A re®ned localization of two deleted regions in chromosome 6q associatedwith salivary gland carcinomas

Lurdes Queimado1,2,4, Anto nio Reis1, Isabel Fonseca1, Carmo Martins1, Michael Lovett2,Jorge Soares1 and Leonor Parreira3

1Departmento de Patologia MorfoloÂgica & Centro de InvestigacË aÄo de Patobiologia Molecular (CIPM), Instituto PortugueÃs deOncologia de Francisco Gentil (IPOFG), Lisboa, Portugal; 2Departments of Otorhinolaryngology, Molecular Biology andOncology, and The McDermott Center, University of Texas Southwestern Medical Center, Dallas, Texas 75235, USA; 3Institutode Histologia e Embriologia, Faculdade de Medicina de Lisboa, Lisboa, Portugal

Deletions within chromosome 6 (6q25 to 6qter) are themost consistent structural change observed in salivarygland carcinomas. To better de®ne the location of thesedeletions we investigated loss of heterozygosity (LOH)for 23 polymorphic markers within 19 salivary glandcarcinomas covering several histological subtypes. LOHwas observed in 47% of tumors, con®rming previousreports that such losses are frequent and occur in almostall histological subtypes of tumors. The highestfrequency of LOH was found at, or distal to, D6S437.Seven tumors had allelic losses for D6S297 and/orD6S37. A second peak of loss was also observed atD6S262 and D6S32. In some tumors we observed LOHin one or the other of these two regions. In other tumorswe observed loss of both regions with retention ofintervening loci. These data suggest that two smalldeletions commonly occur, one between D6S262 andD6S32 (estimated to cover less than 1.5 Mb) andanother between D6S297 and D6S446 (estimated tocover *2 Mb). These results extend previous studies bysublocalizing the regions of LOH and suggest thatinactivation of one or more tumor suppressor geneslocated in these regions may be an important step insalivary gland carcinogenesis.

Keywords: salivary gland carcinomas; chromosome 6q;loss of heterozygosity (LOH); deletions; histologicalsubtypes

Introduction

Cytogenetic and molecular studies have shown thatnon-random deletions and loss of heterozygosity(LOH) within chromosome 6q are frequent events inmany human neoplasms including breast carcinomas(Devilee et al., 1991a), melanomas (Taguchi et al.,1993), head and neck tumors (BockmuÈ hl et al., 1996)and renal carcinomas (Morita et al., 1991), amongothers (Gaidano et al., 1992; Saito et al., 1992;Menasce et al., 1994; Queimado et al., 1995; Cooneyet al., 1996; Tahara et al., 1996). Chromosome transferexperiments indicate that chromosome 6 can suppressthe tumorigenic (Trent et al., 1990) or metastatic

activity (Welch et al., 1994) of melanoma cell lines andcan induce senescence when introduced into immorta-lized cell lines (Hubbard-Smith et al., 1992). Takentogether, these results support the hypothesis that oneor more generally important tumor suppressor genesreside within chromosome 6q.Salivary gland carcinomas are a relatively rare

tumor type. However, deletions of chromosome 6qare frequently observed in these tumors and are oftenthe only cytogenetically visible alteration (Sandros etal., 1990; Jin et al., 1994). Furthermore, these deletionsappear to be the only structural abnormality shared incommon between all of the histological subtypes ofsalivary gland carcinoma, with the exception ofcarcinoma ex-pleomorphic adenoma (Jin et al., 1994),again suggesting that genes within this region play acritical part in the development of these malignanttumors. Previous cytogenetic studies have broadlymapped the 6q deletions into the 6q16-6qter regionwith an apparent clustering in the 6q25-6qter interval(Sandros et al., 1990; Jin et al., 1994). In this report wehave more precisely re®ned the location of thesedeletions by investigating loss of heterozygosity withinchromosome 6q for 21 polymorphic DNA loci in aseries of 19 salivary gland carcinomas. Our resultsindicate that two separate deleted regions occur. Theminimal regions of overlap for each of these deletionssuggest that they span relatively small genomicintervals that may now be tractable by physicalmapping and positional cloning methods.

