Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Identification of a tumor suppressive critical region mapping to 3p14.2 in esophageal squamous cell carcinoma and studies of a candidate tumor suppressor gene, ADAMTS9


A gene critical to esophageal cancer has been identified. Functional studies using microcell-mediated chromosome transfer of intact and truncated donor chromosomes 3 into an esophageal cancer cell line and nude mouse tumorigenicity assays were used to identify a 1.61 Mb tumor suppressive critical region (CR) mapping to chromosome 3p14.2. This CR is bounded by D3S1600 and D3S1285 microsatellite markers. One candidate tumor suppressor gene, ADAMTS9, maps to this CR. Further studies showed normal expression levels of this gene in tumor-suppressed microcell hybrids, levels that were much higher than observed in the recipient cells. Complete loss or downregulation of ADAMTS9 gene expression was found in 15 out of 16 esophageal carcinoma cell lines. Promoter hypermethylation was detected in the cell lines that do not express this gene. Re-expression of ADAMTS9 was observed after demethylation drug treatment, confirming that hypermethylation is involved in gene downregulation. Downregulation of ADAMTS9 was also found in 43.5 and 47.6% of primary esophageal tumor tissues from Hong Kong and from the high-risk region of Henan, respectively. Thus, this study identifies and provides functional evidence for a CR associated with tumor suppression on 3p14.2 and provides the first evidence that ADAMTS9, mapping to this region, may contribute to esophageal cancer development.


Esophageal cancer (EC) is a deadly cancer that is ranked as the seventh leading cause of cancer deaths worldwide. The majority of ECs are classified as esophageal squamous cell carcinoma (ESCC). It has its highest incidence in Northern China including Henan and Shanxi provinces (Li, 1982). The molecular pathogenesis of ESCC is still unclear. Identification of the key genes involved in the development of this disease will provide important information for its diagnosis and treatment.

Many ESCC studies show a high frequency of allelic losses on chromosome 3p by both loss of heterozygosity (LOH) (Wang et al 1996; Ko et al., 2001; Cheung et al., 2005) and comparative genomic hybridization (CGH) (Shiomi et al., 2003; Qin et al., 2005). These results imply that tumor suppressive elements may be located on the short arm of chromosome 3. Up to now, a functional study of chromosome 3 in ESCC has not been reported.

Somatic cell hybridization provides a functional approach to study tumor suppression in cancer cell lines (Saxon et al., 1985). Selected normal human chromosomes can be transferred by microcell-mediated chromosome transfer (MMCT) to complement defects in cancer cells. By comparing the functional impact of the transferred exogeneous chromosome and the tumorigenic potential of different microcell hybrids (MCHs), critical regions (CRs), important for tumorigenic suppression, can be identified. Several CRs in chromosome 3 were previously reported for other cancers including the 3p21.3 and 3p21.2–p14 regions in renal cell carcinoma (Yang et al., 2001), 3p21.3 in nasopharyngeal carcinoma (NPC) (Cheng et al., 1998), and 3p21.3, 3p25, and 3p13 in oral squamous cell carcinoma (Uzawa et al., 1998). Our lab recently identified 9q33–34 and 14q32 as CRs for tumorigenic suppression in ESCC (Ko et al., 2005; Yang et al., 2005).

In the present study, we report the identification of another CR, 3p14.2, using the ESCC cell line SLMT-1. Significant downregulation of a disintegrin-like and metalloprotease with thromobospondin type 1 motif 9 (ADAMTS9), which is located in the CR, was also found in both ESCC cell lines and primary tumor tissues. Promoter hypermethylation and re-expression of ADAMTS9 by demethylation drug treatment suggested that epigenetic inactivation is a major gene silencing mechanism for this new candidate tumor suppressor gene (TSG).


Transfer of human chromosome 3 into ESCC cells

Three chromosome 3 donor cell lines served as sources for human chromosome 3 transfer into the ESCC cell line, SLMT-1. MCHs were selected by geneticin. Transfer of the intact donor chromosome from MCH903.1 cells into SLMT-1 proved to be very difficult. Most hybrid clones were unable to survive in culture. After over 50 fusions, only MCH3.11 retained a nearly intact donor chromosome (Figure 1). Another chromosome 3 hybrid derived from the MCH903.1 fusion, MCH3.15, spontaneously eliminated much of the chromosome. MCH3.22, MCH3.25, and MCH3.26 cell lines were generated from donor MCH910.7 cells containing a chromosome 3 microdeleted in the 3p26–3p21.3 region. MCH3.1, MCH3.2, and MCH3.3 were generated from MCH924.4 donor cells, which only retained the 3p25 and 3q21–3q27 regions. Slot blot hybridization showed all MCHs were free of mouse DNA contamination (data not shown). These eight MCH cell lines were used for subsequent studies.

