Genetic alterations and chemosensitivity profile in newly established human renal collecting duct carcinoma cell lines

Authors

  • Zheng-Sheng Wu,

    1. Department of Pathology, Korea University Ansan Hospital, Ansan, and, Department of Laboratory Medicine, Korea University Guro Hospital, Seoul, Korea
    Search for more papers by this author
  • Ju-Han Lee,

    1. Department of Pathology, Korea University Ansan Hospital, Ansan, and, Department of Laboratory Medicine, Korea University Guro Hospital, Seoul, Korea
    Search for more papers by this author
  • Jung-Ah Kwon,

    1. Department of Pathology, Korea University Ansan Hospital, Ansan, and, Department of Laboratory Medicine, Korea University Guro Hospital, Seoul, Korea
    Search for more papers by this author
  • Seo-Hee Kim,

    1. Department of Pathology, Korea University Ansan Hospital, Ansan, and, Department of Laboratory Medicine, Korea University Guro Hospital, Seoul, Korea
    Search for more papers by this author
  • Sun-Hee Han,

    1. Department of Pathology, Korea University Ansan Hospital, Ansan, and, Department of Laboratory Medicine, Korea University Guro Hospital, Seoul, Korea
    Search for more papers by this author
  • Jung-Suk An,

    1. Department of Pathology, Korea University Ansan Hospital, Ansan, and, Department of Laboratory Medicine, Korea University Guro Hospital, Seoul, Korea
    Search for more papers by this author
  • Ji-Hye Lee,

    1. Department of Pathology, Korea University Ansan Hospital, Ansan, and, Department of Laboratory Medicine, Korea University Guro Hospital, Seoul, Korea
    Search for more papers by this author
  • Eung-Seok Lee,

    1. Department of Pathology, Korea University Ansan Hospital, Ansan, and, Department of Laboratory Medicine, Korea University Guro Hospital, Seoul, Korea
    Search for more papers by this author
  • Heum-Rye Park,

    1. Department of Pathology, Korea University Ansan Hospital, Ansan, and, Department of Laboratory Medicine, Korea University Guro Hospital, Seoul, Korea
    Search for more papers by this author
  • Young-Sik Kim

    1. Department of Pathology, Korea University Ansan Hospital, Ansan, and, Department of Laboratory Medicine, Korea University Guro Hospital, Seoul, Korea
    Search for more papers by this author

  • These two first authors contributed equally to this work

Young-Sik Kim, Department of Pathology, Korea University Ansan Hospital, 516, Gojan-1 Dong, Danwon-Gu, Ansan-Si, Gyeonggi-Do 425-707, Republic of Korea. e-mail: apysk@korea.ac.kr

Abstract

OBJECTIVE

To determine the genetic alterations and chemosensitivity profile of collecting duct carcinoma (CDC) of the kidney, as it is a rare, highly aggressive malignant tumour with frequent distant metastases.

MATERIALS AND METHODS

We first established and characterized two human CDC cell lines designated AP3 and AP8, respectively. The CDC cell lines were assessed using microarray-based comparative genomic hybridization and chemosensitivity testing.

RESULTS

The CDC cells grew in vitro as an adherent monolayer with epithelial morphology, but had different growth rates. The cell lines had the characteristic immunophenotype of CDC (high molecular weight cytokeratin-+ve/cytokeratin 7-+ve/vimentin-+ve). Both cell lines shared copy number gains in chromosomes 20 and X. The loci showing a copy number gain were SOX22 at 20p tel, topoisomerse I (TOP1) at 20q12-q13.1, TPD52L2 at 20q tel, 20QTEL14 at 20q tel, KAL at Xp22.3, STS 5′ at Xp22.3, OCRL1 at Xq25, AR3′at Xq11-q12, and XIST at Xq13.2, respectively. Immunoblot analysis confirmed that the AP3 and AP8 cell lines showed moderate and high levels of TOP1 expression, respectively. By chemosensitivity testing, the AP8 cells were most sensitive to topoisomerase I and II inhibitors such as topotecan, epirubicin and doxorubicin, but the AP3 cells did not. The chemosensitivity to these drugs was paralleled by cell death via apoptosis.

