A Functional Investigation of Tumor SuppressorGene Activities in a Nasopharyngeal Carcinoma CellLine HONE1 Using a Monochromosome TransferApproach

Yue Cheng,1 Eric J. Stanbridge,2 Heidi Kong,1 Ulla Bengtsson,2 Michael I. Lerman,3 and Maria Li Lung1*1Department of Biology, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China2Department of Microbiology and Molecular Genetics, University of California, Irvine, California3National Cancer Institute, Frederick, Maryland

Monochromosome transfers of selected chromosomes into a nasopharyngeal carcinoma (NPC) cell line were performed todetermine if tumor suppressing activity for NPC mapped to chromosomes 9, 11, and 17. Current information fromcytogenetic and molecular allelotyping studies indicate that these chromosomes may harbor potential tumor suppressor genesvital to NPC. The present results show the importance of CDKN2A on chromosome 9 in NPC development. There was nofunctional suppression of tumor development in nude mice with microcell hybrids harboring the newly transferred chromo-some 9 containing an interstitial deletion at 9p21, whereas transfection of CDKN2A into the NPC HONE1 cells resulted inobvious growth suppression. Whereas intact chromosome 17 transfers into HONE1 cells showed no functional suppressionof tumor formation, chromosome 11 was able to do so. Molecular analysis of chromosome 11 tumor segregants indicated thatat least two tumor suppressive regions mapping to 11q13 and 11q22–23 may be critical for the development of NPC. GenesChromosomes Cancer 28:82–91, 2000. © 2000 Wiley-Liss, Inc.

INTRODUCTION

Nasopharyngeal carcinoma (NPC) is a tumorwith a striking geographical and ethnic distribu-tion. Epstein-Barr virus (EBV), genetic predisposi-tion, and dietary and environmental factors are allbelieved to play a role in the development of thistumor (Armstrong et al., 1983; Lu et al., 1990; Lunget al., 1993). Compared with other common humanmalignancies, little is known about the molecularbasis for NPC development.

There are several cytogenetic and allelotypingstudies reporting frequent chromosome aberrationsand loss of heterozygosity (LOH) involving 3p, 9p,11q, 14q, and 17p in NPC (Huang et al., 1989, 1994;Bernheim et al., 1992; Hui et al., 1996; Mutirangura etal., 1997, 1998). Potential tumor-suppressive regionson the chromosomes have been implicated. How-ever, cytogenetic and LOH studies provide only in-direct evidence for the functional importance of tu-mor suppressor genes (TSGs) in NPC. It is especiallyimportant to obtain functional evidence for tumorsuppression before laborious candidate gene cloningattempts are made. This is particularly importantwhen one is dealing with sporadic cancers whereclearly defined familial predisposition and identifiedfamilies are not available. This is the case with NPC.

The human TP53 TSG at 17p13 is one of thebest characterized genes involved in a variety of

human cancers. However, it is widely reported thatTP53 mutation is an infrequent event in NPC (Ef-fert et al., 1992; Spruck et al., 1992; Sun et al., 1992;Lung et al., 1998), even though TP53 protein isoverexpressed in more than 60% of primary NPCspecimens (Sheu et al., 1995; Gulley et al., 1998a).Another human TSG, CDKN2A at 9p21, was re-cently reported to be involved in the developmentof NPC by molecular analysis (Lo et al., 1995;Gulley et al., 1998b). Both TP53 and CDKN2A genestudies so far lack strong functional evidence todemonstrate their potential roles in NPC carcino-genesis.

Multiple genetic events are important in thedevelopment and progression of many human tu-mors. While for some well-studied cancers, such ascolorectal carcinomas, some commonly involvedoncogenes and TSGs have been identified in thismultistep progression (Fearon et al., 1990), forNPC the specific genetic events involved are stillunclear. Microcell-mediated chromosome transfer

Supported by: Research Grants Council of Hong Kong (MLL);NCI (EJS); Grant number: CA19401.

*Correspondence to: Dr. Maria Li Lung, Department of Biology,Hong Kong University of Science and Technology, Clear WaterBay, Kowloon, Hong Kong, China.

Received 3 June 1999; Accepted 26 October 1999

GENES, CHROMOSOMES & CANCER 28:82–91 (2000)

© 2000 Wiley-Liss, Inc.

(MMCT) may provide functional evidence for can-didate TSG activities in cancer cells (Goyette et al.,1992; Anderson et al., 1993). By using the MMCTapproach, a putative NPC TSG was recentlymapped to the 3p21.3 region (Cheng et al., 1998).In the present study, the NPC cell line HONE1was utilized for additional investigation of possibleTSGs on chromosomes 9, 11, and 17.

MATERIALS AND METHODS

Cell Lines and Culture Conditions

The NPC HONE1 cells (Yao et al., 1990) usedas the recipient for monochromosome fusion andgene transfection experiments were grown in Dul-becco’s Modified Eagle Medium (DMEM) supple-mented with 10% fetal calf serum (DMEM/10%FCS; Life Technologies, NY). The donor mousehybrid cells (A9) containing a human chromosometagged with the selectable marker, neomycin resis-tance, were grown in DMEM/10% FCS/G418 (Ge-neticin) at 600 or 800 mg/ml. All HONE1/newlytransferred chromosome hybrids were selected ingrowth medium containing 400-mg/ml G418. Thecells were regularly monitored for mycoplasmapresence and were consistently negative.