Figure 1

Microsatellite analysis of chromosome 3 donor cells, MCHs, and MCH3.11 and MCH3.15 tumor segregants. A total of 32 microsatellite markers were utilized. The presence () and absence (•) of the marker, expected absence from hybrid cell (X) due to deletion in donor chromosome, and uninformative (U) markers of donor and recipient cells are as indicated. Analysis of these hybrids and their matched TSs showed nonrandom loss localized to a 1.61 Mb critical region (CR) bounded by markers D3S1600 and D3S1285. The CR and location of BAC 129B22 are indicated on the right-hand side of the figure.

Microsatellite typing and fluorescence in situ hybridization analyses

A total of 32 microsatellite markers spanning the whole chromosome 3 were used to verify the successful transfer of exogenous chromosome 3 (Figure 1). MCH3.11 shows the presence of all tested markers except one at 3p22.2 (D3S3521) and one at 3q27 (D3S2418). MCH3.15, on the other hand, shows extensive loss, with only an intact 3p14 region and most of the q arm being retained. MCH3.22 shows the entire truncated chromosome 3 was successfully transferred from MCH910.7 donor cells. This is in contrast to MCH3.25 and MCH3.26, which both show loss of a 3p21.1–3p14 region covered by seven markers and a 3p11.2–3q13.1 region showing loss of four markers. MCH3.1, MCH3.2, and MCH3.3 show the successful transfer of the entire truncated chromosome region from the donor cells, except for one marker (D3S1593) at 3q21.3–3q25.2.

After MMCT the exogenous transferred chromosome 3 was visualized by fluorescence in situ hybridization (FISH) to cytogenetically confirm its successful transfer. A human-specific whole chromosome 3 painting (WCP) probe was hybridized to the metaphase spreads. All MCHs show one extra fluorescence signal in more than 90% of the metaphase spreads, confirming successful transfer of the exogenous copy of chromosome 3 in the hybrids.

Tumorigenicity of MCH cell lines

The recipient ESCC cell line, SLMT-1, is highly tumorigenic and forms tumors larger than 150 mm3 within 28 days after nude mouse inoculation in all injection sites. The tumorigenicity of the various MCH cell lines containing different regions of the transferred normal exogenous chromosome was compared to SLMT-1 in the nude mouse model (Table 1). MCH3.11 and MCH3.15 induced suppression of tumor growth, with longer latency periods of up to 70 days and the results were statistically significant (P0.0003). The tumor growth kinetics show significant differences between the hybrids and recipient cell lines (Figure 2a). Of the MCH cell lines generated from MCH910.7 donor cells, MCH3.22 exhibits a longer latency period of up to 53 days. However, the other two hybrids, MCH3.25 and MCH3.26, do not show significant differences in tumor formation kinetics or latency periods, compared to recipient ESCC cells (Figure 2b). No observed tumor suppression ability was observed for all hybrids generated from MCH924.4, the most truncated donor chromosome 3 cell line (Figure 2c).

Table 1 Summary of tumorigenicity assay of recipient SLMT-1 and chromosome 3 MCH cell lines
Figure 2

Tumor growth kinetics of the ESC SLMT-1 recipient cell line and chromosome 3 MCHs. The tumor growth kinetics of SLMT-1 was compared against the MCHs derived from the donor cell lines (a) MCH903.1, (b) MCH910.7, and (c) MCH924.4. Each data point represents an average volume of six sites inoculated for each cell line. Three MCHs, MCH3.11, MCH3.15, and MCH3.22, were able to suppress tumor formation, whereas MCH3.25, MCH3.26, MCH3.1, MCH3.2, and MCH3.3 did not.

