Deletion mapping using quantitative real-time PCR identifies two ...

Oncogene (2003) 22, 2984–2992

& 2003 Nature Publishing Group All rights reserved 0950-9232/03 $25.00 www.nature.com/onc

Deletion mapping using quantitative real-time PCR identifies two distinct 3p21.3 regions affected in most cervical carcinomas Vera Senchenko1,2,7, Jian Liu1,7, Eleonora Braga1,3, Natalia Mazurenko4, Witaly Loginov3, Yury Seryogin3, Igor Bazov1,3, Alexei Protopopov1, Fedor L Kisseljov4, Vladimir Kashuba1, Michael I Lerman5, George Klein1 and Eugene R Zabarovsky*,1,6

ONCOGENOMICS

1

Microbiology and Tumor Biology Center, Center for Genomics and Bioinformatics, Karolinska Institute, Stockholm, 17177 Sweden; Center ‘Bioengineering’, Russian Academy of Sciences, Moscow, 117312 Russia; 3Russian State Genetics Center, Moscow, 113545 Russia; 4Blokhin Cancer Research Center, Russian Academy of Medical Sciences, Moscow, 115478 Russia; 5Cancer-Causing Genes Section, Laboratory of Immunobiology, Center for Cancer Research, National Cancer Institute, Frederick, MD 21702, USA; 6 Engelhardt Institute of Molecular Biology, Russian Academy of Sciences, Moscow 119991, Russia 2

We report chromosome 3p deletion mapping of 32 cervical carcinoma (CC) biopsies using 26 microsatellite markers located in frequently deleted 3p regions to detect loss of heterozygosity and homozygous loss. In addition, two STS markers (NLJ-003 and NL3-001) located in the 3p21.3 telomeric (3p21.3T) and 3p21.3 centromeric (3p21.3C) regions, respectively, were used for quantitative real-time PCR as TaqMan probes. We show that quantitative real-time PCR is reliable and sensitive and allows discriminating between 0, 1 and 2 marker copies per human genome. For the first time, frequent (five of 32 cases, i.e. 15.6%) homozygous deletions were demonstrated in CCs in both 3p21.3T and 3p21.3C regions. The smallest region homozygously deleted in 3p21.3C was located between D3S1568 (CACNA2D2 gene) and D3S4604 (SEMA3F gene) and contains 17 genes previously defined as lung cancer candidate Tumor suppressor genes (TSG(s)). The smallest region homozygously deleted in 3p21.3T was flanked by D3S1298 and NL1-024 (D3S4285), excluding DLEC1 and MYD88 as candidate TSGs involved in cervical carcinogenesis. Overall, this region contains five potential candidates, namely GOLGA4, APRG1, ITGA9, HYA22 and VILL, which need to be analysed. The data showed that aberrations of either NLJ-003 or NL3-001 were detected in 29 cases (90.6%) and most likely have a synergistic effect (Po0.01). The study also demonstrated that aberrations in 3p21.3 were complex and in addition to deletions, may involve gene amplification as well. The results strongly suggest that 3p21.3T and 3p21.3C regions harbor genes involved in the origin and/or development of CCs and imply that those genes might be multiple TSG(s). Oncogene (2003) 22, 2984–2992. doi:10.1038/sj.onc.1206429

*Correspondence: ER Zabarovsky, Microbiology and Tumor Biology Center, Center for Genomics and Bioinformatics, Karolinska Institute, Stockholm, 17177 Sweden; E-mail: [email protected]; [email protected] 7 These authors contributed equally to this work Received 22 July 2002; revised 23 January 2003; accepted 28 January 2003

Keywords: quantitative real-time PCR; human chromosome 3p; tumor suppressor genes; NotI linking clone; Loss of heterozygosity; cervical carcinoma

Introduction Tumor suppressor genes (TSG(s)) represent one of the main classes of cancer-associated genes and their identification constitutes one of the major efforts in cancer research today. Deletions in chromosome 3p are frequently associated with different epithelial tumors, and this chromosome is harboring several TSG(s) (Zabarovsky et al., 2002). Many studies have shown abnormalities on the short arm of this chromosome in carcinomas of the kidney, lung, breast, ovary, cervix, testis, head and neck and other (Kok et al., 1997; Braga et al., 1999; Lazo, 1999). However, the search for resident TSG(s) was hampered by the large size of the region that covers practically the whole chromosome 3p (about 100 Mb). It is still not clear which TSG(s) located on chromosome 3 are cancer type specific or common to different epithelial tumors. According to several studies, the most frequently affected regions (FAR) in sporadic renal cell carcinoma (RCC) and small cell lung carcinoma (SCLC) are 3p13– p14 and 3p21.2–3p21.3 (Kok et al, 1997; Van den Berg and Buys, 1997; Alimov et al., 2000; Zabarovsky et al., 2002). Recently, using deletion mapping we have shown that the 3p21.3 region is the most frequently deleted region not only in RCC and SCLC but also in nonsmall cell lung (NSCLC), breast and cervical carcinomas (CCs). Moreover, we have shown that this region could be subdivided into centromeric 3p21.3 (3p21.3C) and telomeric 3p21.3–p22 (3p21.3T) subparts (Alimov et al., 2000; Lerman et al., 2000; Braga et al., 2002). Others and we identified several candidate TSG(s) from the centromeric 3p21.3 region, however, no clear candidate TSG(s) from the telomeric 3p21.3 part have been isolated yet (Zabarovsky et al., 2002). Importantly, both telomeric and centromeric 3p21.3 critical regions

3p deletion hot spots in cervical carcinomas V Senchenko et al

2985

contain several genes that justified further fine deletion mapping of these regions. Loss of heterozygosity (LOH) is frequently used as an indicator of genetic losses associated with tumor development, and the microsatellite repeat analysis is often the method of choice for LOH detection. LOH studies, however, have several disadvantages discussed previously (Liu et al., 1999; Zabarovsky et al., 2002). Real-time PCR using TaqMan probes is a well-known, precise, and reproducible quantitative nucleic acid assay with a high throughput and large dynamic range of applications (Livak et al., 1995). The method overcomes some limitations of LOH analysis, for example, it does not require polymorphism, so any unique marker can be used. In contrast to LOH with microsatellite markers that detect only allelic imbalance without discrimination between deletions and amplifications, real-time PCR yields information about copy number changes. In this study, real-time PCR was used to assess the presence and frequency of homozygous deletions in the centromeric and telomeric 3p21.3 regions.