Microcell-Mediated Chromosome Transfer

The MCH556.15 and MCH313.4 cell lines con-tain intact human chromosomes 11 and 17, respec-

tively. Another donor cell line, Neo.9, contains atruncated human chromosome 9 and is null for theCDKN2A gene. Microcell fusions were performedas previously described (Goyette et al., 1992).

DNA Slot Blot Assay

Mouse A9 genomic DNA (50 ng) was used as aprobe. It was labeled by random priming in thepresence of (a-32P) dCTP (Amersham, U.K.) anddetects mouse DNA in the microcell hybrids. DNAwas transferred directly to a nylon membrane usinga Bio-Rad slot blot apparatus. The blots were hy-bridized at 42°C and washed in 0.2 3 SSC and0.1% SDS at 42°C for 30 min before autoradiog-raphy.

PCR Microsatellite Assay

Microsatellite analysis of recipient, donor, micro-cell hybrid, and tumor segregant cell lines wasperformed using semiautomated fluorescent PCR-based analysis on an Applied Biosystems Prism 310Genetic Analyzer using specific oligonucleotideprimers, (F)dNTP reagents, and AmpliTaqGold(Perkin-Elmer Applied Biosystems, Foster City,CA). The Genescan software and mapping condi-tions used were as recommended by the manufac-turer. The PCR conditions and primers used arelisted in Table 1. Each sample was analyzed atleast twice.

TABLE 1. PCR Primers*

Chromosome locus(location) Primer sequence Annealing temperature, °C

D9S105 (9q22.3–31) GATCATATTGCTTACAACCC 55ACTTACTCATTAAATCTAGGG

D11S1251 (11p15.5) TCCTCTGTATGAAGGCTTCCAGCAGCATCA 60CAAGATTTGGTCTCAGGCAGTCTGGCTCCA

D11S325 (11p13) GACAGACACAGAGGAGAGAATGAATATAT 55CCAGTGCAGCAGAAGCAAAGCGCGG

PYGM (11q13.1) CTAGCAGAGTCCACCTACTG 55CTCTCTCTCTCTCTCTCTGTG

INT2 (11q13.3) TTTCTGGGTGTGTCTGAAT 55ACACAGTTGCTCTAAAGGGT

D11S901 (11q14) TCAGAGGCACAAAAAATATTGGAAG 55CTGGGTGTTGAAGAAGTGAAAATG

D11S2000 (11q22–23) AGTAGAGAACAAAACACTGTGGC 54TTTGAAGATCTGTGAAATGTGC

D11S1647 (11q22–23) AGCTAACCATCTCATTGTAT 46CCCCTCTATTTACCATATG

D11S912 (11q23–25) TCGTGAGANTACTGCTTTGG 56TTTTGTCTAGCCATGATTGC

TP53 (17p13.1) GCACTTTCCTCAACTCTACA 55AACAGCTCCTTTAATGGCAG

*All primer sequences, except PYGM, were obtained from the Genome Data Base (GDB, http://www.gdb.org). The PYGM primer sequences were asreported by Iwasaki et al. (1992).

83TUMOR SUPPRESSOR GENES IN NPC

Fluorescence In Situ Hybridization (FISH)

Biotinylated whole-chromosome 9, 11, and 17FISH probes were purchased from ALTechnolo-gies (Arlington, VA) and Oncor (Gaithersburg,MD). The painting procedure, detection with avi-din antibiotin antibodies, and image processing ona Zeiss Axiophot epifluorescence microscope wereconducted as described previously (Cheng et al.,1998).

PCR-SSCP and DNA Sequencing

Single-strand conformation polymorphism (SSCP)analysis of TP53 exons 5–8 was performed using Taqpolymerase for PCR amplification as previously de-scribed (Lung et al., 1998). The radiolabeled PCRproducts were analyzed for conformational changes innondenaturing polyacrylamide gels. The DNAs fromcell lines were sequenced on an Applied BiosystemsPrism 310 Genetic Analyzer according to the manu-facturer’s protocol.

Tumorigenicity Assay

The tumorigenicity of each cell line was assayedby subcutaneous injection of 1 3 107 cells (0.2 ml)into 4–8-week-old female athymic Balb/c Nu/Numice. A total of six sites were tested for each cellline, with two sites injected per animal. Tumorgrowth in animals was checked regularly and wasmeasured with calipers for comparison of tumorgrowth characteristics for the different injectedcell lines. If tumor formation was noted, then rep-resentative tumors were reconstituted into cell cul-ture for subsequent cytogenetic and molecularanalyses.

DNA Transfection and Colony Formation Assay

The pREP4 expression plasmid constructs con-taining a selectable hygromycin-resistance geneand the 518-bp wild-type CDKN2A gene cDNAand control pREP4 vector (10.2 kb) were trans-fected into HONE1 cells by standard calciumphosphate precipitation methods (Wigler et al.,1978). A total of 2 3 106 cells were plated in100-mm dishes and were transfected with 10 mg ofplasmid DNA for 16 hr. The cells were subse-quently split into four dishes. Four separate exper-iments were performed. After 18 days of selectionin DMEM/10% FCS containing 100-mg/ml hygro-mycin, colonies were fixed and stained withGiemsa to assess the transfection efficiency.