Tumor segregant analyses

Tumor segregant (TS) cell lines were established from MCH3.11- and MCH3.15-induced tumors and were labeled accordingly (ex. MCH3.11TS1L). Microsatellite analysis was performed to verify the presence of the exogenous chromosome in the TSs. Chromosome 3p was lost and some regions of the 3q arm were retained in all TSs. By comparing tumorigenicity and chromosome loss determined by microsatellite analysis, a 1.61 Mb CR between the markers D3S1600 and D3S1285 was found to be essential in suppressing tumorigenicity in the SLMT-1 cell line (Figure 1). Representative microsatellite typing results with marker D3S1600 show that the exogenous chromosome is present in tumor suppressive hybrids, but is absent in tumorigenic hybrids and tumor segregants (Figure 3a). A WCP chromosome 3 probe verified the successful transfer of the exogenous chromosome in MCH3.11 and MCH3.22 (Figure 3b). With a BAC 129B22 probe mapping to 64.1 Mb, we further verified the cytogenetic loss of one copy of the exogenous chromosome in the CR of two representative TS cell lines, MCH3.11TS1L and MCH3.15TS1L, compared to their parental MCH cell lines. A representative BAC FISH is shown in Figure 3b.

Figure 3

(a) Representative results of PCR-microsatellite typing for the marker D3S1600 for the recipient SLMT-1 cell line, donor cell lines, MCHs, and corresponding TSs. The arrows designate exogenous donor allele transfer () and loss (). (b) Representative FISH results of (i) whole chromosome 3 painting probe in donor cell (MCH903.1), recipient cell (SLMT-1), and two MCHs (MCH3.11 and MCH3.22) and (ii) 129B22 BAC probe in MCH3.11 and its corresponding TS. Arrow () indicates exogenously transferred chromosome 3 and BAC signals.

Candidate TSG, ADAMTS9, gene expression in MCH cell lines

A search of the genome database identified a candidate TSG, ADAMTS9, mapping to the chromosome 3 CR (Figure 4a). The expression levels of this gene in all MCHs and their corresponding TSs were examined by real-time quantitative polymerase chain reaction (PCR). Normal RNA transcript expression was found in all tumor suppressive hybrids, but was downregulated in one tumorigenic hybrid and the TSs, when compared to NE1 (Figure 4b).

Figure 4

(a) Map of tumor suppressive region at 3p14.2 determined after analysis by PCR-microsatellite typing of allelic loss in MCHs and TSs. The markers (D3S1600 and D3S1285) defining the 1.61 Mb critical region (CR) are in bold font. Regions and genes in between are drawn to scale based on information available at the UCSC Human Genome Browser Gateway ( (b) Real-time quantitative *RT–PCR analysis of ADAMTS9 in MCHs and TSs. Positive gene expression was observed for all tumor suppressive hybrids, but not for one tumorigenic hybrid and TSs.

Quantitative RNA expression levels of ADAMTS9 in ESCC cell lines and primary tumors

The RNA transcript levels of the candidate gene in a panel of 16 ESCC cell lines and 44 Chinese patient primary tumors were examined by real-time quantitative PCR. The expression levels in cell lines were normalized against the NE1 cell line, an immortalized normal human esophageal cell line (Deng et al., 2004) (Figure 5a). Downregulated RNA expression of ADAMTS9 was found in 15 out of 16 tested cell lines (93.8%); only KYSE270 showed a normal expression level. Four cell lines (KYSE30, KYSE140, KYSE510, and KYSE520) showed no ADAMTS9 expression, even after 50 cycles of PCR amplification. The RNA transcript level in tumor tissues was compared to the patient's own normal tissue. A total of 44 pairs of ESCC primary tumor/normal tissues was examined. These included 23 pairs from a moderate risk area, Hong Kong, and 21 pairs from a high-risk area, Henan and nearby counties in Northern China. The incidences of downregulated expression in the Hong Kong moderate-risk region and the Henan high-risk region were 43.5 and 47.6%, respectively (Figure 5b and c). Downregulated expression in the Hong Kong specimens was found to be significantly associated with the patient's age (P=006). Interestingly, this association was not observed in the Henan specimens. No correlation was found between ADAMTS9 expression and gender, TNM stage, or lymph node metastasis in either patient cohort (Table 2).

Figure 5

Real-time quantitative RT–PCR analysis of ADAMTS9 in (a) 16 ESC cell lines, (b) 23 pairs Hong Kong specimens, and (c) 21 pairs Henan specimens. No ADAMTS9 expression was detected in cell lines KYSE30/140/510/520. The percentage of cell lines showing at least 10-fold downregulated expression, when compared to NE1, for ADAMTS9 was 50%. In primary tumor specimens, 43.5 and 47.6% showed at least a 50% reduction in expression, when compared to normal tissue in Hong Kong and Henan specimens, respectively.

