Detection of Genetic Alterations in Esophageal Squamous ......ISSR-PCR DNA Fingerprint Profiling. For ISSR-PCR DNA fingerprint profiling, 50 ng of genomic DNA were used in each 10-ml

Advances in Brief

Detection of Genetic Alterations in Esophageal SquamousCell Carcinomas and Adjacent Normal Epithelia byComparative DNA Fingerprinting UsingInter-Simple Sequence Repeat PCR1

Johnny C. O. Tang, King Y. Lam, Simon Law,John Wong, and Gopesh Srivastava2

Departments of Pathology [J. C. O. T., K. Y. L., G. S.] and Surgery[S. L., J. W.], The University of Hong Kong, Hong Kong, andDepartment of Applied Biology and Chemical Technology, The HongKong Polytechnic University, Hong Kong [J. C. O. T.]

AbstractIn this study, we screened 19 esophageal squamous cell

carcinomas (ESCCs) for the detection of genetic alterationsusing inter-simple sequence repeat PCR, a DNA fingerprint-ing approach. Three simple repetitive unanchored primersrepresenting tri- and tetranucleotide repeats [(GTG)5,(GACA)4, and (GATA)4] were used, and evidence of gainsand losses of chromosomal sequences were detected in alltumors (19 of 19 cases) for at least one of the primers. In 13of these cases, apparently normal marginal epithelia adja-cent to the tumors were also collected and examined. Eightof the 13 (62%) patients showed matching somatic muta-tions in the marginal epithelia adjacent to the tumors. Fiveof these 8 (63%) marginal epithelial samples were histolog-ically normal, two were dysplastic, and one had extremelyrare tumor cells. In 3 of these 13 (23%) cases, the profilebands were also seen to quantitatively increase in intensity,progressing from normal epithelia to marginal epithelia totumors. Ten profile bands showing gains and one profileband showing loss in tumors compared with the correspond-ing normal epithelia were cloned, and their origins weredetermined by sequencing. The DNA sequence of one of theprofile bands showing gain in the tumor could be matched toan expressed sequence tag sequence that has been mapped tothe 7q22 region, a genomic amplification novel to ESCC.The sequence of the other profile band showing gain in thetumor could be matched to a nonexonic sequence of chro-mosome 20, whereas the sequences of the remaining profilebands could not be matched with any known sequences aftercomparison with the genomic sequence data in the European

Molecular Biology Laboratory and GenBank databases. Thebona fidenature of the gains or losses of 11 profile bands inthe original cases was confirmed by direct genomic PCRamplification. The frequencies of these specific gene alter-ations in tumors were then analyzed in a total of 60 ESCCs,which included 41 additional cases of ESCC. Significantly,26 of 60 (43%) tumors showed the DNA amplification forthe expressed sequence tag sequence of chromosome 7,whereas the frequency of other individual gene alterationsranged from 7% to 15%. It is concluded that the inter-simple sequence repeat PCR strategy is adequate for thedetection of somatic mutations in tumors, most of which arequantitative alterations in anonymous genomic sequences.This approach is also suitable for detection of somatic mu-tations preceding the onset of morphologically detectableneoplasia in ESCC.

IntroductionTumors are composed of populations of abnormal cells

capable of altering their genomes in response to changes in theirmilieu (1, 2). The abnormal biological properties of tumor cellsare acquired by a microevolutionary process that can enhancetheir overall growth, survival, and metastatic potential (1). Tu-mor progression is driven by clonal evolution in which a fewneoplastic clones are selected from a variety of randomly gen-erated cell variants. Rather than a few specific mutations leadingto genetic instability, mechanisms of somatic mutation generatean evolutionary diversity, resulting in tumor heterogeneity, andeventually clonal predominance can be observed (1, 2). Thus,the critical genetic alterations are those that facilitate tumorheterogeneity, and these are necessary for tumor progressionand development (3). Subsequently, or concurrently, a set ofmutations necessary for malignancy is generated and expandedby clonal selection. For gastrointestinal cancers, tumor progres-sion has been proposed to follow a stepwise acquisition ofmultiple changes that involve multiple oncogenes, tumor sup-pressor genes, and cell cycle-regulating genes (4, 5).

Different methods have been used to examine the alter-ations of cancer genomes. Typical examples are the observationof clonal cytogenetic abnormalities (6), detection of amplifica-tions and/or deletions at the gene or transcript level (7), local-ization of point mutations for cancer-related genes (4), andexamination of allelic losses or replication error phenotypes atmicrosatellite loci that are relevant to carcinogenesis (8). An-other approach to detect the alterations of cancer genomes is thegeneration of DNA fingerprint profiles that can be produced byPCR-based methods. Examples include the detection of geneticalterations in colorectal, lung, and ovarian cancers by arbitrarily

Received 9/18/00; revised 3/6/01; accepted 3/8/01.The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisementin accordance with 18 U.S.C. Section 1734 solely toindicate this fact.1 Supported by Grant CRCG 335/046/0086 from The University ofHong Kong.2 To whom requests for reprints should be addressed, at the Departmentof Pathology, The University of Hong Kong, Queen Mary HospitalCompound, 102 Pok Fu Lam Road, Hong Kong. Phone: 852-2855-4859; Fax: 852-2872-5197; E-mail: [email protected].