Results

In this study we investigated the presence of allelicimbalance within chromosome 6q in paired constitu-tional and tumor DNA samples from 19 patients withsalivary gland carcinomas. Twenty-three polymorphicmarkers were used, 21 of which map within the longarm of chromosome 6, and these are listed in Table 1.All samples were heterozygous for at least one locuswithin 6q, although the incidence of heterozygosityvaried among di�erent loci (ranging from 36% in thecase of MYB to 100% for D6S446). The majority ofthe polymorphic markers used in this study weremicrosatellites, therefore evidence for allelic imbalancewas in most cases derived from PCR analysis of tumorand constitutional DNA. Examples of allelic balanceand imbalance for four representative cases are shownin Figure 1. Thus, by PCR analysis, locus D6S262 is inallelic balance in case 11, but is in allelic imbalance in

Correspondence: L Queimado4 Present address: McDermott Center, NB10.106, University of TexasSouthwestern Medical Center, Dallas, Texas 75235.Received 30 June 1997; revised 5 August 1997; accepted 6 August1997

Oncogene (1998) 16, 83 ± 88 1998 Stockton Press All rights reserved 0950 ± 9232/98 $12.00

case 4. Southern blot analysis using probes speci®c toloci D6S32 and D6S37 are also shown in Figure 1. ForD6S37, this analysis indicates that cases 2 and 16 bothexhibit allelic imbalance. For D6S32, case 4 is in allelicbalance and case 11 exhibits allelic imbalance. Controlhybridizations to the same Southern blots (data notshown) indicated that the observed imbalances weredue to true allelic loss rather than allele ampli®cation.

We therefore conclude that the observed allelicimbalance is due to true loss of heterozygosity forone or more loci on chromosome 6q.Taken together, the data from Southern blot and

PCR analysis showed that allelic imbalance at 6q waspresent in nine of the 19 tumors (47%) and occurred at12 of the 21 6q loci studied (Table 1 and Figure 2). Incontrast, no allelic imbalance was found for the 6p

Table 1 Frequency of allelic imbalance observed with 23 polymorphic markers for chromosome 6 in 19salivary gland carcinomas

Chromosomal Informative/ Allelic imbalance/region Locus Marker total cases informative cases (%)

6p216cen6q16.3-216q21-236q21-236q16.3-23.26q21-22.16q23.36q16.3-23.26q23.3-246q21-23.36q6q6q25.26q6q6q25.2-276q25.2-276q276q26-276q6q276q

D6S29D6Z1D6S268D6S261D6S267D6S407D6S262D6S32D6S472MYBD6S311D6S441D6S473D6S255D6S415D6S437D6S305D6S264D6S297D6S37D6S446D6S281234YF4P

pHHH157a

p308a

AFM115xh2b

AFM059xh8b

AFM114xd12b

AFM198wg11b

AFM059yd6b

p3B10b

AFMa128ydb

pHM2.6a

AFM276xf1b

AFM269ze1b

AFM133xa5b

D6S255b

AFM211yb10b

AFM266yb5b

AFM242zg5b

AFM079zb7b

AFM212yf6b

pJCZ30c

AFM290xf5b

AFM176xh8b

AFM234yf4pb

6/116/1114/1612/177/199/1611/189/136/64/119/105/916/1811/1512/133/814/175/1210/196/125/1210/185/12

0/60/60/140/120/70/9

3/11 (27)1/9 (11)0/60/40/90/5

2/16 (12.5)0/11

3/12 (25)2/3 (67)2/14 (14)1/3 (33)4/10 (40)4/6 (67)1/5 (20)1/10 (10)1/5 (20)

aRFLP marker; bmicrosatellite marker; c VNTR marker

N T

N TN TN TN T

N TN TN TD6S262 D6S262 D6S297 D6S37

D6S32 D6S32 D6S37 D6S446

CASE 11 CASE 4 CASE 16 CASE 2

Figure 1 Representative autoradiograms of 6q allelic imbalance in cases 2, 4, 11 and 16. Patient 11 shows retention ofheterozygosity at locus D6S262 but absence of the upper allele at D6S32 (arrowhead). Patient 4 shows allelic imbalance at locusD6S262 (arrowhead) and retention of heterozygosity at locus D6S32. Patients 2 and 16 have absence of one allele at locus D6S37(arrowhead) but show no allelic imbalance at loci D6S446 and D6S297, respectively. T, tumor DNA; N, constitutional DNA