Table 2 Correlation between ADAMTS9 expression and ESCC patient clinicopathological features

Methylation status of ADAMTS9 and demethylation treatment of ESCC cell lines

The promoter and exon 1 sequences of ADAMTS9 are shown in Figure 6a. The promoter region of ADAMTS9 in three cell lines showing either no expression or decreased expression (KYSE510, KYSE520, and TTn6) was amplified using both methylated and unmethylated specific primers. Both ADAMTS9 unmethylated and methylated alleles were detected in all tested cell lines (Figure 6b). To confirm the significance of epigenetic inactivation in these cell lines, a demethylation drug, 5-aza-2-deoxycytidine, was used to treat the cell lines. Expression of ADAMTS9 was restored to normal levels in all treated cell lines (Figure 6c).

Figure 6

Methylation status of ADAMTS9 and re-expression of ADAMTS9 after demethylation treatment. (a) The promoter and exon 1 region of ADAMTS9. The 5′UTR is in lower case and exon 1 region is in upper case. The arrow () indicates the transcription start site. The locations of MSP primers are indicated by arrows. The suggested CpG sites are underlined. The expected size of the amplicon is 171 bp. (b) Methylation-specific PCR results showed both hypermethylated and unmethylated alleles in the ADAMTS9 promoter region of KYSE510, KYSE520, and TTn-6. (c) ADAMTS9 expression in the above cell lines was restored after demethylation drug treatment. GAPDH served as the internal loading control.


Numerous studies in lung (Sekine et al., 2005), hepatocellular (Tischoff et al., 2005), NPC (Xiong et al, 2004), glioma (Hesson et al., 2004), and brain tumors (Horiguchi et al., 2003) suggest that chromosome 3 harbors TSGs. Positional cloning and molecular and cytogenetic approaches identified possible tumor suppressive regions mapping to 3p23 (Wang et al., 1996), 3p14.2 and 3p21.3 (Ko et al., 2001), and 3p21.1–p21.2 (Shiomi et al., 2003) in ESCC. A functional MMCT approach also identified a CR at 3p21.3 in NPC (Cheng et al., 1998) and at 3p21.2–p21.3 in oral squamous cell carcinoma (Uzawa et al., 1998).

In the present study, an intact and two truncated human chromosomes 3 were transferred into the highly tumorigenic ESCC SLMT-1 cell line. The ability of these transferred chromosomes to functionally complement defects in the ESCC cell line was assessed by examining the impact of this transfer on tumorigenic potential in nude mice. The eight MCH cell lines established in this study contain different regions of the exogenous transferred chromosomes and differ dramatically in their tumorigenicity. MCH3.11 and MCH3.15, derived from transfer of the intact donor cell line, MCH903.1, and MCH3.22, derived from transfer of the truncated donor cell line, MCH910.7, show significant tumor suppressive ability, with an increased latency period of up to 70 days (Table 1). In contrast, five MCHs derived from transfer of truncated donor cell lines, MCH910.7 or MCH924.4, did not show tumor suppressive ability in the nude mouse assay (Figure 2).

Difficulties in successful establishment of stable MCH cell lines after transfer of chromosomes harboring growth-related and apoptotic genes, such as p53 on chromosome 17, have been reported previously (Goyette et al., 1992). Indeed, the transfer of chromosome 3 induced senescence and growth arrest in ovarian cancer cells (Rimessi et al., 1994). It is notable that a study of renal cell carcinoma found the 3p14.2–p21.1 region to possibly contain a telomerase repressor gene (Tanaka et al., 1998). In the present study, we observed over a hundred initial MCH clones; however, most of the clones were unable to become established into stably growing cell lines. Typically, the early clones died after 4–8 weeks selection and exhibited cellular enlargement with a vacuolated and flat morphology. It is possible that a senescence-related gene affected their growth and viability after chromosome 3 transfer. The small number of expandable clones that we obtained suggests that transfer of chromosome 3 can generally suppress in vitro growth.

PCR-microsatellite and BAC FISH analyses were used to narrow down and identify the CR associated specifically with tumor suppression. A 1.61 Mb CR located between markers D3S1600 and D3S1285 was found to be necessary for the tumorigenic suppression of ESCC. These findings further suggest that the CR present in the exogenous chromosomes contains functional tumor suppressive elements. Our previous studies using similar approaches successfully identified candidate TSGs such as TSLC1 and THY1 at 11q22–q23 in NPC (Lung et al., 2004, 2005) and DEC1 at 9q33–q34 in ESCC (Yang et al., 2005).