1539Vol. 7, 1539–1545, June 2001 Clinical Cancer Research

Research. on August 23, 2021. © 2001 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

primed PCR genomic fingerprinting (9–11) and the applicationof ISSR-PCR3 analysis on sporadic colorectal cancer (12).

ISSR-PCR has also been described as microsatellite-primed PCR in the literature (13, 14) because it specificallyamplifies regions of the genome between microsatellites. Themethod of comparing inter-repeat profiles from the whole ge-nome was first used in the identification and differentiation ofdifferent eukaryotic species because of its accuracy and repro-ducibility (13, 14). The significance of the changes in theinter-repeat sequences for carcinogenesis was subsequently sug-gested by Basiket al. (12). In the present study, ISSR-PCR wasapplied to ESCC and premalignant lesions. By observing thefingerprinting profiles generated from the malignant and prema-lignant populations compared with the normal genomes, wehoped to follow the process of tumor progression in ESCC and

compare the observable alterations at the genome level with theclinicopathological features of the disease.

Materials and MethodsCollection of Samples and DNA Extraction. For the

initial ISSR-PCR DNA fingerprint profiling, ESCC sampleswith gross tumor appearance were collected prospectively from19 surgically resected specimens after esophagectomy betweenJune 1996 and March 1997 in Queen Mary Hospital, HongKong. All of the selected ESCC samples had more than 80%viable tumor cells as shown by histological assessment. Mor-phologically normal epithelial tissues at least 10 cm away fromthe tumors were also sampled from each patient. Apparentlynormal marginal epithelia that were adjacent to the tumors by 1cm were collected from 13 patients. Sterile equipment was usedfor every sampling of tumor, normal, and marginal epithelia.One-half of each sample was fixed in 10% formalin for histo-logical assessment as described previously (15), and the otherhalf was snap-frozen in liquid nitrogen and stored at285°C forlater DNA extraction using the method described previously(16). For every five cryostat sections cut from the frozen blocksfor DNA extraction, an additional H&E-stained section wasprepared for histological assessment to confirm the presence orabsence of tumor cells. To determine the frequency of occur-rence of the specific gene alterations by direct genomic PCRanalysis for the ISSR-PCR profile bands in ESCC, an additionalbatch of 41 archival ESCCs collected between 1989 and 1993 inQueen Mary Hospital that were not pretreated with chemora-diotherapy were also included in the present study. All theseESCC samples had more than 80% viable tumor cells as shownby histological assessment.

ISSR-PCR DNA Fingerprint Profiling. For ISSR-PCRDNA fingerprint profiling, 50 ng of genomic DNA were used ineach 10-ml reaction as the template. The sequences of the threeprimers are as follows: (a) (GTG)5, 59-GTGGTGGTGGTG-GTG-39; (b) (GACA)4, 59-GACAGACAGACAGACA-39; and(c) (GATA)4, 59-GATAGATAGATAGATA-3 9. In each reac-tion, 0.2 mM each primer was end-labeled with [g-33P]ATP(Amersham, Aylesbury, United Kingdom) using T4 polynucle-otide kinase (Promega, Madison, WI). Each PCR reaction con-tained the same amount of labeled and unlabeled primer. ThePCR protocol (with an annealing temperature of 56°C) and theprocedures for electrophoresis and autoradiography were asdescribed previously (12).33P-end-labeled molecular weightmarkers (HaeIII-cutf174 DNA) were also run on the side lanesof the gels for the size comparison.

DNA Sequencing of Selected ISSR-PCR Profile Bands.Eleven selected bands from the ISSR-PCR profiles were cut outof the polyacrylamide gel based on their discrete and nonover-lapping appearance with other bands in the autoradiographs.DNA was extracted by boiling the gel slices in 30ml of waterfor 10 min. Eightml of each DNA fragment were then clonedinto 1mg of pGEM-T Easy vector (Promega). Cycle sequencingwas done with the Big-Dye sequencing kit (Perkin-Elmer) usingthe forward and reverse primers as suggested by the supplier ofthe vector, and the sequences were then analyzed by an ABI 310Genetic Analyzer (Perkin-Elmer). The sequences were matched

3 The abbreviations used are: ISSR-PCR, inter-simple sequence repeatPCR; ESCC, esophageal squamous cell carcinoma.

Fig. 1 Reproducibility of ISSR-PCR DNA fingerprint profiles. Identi-cal profiles were generated by ISSR-PCR performed on two independ-ent DNA preparations (IandII) from the normal esophageal epithelia ofthe same patient using the (GATA)4, (GACA)4, and (GTG)5 repeatprimers. The sequence of the tri- or tetranucleotide repeat primer used isshown below each block.33P-end-labeled molecular weight markers(HaeIII-cut f174 DNA) were run on the side lanes of the gels for thesize comparison, and the molecular size ranges of the profiles are shownon theleft of each panel.