RESULTS

TP53 Status of HONE1 and MCH313.4 Cells

A mutation in exon 8 was found by the PCR-SSCP analysis of the HONE1 TP53 gene. Thepreviously reported heterozygous G-to-C pointmutation at codon 280 in exon 8 was reconfirmedby DNA sequencing. However, we and others haveshown that wild-type TP53 was expressed inHONE1 cells by Western blot analysis (data notshown) and mRNA Northern blot analysis (Sun etal., 1992). The donor chromosome 17 cells,MCH313.4, were also confirmed to harbor wild-type TP53 (Goyette et al., 1992).

Analysis of A9 and Chromosome 9, 11, and 17Microcell Hybrids

The cytogenetic analyses of donor A9/humanchromosome hybrids were carried out by G-band-ing and FISH analysis. MCH556.15 andMCH313.4 contain intact human chromosomes 11and 17, respectively. Neo.9 contains an intact hu-man chromosome 9 by gross appearance, but fur-ther molecular analysis shows an interstitial dele-tion at 9p21 (data not shown).

Transfer of Chromosomes 9, 11, and 17 IntoHONE1 Cells

The G418-resistant microcell fusion clones werepicked up and expanded for further molecular andcytogenetic analyses. DNA from all hybrids wassubjected to slot blot assays. Mouse A9 DNA wasused as a probe to exclude any possible contami-nation of residual mouse DNA during the chromo-some transfers (data not shown). Only mouseDNA-free clones were utilized for further screen-ing. PCR microsatellite typing was conducted withspecific polymorphic markers to determine thepresence of both the recipient and donor chromo-somes in hybrid cells (Fig. 1). Verification of thecopy number of transferred chromosomes in thehybrid cells was assessed by FISH analyses asshown in Figure 2.

Tumorigenicity Assays and Tumor SegregantAnalyses

No significant differences in growth kineticswere found in chromosome 9 and 17 transfers ascompared with parental HONE1 cells. However,tumorigenicity was suppressed by an extra intactcopy of chromosome 11 with every hybrid clone(Table 2). The HONE1 cells are highly tumori-genic; palpable tumors form within 15–25 days ofinjection of these cells in 100% of nude mice. All

84 CHENG ET AL.

chromosome 11 hybrid cells exhibited a delayedlag period in forming tumors, one of which was notobserved until 90 days post-injection. The tumorgrowth kinetics for chromosome 9, 11, and 17 hy-brids are shown in Figure 3.

To check the stability of newly transferred chro-mosomes following tumor formation, cells from tu-mors were reconstituted in both selective (G418)and nonselective medium. Regardless of whichmedium was used, all tumor segregants examinedretained exogenous chromosomes 9 and 17 as ob-served by cytogenetic analysis (data not shown). Incytogenetic analysis of chromosome 11 tumor seg-regants, however, some showed partial or completeloss of the transferred chromosomes (data notshown). These losses correlated with the growth oftumors in animals, consistent with the hypothesisthat small critical regions, which could not be de-tected by cytogenetics, were selectively lost in tu-mor reconstitutes. A detailed molecular analysis ofchromosome 11 tumor segregants was performed.Eight microsatellite markers (Table 1) were usedto identify the candidate TSG regions presumablylost in 11 tumor segregants obtained as comparedto the HONE1/chromosome 11 hybrid cells origi-nally injected into the mice. Seven of 11 (64%), 9 of11 (82%), and 8 of 9 (89%) of the informative tumorsegregants examined showed consistent loss at

11q13.1, 11q13.3, and 11q22-23, where the markersPYGM, INT2, and D11S2000 map, respectively.Representative results of this molecular analysisare shown for hybrids and tumor segregants at locusD11S2000 (Fig. 4). No obvious alterations weredetected with the other markers. A detailed map ofthe two critical regions, 11q13 and 11q22–q23, ob-served in these studies is illustrated in Figure 5.

CDKN2 Transfection and Colony Formation Assay

Four separate transfection experiments wereperformed in an attempt to express exogenouswild-type CDKN2 cDNA in HONE1 cells. Table 3shows that transfection efficiencies demonstrated asignificant decrease in the number of hygromycin-resistant colonies, which were transfected with thewild-type CDKN2A cDNA as compared to vectoralone and controls.

DISCUSSION

NPC is unique among epithelial malignanciesbecause of its epidemiological and biological char-acteristics. This cancer is particularly prevalentamong populations from southern China and south-east Asia. Genetic predisposition to this disease hasbeen well documented by epidemiological studies(Lu et al., 1990). EBV infection has been associ-ated with the disease (Lung et al., 1993), but therole the virus plays in carcinogenesis is as yet un-clear. Multistep progression, including multiple ge-netic alterations, is a common feature for manyhuman cancers. The activation of oncogenes andinactivation of TSGs are thought to be importantmechanisms in tumor initiation and progression,but are still poorly understood for NPC.