In the present study, we identified a candidate TSG, ADAMTS9, and one ncRNA, ENST351926 mapping to 64.48 Mb at 3p14.2, which are located in the CR. The expression of ADAMTS9 in tumor suppressive MCHs was confirmed by reverse transcription (RT)–PCR. The positive expression of the gene was observed in all tumor suppressive MCHs, but was not found in tumorigenic MCHs and tumor segregants, strongly suggesting that ADAMTS9 plays an important role in tumor suppression. The pseudogene is located upstream of the ADAMTS9 promoter region and whether it can serve as a riboregulator or gene expression regulator remains to be determined.

ADAMTS9 belongs to the ADAM-TS/metallospondin family of proteins, structurally homologous to the ADAM proteins, but containing an additional 15 C-terminal thrombospondin type 1 (TSP1) repeat (Somerville et al., 2003). Members of the ADAM-TS family have been implicated in the cleavage of proteoglycans, control of organ shape during development, and the inhibition of angiogenesis (Clark et al., 2000). Downregulated expression of ADAMTS9 was also found in breast carcinoma (Porter et al., 2004). We further studied ADAMTS9 gene expression in 16 ESCC cell lines and 44 primary ESCC tissues by real-time quantitative RT–PCR. Among the 16 ESCC cell lines, 50% showed at least a 10-fold reduction of gene expression, compared with the immortalized esophageal cell line, NE1. Complete loss of ADAMTS9 expression was observed in four of 16 cell lines. In ESCC primary tumors obtained from Hong Kong and Henan, there was 43.5 and 47.6% of specimens showing downregulation. The significant downregulated expression in both ESCC cell lines and primary tumors again suggests that expression of ADAMTS9 is involved in ESCC tumorigenesis. An association between downregulated ADAMTS9 expression in moderate-risk Hong Kong specimens with older age patients was observed; an explanation for this association awaits further investigation.

One common mechanism of TSG silencing is by epigenetic inactivation (Chen and Baylin, 2005). The methylation status of the ADAMTS9 promoter was analysed by methylation-specific PCR (MSP) in the KYSE510 and KYSE520 cell lines, which show no expression of the gene, and in TTn-6, which shows reduced expression. The hypermethylated allele was found in each of these cell lines. To further confirm the role of methylation in ADAMTS9 downregulation, the demethylation drug, 5′aza-2′-deoxycytidine, was used to treat the cell lines. ADAMTS9 gene expression was restored in all tested cells, providing evidence that epigenetic inactivation was controlling downregulated gene expression.

In conclusion, we identified a 1.61 Mb CR at 3p14.2 associated with tumor suppression of the ESCC cell line, SLMT-1. A candidate TSG mapping to this CR, ADAMTS9, is likely involved in this tumor suppression, as we observed no or reduced gene expression levels of this candidate TSG in a panel of Asian ESCC cell lines and in ESCC primary tumors. Promoter hypermethylation and re-expression of the gene after demethylation drug treatment suggested that epigenetic inactivation was involved in silencing the gene. This is the first report to provide functional evidence of the tumorigenic suppressive ability of chromosome 3 in ESCC. Furthermore, we have now identified a candidate TSG, ADAMTS9, which may be associated with ESCC tumorigenesis. Further studies are underway to elucidate the functional role and potential diagnostic significance of this gene in ESCC and to explore its possible impact on the pathways involved in ESCC tumorigenesis.