1540Genetic Alterations in Esophageal Cancer

Research. on August 23, 2021. © 2001 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

against the genomic sequence data in the European MolecularBiology Laboratory and GenBank databases.

Direct Genomic PCR Analysis of ISSR-PCR ProfileBands. Two specific primers were designed for the directgenomic PCR analysis of the 11 ISSR-PCR profile bands in 60ESCCs. The primer sequences for the amplification of selectedprofile bands by PCR are summarized in Table 2. Patient DNA(30 ng) was used for each 20-ml PCR reaction with 10 pmol ofeach primer, 1 unit of Taq polymerase (Promega), 0.2 mM eachdeoxynucleotide triphosphate, and 2 mM MgCl2. The reactionwas carried out for 35 cycles after an initial denaturation at 95°Cfor 4 min under the following conditions: 95°C for 30 s; 52°Cfor 30 s; and 72°C for 30 s. In addition, theb-globin genesequences (17) were also amplified simultaneously in the samereaction tube as the internal PCR controls to normalize theamount of DNA of each sample analyzed. The products of themultiplex PCR amplification were run in a 3% agarose gel andvisualized under UV light after staining with ethidium bromide(Fig. 2). Molecular weight markers (HaeIII-cut f174 DNA)were run on the side lanes of the gels for the size comparison.

ResultsISSR-PCR DNA Fingerprint Profiling of ESCC

Tumors and Adjacent Normal Epithelia. First, to ensure thereproducibility of ISSR-PCR DNA fingerprint profiling in ourlaboratory, the method was performed on two independent DNApreparations from the normal esophageal epithelia of the samepatient. Identical profiles were generated from the two inde-pendent DNA preparations from the same patient (Fig. 1).Tissue samples collected from 19 patients with ESCC were then

analyzed by ISSR-PCR. The clinicopathological features ofthese patients are summarized in Table 1. In the 13 apparentlynormal marginal epithelia collected, extremely rare tumor cellswere found in 2 samples (patients 2 and 7) by histologicalassessment, dysplasia was noted in 2 other samples (patients 5and 17), and the rest were histologically normal. The profilesgenerated by ISSR-PCR for each individual patient, using eachof the three repeating primers, were aligned together with thosegenerated from the corresponding normal epithelium. Evidenceof amplification and/or deletion of chromosomal sequences wasobserved as shown by changes in intensities or gains and/orlosses of profile bands of tumors compared with the correspond-ing normal epithelia. All of the 19 ESCCs showed evidence ofgenetic alteration compared with their normal epithelia controlsusing at least one of the repeating primers (Table 1). Six of the19 patients (patients 2, 5, 6, 17, 18, and 19; 32%) showedabnormal profile bands for all three primers. Representativeexamples of autoradiographs are shown in Fig. 2I.

Analysis of Selected ISSR-PCR Profile Bands. Tenprofile bands showing gains and one profile band showing lossin tumors compared with the corresponding normal epitheliawere cloned, and their origins were determined by sequencing(Table 2). Profile band A from patient 6 shows 98% homologyto the GenBank sequence that can be matched against a nonex-onic genomic sequence (97% homology) with accession numberAL050325 isolated from chromosome 20. Profile band C show-ing gain in the tumor specimen from patient 19 shows 98%homology to an expressed sequence tag sequence with accessionnumber AA078562 that has been mapped to the chromosome7q22 region. The other profile bands were anonymous and could

Table 1 Summary of ISSR-PCR analysis and clinical data of ESCC patients

Patientno. Age/sex

Histopathologicaltypes

Tumorstage

Histologicalstage

Preoperativechemoradio-

therapyreceived

Types of repeating primers useda

(GTG)5 (GACA)4 (GATA)4

N M T M T M T M T

1 63/M NT NT SCC (M)* T3N1M0 III None Gain Gain11 n n n n2 47/M NT SCC SCC (M) T3N1M0 III None 1 11 Gain Loss Gain n3 59/M NT NT SCC (M) T3N1M0 III None n Loss Gain Loss n n4 68/M NT – SCC (M) T3N1M0 III Received – n – Loss – n5 68/M NT D SCC (M) T3N0M0 IIA None Gain Gain11, Loss Loss Loss n Loss6 66/M NT – SCC (M) T3N1M0 III Received – Loss – Gain – Loss7 66/M NT SCC SCC (P) T3N1M1 IV None n n Gain Gain n Gain8 64/M NT NT SCC (M) T3N1M0 III None n n n n Gain Gain9 68/M NT – SCC (M) T3N1M0 III None – n – n – Loss