Deletions at 6q in salivary gland carcinomasL Queimado et al

84

control locus D6S29. The histological subtypes oftumors that were used in this study are shown in Table2. This list indicates that the nine tumors showingallelic imbalance covered the entire spectrum of studiedsalivary gland carcinoma subtypes, with the exceptionof myoepithelial carcinoma and low grade adenocarci-noma.Figure 2 summarizes the LOH results for the nine

tumors that showed allelic imbalance, relative to themarker order on chromosome 6q. This deletion mapindicates that the lowest frequency of LOH occurs inthe centromere proximal region of chromosome 6q(proximal to D6S407). Another interval that appears tobe retained relatively intact is the region surroundingMYB and D6S311. In four tumors (cases 4, 11, 12, 19)a peak of allelic loss is present at loci D6S262/D6S32.This appears to be an interstitial deletion in all of thefour tumors since proximal and distal markerssurrounding the lost markers are retained. In fact, inone of the tumors (case 11) this peak of LOH is theonly observed loss within chromosome 6q. The smallestregion of overlap includes D6S262 and D6S32. The¯anking markers for this deletion, which do notthemselves show LOH, are D6S407 and D6S472.However, the actual minimal deleted interval appearsto be ¯anked by D6S262 and D6S32 since either one orthe other, but not both, are deleted in these tumors (seethe discussion for size estimates of this region). Themost commonly deleted region in this panel of ninetumors occurs closer to the 6q telomere and peaks atD6S297/D6S37. Seven out of the nine tumors (cases 2,4, 5, 6, 12, 16 and 17) show LOH in this interval(Figure 2). Two of these cases (cases 4 and 12) alsoshow LOH at the more proximal location (D6S262/D6S32) with apparent retention of markers in betweenthese two deleted regions. It is therefore unlikely that

the entire end of 6q is lost in these cases. However, it ispossible that two separate deleted chromosomes 6 arepresent in these tumors. Of the seven tumors that showLOH at D6S297/D6S37, three (cases 2, 16, 17) havewhat appear to be relatively small interstitial deletionsthat encompass D6S297 and/or D6S37. The ¯ankingmarkers for this peak of LOH are D6S297 andD6S447. Thus, in one case (cases 16), D6S297 doesnot shown LOH (see Figures 1 and 2) and in threecases (cases 2, 4, 17), D6S446 does not shown LOH(see Figures 1 and 2).

Discussion

Previous cytogenetic studies have observed frequentdeletions within the 6q25-qter region in salivary glandcarcinomas (Sandros et al., 1990; Jin et al., 1994).However, the one previous study of LOH onchromosome 6 in salivary gland carcinomas (Johns etal., 1996) failed to con®rm this high frequency at themolecular level. LOH was found in only two out of 15

Figure 2 Schematic representation of 6q deletions in nine salivary gland carcinomas. On the left, is represented part ofchromosome 6 and the 23 polymorphic loci studied. Patient identi®cation is indicated above each column. The smallest regions ofoverlapped deletion (SRO) are indicated on the right of the diagram; loss of one allele is indicated by a ®lled circle; retention ofboth alleles is indicated by an open circle; uninformative (homozygous) typings are indicated by hatched circles; gaps in the diagramindicate untyped loci

Table 2 Histopathological diagnosis and LOH at 6q in 19 cases ofsalivary gland carcinomas

Histopathological diagnosis 6q LOH/total cases

Adenoid cystic carcinomaCarcinoma ex-pleomorphic adenomaAcinic cell carcinomaEpithelial-myoepithelial carcinomaMucoepidermoid carcinomaMyoepithelial carcinomaDuctal carcinomaLow grade adenocarcinomaUndi�erentiated carcinoma

2/62/41/21/21/10/11/10/11/1

Deletions at 6q in salivary gland carcinomasL Queimado et al

85

malignant tumors (13%), and these deletions wererestricted to adenoid cystic carcinomas. The resultsreported here, by contrast, agree well with thecytogenetic data and indicate that LOH in 6q is afrequent event in these types of tumors. The apparentdiscrepancy between the two reports is probably due tothe fact that the previous study employed only twopolymorphic markers within 6q, and both of thesemarkers were located proximal to 6q25.2; i.e. withinthe region where we observe the lowest incidence ofLOH (0 ± 27%). In the work described here we used asubstantial panel of polymorphic markers spanning thelong arm of chromosome 6 and this allowed us todetect LOH at 6q in practically all histologicalsubtypes of salivary gland tumors.In addition to con®rming previous cytogenetic