Materials and methods

Cell lines and culture conditions

The SLMT-1 cell line was established from a well-differentiated ESCC of the lower esophagus from a Hong Kong male patient (Tang et al., 2001). It was used as the recipient cell for all MMCT experiments. EC1 (Mok et al., 1987), EC18 (Pan, 1989), and HKESC-2 (Hu et al., 2002) were established from Hong Kong Chinese patients. The 81T cell line was established from a Taiwanese patient (Hu et al., 1984). KYSE30/70/140/150/180/270/410/450/510/520 (Shimada et al., 1992) and TTn (Takahashi et al., 1990) came from Japanese patients and are available commercially from DSMZ (Braunschweig, Germany). NE1 is an immortalized esophageal epithelial cell line established by retroviral expression of HPV16 E6/E7 and served as the positive control in expression studies (Deng et al., 2004). MCH903.1, MCH910.7, and MCH924.4 are the donor cells used for the fusion experiments. They are mouse A9 fibroblast cells containing a single copy of either intact or truncated human chromosome 3 (Cheng et al., 1998). All cell lines were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 5% fetal bovine serum, 5% newborn calf serum, and 1% penicillin/streptomycin (Invitrogen, Carlsbad, CA, USA), except NE1, which was cultured in a defined keratinocyte serum-free medium supplemented with growth factors (Invitrogen). Donor cells were cultured in growth medium with 800 μg/ml geneticin (CalBiochem, Darmstadt, Germany).

Patients and specimens

The 23 pairs of ESCC primary tumor/normal tissues were obtained from Hong Kong patients diagnosed at Queen Mary Hospital between 2001 and 2003. They included 17 male (mean age=61±3) and six female (mean age=74±5) subjects. The tissues were frozen at −80°C immediately after surgical resection until use. Another 21 of pairs ESCC primary tumor/normal tissues from the high-risk area in Henan province and nearby counties were collected in Linzhou Center Hospital and Yaocun Esophageal Cancer Hospital in 2005. They included 11 male (mean age=51±2) and 10 female (mean age=60±1) subjects. To facilitate transport of these specimens to Hong Kong, they were submerged in RNAlater (Ambion, Woodlands, TX, USA) according to the manufacturer's recommendation, and stored at 4°C until use. None of the patients had received radiotherapy or chemotherapy before the surgery.

Microcell-mediated chromosome transfer

A mouse A9 donor cell containing a single intact (MCH903.1) or truncated (MCH910.7 and MCH924.4) human chromosome 3 was transferred to the recipient, SLMT-1 cell line, as described previously (Goyette et al., 1992; Ko et al., 2005; Yang et al., 2005). The MCHs were selected in culture medium containing 500 μg/ml geneticin and hypoxanthine/aminopterin/thymidine (HAT) (Sigma, St Louis, MO, USA) for 4 weeks.

DNA and RNA extraction

Genomic DNAs from different cell lines were extracted by standard protocols (Sambrook et al., 1989). Cells were harvested and lysed by homogenizing buffer. DNA was extracted by phenol/chloroform extraction followed by ethanol precipitation. For the ESC cell lines and Hong Kong tissue specimens, total RNA was extracted using the RNeasy midi kit (Qiagen, Valencia, CA, USA). All procedures were performed according to the manufacturer's manual. For the Henan specimens, the tissues were homogenized in TRIzol reagent (Invitrogen) followed by chloroform extraction, and ethanol precipitation, as described in the manufacturer's manual.

Slot blot hybridization

Mouse DNA contamination of MCHs was detected by DNA slot blot hybridization, as previously described (Cheng et al., 1998).

Fluorescence in situ hybridization

Visualization of the exogenous copy of chromosome 3 and validation of the 3p14.2 CR in MCHs were carried out by FISH. The metaphase spread was prepared on glass slides by a standard protocol (Rooney and Czepulkowski, 1986). Bacterial artificial chromosome clone 129B22 (RPCI-11 Human BAC Library) from Roswell Park Cancer Institute (Invitrogen) located at 3p14.2 was labeled for use as a FISH probe by Nick translation reagent kit (Vysis, Downers Grove, IL, USA). DNA extraction, labeling, and hybridization were carried out as described previously (Ko et al., 2005). The whole chromosome 3 painting SpectrumGreen™ probe (Vysis) and BAC probe were hybridized to metaphase spreads according to the manufacturer's manual. FISH and fluorescence signal capture were performed as described previously (Yang et al., 2005).

Polymerase chain reaction microsatellite assay

A total of 32 pairs of fluorescence-labeled microsatellite markers was used to genotype the recipient, donor, MCHs, and TSs and verify the presence or absence of the exogenous allele, as previously reported (Yang et al., 2005). The PCR amplicons were analysed on the ABI PRISM™ 3100 Genetic Analyzer using GeneScan and Genotyper software (Applied Biosystems, Foster City, CA, USA). An exogenous allele in TSs was classified as lost, when the intensity of the donor-to-recipient allele ratio in TS versus its corresponding MCH was equal to or less than 0.8. The mapping information and primer sequences were obtained from the UCSC Genome Bioinformatics ( and NCBI ( genome databases. Details of marker locations are shown in Figure 1.