10 58/M NT – SCC (M) T2N0M0 IIA Received – Gain, Loss – n – Gain11 66/M NT – SCC (M) T3N0M0 IIA Received – Loss – Gain – n12 70/M NT – SCC (W) T3N1M1 IV None – n – n – Gain, Loss13 63/M NT NT SCC (W) T3N1M0 III None n n Gain Gain n Loss14 70/F NT NT SCC (M) T2N0M0 IIA Received n n Gain Gain Loss Loss15 48/F NT NT SCC (W) T3N1M0 III None n n n n n Loss16 57/M NT NT SCC (W) T3N1M0 III None n n n n Gain Loss17 70/M NT D SCC (M) T3N1M0 III None Loss Loss - Loss Gain Gain18 49/M NT NT SCC (M) T4N1M1 IV Received Loss Loss n Gain, Loss n Gain19 72/F NT NT SCC (W) T3N1M0 III None n Gain Loss Loss n Loss

a All the gains, losses, or increase in intensity of bands reported in both marginal epithelium and tumor of the same patient are of the same sizes;D, dysplasia; Gain, gain of an extra band compared to normal epithelium; Loss, loss of a band compared to normal epithelium; (M), moderatelydifferentiated type of tumor; M, marginal epithelium; n, same pattern as that found in normal epithelium; N, normal epithelium; NT, no tumor; (P),poorly differentiated type of tumor; SCC, squamous cell carcinoma; T, tumor; (W), well differentiated type of tumor; -, decreased intensity of a bandcompared to normal epithelium; –, unavailable;1, increased intensity of a band compared to normal epithelium;11, increased intensity of a bandcompared to marginal epithelium; *, with mucin-producing component.

1541Clinical Cancer Research

Research. on August 23, 2021. © 2001 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

Fig. 2 I, comparative DNA profiling analysis using ISSR-PCR to detect genetic alterations in ESCC patients.A, comparison between the profilesof normal epithelium (N) and tumor (T).B, comparison between the normal epithelium (N), marginal epithelium (M), and tumor (T). The sequenceof the PCR primer used is shownbeloweach block.Arrowheadsindicate the abnormal positions of the bands in the fingerprinting profiles whencompared with normal epithelium. In patient 9, genetic alteration of the tumor sample was shown by the loss of a band (arrowhead) in the profilecompared with the normal epithelium. In patients 12 and 18, genetic alterations of tumor samples were detected by both gains and losses of bands(arrowheads). Intensity changes of bands in the profiles were found in patients 2 and 17, and the alterations were shown by the spectra of changes

1542Genetic Alterations in Esophageal Cancer

Research. on August 23, 2021. © 2001 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

not be matched significantly with any known sequences in theEuropean Molecular Biology Laboratory and GenBank data-bases.

Validation of ISSR-PCR Results and Frequency of In-dividual Gene Alterations of ISSR-PCR Profile Bands inESCC. To show that the genetic alterations observed byISSR-PCR are true somatic mutations occurring in the tumorsand represent true amplifications in the genome, specific prim-ers were designed for 11 ISSR-PCR profile bands based on thesequencing results (Table 2). Using the primers specific forthese gene alterations, direct genomic PCR amplification wasperformed on the original patient DNA samples and confirmedthebona fidenature of the gains or losses of 11 profile bands inthe original cases. Selected examples are shown in Fig. 2II.

To determine the frequency of occurrence of these specificgene alterations in ESCC, DNA from the tumor and the corre-sponding normal epithelium from 60 ESCCs, including 41 ad-ditional cases of ESCC, was analyzed by direct genomic PCR

using the specific PCR primers for ISSR-PCR profile bands(Table 2). For normalization of the DNA quantity of the tumorand the corresponding normal epithelium samples for the PCRanalysis,b-globin gene amplification was used as the internalcontrol (17). The frequencies of the occurrence of these specificgene alterations are summarized in Table 3. Significantly, 26 of60 (43%) tumors showed DNA amplification for the expressedsequence tag sequence of chromosome 7 (the band intensity wasat least two times stronger than that of the corresponding normalcontrol), whereas the frequency of other individual gene alter-ations ranged from 7% to 15%.