Materials and methods DNA samples and Southern blot analysis In all, 32 cervical squamous CCs were studied. For histological verification, top and bottom sections (5-mm thick) cut from frozen tumor tissues were examined after staining with hematoxylin and eosin. Selected samples containing 60% or more tumor cells, and matched normal tissues were stored at
701C. High-molecular-weight DNA was isolated with guanidinium thiocyanate followed by centrifugation through a CsCl cushion (Samoylova et al., 1995). Female and male DNA samples for testing X chromosome copy numbers were obtained from the blood of healthy individuals. High-molecular-weight DNA was isolated by overnight treatment with proteinase K at 501C followed by phenol/chloroform extraction and precipitation with ethanol. The DNA preparations were examined by electrophoresis using 0.8% agarose gels. Chromosome 3 specific NotI linking clones NLJ-003/ D3S1642 and NL3-001/D3S3874 (mapped to 3p21.3T and to 3p21.3C, respectively) were described previously (Kashuba et al., 1999). Southern blotting and hybridization were performed according to the standard procedures. Band intensities were determined using a Molecular Dynamics Personal Densitometer SI (Sunnyvale, CA, USA) according to the manufacturer’s protocol. NotI linking clone NL1-290/D3S4293 mapped to the 3q13.3–q21 (Kashuba et al., 1999) was used as a reference control because this region usually does not reveal frequent allelic imbalances in CCs (Rader et al., 1996; Lazo, 1999; Matthews et al., 2000). The ratio of band intensities showing copy number changes of the probe NLJ003 was calculated according to the formula: PT  RN/ RT  PN, where PT and PN were band intensities for the NLJ-003 bands, and RT and RN represented band intensities of NL1-290 in tumor and normal samples, respectively. Real-time quantitative PCR Selection of primer and TaqMan probe sequences for NLJ003/D3S1642 and NL3-001/D3S3874 was performed using the

ABI Primer Express Software (version 1.5). They were as follows: NLJ-003 forward 50 -CAG AGT GCG TGT GCC GAC T-30 , reverse 50 -ACA ACT TCT CTG CGG GCG T-30 and probe 50 -CTG GCG GAG AGA CTG GGA GCG A-30 (125 bp amplicon); NL3-001 forward 50 -CTT GCC ATC TGC AAT TCC CT-30 , reverse 50 -CTC CAT GAG GCT GTG GGA AG-30 and probe 50 -CCC CAG AAA CGC GCG GGC30 (60 bp amplicon). The sequences of primers and TaqMan probes for phospho-fructo-2 kinase (PF2K) gene were: forward 50 -ATG CCC TGG CCA ACT CA-30 , reverse 50 TGC GAC TGG TCC ACA CCT T-30 and probe 50 -FAMTCA GTC CCA GGG CAT CAG CTC CC-TAMRA-30 (Boulay et al., 1999). Beta-actin gene (ACTB) was used as a reference (Applied Biosystems, Foster City, CA, USA). NLJ003, PF2K and ACTB probes were labeled with FAM (6carboxy-fluoroscein) and NL3-001 contained JOE (2,7-dimethoxy-4,5-dichloro-6-carboxy-fluoroscein) as reporter dyes, located at the 50 -ends. All of the reporters were quenched by TAMRA (6-carboxy-N,N,N0 N0 -tetramethyl-rhodamine), conjugated to their 30 -terminal nucleotides. All probes and primers were purchased from Applied Biosystems. PCRs were carried out in triplicate in 25 ml consisting of 1  PCR buffer A (Applied Biosystems), 3.5 mm MgCl2, 0.2 mm dATP, dGTP, dCTP, and 0.4 mm dUTP, 100 nm TaqMan probe, forward and reverse primers in appropriate concentrations (150–200 nm), 0.025 U/ml Taq Gold DNA polymerase (Applied Biosystems), 0.01 U/ml AmpErase and 5 ml of DNA template (20–60 ng). PCRs were carried out according to the thermal profile: 2 min at 501C, 10 min at 951C, followed by 40 cycles of 15 s at 951C and 1 min at 601C. Comparative CT method (DDCT method) was used for quantification of marker copy numbers. The parameter CT (threshold cycle) is defined as the cycle number required for reporter dye fluorescence to become higher than background fluorescence level and automatically determined using ABI Prisms Model 7700 Sequence Detector (Applied Biosystems). The method is based on the inverse exponential relation that exists between the initial quantity (copy number) of target sequence copies in the reactions and the corresponding CT determinations – the higher the starting copy number of DNA target, the lesser the CT value. This method was used to determine target sequence copy number in tumor DNA sample relative to the normal DNA from the same patient (calibrator) and relative to an endogenous control sequence (reference) – ACTB in both samples. The starting relative copy number DNA (at each locus in tumor sample) is given by 2
DDCT, where DDCT ¼ DCTtumor
DCTnormal and each DCT ¼ CTtarget
CTreference. Data were analysed using ABI Prisms 7700 Sequence Detection System software (version 1.7). The range given for the probes was determined as 2
DDCT with DDCT +s and DDCT
s, where s ¼ the standard deviation of the DDCT value. The absence of nonspecific amplification was confirmed by analysing the PCR amplification by 2.5% agarose gel electrophoresis and ethidium bromide staining. For the DDCT calculation to be valid, the efficiency of the target amplification and the efficiency of the reference amplification must be approximately equal. Before using the DDCT method for quantitative assessment, a validation experiment was performed. The validation experiments (Figure 1) demonstrated that efficiencies of the targets and reference are approximately equal for the chosen dilutions. The absolute value of the slope of log input amount vs DCT should be o0.1. For the NL3-001, the slope value is 0.0848, for NLJ-003 it is 0.0133 and for PF2K it is –0.0577. Thus, the DDCT calculations can be used for the relative quantification of targets without using standard curves. Oncogene