Reverse transcription–PCR and real-time quantitative RT–PCR analyses

The gene expression levels in different cell lines were compared by RT–PCR. In brief, total RNAs were reverse transcribed by M-MLV (GE Healthcare, Littlefont, UK) and amplified by AmpliTaq DNA polymerase (Applied Biosystems). The real-time quantitative PCRs were performed using Mx3000P real-time PCR system (Stratagene, La Jolla, CA, USA). The ADAMTS9 and GAPDH Taqman probes were purchased from Applied Biosystems. The reaction conditions and analysis methods have been previously reported (Lo et al., 2006).

Tumorigenicity assay and tumor segregant analyses

The tumor suppression abilities of different MCHs were compared against the recipient cell line by standard nude mouse tumorigenicity assays, according to protocols previously described (Ko et al., 2005). For each cell line, a total of six sides of three 6- to 8-week-old female athymic Balb/c Nu/Nu mice were injected with 2 × 106 cells subcutaneously. Tumor volume was measured weekly using calipers. The number of days required for tumors to reach a volume of 150 mm3 was defined as the latency period. The P-value was obtained by comparing the tumor size of each cell line with SLMT-1 on day 40 after injection. The tumors arising from tumor suppressive hybrids were reconstituted in cell culture for further analysis.

Methylation-specific PCR and demethylation analysis

Genomic DNA (1 μg) from each sample was modified with CpGenome DNA Modification Kit (Chemicon, Temecula, CA, USA). Modified DNAs were subjected to PCR using methylated and unmethylated primers at location −254 to −83. Primers sequences are listed in Table 3. A quantity of 1 × 105 cells was grown in T75 culture flask 16 h before treatment. Cells were treated with 5 μmole/l 5-aza-2′deoxycytidine (Sigma) for 5 days and were then harvested for RNA extraction (Lung et al., 2004).

Table 3 RT–PCR and MSP primers for candidate ADAMTS9 and GAPDH

Statistical analysis

Statistical analysis was performed using Epi 5.01 statistical software. Associations between clinicopathological variables and expression of ADAMTS9 were analysed using the χ2 test and Fisher's exact test, as appropriate. A P-value of <0.05 was considered as significant.


