Establishment and characterization of two distinct malignant mesothelioma cell lines with common clonal origin

Cancer Genetics and Cytogenetics 176 (2007) 35e47

Establishment and characterization of two distinct malignant mesothelioma cell lines with common clonal origin Hans Ju¨rgen Schultena, Christina Perskea, Paul Thelenb, Andreas Poltenc, Christoph Borstd, Bastian Gunawana, Holger Nagela,e,* a

Institute of Pathology, bDepartment of Urology, Georg-August-University of Go¨ttingen, Robert-Koch-Strabe 40, 37099 Go¨ttingen, Germany c Agilent Technologies, Hewlett-Packard-Strabe 8, 76337 Waldbronn, Germany d Institute of Hematology/Oncology and Nephrology, Friedrich Ebert Hospital, Friesenstrabe 11, 24534 Neumu¨nster, Germany e Department of Pathology, Cantonal Hospital Lucerne, Spitalstrasse, CH-6000 Lucerne 16, Switzerland Received 18 December 2006; received in revised form 16 March 2007; accepted 29 March 2007

Abstract

We describe two newly established malignant mesothelioma (MM) cell lines derived from a pleural effusion of a male. One cell line, designated as MM-Z03E, reveals an epithelioid cobblestone morphology, while the second one, designated as MM-Z03S and subcloned after in vivo selection, exhibits a sarcomatoid storiform growth pattern. Both cell lines showed the immunologic profile characteristic for MM (i.e., expression of cytokeratin, CK18, calretinin, and vimentin in both phenotypes). Cytogenetics, multicolor fluorescence in situ hybridization, comparative genomic hybridization, and oligonucleotide array CGH were performed on both cell lines. Aberrations shared by both cell lines included chromosomal losses of 1q34~qter, 4, 9p, 10p, 13, 14, 16q, 18, and 22, as well as a complex structural aberration involving chromosome 17. Aberrations exclusive to MMZ03E included gains of 3q11q27 and 5p, while gain of 9q and losses of 3q27qter, 11q, and 18 in MM-Z03S were exclusive to MM-Z03E. Both cell lines were able to develop solid transplant tumors in nude mice within 16 weeks, and immunophenotyping of tumor xenografts revealed an overall retained expression profile of the markers used. Remarkably, one xenograft from MMZ03E revealed overexpression of p53 and widely invasive growth. In conclusion, both cell lines are useful in vivo and in vitro model systems to study the underlying genetic mechanisms of biphasic differentiation in MM, which can be of certain value considering the increasing relevance of assessing MM tumor biology for the clinical management of this disease. Ó2007 Elsevier Inc. All rights reserved.

1. Introduction Malignant mesothelioma (MM) is a highly malignant tumor preferentially arising from the pleural mesothelium, although it is unclear whether the exact origin is either surface mesothelial or submesothelial cells. The current World Health Organization classification distinguishes four major histologic subtypes (epithelioid, sarcomatoid, desmoplastic, and biphasic), of which the biphasic type represents about 30% of cases, provided that each component is present with a fraction of at least 10% [1]. MM is predominantly an asbestos-associated cancer and its incidence is expected to rise in the next years because of the long-term onset of MM after asbestos exposure. The long delay period most likely * Corresponding author. Tel.: þ41-412053491; fax: þ41-412053496. E-mail address: [email protected] (H. Nagel). 0165-4608/07/$ e see front matter Ó 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.cancergencyto.2007.03.005

implies that multiple genetic steps are required for tumorigenesis. Chromosomal imbalances are already present in early-stage mesothelioma [2]. In therapy for MM, some new chemotherapeutic agents such as pemetrexed seem to be a little more effective [3], and new promising therapeutic strategies are currently investigated to improve patient outcome [4]. However, the prognosis of MM, with a median survival of about 6e15 months after diagnosis, is still poor, even in most of those patients with early diagnosis [5]. Patient survival is influenced by histologic subtype, with sarcomatoid histology having the worst prognosis [6]. Tumor cell lines are essential for in vitro model systems to study the biologic features of tumor cells, including sensitivity and response to various agents and physical conditions. Although several MM cell lines have been reported to date, the herein presented MM cell lines have unique features in that they show a common cytogenetic background