3p deletion hot spots in cervical carcinomas V Senchenko et al

2986 Polymorphic and STS markers and microsatellite analysis In all, 26 polymorphic and two STS markers of 3p were applied. A total of 23 microsatellite markers were selected from genome- (GDB) and CHLC databases. Three polymorphic (NL1-024/D3S4285, D3S4604 and D3S4597) and two STS (NLJ-003/D3S1642 and NL3-001/D3S3874) markers were designed and located earlier (Braga et al., 1997; Kashuba et al., 1999; Wistuba et al., 2000). The order of all 28 selected markers according to the location database (http://www.cedar.genetics.soton.ac.uk/pub/chrom3/gmap), NCBI human GDB (http://www.ncbi.nlm.nih.gov/genome/guide/human) and our own data was as follows: 3p-tel-D3S2405/GGAT2A11D3S1317-D3S1038-D3S1286-D3S3047/GATA85F02-D3S1283D3S2420 / ATA25A07-NL1-024/D3S4285-NLJ-003/D3S1642D3S1298-D3S3527-p33715/GAAT12D11-D3S2456/GATA63 E04-D3S1767/GATA7A01-D3S2409/ATA10H11-D3 S3615D3S1573-D3S4597-D3S4604-NL3-001/D3S3874-D3S1568-D3 S3667-D3S1766/GATA6F06-D3S1300-D3S1481-D3S1285-D3 S2454/GATA52H09-D3S2406/GGAT2G03-3p-cen. The PCR primers were purchased from Life Technologies/ Invitrogen (Carlsbad, CA, USA). The PCR and PAG electrophoresis was carried out as described earlier (Braga et al., 1999). Band intensities for PCR products of high- and lowmolecular-weight alleles (H- and L alleles) were compared using a Molecular Dynamics Personal Densitometer SI. The LOH was scored when a ratio of intensities for matched tumor and normal DNAs differed twofold or more.

Results Testing sensitivity of real-time PCR using TaqMan probes Figure 1 Relative efficiency plot of log input amount vs DCT for NL3-001 (a), NLJ-003 (b) and ACTB (c). For detailed description see Applied Biosystems guidelines. The absolute value of the slopes for all three probes is less than 0.1

Comparative CT method can be used in separate tubes and in the same tube. Control experiments (data not shown) demonstrated that both methods gave the same results. Quantification of NLJ-003 and PF2K was carried out in separate tubes and quantification of NL3-001 was carried out in the same tube. For an accurate CT quantification using multiplex PCR in the same tube, it is important that two independent reactions do not compete. The primer limitation experiments were performed using a matrix of forward and reverse primer concentrations (15–100 nm). The chosen concentrations were those that have shown a reduction in DRn but little effect on CT. The primer and probe concentrations for multiplex PCRs in the same tubes were found: for NL3-001– 20 nm (probe), 90 nm (primers) and ACTB
20 nm (probe), 60 nm (primers). Sample sex determination using of PF2K gene dosage in X chromosome Real-time PCRs were performed on the 15 male samples and 15 female samples using two different sets of primers, one set for X-linked housekeeping gene PF2K (Boulay et al., 1999), and another for ACTB gene used as a reference. The reactions were carried out in triplicate for each sample. For each individual, CT value obtained for ACTB was subtracted from that of PF2K, thus defining DCT. Oncogene

In previous experiments using LOH deletion mapping, we have shown that regions surrounding NLJ-003 and NL3-001 are the most FARS in RCC (see Figure 2). It was also shown that these clones were homozygously deleted in SCLC and breast cancers (Murata et al., 1994; Wei et al., 1996; Sekido et al., 1998; Alimov et al., 2000; Lerman et al., 2000). Recently, the same was demonstrated for NSCLC, breast and cervical cancers (Braga et al., 2002). The data strongly suggest that both regions (telomeric 3p21.3T and centromeric 3p21.3C) contain multiple tumor suppressor genes. The focus of the present study was to compare LOH data with another independent deletion mapping method and to check whether these regions are also homozygously deleted in CCs. Comparative genomic hybridization (CGH) data could verify and validate the LOH mapping (Alimov et al., 2000); however, the resolution of this method is rather low. We decided to use a much more sensitive, rapid and quantitative method, termed real-time PCR (Heid et al., 1996) to assay genetic changes in tumorigenesis. This technology does not require polymorphic markers, and any marker is informative for any cancer case. Real-time PCR permits to evaluate and compare a set of samples at different loci to identify deletions, retentions and amplifications. As described in Material and methods, all control experiments demonstrated that primers and probes designed for the NLJ-003 and NL3-001 loci were appropriate for the comparative CT method.