DiscussionThe three repeating primers [(GTG)5, (GACA)4, and

(GATA)4] were chosen for ISSR-PCR profiling because theserepeating sequences are somatically stable in normal humangenomes (18, 19), and they are diverse because they can be

(arrowheads) from normal epithelia to marginal epithelia to tumors.33P-end-labeled molecular weight markers (HaeIII-cutf174 DNA) were run onthe side lanes of the gels for size comparison, and the molecular size ranges of the profiles are shown on theleft of each panel.II, direct genomicPCR analysis for the ISSR-PCR profile bands. The primer sequences for the PCR amplifications are shown in Table 2.a, a 520-bp fragment (A) wasisolated from the tumor profile of patient 6 that shows a gain using (GACA)4 as the PCR profiling primer compared with the corresponding normalepithelium. Band A from tumor (T) shows an increase in copy number compared with the normal epithelium (N). b, a 580-bp fragment (B) was isolatedfrom the profile of normal epithelium of patient 11 that shows a loss using (GTG)5 as the PCR profiling primer compared with the correspondingtumor. Band B in tumor (T) shows a decrease in copy number compared with the normal epithelium (N).c, a 270-bp fragment (C) was isolated fromthe tumor profile of patient 19 that shows a gain using (GTG)5 as the PCR profiling primer compared with the corresponding normal epithelium. BandC in tumor (T) shows an increase in copy number compared with the normal epithelium (N). Theb-globin sequences G1 (268 bp) and G2 (110 bp)were also amplified simultaneously in the same reaction tube as the internal PCR controls to normalize the amount of DNA in each sample analyzed.The products of the multiplex PCR amplification were run in a 3% agarose gel and visualized under UV light after staining with ethidium bromide.Molecular weight markers (HaeIII-cutf174 DNA) were run on the side lanes of the gels for size comparison, and selected molecular sizes are shownon theleft of each panel.

Table 2 Sequences of the primers used for direct genomic PCR analysis of the 11 ISSR-PCR profile bands in 60 ESCCs

ISSR-PCR profile bands(direct PCR product size) PCR primers Primer sequences (59to 39)

A (520 bp) A1 ATG-AAT-TCT-CCA-GAT-GCA-CAA2 TAC-GGA-GGG-TAA-AGC-AAG-AC

B (580 bp) B1 TAT-GGT-CAT-GTT-GGG-ACT-CAB2 ACC-CAG-CCA-CCT-AAT-GAC-TT

C (270 bp) C1 GGA-AGA-ACA-AGT-TCC-AGG-AGC2 GCT-TCC-TCC-TTT-CTT-CCC-TA

D (380 bp) D1 ATG-CAG-ACT-GAC-TCC-GAT-TTD2 CAA-AGG-CAG-CAA-GGT-GAG

E (353 bp) E1 TGT-TCT-TCC-TTC-CCC-TCA-CE2 TGA-TGC-CAA-GCA-CTC-ATA-TC

F (989 bp) F1 TTT-GAG-ATA-CGA-CAC-CTG-GAF2 TCG-CAT-CAC-CAT-AAC-TGA-CT

G (341 bp) G1 AAC-AGG-GGA-CCC-AAT-GACG2 AAG-TAG-CGA-GCA-ACA-GGA-AG

H (626 bp) H1 CTC-GGA-TTA-CTG-GGA-GTG-ACH2 CTA-CTG-TTG-CAC-CGA-GCA-T

I (477 bp) I1 AAG-TAG-CGA-GCA-ACA-GGA-AGI2 CTC-CAG-ATG-CAA-ACA-GTG-AC

J (469 bp) J1 ACA-CAC-TAG-GTT-GGG-TAG-GGJ2 AAT-GCA-GGA-AAA-GGA-GAG-TG

K (954 bp) K1 ACG-TTT-GAT-GGG-TTG-AGT-CTK2 GCA-GAT-TCA-ATG-TCT-CCA-CA

G1 (268 bp) Globin-1 GAA-GAG-CCA-AGG-ACA-GGT-ACGlobin-2 CAA-CTT-CAT-CCA-CGT-TCA-CC

G2 (110 bp) Globin-3 ACA-CAA-CTG-TGT-TCA-CTA-GCGlobin-2 CAA-CTT-CAT-CCA-CGT-TCA-CC

1543Clinical Cancer Research

Research. on August 23, 2021. © 2001 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

found in both autosomes and sex chromosomes (20, 21). ISSR-PCR amplifies the inter-repeat sequences (13, 14), and theobserved alterations of the profiles from tumors in the presentstudy and in that of Basiket al. (12) include changes in inten-sities, gains and/or losses, which, as shown here, correlate withdeletions and amplifications of genomic fragments. The profilebands with increased or decreased intensities in tumor speci-mens must reflect spontaneous somatic mutations consisting ofgains and losses of chromosomal sequences occurring in tumorsthat became detectable by tumor clonality. These genetic alter-ations can be regarded as markers for the clonal evolution oftumor progression, showing the emergence and/or deletion oftumor clones (1, 2, 16).