H.J. Schulten et al. / Cancer Genetics and Cytogenetics 176 (2007) 35e47

36

with monoclonal origin but divergent cytogenetic evolution and different phenotypes. These characteristics may be of interest when studying the underlying genetic mechanisms of biphasic differentiation. 2. Materials and methods 2.1. Establishment of cell lines An 83-year-old Caucasian male presented in June 2003 with clinically and radiologically (CT scan) suspected malignant mesothelioma of the left pleura. Cytomorphologic and immunocytologic examinations of pleural effusion confirmed the clinical diagnosis, and since clinical and cytologic findings were conclusive, invasive surgical biopsy for histologic examination was abandoned. The patient was treated symptomatically and died of disease 6 months after diagnosis, without subsequent autopsy. The cell lines were established from the malignant pleural effusion on which the morphologic and immunohistochemical diagnosis of MM was based. Briefly, cell sediments were washed in RPMI-1640 (Gibco, Karlsruhe, Germany) supplemented with 10% fetal calf serum (PAA, Pasching, Austria), L-glutamine, and antibiotic mixture (Roche, Mannheim, Germany), and then subsequently cultured at 37 C in humidified 5%CO2/95% air atmosphere. After 2 days, viable cells were firmly attached to the bottom of the culture flasks. The supernatant was carefully replaced by fresh culture medium, and the medium was subsequently replaced at least twice a week. After 4 weeks of cultivation, colonies of polygonal tumor cells with a cobblestone appearance predominated (MM-Z03E). In an early cell culture passage, a minor colony of spindle-shaped cells with a storiform growth pattern could be isolated under microscopic control by careful scraping and were cultured separately (MMZ03S). For long-term cell culture, both cell clones were maintained in 25-cm2 culture flasks. Confluent cells were

Antibody

Code/species

Pretreatment Dilution Source

EMA Cytokeratin CK 5/6 CK 18 BerEP4 Vimentin SM-actin CD 34

E29, mouse, mc MNF-116, mouse, mc M 7237, mouse, mc NCL-5D3, mouse, mc BerEP4, mouse, mc V9, M7019, mouse, mc 1A4, mouse, mc My 10, mouse, mc

Heat Protease Heat Protease Protease Heat Heat Heat

1:50 1:50 1:50 1:100 1:80 1:20 1:400 1:20

P53 CD 117 E-cadherin MIB-1

DO-7, mouse, mc A4502, rabbit, mc NCH-38, mouse, mc M 7240, mouse, mc

Heat Heat Heat Heat

1:10 1:50 1:10 1:20

Dako Dako Dako Novocastra Dako Dako Dako Becton Dickinson Dako Dako Dako Dako

dissociated with 0.25% trypsin (Sigma-Aldrich, Taufkirchen, Germany) in 0.25% ethylenediamine-tetraaceticacid solution and subcultured at a split ratio of 1:2 every 3 days using trypsin/ethylenediamine-tetraaceticacid. Cells were repeatedly examined by phase-contrast microscopy. At various passage numbers, aliquots of cells were resuspended in complete fresh culture medium with 10% dimethyl sulfoxide and then frozen in liquid nitrogen. The study was approved by the local ethics committee. 2.2. Immunohistochemical analyses Immunohistochemistry with primary antibodies listed in Table 1a was performed on cytospin preparations of MM cells in advanced passages and on tumor sections of xenografts. Cytospin preparations were fixed with ice-cold acetone, and for staining, the standard alkaline phosphatase anti-alkaline phosphatase method and new fuchsin as chromogen were used. Paraffin sections from mouse xenografts were dewaxed in xylene and graded alcohol. Slides were stained with an immunostainer (Ventana Nexes, Illkirch Table 2 Immunocytologic features of cell lines

Table 1a Primary antibodies applied for immunocytology Antibody

Code/species

Dilution

EMA Cytokeratin CK 5/6 CK 18 BerEp4 Calretinin Vimentin S100 SM-actin CD 34 CD 117 CEA TTF-1 D2/40 P53 E-cadherin MIB-1