CONCLUSION

The results suggest that TOP1 might be one of the molecular targets in AP8 CDC cells. Thus, these novel CDC cell lines will be useful for discovering therapeutic targets and developing effective anticancer drugs against CDC.

Abbreviations
CDC

collecting duct carcinoma

CGH

comparative genomic hybridization

IC50

50% inhibitory concentration

FBS

fetal bovine serum

HMW

high molecular weight

TOP

topoisomerase

5-FU

5-fluorouracil.

INTRODUCTION

Collecting duct (Bellini duct) carcinoma (CDC) is a rare aggressive form of RCC, accounting for <1% of all RCCs [1]. Of patients with CDC, ≈44% present with nodal metastasis and 32% have distant metastasis, including to the lung and bone [2]. A typical CDC is a grey-white solid neoplasm with infiltrative borders centred on the renal medulla. The common histological pattern is that of a tubular or tubulopapillary carcinoma with a desmoplastic stroma [3].

The chemotherapeutic drugs most commonly used in patients with advanced CDC are methotrexate, vinblastine, doxorubicin and cisplatin [2]. However, it has been reported that combined cisplatin and gemcitabine [4], doxorubicin and gemcitabine [5] or paclitaxel and carboplatin therapy [6,7] show a partial response. Thus, understanding the genetic basis or the gene expression profile of this rare type of kidney cancer will provide the opportunity to develop target-specific therapy and to understand its aggressive biological behaviour. The establishment of a permanent cell line from CDC is mandatory as a first step to fulfil this goal.

However, there have been great difficulties in establishing a permanent cell line from CDC, when considering its lack of clear pathological criteria and its aggressive clinical course, as well as its very low incidence. To the best of our knowledge, we have, for the first time, established and characterized two human CDC cell lines, designated AP3 and AP8. To identify the genetic alterations of these cell lines, we carried out conventional karyotyping and microarray-based comparative genomic hybridization (CGH). For chemosensitivity profiling of the CDC cell lines, we determined the 50% inhibitory concentration (IC50) values of anticancer drugs using a cell proliferation assay, and measured the caspase 3/7 activity.

MATERIALS AND METHODS

The AP3 and AP8-CDC cancer cell lines were established from a 66-year-old CDC male patient and a 61-year-old CDC female patient, respectively. The former patient was admitted due to a whitish large solid mass in the left kidney and multiple liver metastases. The patient died 4 months after partial resection of the left kidney. The latter patient was admitted due to a huge solid mass in the left kidney, but died 1 month after radical nephrectomy. This study was conducted in accordance with the guidelines of the Institutional Human Tissue Committee.

To establish the cell lines, the fresh tumour tissue was finely minced and suspended with 2 mL of collagenase I (Sigma-Aldrich, St. Louis, MO, USA). The pellet was seeded into 25-cm2 culture flasks in a 5% CO2 humidified incubator. The tumour cells in Dulbecco modified Eagle medium were supplemented with 20% fetal bovine serum (FBS, Invitrogen, Carlsbad, CA, USA). There was exponential growth of the cells ≈3 weeks after primary culturing. Subsequently, these cells were maintained in RPMI 1640 with 10% FBS.

The surgical specimen and cultured cells were evaluated by immunohistochemistry using a ChemMate EnVision Kit (Dako, Carpinteria, CA, USA). The primary antibodies were anti-high molecular weight (HMW) cytokeratin (1:100 dilution, clone 34βE12, Dako), cytokeratin (1:100, clone AE1/AE3, Zymed Laboratory, CA, USA), CD10 (1:100, clone 56C6, Novocastra, Newcastle, UK), vimentin (1:100, clone V9, Zymed Laboratory), cytokeratin 7 (1:100, clone DV-TL 12/30, Dako), cytokeratin 20 (1:100, clone Ks20.8, Dako), TFE-3 (1:300, polyclonal, Santa Cruz Biotech. Santa Cruz, CA, USA), and c-KIT (1:250, polyclonal, Dako). The Ki-67 labelling indices of original tumours were quantified with ImageJ software.