Eight of 13 patients (patients 5, 7, 8, 13, 14, 17, 18, and 19;62%) were found to have matching gains or losses of profilingbands in the ESCC and the corresponding apparently normalmarginal epithelia adjacent to the tumors for at least one primerused. For the marginal epithelia of these eight patients, twopatients (patients 5 and 17) had dysplastic epithelia, one patient(patient 7) had extremely rare tumor cells, and the other fivepatients were histologically normal. This observation suggeststhat the generation of widespread genetic alterations can be anearly event in tumor progression and that these alterations canbe detected by ISSR-PCR in premalignant stages. However, itshould be pointed out that although all of the necessary precau-tions were taken to avoid any contamination of histologicallynormal marginal epithelia with tumor, one could never be surethat absolutely no tumor cells were present in the nontumoroussample. PCR reactions are very sensitive, almost certainly moresensitive than histopathological assessment. Nevertheless, ourfindings are consistent with previous reports showing the pres-ence of genomic abnormalities in premalignant lesions beforethey are morphologically recognized (22, 23) and the concept offield cancerization, by which early genetic events are shared bycells in a local anatomical area before they appear to be mor-phologically neoplastic (24).

Of the 13 patients providing marginal epithelia, spectra ofincreasing intensities were noted in 3 patients (patients 1, 2, and5; 23%). The intensities of the amplified bands in a fingerprint-ing profile are semiquantitative, and the intensity of an ampli-fied band is proportional to the concentration of its correspond-

ing template sequence (12). For patients 1 and 5, the extra bandswere detected in their marginal epithelia, and the intensities ofthe bands were further increased in their tumor samples (Table1), indicating further genomic amplification in tumors. It isworth noting that the marginal epithelia of patients 1 and 5 werehistologically normal and dysplastic, respectively. Thus, theobserved genetic alterations preceded even the appearance ofpremalignant lesions in patient 1. For patient 2, the spectrum ofincreasing intensity was observed in the normal, marginal, andtumor samples. Because squamous cell carcinoma was alsopresent in the marginal epithelium of patient 2 (Table 1), such anincreased band intensity in the tumor profile might be related toeither an increased amount of tumor genomes in the tumorsample or an increase in the copy number of certain genomicsequences.

The finding of genetic alterations in all of the ESCC tumorsand 62% of epithelia surrounding ESCC derives from the abilityof the ISSR-PCR method to detect abnormalities in one reaction,throughout the genome, and in the absence of any informationabout loci-specific sequences. Moreover, the comparative DNAprofiling analysis appears to be more sensitive than histologicalassessments in showing early genetic alterations in preneoplas-tic epithelia because five of eight (63%) samples of the marginalepithelia that matched tumor abnormalities are histologicallynormal. The sensitivity and reproducibility of the method sug-gest the application of ISSR-PCR for early detection of ESCCand also for the detection of early tumor recurrence. Anotherinteresting means to test the usefulness of ISSR-PCR for earlydetection of cancer would be to investigate Barrett’s esophagealtissues, Barrett’s dysplasia, and adenocarcinomas (25).

As shown in the present study, the ISSR-PCR method alsooffers an opportunity to identify the genomic fragments thatshow amplifications or deletions in ESCC. Several amplifiedchromosomal regions and genes in ESCC have been implicatedin the development and progression of ESCC, such as chromo-somal regions 11q13 (5) and 13q34 (26),c-myconcogene, andepidermal growth factor receptor genes (27). The overexpres-sions of these genes have also been suggested to be involved intumorigenesis (4). In the present study, the detections of ampli-fied exonic fragments of 7q22 from patient 19 with a highfrequency (43%) are novel to ESCC. The chromosomal over-

Table 3 Frequencies of the specific gene alterations detected by direct genomic PCR for the 11 ISSR-PCR profile bands in ESCCs

ISSR-PCR profile bands Patient no.ISSR-PCR primer

sequences usedSizes of direct

PCR products (bp)

No. of ESCCs with gainsor losses of the

profile band (n5 60)

A 6 (GACA)4 520 4 (7)a

Bb 11 (GTG)5 580 4 (7)C 19 (GTG)5 270 26 (43)D 11 (GACA)4 380 5 (8)E 13 (GACA)4 353 4 (7)F 16 (GATA)4 989 7 (11)G 7 (GATA)4 341 8 (13)H 10 (GTG)5 626 5 (8)I 5 (GTG)5 477 9 (15)J 14 (GACA)4 469 4 (7)K 17 (GATA)4 954 4 (7)

a Data are numbers, with percentages in parentheses.b The profile band showed loss, whereas all other profile bands showed gains in tumor compared with corresponding normal epithelia.

1544Genetic Alterations in Esophageal Cancer

Research. on August 23, 2021. © 2001 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

representation of 7q22 has also been shown by comparativegenomic hybridization to be involved in pancreatic carcinomas(28) and nonpapillary renal cell carcinomas (29). The involve-ment of exonic amplification in carcinogenesis of ESCC iscurrently being investigated and will be reported later.