Previous cytogenetic and genotyping analysesidentified a number of chromosome regions harbor-ing potential TSGs involved in the development ofNPC. To determine whether TSGs on chromo-somes 9, 11, and 17 directly control the tumorigenicphenotype of the NPC cell line, HONE1, the in-tact chromosomes 11 and 17 and a chromosome 9containing a microdeletion in the 9p were trans-ferred into HONE1 cells to determine the func-tional significance of the chromosomes.

The donor chromosome 9 lacking 9p21 regioncould not suppress the tumorigenicity of HONE1cells. The role of a known TSG, CDKN2A, map-ping to this region was investigated. CDNK2A in-hibits cell cycle progression by binding directly tocyclin-dependent kinases to induce a G1-S block incell cycle kinetics. One of the most widely usedfunctional assays for CDKN2A involves suppression

Figure 1. PCR microsatellite typing of a representative donor, re-cipient, and hybrid cell line. The recipient HONE1, donor MCH313.4,and hybrid HK17.3 cell lines were analyzed with the TP53 primer. Thecombined peak patterns of the hybrid cell line show successful transferof chromosome 17.

85TUMOR SUPPRESSOR GENES IN NPC

of colony formation in transformed cells. The G1-Sblock induced by CDKN2A is very rapid and just 12hr after transfection, it may be demonstratedwhether or not CDKN2A has lost its growth inhib-itory function in tumor cells (Liu et al., 1995; Pou-los et al., 1996; Ruas et al., 1998). In HONE1 cells,it is reported that no detectable mutation ofCDKN2A gene was found by sequencing 97.5% ofthe entire open reading frame. However, CDKN2Ais downregulated in this cell line, as seen by North-

ern blot analysis (Sun et al., 1995). While CDKN2Amutations are uncommon in NPC, hypermethyl-ation is frequently observed in NPC specimenslacking CDKN2A TSG protein (Lo et al., 1995,1996; Gulley et al., 1998b). Based on the down-regulation of CDKN2A in HONE1 cells, we pre-dicted that the inhibition of cell cycle kinetics bytransfected wild-type CDKN2A would be directlyreflected by a decrease in the number of hygromy-cin-resistant colonies. The results of CDKN2A

Figure 2. FISH analysis of the NPC recipientHONE1 cells and HONE1/chromosome 9, 11, and17 microcell hybrids. A minimum of 20 metaphasespreads for each cell line was analyzed. An addi-tional copy of the chromosome was on averagefound in 93%, 88%, and 97% of chromosomes 9, 11,and 17 hybrid cells, respectively. The arrowheadsindicate the presence of an extra copy of the newlytransferred chromosomes.

86 CHENG ET AL.

transfection confirm that exogenous wild-typeCDKN2A suppresses the growth of NPC HONE1cells and provide strong functional proof of theinvolvement of CDKN2A in the development ofNPC. Further efforts may focus on the analysis of

expression of the CDKN2A gene in these trans-fected clones in order to obtain a better under-standing of the role of CDKN2A and its relationship

TABLE 2. Tumorigenicity Assays of Parental HONE1 and HONE1/Chromosomes 9, 11, and 17 MicrocellHybrid Clones in Nude Mice

Cell line IdentificationTumor formation, number of

tumors/number of sitesTime to appearance

of tumors (days)

HONE1 Parental NPC cells 18/18 15–25HK9.11 HONE1 3 Neo.9 6/6 25–35HK9.15 HONE1 3 Neo.9 6/6 10–15HK9.19 HONE1 3 Neo.9 6/6 20–25HK11.1 HONE1 3 MCH556.15 4/6 30–35HK11.8 HONE1 3 MCH556.15 2/6 30–40HK11.12 HONE1 3 MCH556.15 4/6 25–35HK11.13 HONE1 3 MCH556.15 2/6 25–35HK11.19 HONE1 3 MCH556.15 4/6 30–90HK17.1 HONE1 3 MCH313.4 6/6 20–30HK17.3 HONE1 3 MCH313.4 6/6 15–25HK17.8 HONE1 3 MCH313.4 6/6 20–25

Figure 3. Tumor growth kinetics of the NPC recipient HONE1 cellsand chromosome 9, 11, and 17 microcell hybrids in nude mice. Thecurves represent an average tumor volume of all sites inoculated foreach cell population. The vertical bars indicate standard error.

Figure 4. Nonrandom loss of DNA of chromosome 11 tumorsegregants represented by D11S2000 (11q22–23) PCR microsatellitetyping analysis. The recipient HONE1 cells have three pairs of peaks, of195/196, 201/202, and 211/212 base pairs, designated A (left) and B(right); the donor cells MCH 556.15 have only the 211/212 base pairpeaks. In HONE1 cells, all peaks A occur at a higher peak intensity thanpeaks B. After the exogenous chromosome 11 was introduced intoHONE1 cells by MMCT, the resultant hybrid cells, HK11.8 andHK11.12, show a change in the peak patterns. The intensity of peaks Bwithin each pair is greater than that of peaks A. However, this occurrednot only for the pair of peaks mapping to 211/212 base pairs, as seen inboth HONE1 and MCH556.15, but also for the 195/196 and 201/202base pair peaks occurring only in HONE1. Most likely this was due tononspecific competition occurring with these particular PCR primersequences. The peak patterns in all tumor segregants (HK11.8-2TS andHK11.12-2TS) appear similar to HONE1 cells, presumably due toselective loss of critical regions on 11q22–23.