To assess the growth kinetics, each of the 33 replicate 25-cm2 culture flasks received inocula of 4.2 × 105 cells of the AP3 cell line and 5.5 × 105 cells of the AP8 cell line, respectively. Cells from three culture flasks were harvested separately from 3 days after inoculation. The viable cells were counted. The doubling time of the cell population was estimated in a logarithmic growth phase.

For chromosome karyotyping, the cytogenetics were analysed in the AP3 cell line at passage 9 and in the AP8 cell line at passage 15. The cells were prepared for karyotyping and the chromosome spreads were banded by the Giemsa-trypsin method. The karyotypes were described according to guidelines of the International System for Human Cytogenetic Nomenclature.

Microarray CGH was conducted and analysed using the GenoSensor Array 300 system, according to the manufacturer’s instructions (Abbott-Vysis, Downers Grove, IL, USA). The median of the three spots of each probe in the array was calculated and its log2 transformed value was used for further analysis. A fluorescence ratio >1.25 (log2 ratio = 0.32) was considered as a DNA gain, while DNA losses were scored when the ratio was <0.75 (log2 ratio = −0.41). A ratio of >2 (log2 ratio = 1) was considered as a high copy number amplification. For each target, a P value was calculated automatically by the GenoSensor Array 300 Reader Software, and P < 0.01 was considered to indicate significance.

For Western blot analysis, cell pellets were lysed using radio-immunoprecipitation assay buffer. The protein from each sample (40 µg per lane) was loaded and resolved by 8% SDS-PAGE. Primary monoclonal antibodies to topoisomerase I (TOP1, C-21, Santa Cruz Biotech.) and β-actin (Sigma-Aldrich) were used. The blots were visualized with chemiluminescence luminol reagent (Santa Cruz Biotech).

The anticancer drugs used were as follows; 5-fluorouracil (5-FU, Choongwae Pharma, Seoul, Korea), gemcitabine (Dong-A Pharma, Seoul, Korea), methotrexate (Choongwae Pharma), oxaliplatin (Sanofi-Aventis Korea, Seoul, Korea), cisplatin (Dong-A Pharma), carboplatin (Boryung Pharma, Seoul, Korea), irinotecan (CJ Pharm., Seoul, Korea), topotecan (GSK Korea, Seoul, Korea), etoposide (Dong-A Pharma), doxorubicin (Dong-A Pharma), epirubicin (Ildong Pharma, Seoul, Korea), paclitaxel (Hanmi Pharm, Seoul, Korea), and vinblastine (Myung Ji Pharm, Seoul, Korea). Trichostatin A and sodium butylate were obtained from Sigma.

In vitro chemosensitivity was determined by a cell proliferation assay using CellTiter 96® AQueous One Solution Reagent, according to the manufacturer’s instructions (Promega, Madison, WI, USA). The CDC cells were seeded at a concentration of 5 × 103 cells/well. After 24 h, the cells were incubated in the presence of each concentration of 0 (control), 0.05, 0.5, 2.5, 5, 10, 25, 50 and 100 µm of anticancer drugs for another 48 h in 5% CO2 humidified incubator. After treatments with the CellTiter reagent, the absorbance at 490 nm was measured. Each condition was applied in eight wells and each experiment was repeated twice. The antiproliferative activities of the anticancer drugs were expressed as the IC50 (compared to the controls). The IC50 values were calculated by regression analyses.