3p deletion hot spots in cervical carcinomas V Senchenko et al

2987

approach permits detection of the difference in X chromosome copy number between males (one copy) and females (two copies). Analysis of two critical regions in 3p21.3

Figure 2 Schematic maps of 3p21.3C and 3p21.3T regions. For homozygous deletions found in SCLC cell lines (Wei et al., 1996; Ishikawa et al., 1997; Lerman et al., 2000; Protopopov et al., 2003) and for flanking regions, the genes are represented by pointed arrows, indicating the orientations of transcription. Physical position from 3p telomere is according to http://www.ncbi.nlm.nih.gov/cgi-bin/Entrez/hum_srch?chr ¼ hum_chr.inf&query (build 30)

Before doing deletion mapping of cervical carcinomas, we tested real-time PCR with human X chromosome, thus requiring differentiation between one (male) and two (female) copies per genome. The DCT values obtained for 30 samples indicated that samples could be divided into two nonoverlapping groups. The DCT value for one group was 6.0170.40 (av.7s.d.) and it was approximately one unit higher than the DCT for another group: 4.9170.44. The difference of one unit reflects a twofold value of initial gene copy number for females compared to males. The amount of target (PF2K) for females, normalized to a reference (ACTB) and relative to target (PF2K) for males was: N Female =N Male ¼ 2
DDCT ¼ 2.14(2.08–2.20). Thus, this

To test two critical sites in 3p for copy number changes and homozygous deletions, two TaqMan probes were designed: the NLJ-003 (3p21.3T), residing in the AP20 homozygous deletions site (Kashuba et al., 1995; Ishikawa et al., 1997; Protopopov et al., 2003) and NL3-001 (3p21.3C), mapped to the LUCA homozygous deletion site (Wei et al., 1996; Lerman et al., 2000, see Figure 2). Frequent aberrations were shown in these regions in previous studies (Alimov et al., 2000; Wistuba et al., 2000; Braga et al., 2002). The results of the experiments are shown in Table 1. According to histology analysis, contamination of tumor samples with normal stroma and lymphocytes can reach up to 30–40%. Therefore, alleles were taken as homozygously deleted if the highest value of standard deviation is below 0.5 and hemizygously deleted if this value is below 1.0. An allele was considered as amplified/ multiplied, if the lowest value of standard deviation is over 1.0. In this case, the results of quantitative PCR analysis suggest that seven samples (21.8%) contained amplified NLJ-003 marker and nine samples (28.1%) had amplified NL3-001. Homozygous deletions were detected in five cases (15.6%) for both markers, of which four cases were in the same tumor samples. Frequencies of simultaneous homozygous deletions in both loci were highly significant (P ¼ 0.0009). Therefore, some aberrations in 3p21.3T and 3p21.3C often occurred in the same tumor. This might mean that TSG(s) in these regions can have a synergistic effect. Hemizygous deletions of NLJ-003 were found in 14 samples (43.8%) and of NL3-001 in 10 (31.3%). Altogether, these data showed that copy number changes (homo- and hemizygous deletions and amplifications) of either NLJ3-001 or NL3-001 were detected in 29 cases (90.6%). This is significantly higher than it was reported earlier for the whole 3p – around or some more than 70% (Larson et al., 1997a; Wistuba et al., 1997; Braga et al., 1999; 2002; Helland et al., 2000). These results lend strong support to the conclusion that these two regions are really hot spots for rearrangements in CCs. Only in three cases (9.4%), no aberrations were detected in either of these loci. This study has highlighted the complex character of chromosomal rearrangements in 3p during development of CC. In addition to homo- and hemizygous deletions, duplication/amplification of the whole chromosome 3 or some regions was also detected. One tentative explanation of these phenomena might be that inactivated TSG(s) were amplified together with neighboring oncogenes. To confirm that allelic imbalance could be because of the amplification, we performed additional experiments with five RCC cell lines showing LOH and considered to have 3p deletions (Alimov et al., 2000). However, Oncogene

3p deletion hot spots in cervical carcinomas V Senchenko et al

2988 Table 1

Results of real-time PCR for cervical cancer samples b

Relative DNA copy number a

Sample no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

NLJ-003 (3p21.3) 1.34 1.16 0.98 1.20 2.75 0.61 1.56 1.10 0.36 0.78 0.80 2.39 0.79 0.38 0.66 0.61 0.70 0.17 0.15 0.78 0.44 1.37 2.74 0.98 0.50 0.20 0.61 0.88 1.08 2.79 0.48 0.57

(1.24–1.45) (0.90–1.42) (0.91–1.06) (1.13–1.27) (2.46–3.04) (0.50–0.71) (1.32–1.80) (0.97–1.23) (0.28–0.43) (0.66–0.90) (0.61–1.00) (2.19–2.59) (0.66–0.92) (0.30–0.46) (0.61–0.71) (0.51–0.71) (0.56–0.85) (0.13–0.22) (0.11–0.19) (0.68–0.89) (0.38–0.51) (1.25–1.50) (2.66–2.82) (0.94–1.03) (0.40–0.61) (0.19–0.22) (0.55–0.66) (0.87–0.89) (0.98–1.14) (2.44–3.13) (0.39–0.58) (0.54–0.61)

NL3-001 (3p21.2–21.3) 0.51 0.71 1.04 1.43 1.65 2.85 1.15 1.09 0.18 1.03 1.06 2.55 1.45 0.49 0.50 0.63 0.57 0.06 0.04 1.51 0.31 0.62 2.80 1.26 2.03 0.10 1.07 1.54 0.60 1.05 1.14 0.80