In summary, here we describe our results of the applicationof ISSR-PCR to detect somatic mutations in different tissuesfrom ESCC patients including tumor and marginal epithelia incomparison with the corresponding normal epithelium, and weshow that ISSR-PCR profiling offers a powerful approach toidentify novel genomic fragments that may be involved in thecarcinogenesis of esophageal cancers. Analysis of 19 ESCCs bythis technique detected frequent changes in the intensities of theprofile bands not only in tumor samples but also in apparentlynormal marginal epithelia. From these data, it is concluded that(a) widespread genetic alterations can be an early event in tumorprogression in ESCC, and these alterations can be detected byISSR-PCR in tumors and premalignant stages; and (b) theobserved genetic alterations can be detected before the appear-ance of premalignant lesions and may provide the basis for ageneral screening method for predisposition to the developmentof patent disease.

References1. Nowell, P. C. Mechanisms of tumor progression. Cancer Res.,46:2203–2207, 1986.2. Tang, J. C. O., Ho, F. C. S., Chan, A. C. L., Chow, E. Y. W., andSrivastava, G. Progression of spontaneous lymphomas in SJL mice:monitoring in vivo clonal evolution with molecular markers in sequen-tial splenic samples. Lab. Investig.,78: 1459–1466, 1998.3. Ionov, Y., Peinado, M. A., Malkhosyan, S., Shibata, D., and Perucho,M. Ubiquitous somatic mutations in simple repeated sequences reveal anew mechanism for colonic carcinogenesis. Nature (Lond.),363: 558–561, 1993.4. Tahara, E. Genetic alterations in human gastrointestinal cancers: theapplication to molecular diagnosis. Cancer (Phila.),75: 1410–1417,1995.5. Schuuring, E. The involvement of the chromosome 11q13 region inhuman malignancies. Cyclin D1 and EMS1 are 2 new candidate onco-genes: a review. Gene (Amst.),159: 83–96, 1995.6. Nowell, P. C. Cytogenetic approaches to human cancer genes.FASEB J.,8: 408–413, 1994.7. Maelandsmo, G. M., Berner, J. M., Florenes, V. A., Forus, A., Hovig,E., Fodstad, O., and Myklebost, O. Homozygous deletion frequency andexpression levels of the cdkn2 gene in human sarcomas: relationship toamplification and messenger RNA levels of CDK4 and CCND1. Br. J.Cancer,72: 393–398, 1995.8. Nawroz, H., Koch, W., Anker, P., Stroun, M., and Sidransky, D.Microsatellite alterations in serum DNA of head and neck cancer pa-tients. Nat. Med.,2: 1035–1037, 1996.9. Peinado, M. A., Malkhosyan, S., Velazquez, A., and Perucho, M.Isolation and characterization of allelic losses and gains in colorectaltumors by arbitrarily primed polymerase chain reaction. Proc. Natl.Acad. Sci. USA,89: 10065–10069, 1992.10. Sood, A. K., and Buller, R. E. Genomic instability in ovariancancer: a reassessment using an arbitrarily primed polymerase chainreaction. Oncogene,13: 2499–2504, 1996.11. De Juan, C., Iniesta, P., Cruces, J., Sanchez, A., Massa, M. J.,Gonzalez-Quevedo, R., Torres, A. J., Balibrea, J. L., and Benito, M.DNA amplification on chromosome 6p12 in non-small cell lung cancerdetected by arbitrarily primed polymerase chain reaction. Int. J. Cancer,84: 344–349, 1999.

12. Basik, M., Stoler, D. L., Kontzoglou, K. C., Rodriguez Bigas,M. A., Petrelli, N. J., and Anderson, G. R. Genomic instability insporadic colorectal cancer quantitated by inter-simple sequence repeatPCR analysis. Genes Chromosomes Cancer,18: 19–29, 1997.

13. Perring, T. M., Cooper, A. D., Rodriguez, R. J., Farrar, C. A., andBellows, T. S. Identification of a whitefly species by genomic andbehavioral studies. Science (Wash. DC),259: 74–77, 1993.

14. Weising, K., Atkinson, R. G., and Gardner, R. C. Genomic finger-printing by microsatellite-primed PCR: a critical evaluation. PCR Meth-ods Applications,4: 249–255, 1995.

15. Lam, K. Y., Law, S., Ma, L. T., Ong, S. K., and Wong, J. Pre-operative chemotherapy for squamous cell carcinoma of the oesopha-gus: do histological assessment and p53 overexpression predict chemo-responsiveness? Eur. J. Cancer,33: 1221–1225, 1997.