To assess caspase 3/7 activity, the CDC cells were seeded on 96-well white tissue culture plates at a concentration of 1 × 104 cells/well. After 24 h, the cells were incubated in the presence of each concentration of 0 (control), 0.05, 0.5, 2.5, 5, 10, 25, 50 and 100 µm of anticancer drugs for another 24 h in 5% CO2 humidified incubator. After treatment with Caspase-Glo 3/7 reagent (Promega), luminescence was measured using a GloMax luminometer (Promega). The relative luminescence units were calculated based on the activity measured from the untreated controls.

RESULTS

Grossly, the original AP3 tumour (2.5 × 2.5 × 1 cm) was partially resected. The tumour had a whitish solid cut surface with focal necroses. The original AP8 tumour, (8 × 6.5 × 4 cm) was a yellowish-white solid mass with irregular infiltrative borders. The tumour involved the upper four-fifths of the left kidney and invaded the renal pelvis, sinus, perinephric fat, and renal vein. Microscopically, both tumours consisted of irregular angulated tubules and papillae in a desmoplastic stroma. The CDC cell lines grew in vitro as an adherent monolayer with a polyhedral or oval epithelial appearance (Fig. 1).

Figure 1.

Histological examination and immunohistochemistry of the primary CDC tissues and cell lines. Haematoxylin and eosin staining of the original tumour shows irregular angulated tubules and papillae in a desmoplastic stroma (original × 200). The culture cells show an epithelial morphology (original × 400). The tumour cells from the primary tumour and the cell lines are all strongly positive for HMW cytokeratin, cytokeratin 7 (CK7) and vimentin (Vim); (original × 400).

Immunohistochemically, the tumour cells in the tissue sections and in the cultured cell lines were positive for cytokeratin (AE1/AE3), HMW cytokeratin, cytokeratin 7 and vimentin (Fig. 1), but negative for c-KIT, CD10, TFE-3, and cytokeratin 20. The Ki-67 labelling indices of AP3 and AP8 tumours were 26% and 33%, respectively.

The AP8 cells had a shorter mean population doubling time (29 h) than the AP3 cells (72 h) (Fig. 2). The permanent cell lines were established by culturing over >20 consecutive passages.

Figure 2.

Growth curves of AP3 and AP8 cells. Each value represents the mean (sd) of three samples.

On chromosomal analysis, we counted 20 metaphases of the AP3 cell line at passage 9. The karyotype was 52, XY, +X, +1, i(1)(q10), +3, −4, +7, del(7)(q21q31),?del(9)(p22)×2, + der(9)t(9;?)(q22;?), dic(15;16)(p11.2;q34), del(17)(q25), +20, +21 and +r (Fig. 3A). Twenty-five metaphases of the AP8 cell line at passage 15 were counted and two populations were found. The karyotype of the main population (22 cells) was 49, XX, +X, dup(6)(p21.3p25), del(9)(p22), der(9)t(9;?)(p22;?), del(10)(q24), i(10)(q10), + 11,?del(13)(q14), der(17)t(17; 17)(p13;q21), + 20 and der(21)t(21;?)(q22;?) (Fig. 3B). The karyotype of the other population (three cells) was 48, X, +X, dic(X;?8)(p11.2;?p11.2), dup(6)(p21.3p25)×2, der(7)t(3;7)(p25;p22),?dic(8;9)(p11.2;p13), del(9)(p22), der(9)t(9;?)(p22;?), del(10)(q24), i(10)(q10),?del(13)(q14), +20 and der(21)t(21;?)(q22;?).

Figure 3.

Representative G-banded karyotype of AP3 (A) and AP8 (B) cells. Arrows indicate structural abnormalities. A ring chromosome is denoted by the symbol r.

The copy number abnormalities of the AP3 and AP8 genomes are shown in Tables 1 and 2, respectively. The loci that shared a chromosomal copy number gain or loss in the two cell lines are shown in bold type. The AP3 and AP8 cell lines contained common copy number gains in chromosomes 20 and X. The regions and loci showing a copy number gain were detected by array CGH in both cell lines, and these were as follows: SOX22 at 20p tel, TOP1 at 20q12-q13.1, TPD52L2 at 20q tel, 20QTEL14 at 20q tel, KAL at Xp22.3, STS5′ at Xp22.3, OCRL1 at Xq25, AR3′ at Xq11-q12, and XIST at Xq13.2, respectively. However, no loci that showed common copy number losses in both cell lines were found. The AP3 cell line had copy number losses in chromosomes 4 and 9; the AP9 cell line had copy number losses in chromosomes 9, 10, and 13.