(0.34–0.67) (0.54–0.89) (0.99–1.09) (1.19–1.67) (1.53–1.77) (2.48–3.22) (0.99–1.31) (0.83–1.35) (0.10–0.26) (0.98–1.08) (0.97–1.15) (2.50–2.60) (1.37–1.53) (0.43–0.55) (0.42–0.58) (0.52–0.75) (0.52–0.63) (0.04–0.08) (0.03–0.05) (1.19–1.82) (0.26–0.36) (0.58–0.66) (2.38–3.21) (1.00–1.51) (1.82–2.24) (0.09–0.11) (0.98–1.17) (1.08–2.01) (0.53–0.67) (1.00–1.09) (0.99–1.29) (0.65–0.95)

a Gray samples did not show any aberrations and bold contained homozygous deletions in one or both of the loci bSamples showed multiplication are in bold italics and samples with white letters contained hemizygous deletions

cytogenetic analysis demonstrated that among them only A498 had two copies of chromosome 3 and four lines had from three to five copies (KRC/Y, HN51, CAKI1 and TK164). Nevertheless, all these five lines revealed only one marker in LOH analysis arguing that one 3p arm was lost. These data proved that amplification associated with 3p deletion is a common event in cancer cells. Real-time PCR data for NLJ-003 and NL3-001 loci were compared with allelic alterations observed for neighboring polymorphic markers (Figure 3). In all, 20 cancer samples were analysed in this work. Significant correlation between real-time PCR and LOH data was observed in both loci for many cancer samples. For example, relative copy number values for NLJ-003 locus in CC samples #6, #17 and #25 suggested allelic imbalance, and flanking microsatellite markers supported this finding. In cases #2, #8 and #24 quantitative real-time PCR suggested for NLJ-003 retention and this result coincided with the LOH in neighboring polymorphic markers. However, in other cases discordance was apparent. For instance, NLJ-003 demonstrated hemizygous deleOncogene

tions in cases #10 and #15, but neighboring markers showed retention. This disagreement can be explained by different factors. On the one hand, it can mean that NLJ-003 is a hot spot for the rearrangements and flanking regions were unaffected; on the other, it is clear that not every copy number change will result in allelic imbalance. Moreover, application of real-time PCR allowed detecting homozygous deletions that is very difficult with microsatellite markers. From the data shown in Figure 3, it is possible to conclude that the smallest homozygously deleted region in 3p21.3T is bordered with NL1-024 (D3S4285) and D3S1298. Additional data with primers D3S2968, D3S4597 and D3S4604 (not shown) suggested that the similar region in 3p21.3C is limited with D3S4604 and D3S1568. It is important to mention that these two regions overlap with previously described homozygous deletions in lung and breast carcinomas (Murata et al., 1994; Wei et al., 1996; Sekido et al., 1998; Lerman et al., 2000). Owing to the limited amount of tumor material, only some samples could be tested by Southern hybridization. Results for the available cases are shown in Figure 4. This figure clearly showed that the realtime PCR data were consistent with the Southern data. Discussion According to LOH and CGH data (Kisseljov et al., 1996; Rader et al., 1996, 1998; Lazo, 1999; Mazurenko et al., 1999; Matthews et al., 2000) chromosome arms 3p, 6p and 11q were most frequently affected in squamous CC of the uterine cervix. Moreover, 3p and 6p were also involved at the early precancer stages and during progression of cervical intraepithelial neoplasia to invasive cervical cancer (Larson et al., 1997b; Wistuba et al., 1997; Fouret et al., 1998; Guo et al., 1998, 2000, 2001; Umayahara et al., 2002). Different studies reported various 3p LOH rates in the range 50– 76% (Wistuba et al., 1997; Braga et al., 1999; Guo et al., 2000; Helland et al., 2000; Herzog et al., 2001; Acevedo et al., 2002). Several reports suggested that the most frequently affected in CC regions were located in 3p21.3–p22 region and near FHIT gene in 3p14.2 (Wistuba et al., 1997; Muller et al., 1998; Helland et al., 2000; Herzog et al., 2001; Acevedo et al., 2002). We have found recently that two regions in 3p21.3 (centromeric – C or LUCA and telomeric – T or AP20) were frequently deleted in major human malignancies (Braga et al., 2002). Homozygous deletions are excellent indicators of the locations of TSG(s); however, microsatellite deletion mapping is not very well suited for the detection of such deletions because of normal cell contamination. Balance between two alleles does not change in such cases and deletion would go undetected. In fact, this can lead to paradoxical results when LOH in homozygously deleted regions may be less frequent than in surrounding regions. Furthermore, this approach in fact detects only allelic imbalance that could result from multiplication as well as

3p deletion hot spots in cervical carcinomas V Senchenko et al

2989

Figure 3 Combination of LOH and real-time PCR data for 3p deletion mapping. Black squares represent LOH and homozygous deletions; white squares, retention of heterozygosity and normal copy number; gray squares, multiplication and dashed boxes are noninformative cases. L and H inside black squares denote the type of allele lost. It was shown (Liu et al., 1999) that the H-allele is less sensitive to normal cell contamination than L-allele

Figure 4 Comparison of real-time PCR and Southern hybridization results. DNA was digested with BamHI and transferred to the nylon filters. Hybridization was carried out simultaneously with two probes NLJ-003 (3p21.3T) and NL1-290 (3q13-q21). Upper numbers show the ratio according band intensities determined by scanning nylon filters and bottom numbers represent results of realtime PCR

from deletions. Real-time PCR in contrast to the LOH can detect precisely copy number changes (Cairns et al., 1998; Chiang et al., 1999; Braga et al., 2002) and in this study we performed detailed analysis of these two critical loci combining real-time PCR with LOH. Several candidate TSG(s) were identified in these two regions; however, none of them showed a frequent mutation rate in tumor samples (Daigo et al., 1999; Lerman et al., 2000; Zabarovsky et al., 2002). Epigenetic inactivation by methylation and homozygous deletions are alternative pathways for TSG inactivation (Baylin and Herman, 2000). Indeed, for example, hypermethylation of promoter region was demonstrated for RASSF1A gene from 3p21.3C (LUCA) region in various epithelial carcinomas (Dammann et al., 2000; Burbee et al., 2001; Dreijerink et al., 2001; Kuzmin et al., 2002). We designed two probes for real-time PCR, one from each of these regions. Clone NL3-001 is located closely to candidate TSG semaphorin 3B (SEMA3B), while NLJ003 is between the integrin alpha 9 (ITGA9) gene involved in cell adhesion and border of homozygous deletion in ACC-LC5 SCLC cell line (Murata et al., 1994). To our knowledge, real-time quantitative PCR was applied here for the first time in a detailed study of copy number changes in chromosome 3p loci in solid tumors, and we demonstrated frequent (15.6%) homozygous deletions in both 3p21.3C and 3p21.3T critical regions in squamous cell CCs. Earlier homozygous deletions were detected in CC only around the FHIT region, and these Oncogene