87TUMOR SUPPRESSOR GENES IN NPC

with retinoblastoma and cyclin-dependent kinasesin the cell growth regulation in NPC cells.

TP53 gene mutations are rarely found in NPCeven with a more sensitive functional yeast assay(Lung et al., 1998). However, p53 protein com-monly accumulates in NPC tissues (Gulley et al.,1998a). Interestingly, the overexpression and raremutation rate of TP53 do not always correlate witheach other (Niedobitek et al., 1993). HONE1 cellscarry a heterozygous point mutation at codon 280,

which was also reported in other cell lines andprimary samples (Sun et al., 1992; Chakrani et al.,1995). It was suggested that this point mutation isa gain of function that appears to operate domi-nantly in tumor progression, when the gene is over-expressed (Sun et al., 1993). However, there is nodetectable overexpressed p53 protein or abnormalRNA in this cell line. The donor cells MCH 313.4contain a copy of wild-type TP53. Transfer of thisintact chromosome into HONE1 cells is, therefore,equivalent to introducing a single copy of wild-typeTP53, but under the control of its own promoterand other regulatory sequences. This donor chro-mosome suppressed tumorigenicity in a breast can-cer cell line, MCF7 (Casey et al., 1993), and con-trolled the growth properties of the fibrosarcomacell line, HT1080 (Anderson et al., 1994). Notsurprisingly, transfer of an intact copy of chromo-some 17 could not reverse the tumorigenicity ofHONE1 cells. The results suggest that the het-

TABLE 3. CDKN2A-Dependent Inhibition of ColonyFormation: Number of Colony Forming Units

Experimentnumber

pREP4–p16(wild-type)

pREP4(vector alone)

Control(no DNA)

1 27 483 02 47 797 03 4 359 04 6 322 0

Figure 5. The ideogram on the left shows chromosome 11. The physical order, genetic, and cytogeneticdistances between the microsatellite loci were derived from the Genome Data Base (GDB) and the FifthInternational Workshop on Chromosome 11 Mapping (Shows et al., 1996). The genotyping results on theright show 5 microcell hybrid cell lines (HK11.1, 11.8, 11.12, 11.13, and 11.19) and their tumor segregants.The presence of a marker (open circle), the absence of the marker (closed circle), and uninformativemarkers (crossed circle) are as indicated. Nonrandom loss of DNA in tumor segregant cell lines was foundat two critical regions (CR), 11q13 and 11q22–23.

88 CHENG ET AL.

erozygous mutation of the TP53 gene is not essen-tial for tumorigenesis of HONE1 cells, and p53protein overexpression may perhaps correlate withEBV infection (Gulley et al., 1998a), which is com-mon in NPC.

Previous LOH analysis demonstrated at leasttwo distinct regions of deletion at 11q13.3–q22 and11q22–q24 in NPC specimens (Hui et al., 1996;Mutirangura et al., 1997). This suggests the pres-ence of multiple possible TSGs on the long arm ofchromosome 11. The donor cell line harboring theintact chromosome 11, MCH556.15, which hadpreviously been shown to suppress tumorigenicityin the breast cancer cell line MCF-7 (Negrini et al.,1994), but not the fibrosarcoma cell line HT1080(Anderson et al., 1994), was transferred intoHONE1 cells. Functional evidence shows the im-portance of this chromosome in NPC. All testedHONE1/chromosome 11 hybrids suppress tumori-genicity in nude mice. Tumors that are not sup-pressed all display an increased lag period beforetheir appearance. It is believed that under the se-lective pressure in the mouse there is critical loss ofvital tumor-suppressive regions and the analysisof emergent tumors is key to proving this hypoth-esis.

We analyzed tumor segregants using cytogeneticand molecular techniques. Interestingly, all tumorsegregants from chromosomes 9 and 17 retainedthe newly transferred chromosomes. However,when tumor segregants from chromosome 11 wereexamined by FISH analysis, some reconstitutedcell lines lost the transferred chromosomes or theirfragments. By PCR microsatellite typing, 64% and82% of the tumor reconstitutes examined showed adeletion at the 11q13 region covered by PYGM andINT2, respectively. Localization of these nonran-dom deletions coincided with previous findings ofLOH studies in primary samples. Previous studiesreported LOH in an extensive region of the chro-mosome mapping from 11q13.3 to 11q22. The11q14 region covered by D11S901, however, wasnot lost in this study, suggesting that the criticalregion may be narrowed down to 11q13. Interest-ingly, this deletion region overlaps with that ofhead and neck (Jin et al., 1998; Venugopalan et al.,1998), endocrine (Chakrabarti et al., 1998), breast(Sanz-Ortega et al., 1995), and ovarian cancers(Foulkes et al., 1993).