Table 1.  Copy number abnormalities of the AP3 genome detected by array-CGH analysis; genes of interest are shown in bold
Chromosome no.Locus nameCytogenetic locationAP3 (23 targets)
Mean mass ratioGain/lossNon-modal P
1SHGC-182901q tel1.5Gain<0.001
11QTEL101q tel1.63Gain<0.001
4D4S1144p16.30.66Loss<0.001
4PDZ-GEF14q32.10.68Loss<0.001
9D9S9139p tel0.65Loss<0.001
XOCRL1Xq251.57Gain<0.001
1LAMC21q25-q311.44Gain<0.001
4SHGC4-2074p tel0.72Loss<0.001
4WHSC14p16.30.69Loss<0.001
9AF1702769p tel0.73Loss<0.001
3RASSF13p21.31.34Gain0.001
4GS10K2/T74p tel0.72Loss0.001
2020QTEL1420q tel1.38Gain0.001
1WI-5663, WI-134141q211.46Gain0.002
XAR 3′Xq11-q121.39Gain0.002
77QTEL207q tel1.33Gain0.005
20SOX2220p tel1.29Gain0.005
20TOP120q12-q13.11.31Gain0.005
2121QTEL0821q tel1.37Gain0.005
XKALXp22.31.3Gain0.005
20TPD52L2, TOM20q tel1.31Gain0.01
XSTS 5′Xp22.31.33Gain0.01
XXISTXq13.21.26Gain0.01
Table 2.  Copy number abnormalities of the AP8 genome detected by array-CGH analysis
Chromosome no.Locus nameCytogenetic locationAP8 (56 targets)
Mean mass ratioGain/lossNon-modal P
3EIF 5A23q26.21.45Gain<0.001
3TP633q27-q291.67Gain<0.001
9MTAP9p21.30.61Loss<0.001
9CDKN2A(p16),MTAP9p210.4Loss<0.001
1010PTEL00610p tel0.63Loss<0.001
10D10S249,D10S53310p150.54Loss<0.001
10GATA310p150.65Loss<0.001
10WI-2389,D10S126010p14-p130.62Loss<0.001
10BMI110p130.66Loss<0.001
10D10S16710p11–10q110.56Loss<0.001
13D13S2513q14.30.61Loss<0.001
17D17S167017q231.45Gain<0.001
20TOP120q12-q13.11.47Gain<0.001
XSTS 3′Xp22.31.71Gain<0.001
XSTS 5′Xp22.31.89Gain<0.001
XKALXp22.31.75Gain<0.001
XDMD exon 45–51Xp21.11.97Gain<0.001
XDXS580Xp11.21.59Gain<0.001
XDXS7132Xq121.56Gain<0.001
XAR 3′Xq11-q122.07Gain<0.001
XXISTXq13.21.59Gain<0.001
XOCRL1Xq252.38Gain<0.001
2020QTEL1420q tel1.43Gain<0.001
3TERC3q261.41Gain<0.001
10SHGC-4425310p tel0.7Loss<0.001
20NCOA3(AIB1)20q121.41Gain<0.001
3PIK3CA3q26.31.35Gain0.001
10DMBT110q25.3-q26.11.35Gain0.001
20JAG120p12.1-p11.231.36Gain0.001
33QTEL053q tel1.37Gain0.002
5D5S235p15.20.74Loss0.002
10EGR210q21.31.32Gain0.002
13RB113q140.75Loss0.002
17282M15/SP617ptel0.74Loss0.002
20SOX2220p tel1.33Gain0.002
8D8S5048p tel0.77Loss0.005
10FGFR210q261.29Gain0.005
13D13S31913q14.20.72Loss0.005
17PPARBP(PBP)17q121.32Gain0.005
17AFM217YD1017q tel1.29Gain0.005
2020PTEL1820p tel1.28Gain0.005
20MYBL220q13.11.3Gain0.005
20ZNF217(ZABC1)20q13.21.3Gain0.005
20CYP2420q13.21.29Gain0.005
20TPD52L2,TOM20q tel1.28Gain0.005
8CTSB8p220.78Loss0.01
11CDKN1C(p57)11p15.51.26Gain0.01
11KAI111p11.21.25Gain0.01
11D11S46111q12.21.27Gain0.01
11RDX11q22.31.24Gain0.01
11MLL11q231.28Gain0.01
18FRA18A(D18S978)18q12.30.79Loss0.01
20CSE1L(CAS)20q131.28Gain0.01
20PTPN120q13.1-q13.21.29Gain0.01
20STK6(STK15)20q13.2-q13.31.24Gain0.01
22PDGFB(SIS)22q13.10.79Loss0.01