3p deletion hot spots in cervical carcinomas V Senchenko et al

2990

results were based on application of LOH and multiplex PCR (Larson et al., 1997a; Muller et al., 1998). Moreover, our results strongly argue that these two regions in 3p21.3 are hot spots for rearrangements in cervical cancer and therefore most likely contain TSG(s) and other cancer-associated genes. Indeed, aberrations in NLJ-003 locus were detected in 81.3% and in NL3001 in 71.9% cases. Furthermore, aberrations in at least one of these loci were detected in 90.6% cases that is significantly higher than was reported before for 3p21.3 (23–57%) and FHIT (30–58%) (Wistuba et al., 1997; Muller et al., 1998; Helland et al., 2000; Herzog et al., 2001; Acevedo et al., 2002). According to histology analysis, contamination of tumor samples with normal stroma and lymphocytes was not more than 40%. Quantitative data are heavily affected by the degree of contamination of tumor samples. Intratumoral heterogeneity related to clonal or subclonal events was reported earlier (Larson et al., 1997a) and observed here by LOH (data not shown). Therefore, quantitative copy number changes determined by real-time PCR could be even more significant. The smallest region homozygously deleted in 3p21.3C is bordered by D3S1568 and D3S4604. Therefore, the centromeric border is inside the CACNA2D2 gene (cases #18 and #19) that is consistent with the data for SCLCs, for example, GLC20 SCLC cell line (Wei et al., 1996; Lerman et al., 2000). This implies that the TSG(s) located in 3p21.3C could be involved in the development of different tumors and genes located centromeric to CACNA2D2 (Lerman et al., 2000) could be excluded from the list of candidate cervical TSG(s). The telomeric border is inside the SEMA3F gene (cases #18, #19 and #26) and this supported suggestion that in addition to RASSF1A, both SEMA genes in this region could be TSG(s) (Zabarovsky et al., 2002). The smallest region homozygously deleted in 3p21.3T is flanked by D3S1298 and NL1-024 (D3S4285). The centromeric border is close to D3S1298 (cases #9 and #26) and this most probably suggests that DLEC1 and MYD88 genes (Daigo et al., 1999; Protopopov et al., 2003) could not be considered as candidate TSG(s) involved in cervical cancer. The telomeric border was not precisely determined in this study and included two candidate TSG(s), namely MLH1 and ITGA9. Keeping in mind that telomeric border of the homozygous deletion in SCLC cell line ACC-LC5 was determined earlier (Ishikawa et al., 1997), we could identify here only five known candidate TSG(s): trans-Golgi p230 (GOLGA4), gene deleted in AP20-region 1 (APRG1), ITGA9, gene homologous to S. pombe YA22 (actually a human ortholog referred to as HYA22) and villin-like VILL genes (Daigo et al., 1999, Protopopov et al., 2003). Additional studies are needed to reveal which of these genes has TSG activity. Our study also demonstrated that aberrations in 3p21.3 have a complex character and in addition to deletions, amplification also could happen in this region with duplication/multiplication of the second allele, which can accompany allelic loss (De Nooij-van Dalen et al., 1998; Varella-Garcia et al., 1998; Thiagalingam Oncogene

et al., 2001). This result further confirmed the usefulness of quantitative real-time PCR in the mapping of important regions in cancer genome research. Amplification of 3p loci in CC was shown here for the first time. Earlier increase in chromosome copy number and loss of one allele with concomitant duplication of the second allele was shown in 3p only for lung cancer. Combination of LOH and FISH was applied in that study (Varella-Garcia et al., 1998). Thus, allelic imbalance can frequently lead to an erroneous conclusion that the other allele had suffered deletion while indeed another allele was multiplied (Zabarovsky et al., 2002). Human chromosome 3p probably contains many genes that can play a dominant oncogenic role. MST1 receptor (RON) and its ligand MST1 gene are located in 3p21 and their overproduction can result in autocrine stimulation and uncontrolled proliferation (Angeloni and Lerman, 2001). Thus, amplification of some 3p21.3 regions can lead to oncogene activation following the deletion of resident TSG(s). To confirm that allelic imbalance may result from frequent multiplication/amplification in tumor cells, we performed additional experiments with RCC cell lines. RCC showed very high level of 3p deletions by LOH studies (>90%) and it has been shown that 3p21.3T and 3p21.3C are the most frequently deleted regions in both RCC and cervical cancers (see for example Braga et al., 2002). For five RCC cell lines we have previously shown 3p deletions by LOH analysis (Alimov et al., 2000). However, cytogenetic analysis demonstrated that among them only one had two copies of chromosome 3 and four lines had from three to five copies. These data proved that amplification associated with 3p deletion is a common event and therefore, without additional independent data, it is more correct to speak about allelic disbalance. Interestingly, homozygous deletions were detected in five cases (15.6%) for both markers, and in four cases both markers were simultaneously homozygously deleted in the same tumor (P ¼ 0.0009). This most probably means that TSG(s) in these regions could have synergistic effects. The estimation of possible interdependency between all aberrations in loci NL3001 and NLJ-003 as different events was carried out using a permutation test (Mantel, 1967). This test demonstrated a significant correlation between different aberrations in these two loci (Po0.01). However, further research is necessary to clarify this situation. Finally, the exceptionally high level of chromosome aberrations in NLJ-003 and NL3-001 loci suggests that cervical TSG(s) could be located very near to these markers.