The second common deletion region detected inthis study maps to 11q22–23. Our data demonstratethat 89% of the informative cases displayed lossesat this region and indicate that a putative TSGmapping within or near the D11S2000 locus is crit-

ical for the development of NPC. A recent reportby Robertson et al. (1999) suggests that a moreprecise location for this locus is 11q22.3–23.1. Theprevious LOH studies suggested that the commondeletion region in NPC is at 11q22–24. In thisstudy, the more centromeric locus D11S901 andthe more telomeric locus D11S912 showed no lossin the tumor segregants. The D11S1647 markeroverlaps with D11S2000, but was not informative inthese cell lines. Other than NPC, a deletion at11q22–23 has been reported in a variety of carci-nomas, including melanoma (Robertson et al.,1999), breast (Carter et al., 1994), ovarian (Foulkeset al., 1993), lung (Rasio et al., 1995), cervical (Be-thwaite et al., 1995), bladder (Shaw et al., 1995),colorectal (Keldysh et al., 1993), and prostate can-cers (Dahiya et al., 1997). Molecular analysis of thechromosome 11 tumor segregants clearly indicatesthe existence of at least two putative TSGs resid-ing on the long arm of chromosome 11, which maybe critical to the development of NPC.

We have previously mapped an NPC tumor sup-pressor activity at 3p21.3 in HONE1 cells. Thispresent study provides proof of the presence ofadditional TSGs on chromosome 11 and function-ally confirms the involvement of a known TSG,CDKN2A, in NPC carcinogenesis. Not surprisingly,taken together, the results suggest that multipletumor suppressor genes contribute to the develop-ment of NPC. While all these experiments havebeen performed in a single NPC cell line, it shouldbe appreciated that there are few well-establishedand well-characterized NPC cell lines available forstudy. We are currently using this monochromo-some transfer approach to analyze a recently estab-lished EBV-positive NPC cell line, C666 (Hui etal., 1998), to establish whether the significance ofthe tumor-suppressive regions already identified inHONE1 is also important for C666.

ACKNOWLEDGMENTS

We thank Haiyan Ge of the Department of Mi-crobiology and Molecular Genetics, University ofCalifornia, Irvine, and Carolyn Choi, ElizabethPang, Stanley Lam, and Josephine Ko of the De-partment of Biology, Hong Kong University of Sci-ence and Technology, Hong Kong, for technicalhelp. We also thank Dr. Mitsuo Oshimura forNeo.9 cells.

REFERENCES

Armstrong RW, Armstrong MJ, Yu MC, Henderson BE. 1983. Saltedfish and inhalants as risk factors for nasopharyngeal carcinoma inMalaysian Chinese. Cancer Res 43:2967–2970.

Anderson MJ, Stanbridge EJ. 1993. Tumor suppressor genes studied

89TUMOR SUPPRESSOR GENES IN NPC

by cell hybridization and chromosome transfer. FASEB J 7:826–833.

Anderson MJ, Casey G, Fasching CL, Stanbridge EJ. 1994. Evi-dence that wild-type TP53, and not genes on either chromosome1 or 11, controls the tumorigenic phenotype of the human fibro-sarcoma HT 1080. Genes Chromosomes Cancer 9:266–281.

Bernheim A, Rousselet G, Massaad L, Busson P, Tursz T. 1992.Cytogenetic studies in three xenografted nasopharyngeal carcino-mas. Cancer Genet Cytogenet 66:11–15.

Bethwaite PB, Koreth J, Herrington CS, McGee JOD. 1995. Loss ofheterozygosity occurs at the D11S29 locus on chromosome 11q23in invasive carcinoma. Br J Cancer 71:814–818.

Carter SL, Negrini M, Baffa R, Gillum DR, Rosenberg AL,Schwartz GF, Croce CM. 1994. Loss of heterozygosity at 11q22–q23 in breast cancer. Cancer Res 54:6270–6274.

Casey G, Plummer S, Hoeltge G, Scanlon D, Fasching C, Stan-bridge EJ. 1993. Functional evidence for a breast cancer growthsuppressor gene on chromosome 17. Hum Mol Genet 11:1921–1927.

Chakrabarti R, Srivatsan ES, Wood TF, Eubanks PJ, Ebrahimi SA,Gatti RA, Passaro E Jr, Sawicki MP. 1998. Deletion mapping ofendocrine tumors localizes a second tumor suppressor gene onchromosome 11q13. Genes Chromosomes Cancer 22:130–137.

Chakrani F, Armand JP, Lenoir G, Ju LY, Liang JP, May E, May P.1995. Mutations clustered in exon 5 of the p53 gene in primarynasopharyngeal carcinomas from Southeastern Asia. Int J Cancer61:316–320.

Cheng Y, Poulos NE, Lung ML, Hampton G, Ou B, Lerman MI,Stanbridge EJ. 1998. Functional evidence for a nasopharyngealcarcinoma tumor suppressor gene that maps at chromosome3p21.3. Proc Natl Acad Sci USA 95:3042–3047.

Dahiya R, McCarville J, Lee C, Hu W, Kaur G, Carroll P, Deng G.1997. Deletion of chromosome 11p15, p12, q22, q23–24 loci inhuman prostate cancer. Int J Cancer 72:283–288.

Effert P, McCoy R, Abdel-Hamid M, Flynn K, Chang Q, Busson P,Tursz T, Liu E, Raab-Traub N. 1992. Alterations of p53 gene innasopharyngeal carcinoma. J Virol 66:3768–3775.

