Pardinas M, Gabbiani G. The contractile fila- mentt system of carcinoma cells: immunofluorescent and electron microscopic study.. Rev Stomatol Chir Maxillofac 1977; 78:371-383. 69.. Savagner P, Yamada KM, Thiery JP. The zinc-finger protein slug causes desmo- somee dissociation an initial and necessary step for growth factor-induced epithelial- mesenchymall transition. J Cell Biol 1997; 137:1403-1419. 70.. Hajra KM, Chen DY, Fearon ER. The SLUG zinc-finger protein represses E-cad- herinn in breast cancer. Cancer Res 2002; 62:1613-1618. 71.. Lu PJ, Sundquist K, Baeckstrom D, Poulsom R, Hanby A, Meier-Ewert S, Jones T, Mitchelll M, Pitha-Rowe P, Freemont P, Taylor-Papadimitriou J. A novel gene (PLU-1)) containing highly conserved putative DNA/chromatin binding motifs is specificallyy up-regulated in breast cancer. J Biol Chem 1999; 274:15633-15645. 72.. Otsuka M, Kato M, Yoshikawa T, Chen H, Brown EJ, Masuho Y, Omata M, Seki N.. Differential expression of the L-plastin gene in human colorectal cancer pro- gressionn and metastasis. Biochem Biophys Res Commun 2001; 289:876-881. 73.. Chow V, Yuen AP, Lam KY, Tsao GS, Ho WK, Wei WI. A comparative study of thee clinicopathological significance of E-cadherin and catenins (alpha beta gamma) expressionn in the surgical management of oral tongue carcinoma. J Cancer Res Clin Oncoll 2001; 127:59-63. 74.. Vesce F, Biondi C. Alpha-L-fucosidase activity in endometrial cervical and ovarian cancer.. Eur J Gynaecol Oncol 1983; 4:135-138. 75.. Shintani S, Mihara M, Terakado N, Nakahara Y, Matsumura T, Kohno Y, Ohyama H,, McBride J, Kent R, Todd R, Tsuji T, Wong DT. Reduction of pl2DOC-l expressionn is a negative prognostic indicator in patients with surgically resected
237 7 Chapterr 13
orall squamous cell carcinoma. Clin Cancer Res 2001; 7:2776-2782. 76.. Henskens YM, Veerman EC, Nieuw Amerongen AV. Cystatins in health and dis- ease.. Biol Chem Hoppe Seyler 1996; 377:71-86. 77.. Iozzo RV, Moscatello DK, McQuillan DJ, Eichstetter, I. Decorin is a biological lig- andd for the epidermal growth factor receptor. J Biol Chem 1999; 274:4489-4492. 78.. Moscatello DK, Santra M, Mann DM, McQuillan DJ, Wong AJ, Iozzo RV. Decorinn suppresses tumor cell growth by activating the epidermal growth factor receptor.. J Clin Invest 1998; 101:406-412. 79.. Harada T, Shinohara M, Nakamura S, Oka M. An immunohistochemical study of thee extracellular matrix in oral squamous cell carcinoma and its association with invasivee and metastatic potential. Virchows Arch 1994; 424:257-266. 80.. Mangeat P, Roy C, Martin M. ERM proteins in cell adhesion and membrane dynamics.. Trends Cell Biol 1999; 9:187-192. 81.. Ichikawa T, Masumoto J, Kaneko M, Saida T, Sagara J, Taniguchi S. Expression of moesinn and its associated molecule CD44 in epithelial skin tumors. J Cutan Pathol 1998;; 25:237-243. 82.. Khanna C, Khan J, Nguyen P, Prehn J, Caylor J, Yeung C, Trepel J, Meltzer P, Helmann L. Metastasis-associated differences in gene expression in a murine model off osteosarcoma. Cancer Res 2001; 61:3750-3759. 83.. Ohtani K, Sakamoto H, Rutherford T, Chen Z, Satoh K, Naftolin F. Ezrin, a mem- branecytoskeletall linking protein is involved in the process of invasion of endome- triall cancer cells. Cancer Lett 1999; 147:31-38. 84.. Fujiwara Y, Sugita Y, Nakamori S, Miyamoto A, Shiozaki K, Nagano H, Sakon M, Mondenn M. Assessment of Stanniocalcin-1 mRNA as a molecular marker for micrometastasess of various human cancers. Int J Oncol 2000; 16:799-804. 85.. Lai A, Peters H, St Croix B, Haroon ZA, Dewhirst MW, Strausberg RL, Kaanders JH,, van der Kogel AJ, Riggins GJ. Transcriptional response to hypoxia in human tumors.. J Natl Cancer Inst 2001; 93:1337-1343. 86.. Kurata T, Oguri T, Isobe T, Ishioka S, Yamakido M. Differential expression of facilitativee glucose transporter (GLUT) genes in primary lung cancers and their liverr metastases. Jpn J Cancer Res 1999; 90:1238-1243. 87.. Schott DR, Chang JN, Deng G, Kurisu W, Kuo WL, Gray J, Smith HS. A candi- datee tumor suppressor gene in human breast cancers. Cancer Res 1994; 54:1393- 1396. . 88.. Schreiber-Agus N, Meng Y, Hoang T, Hou H Jr, Chen K, Greenberg R, Cordon- Cardoo C, Lee HW, DePinho RA. Role of Mxil in ageing organ systems and the regulationn of normal and neoplastic growth. Nature 1998; 393:483-487. 89.. Taj MM, Tawil RJ, Engstrom LD, Zeng Z, Hwang C, Sanda MG, Wechsler DS. Mxill a Myc antagonist suppresses proliferation of DU145 human prostate cells. Prostatee 2001; 47:194-204. 90.. Sanchez-Carbayo M, Socci ND, Charytonowicz E, Lu M, Prystowsky MB, Childs G,, Cordon-Cardo C: Molecular profiling of bladder cancer using cDNA microar-
238 8 Molecularr Profiling in Head and Neck Cancer
rays:: defining histogenesis and biological phenotypes. Cancer Res 2002; 62:6973- 6980. .
Acknowledgments s Wee thank Dr. Thomas Harris for critical review and helpful discussions during the preparationn of this manuscript. We also thank Aldo Massimi and the Albert Einstein Collegee of Medicine Microarray Facility for their assistance.
239 9 240 0