Immunoblot analysis using the whole-cell lysates showed that the expression of TOP1 was increased about 1.5–3 times in the CDC cells (AP3 and AP8 cells) than in the clear cell RCC cells (the Caki-1 and Caki-2 cells) (Fig. 4).

Figure 4.

DNA TOP1 expression in clear cell carcinoma (Caki-1 and Caki-2) and CDC cell lines (AP3 and AP8) by Western blot analysis. The relative band intensities at the bottom of the blot were measured with ImageJ software. β-actin was used as a control for loading.

The CDC cell lines were treated with different concentrations of 15 anticancer drugs for 48 h, and the dose-dependent curves were determined (Figs 5,6). The known modes of actions and the IC50 values of 15 anticancer drugs in the CDC cell lines are summarized in Table 3. The IC50 values were 0.6–>100 µm. Interestingly, both cell lines had rather different patterns of chemosensitivity profiles (Figs 5,6). The AP3 cell line was responded poorly to vinblastine, trichostatin A or epirubicin; the IC50 of these drugs were >1 µm (Fig. 5; Table 3). By contrast, DNA topoisomerase I (TOP1) and II (TOP2) inhibitors such as topotecan, epirubicin and doxorubicin were the most effective drugs for the AP8 cell line (Fig. 6). The IC50 for epirubicin and doxorubicin were <1 µm. The IC50 for topotecan was 2 µm, which was <2.2 µm, the clinically achievable peak plasma concentration of this drug (Table 3). The peak plasma concentrations of investigational chemotherapeutic drugs are very useful to interpret in vitro drug sensitivity data in relation to their relevance for clinical effectiveness [8]. However, although irinotecan and etoposide target TOP1 and TOP2, respectively, they were not effective for inhibiting the growth of the AP8 cell line (Fig. 6C,D). The CDC cell line was also sensitive to vinblastine, paclitaxel (Fig. 6E), trichostatin A (Fig. 6F) and cisplastin (Fig. 6B), in descending order.

Figure 5.

Chemosensitivity profile in AP3 cell line by CellTiter assay. The AP3 cells were treated with different concentrations of 15 anticancer drugs for 48 h, including (A) 5-FU, gemcitabine, methotrexate (B) oxaliplatin, cisplatin, carboplatin (C) irinotecan, topotecan (D) etoposide, doxorubicin, epirubicin (E) paclitaxel, vinblastine (F) trichostatin A and sodium butylate. The dose-dependent curves were then determined. DHFR and HDAC indicate dihydrofolate reductase and histone deacetylase, respectively.

Figure 6.

Chemosensitivity profile in AP8 cell line by CellTiter assay. The AP8 cells were treated with different concentrations of 15 anticancer drugs for 48 h, including (A) 5-FU, gemcitabine, methotrexate (B) oxaliplatin, cisplatin, carboplatin (C) irinotecan, topotecan (D) etoposide, doxorubicin, epirubicin (E) paclitaxel, vinblastine (F) trichostatin A and sodium butylate. The dose-dependent curves were then determined. DHFR and HDAC indicate dihydrofolate reductase and histone deacetylase, respectively.