Fearon ER, Vogelstein B. 1990. A genetic model for colorectaltumorigenesis. Cell 61:759–767.

Foulkes WD, Campbell IG, Stamp GWH, Trowsdale J. 1993. Lossof heterozygosity and amplification on chromosome 11q in humanovarian cancer. Br J Cancer 67:268–273.

Goyette MC, Cho K, Fasching CL, Levy DB, Kinzler KW,Paraskeva C, Vogelstein B, Stanbridge EJ. 1992. Progression ofcolorectal cancer is associated with multiple tumor suppressorgene defects but inhibition of tumorigenicity is accomplished bycorrection of any single defect via chromosome transfer. Mol CellBiol 12:1387–1395.

Gulley ML, Burton MP, Allred DC, Nicholls JM, Amin MB, Ro JY,Schneider BG. 1998a. Epstein-Barr virus infection is associatedwith p53 accumulation in nasopharyngeal carcinoma. Hum Pathol29:252–259.

Gulley ML, Nicholls JN, Schneider BG, Amin MB, Ro JY, GeradtsJ. 1998b. Nasopharyngeal carcinoma frequently lack the p16/MTS1 tumor suppressor protein but consistently express theretinoblastoma gene product. Am J Pathol 152:865–869.

Huang DP, Ho JHC, Chan WK, Lau WH, Lui M. 1989. Cytoge-netics of undifferentiated nasopharyngeal carcinoma xenograftsfrom southern Chinese. Int J Cancer 43:936–939.

Huang DP, Lo KW, van Hasselt CA, Woo JKS, Choi PHK, LeungCSF, Cheung ST, Cairns P, Sidransky D, Lee JCK. 1994. Aregion of homozygous loss on chromosome 9p21–22 in primarynasopharyngeal carcinoma. Cancer Res 54:4003–4006.

Hui ABY, Lo KW, Leung SF, Choi PHK, Fong Y, Lee JCK, HuangDP. 1996. Loss of heterozygosity on the long arm of chromosome11 in nasopharyngeal carcinoma. Cancer Res 56:3225–3229.

Hui ABY, Cheung ST, Fong Y, Lo KW, Huang DP. 1998. Charac-terization of a new EBV-associated nasopharyngeal carcinoma cellline. Cancer Genet Cytogenet 101:83–88.

Iwasaki H, Steward PW, Dilley WG, Holt MS, Steinbrueck TD,Wells SA, Donis-Keller H. 1992. A minisatellite and a microsat-ellite polymorphism within 1.5 kb at the human muscle glycogenphosphorylase (PYGM) locus can be amplified by PCR and havecombined informativeness of PIC 0.95. Genomics 13:7–15.

Keldysh PL, Dragani TA, Fleischman EW, Konstantinova LN,Perevoschikov AG, Pierotti MA, Della Porta G, Kopnin BP. 1993.11q deletions in human colorectal carcinomas: cytogenetics andrestriction fragment length polymorphism analysis. Genes Chro-mosomes Cancer 6:45–50.

Jin Y, Hoglund M, Jin C, Martins C, Wennerberg J, Akervall J,

Mandahl N, Mitelman F, Mertens F. 1998. FISH characterizationof head and neck carcinomas reveals that amplification of band11q13 is associated with deletion of distal 11q. Genes Chromo-somes Cancer 22:312–320.

Liu L, Lassam NJ, Slingerland JM, Bailey D, Cole D, Jenkins R,Hogg D. 1995. Germline p16INK4A mutation and protein dysfunc-tion in a family with inherited melanoma. Oncogene 11:405–412.

Lo KW, Huang DP, Lau KM. 1995. p16 gene alterations in naso-pharyngeal carcinoma. Cancer Res 55:2039–2043.

Lo KW, Cheung ST, Leung SF, van Hasselt A, Tsang YS, Mak KF,Chung YF, Woo JKS, Lee JCK, Huang DP. 1996. Hypermethyl-ation of the p16 gene in nasopharyngeal carcinoma. Cancer Res56:2721–2725.

Lu SJ, Day NE, Degos L, Lepage V, Wang PC, Chan SH, SimonsM, McKnight B, Easton D, Zeng Y, de The G. 1990. Linkage ofa nasopharyngeal carcinoma susceptibility locus to the HLA re-gion. Nature 346:470–471.

Lung ML, Sham J, Lam WP, Choy D. 1993. Analysis of localizedtumors provides further evidence for the direct association ofEpstein-Barr virus with nasopharyngeal carcinoma. Cancer 71:1190–1192.

Lung ML, Hu Y, Cheng Y, Li MF, Tang CMC, O SK, Iggo RD.1998. P53 inactivating mutations in Chinese nasopharyngeal car-cinomas. Cancer Letters 133:89–94.

Mutirangura A, Tanunyutthawongese C, Pornthanakasem W,Kerekhanjanarong V, Sriuranpong V, Yenrudi S, Supiyaphun P,Voravud N. 1997. Genomic alterations in nasopharyngeal carci-noma: loss of heterozygosity and Epstein-Barr virus infection. Br JCancer 76:770–776.