observations, there are two other important conclu-sions from this study. The ®rst is that two distinctareas of LOH exist within 6q in the tumor types thatwe have studied. The ®rst of these lies within 6q23(D6S262/D6S32) and appears to be embedded within amore stable chromosomal region, as assessed by theretention of surrounding proximal and distal loci.Large cytogenetically visible interstitial deletions with-in this region have been occasionally described insalivary gland carcinomas (Sandros et al., 1990;Nordkvist et al., 1994) and interstitial loss of geneticmaterial within 6q21-23 seems to be a relativelycommon event in the development of other types ofbenign tumors (Tahara et al., 1996), malignant tumors(Gaidano et al., 1992; Queimado et al., 1995; Orphanoset al., 1995; Sheng et al., 1996) and in theimmortalization of cell lines (Gualandi et al., 1994).Interestingly, the loss of D6S32 has been previouslyreported in all immortalized clones of monochromoso-mal hybrids (Gualandi et al., 1994) and in all cases of aseries of 51 gastric carcinomas that were informativefor this locus and also exhibited LOH at any other 6qlocus (Queimado et al., 1995). Taken together, the datasupport the hypothesis that this small interstitial regionof deletion at 6q is another good candidate for thelocalization of a tumor suppressor gene.The second, and more frequent region of LOH,

resides within 6q27 (D6S297/D6S37). Again, thisregion has been reported to be a recurrent target forLOH in a wide variety of other benign and malignanttumor types (Noviello et al., 1996; Gaidano et al.,1992; Saito et al., 1996; Queimado et al., 1995; Taharaet al., 1996; Cooke et al., 1996a; Chenevix-Trench etal., 1997). Moreover, a candidate gene (SEN6) thatmight have a role in tumor suppression has beenmapped into this general location (Banga et al., 1997).This gene appears to mediate the immortalization ofhuman cells after infection with SV40. However, theexact sublocalization of SEN6 within this region, andits role, if any, in tumor suppression remain to beinvestigated. Five additional candidate tumor and/ormetastasis suppressor genes have been mapped to thelong arm of chromosome 6: ER at 6q25.1 (Garcia etal., 1992); MnSOD at 6q25.2 (Church et al., 1993;Sa�ord et al., 1994); IGFIIR at 6q25.3 (Hankins et al.,1996); SEN 6A at 6q14-21 (Sandhu et al., 1996) andLOT1 at 6q25 (Abdollahi et al., 1997). None of thesegenes appears to map within the smallest regions ofoverlap de®ned in this study. However, it should benoted that correlations between the cytogenetic and

genetic/physical maps are imprecise. All of thesecandidates will therefore have to be tested forplacement within the physical maps of the regions wehave de®ned.Correlations between the genetic and physical maps

can, however, be quite precisely drawn. When thesmallest regions of overlaps that we have de®ned inthis study are correlated with the available physicalmaps some estimates of true genomic distances can bedrawn. The latest radiation hybrid maps of the entirehuman genome (Stewart et al., 1997) do notincorporate all of the markers that we used in thisstudy. Nevertheless, even when more remote ¯ankingmarkers are used as landmarks the two regions wehave de®ned (D6S262/D6S472 and D6S297/D6S446)appear to encompass only 1 ± 2 Mb in each case. Amore useful estimate can be obtained by searching theWhitehead Institute Center for Genome ResearchYeast Arti®cial Chromosome (YAC) map of chromo-some 6. In the case of D6S297/D6S446 these twomarkers map to adjacent YACs (911c10 for D6S297and 933f7 for D6S446) with a gap of unknown lengthin between. It is unlikely that this re¯ects a region of45 Mb and is certainly not inconsistent with anestimate of *2 Mb for this region. In the case ofD6S262/D6S472 the YAC map provides a much morere®ned estimate. Both of these loci map to the sametwo YACS (831d8 and 816b5) that are eachapproximately 1.5 Mb in length. This stronglysuggests that this region of LOH is less than 1.5 Mbin size and renders it tractable by positional cloningtechnologies.Recently, a link has been observed between the most

active common fragile site in the human genome,3p14.2 and cancer (Ohta et al., 1996). In this regard, itis interesting to note that there are at least 4 fragilesites reported within 6q (Simonic and Gericke, 1996).However, none of them appear to lie within the samechromosome bands that de®ne our regions of minimaldeletion.In conclusion, our results show that loss of genetic

material at 6q is a frequent event in salivary glandcarcinomas and occurs in all histological subtypesincluding carcinoma ex-pleomorphic adenoma. Thedata further extend previous cytogenetic observationsby de®ning two small deleted regions at 6q21-23.3 andat 6q27 and support the hypothesis that one or moretumor suppressor genes are located in these relativelyshort regions of 6q, which are relevant to thedevelopment of malignant salivary gland tumors.