Table 3.  IC50 values of anticancer drugs for CDC cell lines
DrugTarget/mode of actionmean IC50, µm
AP3AP8
  1. HDAC, histone deacetylase; DHFR, dihydrofolate reductase.

5-FUPyrimidine>10012.1
GemcitabinePyrimidine30.1>100
MethotrexateDHFR72.8>100
OxaliplatinDNA cross-linker42.811.7
CisplatinDNA cross-linker26.8 6.9
CarboplatinDNA cross-linker>100>100
IrinotecanTOP I>10039.4
TopotecanTOP I39.4 2.0
EtoposideTOP II79.4>100
DoxorubicinTOP II25.2 0.8
EpirubicinTOP II 3.9 0.6
PaclitaxelTubulin 6.9 3.1
VinblastineTubulin 1.7 2.6
Trichostatin AHDAC inhibitor 3.1 4.1
Sodium butylateHDAC inhibitor>100>100

To confirm whether the responsiveness of the chemotherapeutic agents was due to apoptosis, we also measured the caspase 3/7 activity in the CDC cells (data not shown). The caspase 3/7 activity of AP3 cell line was increased when treated with 0.5 µm trichostatin A, and 2.5 µm of epirubicin and doxorubicin each. By contrast, the caspase 3/7 activity of AP8 cell line was abruptly increased when treated with 0.5 µm of topotecan and 2.5 µm of epirubicin and doxorubicin each. The induction of apoptosis at these concentrations of anticancer drugs was paralleled by the results of the chemosensitivity.

DISCUSSION

The CDC cell lines strongly express CDC-specific antigens, especially HMW cytokeratin, cytokeratin 7 and vimentin [3]. However, there was some variation in HMW cytokeratin and cytokeratin 7 expression between tissue sections and cultured cells. The staining discrepancy for these antigens might be due to the technical difficulties in immunohistochemistry on cultured cells, which heavily depend on the proper cellular fixation and permeabilization. In addition, tissues provide a static picture of protein expression, while cells in in vitro culture provide a dynamic process of protein expression. The main differential diagnosis of CDC includes papillary RCC, urothelial carcinoma with glandular differentiation, and adenocarcinoma arising from the urothelium of the renal pelvis [3]. Papillary RCC is positive for cytokeratin 7 but it is frequently negative for HMW cytokeratin and vimentin. By contrast, CDC usually reacts with antibodies to cytokeratin 7, HMW cytokeratin and vimentin. Urothelial carcinomas express cytokeratin 5/6, cytokeratin 7, cytokeratin 17, cytokeratin 20 and HMW cytokeratin. However, vimentin is rare in urothelial carcinoma [3]. Taken together, the CDC cell lines retained the characteristic immunophenotype of CDC (HMW cytokeratin+/cytokeratin 7+/vimentin+).