Abbreviations CC, cervical carcinoma; CGH, comparative genome hybridization; GDB, genome data base; H allele, high-molecularweight allele; ACTB, human b-actin gene; L allele, lowmolecular-weight allele; LDB, location database; LOH, loss of heterozygosity; NSCLC, non-small cell lung carcinoma; PF2K, phospho-fructo-2 kinase gene; RCC, renal cell carcinoma;

3p deletion hot spots in cervical carcinomas V Senchenko et al

2991 STS, sequence-tagged site; SCLC, small cell lung carcinoma; TSG, tumor suppressor gene.

Acknowledgements This work was supported by research grants from the Swedish Cancer Society, the Swedish Research Council, STINT,

Ingabritt och Arne Lundbergs Forskningsstiftelse, Pharmacia Corporation, Karolinska Institute, the Russian National Human Genome Program and the Russian Foundation for Basic Research (Grants 01-04-48028 and 01-04-49086). MIL was funded in toto with funds from the National Cancer Institute, National Institutes of Health, under Contract No. NO1-CO-56000.

References Acevedo CM, Henriquez M, Emmert-Buck M and Chuaqui RF. (2002). Cancer, 94, 793–802. Alimov A, Kost-Alimova M, Liu J, Li C, Bergerheim U, Imreh S, Klein G and Zabarovsky ER. (2000). Oncogene, 19, 1392–1399. Angeloni D and Lerman MI (2001). Human Lung cancer: Cancer-Causing Genes, Environmental Factors. Sourcebook on Asbestos Diseases, Vol. 23. Peters GA, Peters BJ (eds). Philadelphia: LexisNexis Press, pp 169–209. Baylin SB and Herman JG. (2000). Trends Genet., 16, 168–174. Boulay J-L, Reuter J, Ristchard R, Terracciano L, Herrmann R and Rochlist C. (1999). BioTechniques, 27, 228–232. Braga EA, Kotova EI, Pugacheva EM, Gizatullin RZ, Kashuba VI, Lerman MI, Aksenova MG, Efremov IA, Zabarovsky ER and Kiselev LL. (1997). Mol. Biol. (Mosk.), 31, 985–996. Braga E, Pugacheva E, Bazov I, Ermilova V, Kazubskaya T, Mazurenko N, Kisseljov F, Liu J, Garkavtseva R, Zabarovsky E and Kisselev L. (1999). FEBS Lett., 454, 215–219. Braga E, Senchenko V, Bazov I, Loginov W, Liu J, Ermilova V, Kazubskaya T, Garkavtseva R, Mazurenko N, Kisseljov F, Lerman MI, Klein G, Kisselev L and Zabarovsky ER. (2002). Int. J. Cancer, 100, 534–541. Burbee DG., Forgacs E, Zochbauer-Muller S, Shivakumar L, Fong K, Gao B, Randle D, Virmani A, Bader S, Sekido Y, Latif F, Milchgrub S, Gazdar AF, Lerman MI, Zabarovsky E, White M and Minna JD. (2001). J. Natl. Cancer Inst., 93, 691–699. Cairns JP, Chiang P-W, Ramamoorthy S, Kurnit DM and Sidransky D. (1998). Clin. Cancer Res., 4, 441–444. Chiang P-W, Beer DG, Wei W-L, Orringer MB and Kurnit DM. (1999). Clin. Cancer Res., 5, 1381–1386. Daigo Y, Isomura M, Nishiwaki T, Tamari M, Ishikawa S, Kai M, Murata Y, Takeuchi K, Yamane Y, Hayashi R, Minami M, Fujino MA, Hojo Y, Uchiyama I, Takagi T and Nakamura Y. (1999). DNA Res., 6, 37–44. Dammann R, Li C, Yoon J-H, Chin P L, Bates S and Pfeifer GP. (2000). Nat. Genet., 25, 315–319. De Nooij-vanDalen AG, van Buuren-van Seggelen VHA, Lohman PHM and Giphart-Gassler M. (1998). Genes Chromosomes Cancer, 21, 30–38. Dreijerink K, Braga E, Kuzmin I, Geil L, Duh F-M, Angeloni D, Zbar B, Lerman MI, Stanbridge EJ, Minna JD, Protopopov A, Li J, Kashuba V, Klein G and Zabarovsky ER. (2001). Proc. Natl. Acad. Sci. USA, 98, 7504–7509. Fouret PJ, Dabit D, Mergui JL and Uzan S. (1998). Pathobiology, 66, 306–310. Guo Z, Wilander E, Sallstrom J and Ponten J. (1998). Anticancer Res., 18, 707–712. Guo Z, Hu X, Afink G, Ponten F, Wilander E and Ponten J. (2000). Int. J. Cancer, 86, 518–523. Guo Z, Wu F, Asplund A, Hu X, Mazurenko N, Kisseljov F, Ponten J and Wilander E. (2001). Mod. Pathol., 14, 54–61.