Materials and methods

Tumor specimens

Nineteen primary salivary gland carcinomas and corre-sponding normal tissue samples were collected from freshtissue following surgical resection at IPOFG in Lisbon.None of the patients had received preoperative chemother-apy or radiotherapy. Tumors were classi®ed according toSeifert et al. (1990). Clinical and histopathological datawere available for all cases. Tissue samples were frozenimmediately postoperatively in liquid nitrogen and storedat 7708C until DNA extraction. The percentage of nonmalignant cells in the tumor samples was estimated byscoring adjacent histological cryostat sections at low

Deletions at 6q in salivary gland carcinomasL Queimado et al

86

magni®cation. Microdissection of the specimen wasperformed on a cryostat, in the presence of an experiencedpathologist (IF). Samples with less than 50% tumor cellswere excluded from this study.

DNA extraction and DNA markers

High molecular weight DNA was extracted from 19primary salivary gland carcinomas and respective samplesof normal tissues (normal salivary gland and/or bloodsamples), according to standard procedures (MuÈ llenbach etal., 1989). Twenty-three polymorphic markers (one to theshort arm, one centromeric and 21 to the long arm ofchromosome 6) were used in this study. They are listed,along with their approximate chromosomal locations inTable 1. The relative localization of all microsatellitemarkers are in accord with the Whitehead Institute map(Hudson et al., 1995). The relative localization of the othermarkers was based on Cooke et al. (1996b) (D6S29, D6Z1,MYB, D6S37) and on Buetow et al. (1994) (D6S32). Theseloci included three restriction fragment length polymorph-ism (RFLP), 1 variable number of tandem repeats (VNTR)and 19 microsatellite markers. All the primers used for theampli®cation of microsatellite markers were purchasedfrom Research Genetics (Huntsville, AL). MarkerspHHH157, p3B10 and pJCZ30 were obtained fromAmerican Type Culture Collection (Rockville, MD).Marker p308 was kindly provided by Dr EW Jabs (JohnHopkins University).

LOH analysis

Methods for the analysis of RFLP/VNTR markers(restriction endonuclease digestions, gel electrophoresis,transfer to nylon membranes, probe labeling and hybridi-zation) have been previously described (Queimado et al.,1995). For microsatellite marker analysis, the polymerasechain reaction (PCR) was carried out in 96-well plates in aPerkin-Elmer (Norwalk, CT) 9600 GeneAmp PCR System.PCR reactions were performed in a total volume of 20 mlcontaining 30 ± 100 ng of DNA, 0.5 mM of each primer,100 mM of each deoxynucleotide triphosphate exceptdCTP, 10 mM dCTP, 1 mCi of [a32P]dCTP and 0.5 unitsof Taq DNA polymerase. DNAs were ampli®ed for 30

cycles. Each cycle consisted of 948C for 1 min, annealingfor 1 min, 728C for 1 min, with an initial denaturation stepat 948C for 5 min and a ®nal extension step at 728C for10 min. Annealing temperatures (55 ± 608C) varied accord-ing to the primer pairs used. PCR products were diluted1 : 1 with loading bu�er (98% formamide, 0.1% xylenecyanol �, 0.1% bromophenol blue, and 10 mM EDTA, pH8.0) and denatured for 5 min at 948C. Subsequently, 2 ml ofthis solution was electrophoresed in a 6% polyacrylamidegel containing 6.9 M urea and 32.5% formamide, for 3 ±4 h, at 50 W. After electrophoresis, gels were ®xed in 10%acetic acid, washed, dried and exposed to X-ray ®lm for12 ± 24 h. Allele scorings were conducted independently bythree investigators. Allelic imbalances were con®rmed by asecond PCR and a second electrophoresis run.