Chromosome analyses of the AP3 and AP8 cell lines suggest that they derived from malignant cells with highly complex chromosomal imbalances. AP3 showed losses of chromosomes 4 and 9p and gains of chromosomes 1, 3, 7, 20, 21, and X. AP8 had losses of chromosomes 9p, 10 and 13, and gains of chromosomes 11, 20 and X. The common alterations of these two cell lines include the loss of chromosome 9p and gains of chromosomes 20 and X. The current study has the limitation that cytogenetic analyses were not performed on earlier passages of the CDC cell lines. The chromosome alterations of CDC are known to be highly variable. The only common alterations include losses of chromosomes 1, 6, 14, 15 and 22 and gain of chromosome 3 [9]. In our study, chromosome 3 trisomy in the AP3 cell line was linked to the gain of RASSF1 at 3p21.3. This gain of chromosome 3 is in contrast with the fact that this region of clear cell RCCs is known to be frequently lost [10]. The CGH analysis of AP8 cells also showed the gains of EIF5A2 at 3q26.2 and TP63 at 3q27–29 without structural abnormalities. Previous large- scale studies suggested that CDCs are associated with the losses of chromosomes 1q32, 6p, 8p and 21q [11–13]. Monosomies of chromosomes 18 and 21, loss of chromosome Y, and gains of chromosomes 7, 12, 17 and 20 have also been reported [14,15]. Antonelli et al.[16] proposed that CDCs have many numerical and structural chromosomal aberrations involving chromosome 1 and the autosomes. The loci showing chromosome 9p loss include D9S913 at 9p tel in AP3 cells, and MTAP at 9p21.3 and p16 INK4a at 9p21 in AP8. Absent or low expression of p16INK4a was also found in ≈90% of clear cell RCCs as a poor prognostic factor [17]. In addition, AP8 cells showed losses of GATA3 at 10p15 and D13S25 at 13q14.3. GATA3 expression is mostly lost in all subtypes of RCCs but up-regulated in urothelial carcinomas [18]. Schoenberg et al.[19] reported that tumour suppressor genes on 13q are involved in the development of CDC. The CDC cell lines showed complex cytogenetic alterations including losses of chromosomes 4, 9p, 10, and 13, and gains of chromosomes 1, 3, 7, 11, 20, 21, and X.

The genes in the CDC cell lines showing copy number gains were SOX22, TOP1, TPD52L2, KAL, OCRL1, AR and XIST. Among these, TOP1 relaxes DNA supercoiling by nicking the DNA and enabling the broken strand to rotate around the TOP1-bound DNA strand [20,21]. Camptothecins stabilize the trapped TOP1-DNA cleavage complex. When the replication forks encounter the TOP1-cleavage complexes, the replication machinery induces irreversible double- strand breaks and subsequent cell death via apoptosis [20].

The growth of AP8 cells was inhibited by treatment with topotecan, epirubicin or doxorubicin, whereas the AP3 cells did not respond to all anticancer drugs. Among these, the IC50 of topotecan (2 µm) in this study was also at the clinically relevant concentration (≤1 µg/mL) [22]. Amplification of chromosome 20q (TOP1) and up-regulation of TOP1 gene expression appear to play a role in the chemosensitivity of AP8 cells to topotecan. Topotecan and irinotecan are camptothecin derivatives that inhibit TOP1 [20]. However, irinotecan was not effective in suppressing the growth of AP8 cells. Irinotecan should be converted to its active metabolite, SN-38, by carboxyesterases to exert its anticancer effects [20]. As the activity of carboxyesterase might determine the effect of irinotecan on cancer cells, further studies are needed to explain the different chemosensitivity between topotecan and irinotecan. The AP8 cell line also responded to TOP2 inhibitors such as epirubicin and doxorubicin, but not etoposide. Although these drugs target TOP2, the anthracyclines, including epirubicin and doxorubicin, produce a wide-range of biological reactions in addition to TOP inhibition. By contrast, the cytotoxicity of etoposide is mainly due to its effect on TOP2 [23]. Taken together, this study shows the potential value of the TOP1 expression level or chromosome 20q amplification status in CDC.

In conclusion, the newly established CDC cell lines share copy number gains involving TOP1. TOP1 and TOP2 inhibitors suppress the growth of AP8 cells, suggesting that TOP1 might be one of the molecular targets in AP8 cells. Taken together, the CDC cell lines will be valuable tools to explore its potential therapeutic targets.

ACKNOWLEDGEMENTS

The authors thank Sung-Su Lee, Sang-Ju Lee, Mi-Ran Jeong, Eun-Ja Lee, and Jee-Hye Ok for their technical assistance. This work was supported by a grant from Korea University College of Medicine. Current address of Z-S Wu is the Department of Pathology, Anhui Medical University, P.R. China.

CONFLICT OF INTEREST

None declared.

Ancillary