  1. Chen WY, Baylin SB . (2005). Cell Cycle 4: 10–12.

  2. Cheng Y, Poulos NE, Lung ML, Hampton G, Ou B, Lerman MI et al. (1998). Proc Natl Acad Sci USA 95: 3042–3047.

  3. Cheung AL, Si HX, Wang LD, An JY, Tsao SW . (2005). Cancer Lett (in press).

  4. Clark ME, Kelner GS, Turbeville LA, Boyer A, Arden KC, Maki RA . (2000). Genomics 67: 343–350.

  5. Deng W, Tsao SW, Guan XY, Lucas JN, Si HX, Leung CS et al. (2004). Oncogene 23: 9090–9101.

  6. Goyette MC, Cho K, Fasching CL, Levy DB, Kinzler KW, Paraskeva C et al. (1992). Mol Cell Biol 12: 1387–1395.

  7. Hesson L, Bieche I, Krex D, Criniere E, Hoang-Xuan K, Maher ER et al. (2004). Oncogene 23: 2408–2419.

  8. Horiguchi K, Tomizawa Y, Tosaka M, Ishiuchi S, Kurihara H, Mori M et al. (2003). Oncogene 22: 7862–7865.

  9. Hu CP, Hsieh HG, Chien KY, Wang PY, Wang CI, Chen CY et al. (1984). J Natl Cancer Inst 72: 577–583.

  10. Hu YC, Lam KY, Law SY, Wan TS, Ma ES, Kwong YL et al. (2002). Cancer Genet Cytogenet 135: 120–127.

  11. Ko JM, Wong CP, Tang CM, Lau KW, Lung ML . (2001). Cancer Lett 170: 131–138.

  12. Ko JM, Yau WL, Chan PL, Lung HL, Yang L, Lo PH et al. (2005). Genes Chromosomes Cancer 43: 284–293.

  13. Li JY . (1982). Natl Cancer Inst Monogr 62: 113–120.

  14. Lo PH, Chung Leung AC, Xiong W, Law S, Duh FM, Lerman MI et al. (2006). Cancer Lett 234: 184–192.

  15. Lung HL, Bangarusamy DK, Xie D, Cheung AKL, Cheng Y, Kumaran MK et al. (2005). Oncogene 24: 6525–6532.

  16. Lung HL, Cheng Y, Kumaran MK, Liu ET, Murakami Y, Chan CY et al. (2004). Int J Cancer 112: 628–635.

  17. Mok CH, Chew EC, Riches DJ, Lee JC, Huang DP, Hadgis C et al. (1987). Anticancer Res 7: 409–415.

  18. Pan QQ . (1989). Proc Chin Acad Med Sci Peking Union Med Coll 4: 52–57.

  19. Porter S, Scott SD, Sassoon EM, Williams MR, Jones JL, Girling AC et al. (2004). Clin Cancer Res 10: 2429–2440.

  20. Qin YR, Wang LD, Kwong D, Gao SS, Guan XY, Zhuang ZH et al. (2005). Zhonghua Bing Li Xue Za Zhi 34: 80–83.

  21. Rimessi P, Gualandi F, Morelli C, Trabanelli C, Wu Q, Possati L et al. (1994). Oncogene 9: 3467–3474.

  22. Rooney DE . (1986). In Human Cytogenetics: A Practical Approach. Czeoulkowski BM (ed). IRL Press: Washington, DC.

    Google Scholar 

  23. Sambrook J, Fritsch EF, Maniatis T . (1989). Molecular Cloning. Cold Spring Harbor Laboratory Press: NY.

    Google Scholar 

  24. Saxon PJ, Srivatsan ES, Leipzig GV, Sameshima JH, Stanbridge EJ . (1985). Mol Cell Biol 5: 140–146.

  25. Sekine I, Sato M, Sunaga N, Toyooka S, Peyton M, Parsons R et al. (2005). Oncogene 24: 2735–2738.

  26. Shimada Y, Imamura M, Wagata T, Yamaguchi N, Tobe T . (1992). Cancer 69: 277–284.

  27. Shiomi H, Sugihara H, Kamitani S, Tokugawa T, Tsubosa Y, Okada K et al. (2003). Cancer Genet Cytogenet 147: 50–61.

  28. Somerville RP, Longpre JM, Jungers KA, Engle JM, Ross M, Evanko S et al. (2003). J Biol Chem 278: 9503–9513.

  29. Takahashi K, Kanazama H, Chan CT . (1990). Jpn J Oral Maxillofac Surg 36: 307–316.

  30. Tanaka H, Shimizu M, Horikawa I, Kugoh H, Yokota J, Barrett JC et al. (1998). Genes Chromosomes Cancer 23: 123–133.

  31. Tang JC, Wan TS, Wong N, Pang E, Lam KY, Law SY et al. (2001). Cancer Genet Cytogenet 124: 36–41.

  32. Tischoff I, Markwarth A, Witzigmann H, Uhlmann D, Hauss J, Mirmohammadsadegh A et al. (2005). Int J Cancer 115: 684–689.

  33. Uzawa N, Yoshida MA, Hosoe S, Oshimura M, Amagasa T, Ikeuchi T . (1998). Cancer Genet Cytogenet 107: 125–131.

  34. Wang L, Li W, Wang X, Zhang C, Zhang T, Mao X et al. (1996). Oncogene 12: 699–703.

  35. Xiong W, Zeng ZY, Xia JH, Xia K, Shen SR, Li XL et al. (2004). Cancer Res 64: 1972–1974.

  36. Yang L, Leung AC, Ko JM, Lo PH, Tang JC, Srivastava G et al. (2005). Oncogene 24: 697–705.

  37. Yang Y, Kost-Alimova M, Ingvarsson S, Qianhui Q, Kiss H, Szeles A et al. (2001). Proc Natl Acad Sci USA 98: 1136–1141.

Download references


We acknowledge the financial support of the Research Grants Council of the Hong Kong Special Administration Region, China to MLL (HKUST6106/00M).

Author information



Corresponding author

Correspondence to M L Lung.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Lo, P., Leung, A., Kwok, C. et al. Identification of a tumor suppressive critical region mapping to 3p14.2 in esophageal squamous cell carcinoma and studies of a candidate tumor suppressor gene, ADAMTS9. Oncogene 26, 148–157 (2007).

Download citation


  • esophageal carcinoma
  • tumor suppressor gene
  • chromosome 3
  • microcell-mediated chromosome transfer

Further reading


Quick links