Classi®cation of allelic imbalance

The relative intensity of the polymorphic fragmentsobtained by Southern blot analysis (RFLP/VNTR mar-kers) or PCR (microsatellites markers) was estimated byvisual inspection. Densitometric analysis of the intensity ofthe hybridization signals in cases where changes werevisually detected was accomplished by scanning of radio-graphs and analysis with Ambis core software (AmbisCorporation). The imbalance factor was de®ned as theratio of the allele intensities in the tumor sample relative tothe ratio of the alleles in the constitutional DNA. A factorof 1.3 or lower was considered inconclusive (Devilee et al.,1991b). LOH was de®ned as total loss of one allele orpartial loss (reduction in the intensity of one allele)compared to the retained one. The latter was interpretedas contamination from normal tissue, if consistent with theevaluation of the respective cryostat sections before DNAextraction, or as clonal variation.

AcknowledgementsThis work was supported by grant PEC5/C/SAU/245/95from Junta Nacional de InvestigacË aÄ o Cienõ ®ca e Tecnolo -gica (JNICT). LQ and AR were supported by grantsPRAXISXXI/BD/3216/94 and BIC/1463/95, respectivelyfrom JNICT.

References

Abdollahi A, Roberts D, Godwin AK, Schultz DC, SonodaG, Testa JR and Hamilton TC. (1997). Oncogene, 14,1973 ± 1979.

Banga SS, Kim S, Hubbard K, Dasgupta T, Jha KK, PatsalisP, Hauptschein R, Gamberi B, Dalla-Favera R, KraemerP and Ozer HL. (1997). Oncogene, 14, 313 ± 321.

BockmuÈ hl U, Schwendel A, Dietel M and Petersen I. (1996).Cancer Res., 56, 5325 ± 5329.

Buetow KH, Weber JL, Ludwigsen S, Scherpbier-HeddemaT, Duyk GM, Sche�eld VC, Wang Z and Murray JC.(1994). Nat. Genet., 6, 391 ± 393.

Chenevix-Trench G, Kerr J, Hurst T, Shih YC, Purdie D,Bergman L, Friedlander M, Sanderson B, Zournazi A,Coombs T, Leary JA, Crawford E, Shelling A, Cooke I,Ganesan TS, Searle J, Choi C, Barrett C, Khoo SK andWard B. (1997). Genes Chrom. Cancer, 18, 75 ± 83.

Church SL, Grant JW, Ridnour LA, Oberley LW, SwansonPE, Meltzer PS and Trent JM. (1993). Proc. Natl. Acad.Sci. USA, 90, 3113 ± 3117.

Cooke IE, Shelling AN, Le Meuth VG, Charnock FML andGanesan TS. (1996a). Genes Chrom. Cancer, 15, 223 ± 233.

Cooke IE, Cox SA, Shelling AN, Le Meuth VG, Spurr NKand Ganesan TS. (1996b). Mamm. Genome, 7, 157 ± 159.

Cooney KA, Wetzel JC, Consolino CM and Wojno KJ.(1996). Cancer Res., 56, 4150 ± 4153.

Devilee P, van Vliet M, van Sloun P, Dijkshoorn KN,Hermans J, Pearson PI and Cornelisse CJ. (1991a).Oncogene, 6, 1705 ± 1711.

Devilee P, van Vliet M, Bardoel A, Kievits T, Kuipers-Dijkshoorn N, Pearson PL and Cornelisse CJ. (1991b).Cancer Res., 51, 1020 ± 1025.

Gaidano G, Hauptschein RS, Parsa NZ, O�t K, Rao PH,Lenoir G, Knowles DM, Chaganti RSK and Dalla-FaveraR. (1992). Blood, 8, 1781 ± 1787.

Garcia M, Derocq D, Freiss G and Rochefort H. (1992).Proc. Natl. Acad. Sci. USA, 89, 1538 ± 11542.

Gualandi F, Morelli C, Pavan JV, Rimessi P, Sensi A,Bonfatti A, Gruppioni R, Possati L, Stanbridge EJ andBarbanti-Brodano G. (1994). Genes Chrom. Cancer, 10,77 ± 84.

Hankins GR, De Souza AT, Bentley RC, Patel MR, MarksJR, Iglehart JD and Jirtle RL. (1996). Oncogene, 12,2003 ± 2009.