Heid CA, Stevens J, Livak KJ and Williams PM. (1996). Genome Res., 6, 986–994. Helland A, Kraggerud SM, Kristensen GB, Holm R, Abeler VM, Huebner K, Borresen-Dale AL and Lothe RA. (2000). Int. J. Cancer, 88, 217–222. Herzog CR, Keith AC, Sabourin CLK, Kelloff GJ, Boone CW, Stoner GD and You M. (2001). Mol. Carcinog., 30, 159–168. Ishikawa S, Kai M, Tamari M, Takei Y, Takeuchi K, Bandou H, Yamane Y, Ogawa M and Nakamura Y. (1997). DNA Res., 4, 35–43. Kashuba VI, Gizatullin RZ, Protopopov AI, Li J, Vorobieva NV, Fedorova L, Zabarovska VI, Muravenko OV, KostAlimova M, Domninsky DA, Kiss C, Allikmets R, Zakharyev VM, Braga EA, Sumegi J, Lerman M, Wahlestedt C, Zelenin AV, Sheer D, Winberg G, Grafodatsky A, Kisselev LL, Klein G and Zabarovsky ER. (1999). Gene, 239, 259–271. Kashuba VI, Szeles A, Allikmets RL, Nilsson AS, Bergerheim SR, Modi W, Grafodatsky A, Dean M, Stanbridge EJ, Winberg G, Klein G and Zabarovsky ER. (1995). Cancer Genet. Cytogenet., 81, 144–150. Kisseljov F, Semionova L, Samoylova E, Mazurenko N, Komissarova E, Zourbitskaya V, Gritzko T, Kozachenko V, Netchushkin M, Petrov S, Smirnov A and Alonso A. (1996). Int. J. Cancer, 69, 484–487. Kok K, Naylor SL and Buys CH. (1997). Adv. Cancer Res., 71, 27–92. Kuzmin I, Gillespie JW, Protopopov A, Geil L, Dreijerink K, Yang Y, Vocke CD, Duh F-M, Zabarovsky E, Rhim JS, Emmert-Buck MR, Linehan WM and Lerman ML. (2002). Cancer Res., 62, 3498–3502. Larson AA, Kern S, Curtiss SH, Gordon R, Cavenee W and Hampton GM. (1997a). Cancer Res., 57, 4082–4090. Larson AA, Liao S-Y, Stanbridge EJ, Cavenee WK and Hampton GM. (1997b). Cancer Res., 57, 4171–4176. Lazo PA. (1999). Br. J. Cancer, 80, 2008–2019. Lerman M, Minna J, Sekido Y, Bader S, Burbee D, Fong K, Forgacs E, Gao B, Garner H, Gazdar AF, Girard L, Kamibayashi C, Kondo M, Pande A, Ramalingam V, Randle D, Tomizawa Y, Virmani A, Wistuba I, Macartney D, Honorio S, Maher E, Latif F, Kashuba V, Protopopov A, Li J, Klein G, Zabarovsky E, Duh F-M, Wei M-H, Geil L, Kuzmin I, Ivanov S, Angeloni D, Ivanova A, Zbar B and Johnson B.E. (2000). Cancer Res., 60, 6116–6133. Liu J, Zabarovska VI, Braga E, Alimov A, Klein G and Zabarovsky ER. (1999). FEBS Lett., 462, 121–128. Livak KJ, Flood SJ, Marmaro J, Giusti W and Deets K. (1995). PCR Methods Appl., 4, 357–362. Mantel N. (1967). Cancer Res., 27, 209–220. Matthews CP, Shera KA and McDougall JK. (2000). Proc. Soc. Exp. Biol. Med., 223, 316–321. Mazurenko N, Attaleb M, Gritsko T, Semjonova L, Pavlova L, Sakharova O and Kisseljov F. (1999). Oncol. Rep., 6, 859–863. Oncogene

3p deletion hot spots in cervical carcinomas V Senchenko et al

2992 Muller CY, O’Boyle JD, Fong KM, Wistuba II, Biesterveld E, Ahmadian M, Miller DS, Gazdar AF and Minna JD. (1998). J. Natl. Cancer Inst., 90, 433–439. Murata Y, Tamari M, Takahashi T, Horio Y, Hibi K, Yokoyama S, Inazawa J, Yamakawa A, Takahashi T and Nakamura Y. (1994). Hum. Mol. Genet., 3, 1341–1344. Protopopov A, Kashuba VI, Zabarovska VI, Muravenko OV, Lerman MI, Klein G and Zabarovsky ER. (2003). Cancer Res., 63, 404–412. Rader JS, Gerhard DS, O’Sullivan MJ, Li Y, Li L, Liapis H and Huettner PC. (1998). Genes Chrom. Cancer, 22, 57–65. Rader JS, Kamarasova T, Huettner PC, Li L, Li Y and Gerhard DS. (1996). Oncogene, 13, 2737–2741. Samoylova E, Shihaev G, Petrov S, Kisseljova N and Kisseljov F. (1995). Int. J. Cancer, 61, 337–341. Sekido Y, Ahmadian M, Wistuba II, Latif F, Bader S, Wei MH, Duh F-M, Gazdar AF, Lerman MI and Minna JD. (1998). Oncogene, 16, 3151–3157. Thiagalingam S, Laken S, Willson JKV, Markowitz SD, Kinzler KW, Vogelstein B and Lengauer Ch. (2001). Proc. Natl. Acad. Sci. USA, 98, 2698–2702.

Oncogene

Umayahara K, Numa F, Suehiro Y, Sakata A, Nawata S, Ogata H, Suminami M, Sasaki K and Kato H. (2002). Genes Chromosomes Cancer, 33, 98–102. Van den Berg A and Buys CH. (1997). Genes Chromosomes Cancer, 19, 59–76. Varella-Garcia M, Gemmill RM, Rabenhorst SH, Lotto A, Drabkin HA, Archer PA and Franklin WA. (1998). Cancer Res., 58, 4701–4707. Wei MF, Latif F, Bader S, Kashiba V, Chen JY, Duh FM, Sekido Y, Lee C, Geil L, Kuzmin I, Zabarovsky E, Klein G, Zbar B, Minna JD and Lerman MI. (1996). Cancer Res., 56, 1487–1492. Wistuba II, Behrens C, Virmani AK, Meme G, Milchgrub S, Girard L, Fondon III JW, Garner HR, McKay B, Latif F, Lerman MI, Lam S, Gazdar AF and Minna JD. (2000). Cancer Res., 60, 1949–1960. Wistuba II, Montellano FD, Milchgrub S, Virmani AK, Behrens C, Chen H, Ahmadian M, Nowak JA, Muller C, Minna JD and Gazdar AF. (1997). Cancer Res., 57, 3154–3158. Zabarovsky ER, Lerman MI and Minna JD. (2002). Oncogene, 21, 6915–6935.