E29, mouse mc KL1-1077, mouse mc D5/16B4, mouse, mc KS-B172, mouse, mc BerEp4, mouse, mc 7699/4, rabbit, pc PN 1144, mouse, mc 1071, rabbit, pc 1a4, mouse, mc 1185, mouse, mc A 4502, rabbit, mc Clone 11-7, mouse, mc 8G7G3/1, mouse, mc M 3619, mouse, mc DO-7, mouse, mc NCH-38, mouse, mc M 7240, mouse, mc

1:50 Dako Ready to use Coulter-Immunotech 1:50 Dako 1:50 Sigma 1:50 Dako 1:2000 Swant Ready to use Coulter-Immunotech Ready to use Coulter-Immunotech 1:60 Coulter-Immunotech 1:10 Coulter-Immunotech 1:50 Dako 1:20 Dako 1:100 Dako 1:100 Dako 1:50 Dako 1:25 Dako 1:100 Dako

Abbreviations: mc, monoclonal; pc, polyclonal.

Table 1b Primary antibodies applied for immunohistology

Source

Antibody

MM-Z03S

MM-Z03E

EMA Cytokeratin CK 5/6 CK 18 BerEP4 Calretinin Vimentin S-100 SM-actin CD 34 CD 117 CEA TTF-1 D2/40 P53 E-Cadherin MIB-1 (%)

(þ) þþ e þþþ e þþ þþþ e e e e e e e (þ) e 50

þþ þþþ (þ) þþþ e þþþ þ e e e e (þ) e þ e þ 30

Staining was scored as follows: e,/(þ)/þ/þþ/þþþ, no/very weak/ weak/moderate/strong expression.

H.J. Schulten et al. / Cancer Genetics and Cytogenetics 176 (2007) 35e47

37

Fig. 1. Phaseecontrast microscopy and immunophenotyping of MM cells in cell culture. MM-Z03E cells (A) are characterized by epithelioid cobblestone morphology, while MM-Z03S cells (B) exhibit a sarcomatoid storiform growth pattern. Cytospin preparations of the epithelioid (C) and sarcomatoid (D) type and immunohistochemistry reveal moderate to strong cytokeratin expression in MM-Z03E (E) and in MM-Z03S cells (F).

Cedex, France). Incubation time for each primary antibody (Table 1b) was 30 minutes followed by visualization using the iView DAB detection kit (Ventana). Suppliers of antibodies were Dako, Hamburg, Germany; Coulter-Immunotech, Marseille, France; Sigma, St. Louis, MO; Swant, Bellinzona, Switzerland; Novocastra, Newcastle-uponTyne, UK; and Becton Dickinson, San Jose, CA. The staining was scored as follows: e/(þ)/þ/þþ/þþþ, no/very weak/weak/moderate/strong expression. 2.3. Proliferation assay Proliferation analyses were performed on MM cells in advanced passages using a bromodeoxyuridine-cell

proliferation ELISA (Roche) in 96-well plates (Nunc, Naperville, IL) according to the manufacturer’s instructions. Briefly, 2 103 tumor cells were seeded in a volume of 200 mL and cultured for 3 days. Five hours before the cells were harvested, BrdU was added to a final concentration of 10 mmol/L. Finally, the color reaction was conducted for 1 hour with anti-BrdU peroxidase and terminated with H2SO4 reagents, and absorbance was measured in a Microplate Reader (Bio-Rad Model 680) at 450 nm (control wavelength 655 nm). All measurements were done in triplicate and repeated once. The results were statistically analyzed with the two-site analysis of variance test for repeated measurements using the GraphPad prism program (GraphPad Software, Inc., San Diego, CA).