16. Tang, J. C. O., Ho, F. C. S., Chan, A. C. L., and Srivastava, G.Clonality of lymphomas at multiple sites in SJL mice. Lab. Investig.,78:205–212, 1998.

17. Greer, C. E., Peterson, S. L., Kiviat, N. B., and Manos, M. M. PCRamplification from paraffin-embedded tissues: effects of fixative andfixation time. Am. J. Clin. Pathol.,95: 117–124, 1991.

18. Nurnberg, P., Roewer, L., Neitzel, H., Sperling, K., Popperl, A.,Hundrieser, J., Poche, H., Epplen, C., Zischler, H., and Epplen, J. T.DNA fingerprinting with the oligonucleotide probe (CAC)5/(GTG)5:somatic stability and germline mutations. Hum. Genet.,84: 75–78,1989.

19. Epplen, J. T. On simple repeated GA-C-T-A sequences in animalgenomes: a critical reappraisal. J. Hered.,79: 409–417, 1988.

20. Schafer, R., Ali, S., and Epplen, J. T. The organization of theevolutionarily conserved GATA/GACA repeats in the mouse genome.Chromosoma (Berl.),93: 502–510, 1986.

21. Zischler, H., Nanda, I., Schafer, R., Schmid, M., and Epplen, J. T.Digoxigenated oligonucleotide probes specific for simple repeats inDNA fingerprinting and hybridizationin situ. Hum. Genet.,82: 227–233, 1989.

22. Shimada, M., Yanagisawa, A., Kato, Y., Inoue, M., Shiozaki, H.,Monden, M., and Nakamura, Y. Genetic mechanisms in esophagealcarcinogenesis: frequent deletion of 3p and 17p in premalignant lesions.Genes Chromosomes Cancer,15: 165–169, 1996.

23. O’Connell, P., Pekkel, V., Fuqua, S. W., Osborne, C. K., Clark,G. M., and Allred, D. C. Analysis of loss of heterozygosity in 399premalignant breast lesions at 15 genetic loci. J. Natl. Cancer Inst.(Bethesda),90: 697–703, 1998.

24. Califano, J., van der Riet, P., Westra, W., Nawroz, H., Clayman, G.,Piantadosi, S., Corio, R., Lee, D., Greenberg, B., Koch, W., and Sidran-sky, D. Genetic progression model for head and neck cancer: implica-tions for field cancerization. Cancer Res.,56: 2488–2492, 1996.

25. Reid, B. J. Barrett’s esophagus and esophageal adenocarcinoma.Gastroenterol. Clin. N. Am.,20: 817–834, 1991.

26. Shinomiya, T., Mori, T., Ariyama, Y., Sakabe, T., Fukuda, Y.,Murakami, Y., Nakamura, Y., and Inazawa, J. Comparative genomichybridization of squamous cell carcinoma of the esophagus: the possibleinvolvement of theDPI gene in the 13q34 amplicon. Genes Chromo-somes Cancer,24: 337–344, 1999.

27. Lu, S. H., Hsieh, L. L., Luo, F. C., and Weinstein, I. B. Amplifi-cation of the EGF receptor and c-myc genes in human esophagealcancers. Int. J. Cancer,42: 502–505, 1988.

28. Solinas Toldo, S., Wallrapp, C., Muller Pillasch, F., Bentz, M.,Gress, T., and Lichter, P. Mapping of chromosomal imbalances inpancreatic carcinoma by comparative genomic hybridization. CancerRes.,56: 3803–3807, 1996.

29. Yang, Z. Q., Yoshida, M. A., Fukuda, Y., Kurihara, N., Nakamura,Y., and Inazawa, J. Molecular cytogenetic analysis of 17 renal cancercell lines: increased copy number at 5q31–33 in cell lines from non-papillary carcinomas. Jpn. J. Cancer Res.,91: 156–163, 2000.

1545Clinical Cancer Research

Research. on August 23, 2021. © 2001 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

2001;7:1539-1545. Clin Cancer Res   Johnny C. O. Tang, King Y. Lam, Simon Law, et al.   DNA Fingerprinting Using Inter-Simple Sequence Repeat PCRCarcinomas and Adjacent Normal Epithelia by Comparative Detection of Genetic Alterations in Esophageal Squamous Cell

  Updated version

  http://clincancerres.aacrjournals.org/content/7/6/1539

Access the most recent version of this article at:

  Cited articles

  http://clincancerres.aacrjournals.org/content/7/6/1539.full#ref-list-1

This article cites 28 articles, 5 of which you can access for free at:

  Citing articles

  http://clincancerres.aacrjournals.org/content/7/6/1539.full#related-urls

This article has been cited by 3 HighWire-hosted articles. Access the articles at:

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  [email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/7/6/1539To request permission to re-use all or part of this article, use this link

Research. on August 23, 2021. © 2001 American Association for Cancerclincancerres.aacrjournals.org Downloaded from