Mutirangura A, Pornthanakasem W, Sriuranpong V, Supiyaphun P,Voravud N. 1998. Loss of heterozygosity on chromosome 14 innasopharyngeal carcinoma. Int J Cancer 78:153–156.

Negrini M, Sabbioni S, Possati L, Rattan S, Corallini A, Barbanti-Brodano G, Croce CM. 1994. Suppression of tumorigenicity ofbreast cancer cells by microcell-mediated chromosome transfer:studies on chromosomes 6 and 11. Cancer Res 54:1331–1336.

Niedobitek G, Agathanggelou A, Barber P, Smallman LA, Jones EL,Young LS. 1993. p53 overexpression and Epstein-Barr virus in-fection in undifferentiated and squamous cell nasopharyngealcarcinomas. J Pathol 170:457–461.

Poulos NE, Farmer AA, Chan KWK, Stanbridge EJ. 1996. Design ofa novel bicistronic expression vector with demonstration of ap16INK4-induced G1-S block. Cancer Res 56:1719–1723.

Rasio D, Negrini M, Manenti G, Dragani TA, Croce CM. 1995. Lossof heterozygosity at chromosome 11q in lung adenocarcinoma:identification of three independent regions. Cancer Res 55:3988–3991.

Robertson GP, Goldberg EK, Lugo TG, Fountain JW. 1999. Func-tional localization of a melanoma tumor suppressor gene to a small(# 2 Mb) region on 11q23. Oncogene 18:3173–3180.

Ruas M, Peters G. 1998. The p16INK4a/CDKN2A tumor suppressorand its relatives. Biochim Biophys Acta 1378:F115–F177.

Sanz-Ortega J, Chuaqui R, Zhuang Z, Sobel ME, Sanz-Esponera J,Liotta LA, Emmert-Buck MR, Merino MJ. 1995. Loss of het-erozygosity on chromosome 11q13 in microdissected human malebreast carcinomas. J Natl Cancer Inst 87:1408–1410.

Shaw ME, Knowles MA. 1995. Deletion mapping of chromosome 11in carcinoma of the bladder. Genes Chromosomes Cancer 13:1–8.

Sheu LF, Chen A, Tseng HH, Leu FJ, Lin JK, Ho KC, Meng CL.1995. Assessment of p53 expression in nasopharyngeal carcinoma.Hum Pathol 26:380–386.

Shows TB, Alders M, Bennett S, Burbee D, Cartwright P, Chan-drasekharappa S, Cooper P, Courseaux A, Davies C, DevignesMD, Devilee P, Elliott R, Evans G, Fantes J, Garner H, GaudrayP, Gerhard DS, Gessler M, Higgins M, Hummerich H, James M,Lagercrantz J, Litt M, Little P, Mannens M, Munroe D, NowakN, O’Brien S, Parker N, Perlin M, Reid L, Richard C, Sawicki M,Swallow D, Thakker R, van Heyningen V, van Schothorst E,Vorechovsky I, Wadelius C, Weber B, Zabel B. 1996. Report ofthe Fifth International Workshop on Human Chromosome 11Mapping (1996). Cytogenet Cell Genet 74:1–56.

Spruck CH, Tsai YC, Huang DP, Yang AS, Rideout WM, Gonzalez-Zulueta M, Choi P, Lo KW, Yu MC, Jones PA. 1992. Absence ofp53 gene mutations in primary nasopharyngeal carcinomas. Can-cer Res 52:4787–4790.

Sun Y, Hegamyer G, Cheng YJ, Hildesheim A, Chen JY, Chen IH,Cao Y, Yao KT, Colburn NH. 1992. An infrequent point mutationof the p53 gene in human nasopharyngeal carcinoma. Proc NatlAcad Sci USA 89:6516–6520.

Sun Y, Nakamura K, Wendel E, Colburn N. 1993. Progression

90 CHENG ET AL.

toward tumor cell phenotype is enhanced by overexpression of amutant p53 tumor suppressor gene isolated from nasopharyngealcarcinoma. Proc Natl Acad Sci USA 90:2827–2831.

Sun Y, Hildesheim A, Lanier AEP, Cao Y, Yao KT, Raab-Traub N,Yang CS. 1995. No point mutation but decreased expression ofthe p16/MTS1 tumor suppressor gene in nasopharyngeal carcino-mas. Oncogene 10:785–788.

Venugopalan M, Wood TF, Wilczynski SP, Sen S, Peters J, Ma GC,Evans GA, Srivatsan ES. 1998. Loss of heterozygosity in squa-mous cell carcinomas of the head and neck defines a tumor

suppressor gene region on 11q13. Cancer Genet Cytogenet 104:124–132.

Wigler M, Pellicer A, Silverstein S, Axel R. 1978. Biochemicaltransfer of single-copy eucaryotic genes using total cellular DNAas donor. Cell 14:725–731.

Yao K, Zhang HY, Zhu HC, Wang FX, Li GY, Wen DS, Li YP, TsaiCHA, Glaser R. 1990. Establishment and characterization of twoepithelial tumor cell lines (HNE-1 and HONE-1) latently in-fected with Epstein-Barr virus and derived from nasopharyngealcarcinomas. Int J Cancer 45:83–89.

91TUMOR SUPPRESSOR GENES IN NPC