Hudson TJ, Stein LD, Gerety SS, Ma J, Castle AB, Silva J,Slonim DK, Baptista R, Kruglyak L, Su SH et al. (1995)Science, 270, 1945 ± 1954.

Hubbard-Smith K, Patsalis P, Pardinas JR, Jha KK,Henderson AS and Ozler HL. (1992). Mol. Cell. Biol. 12,2273 ± 2281.

Deletions at 6q in salivary gland carcinomasL Queimado et al

87

Jin Y, Mertens F, Limon J, Mandahl N, Wennerberg J,Dictor M, Heim S and Mitelman F. (1994). Br. J. Cancer,70, 42 ± 47.

Johns III MM, Westra WH, Califano JA, Eisele D, KochWM and Sidransky D. (1996). Cancer Res., 56, 1151 ±1154.

Menasce LP, Orphanos V, Santibanez-Koref M, Boyle JMand Harrison CJ. (1994). Genes Chrom. Cancer, 10, 286 ±288.

Morita R, Saito S, Ishikawa J, Ogawa O, Yoshida O,Yamakawa K and Nakamura Y. (1991). Cancer Res., 51,5817 ± 5820.

MuÈ llenbach R, Lagoda PJL and Welter C. (1989). TrendsGenet., 5, 391.

Nordkvist A, Mark J, Gustafsson H, Bang G and StenmanG. (1994). Genes Chrom. Cancer, 10, 115 ± 121.

Noviello C, Courjal F and Theillet C. (1996). Clin. CancerRes., 2, 1601 ± 1606.

Ohta M, Inoue H, Cotticelli MG, Kastury K, Ba�a R,Palazzo J, Siprashvili Z, Mori M, McCue P, Druck T,Croce CM and Huebner K. (1996). Cell, 84, 587 ± 597.

Orphanos V, McGown G, Hey Y, Thorncroft M, Santiba-nez-Koref M, Russel SHE, Hickey I, Atkinson RJ andBoyle JM. (1995). Br. J. Cancer, 71, 666 ± 669.

Queimado L, Seruca R, Costa-Pereira A and Castedo S.(1995). Genes Chrom. Cancer, 14, 28 ± 34.

Sa�ord SE, Oberley TD, Urano M and St Clair DK. (1994).Cancer Res., 54, 4261 ± 4265.

Saito S, Saito H, Koi S, Sagae S, Kudo R, Saito J, Noda Kand Nakamura Y. (1992). Cancer Res., 52, 5815 ± 5817.

Saito S, Sirahama S, Matsushima M, Suzuki M, Sagae S,Kudo R, Saito J, Noda K and Nakamura Y. (1996).Cancer Res., 56, 5586 ± 5589.

Sandhu AK, Kaur GP, Reddy DE, Rane NS and Athwal RS.(1996). Oncogene, 12, 247 ± 252.

Sandros J, Stenman G and Mark J. (1990). Cancer Genet.Cytogenet., 44, 153 ± 160.

Seifert G, Brocheriou C, Cardesa A and Eveson JW. (1990).Path. Res. Pract., 186, 555 ± 581.

Sheng ZM, Marchetti A, Buttitta F, Champeme MH,Campani D, Bistocchi M, Lidereau R and Callahan R.(1996). Br. J. Cancer, 73, 144 ± 147.

Simonic I and Gericke GS. (1996). Hum. Genet., 97, 524 ±531.

Stewart EA, McKusic KB, Aggarwal A, Bajorek E, Brady S,Chu A, Fang N, Hadley D, Harris M, Hussain S et al.(1997). Genome Res., 7, 422 ± 433.

Taguchi T, Jhanwar SC, Siegfried JM, Keller SM and TestaJR. (1993). Cancer Res., 53, 4349 ± 4355.

Tahara H, Smith AP, Gaz RD, Cryns VL and Arnold A.(1996). Cancer Res., 56, 599 ± 605.

Trent JM, Stanbridge EJ, McBride HL, Meese EU, Casey G,Araujo DE, Witkowski CM and Nagle RB. (1990).Science, 247, 568 ± 571.

Welch DR, Chen P, Miele ME, McGary CT, Bower JM,Stanbridge EJ and Weissman BE. (1994). Oncogene, 9,255 ± 262.

Deletions at 6q in salivary gland carcinomasL Queimado et al

88