H.J. Schulten et al. / Cancer Genetics and Cytogenetics 176 (2007) 35e47

38

Table 3 Morphologic and immunohistologic features of tumors in nude mice MM-Z03S

MM-Z03E

Features

Mouse 1

Mouse 2

Mouse 3

Mouse 4

Max. diameter (mm) Tumor shape Cell shape Invasion Inflammation Fibrosis/hyalinosis Necrosis EMA Cytokeratin CK 5/6 CK 18 BerEP4 Vimentin SM-actin CD 34 P53 CD 117 E-cadherin MIB-1 (%)

11 Round Polygonal (þ) (þ) (þ) e þþ þþþ e þ e þþ e e þ e e 15

12 Oval Polygonal (þ) þ (þ) e þ þþ e þþ e þþþ e e þ e e 20

11 Irregular Polygonal þþþ þ (þ) e þþ þþþ þ þþþ e þþþ e e þþ (þ) e 50

7 Irregular Polygonal (þ) þþ þ e þ þþ (þ) þþ e þþ e e þ (þ) e 35

Staining was scored as follows: e/(þ)/þ/þþ/þþþ, no/very weak/weak/moderate/strong expression.

2.4. Xenotransplantation

2.6. CGH and array CGH

Four male athymic nude mice (BALB/c-nu; Charles River, Sulzfield, Germany) were each inoculated subcutaneously with 1 106 MM cells that were suspended in 200 mL matrigel solution, from either MM-Z03S (mice nos. 1 and 2) or MM-Z03E (mice nos. 3 and 4). Tumor growth was measured weekly during a 16-week period, and solid tumors were resected from mice that had been killed by CO2 inhalation. One part of a tumor specimen was subsequently snap-frozen in liquid nitrogen for comparative genomic hybridization (CGH) analysis, and the other part was formalin-fixed and paraffin-embedded for histologic examination.

CGH was performed on spin columnepurified (Qiagen, Hilden, Germany) DNA from cell culture material obtained from the same passages as used for cytogenetic analysis (53rd and 27th passages, respectively) and from mouse xenograft tumor tissue. All CGH steps were carried out as reported in Gunawan et al. [8], using CGH karyotype and interpretation software (Abbot Molecular/Vysis, Des Plaines, IL). To control the quality of hybridization conducted on xenograft specimens, reference DNA form the opposite gender was used. Oligonucleotide array CGH (aCGH) was performed on both cell lines (53rd and 27th passages, respectively) with human 244A microarrays (Agilent Technologies, Waldbronn, Germany) containing 243,504 features of in situesynthesized 60-mer oligonucleotides. The average spatial resolution of features was approximately 6.4 kb, covering coding as well as noncoding sequences. Labelling of test DNA with Cy5 and a pool of normal reference DNA with Cy3, as well as hybridization and post-washing, were carried out according to the manufacturer’s protocol (Agilent Technologies). Scanning of arrays, analysis, visualization, and comparison of the data set was done with the manufacturer’s hardware and software package (Agilent Scanner, feature extraction v. 9.1, and CGH-analytics v. 3.2).

2.5. Cytogenetic analysis Metaphases for cytogenetic analysis were prepared from the 5th culture passage of the malignant pleural effusion, as well as from the 53rd and 85th passages of MM-Z03E and the 27th and 74th passages of MM-Z03S. The preparations were obtained by standard cytogenetic techniques. In brief, cells were arrested in the metaphase with colchicine reagents (Sigma-Aldrich, Munich, Germany), subsequently swollen in hypotonic KCl solution, and finally fixed in 3:1 methanol/acetic acid and dropped on water-rinsed slides. Metaphase spreads were stained with 4’,6-diamidino2-phenylindole (DAPI, 2.5 mg/mL; Sigma-Aldrich) and actinomycin (12.5 mg/mL; Sigma-Aldrich) in Vectashield medium (Vector Laboratories, Burlingame, CA) and karyotyped according to the International System for Human Cytogenetic Nomenclature (ISCN 2005) [7] at a Quips Genetics Workstation using the Quips Karyotyping software (distributed through Applied Imaging, Newcastle-uponTyne, UK).

2.7. Multicolor fluorescence in situ hybridization and chromosome 3especific multicolor banding Multicolor fluorescence in situ hybridization (M-FISH) analysis was performed on metaphase preparations from the same passages as used for karyotype analysis using the SpectraVysion 24-color karyotyping assay according

H.J. Schulten et al. / Cancer Genetics and Cytogenetics 176 (2007) 35e47

39

Fig. 2. Xenograft in nude mice. Tumor growing subcutaneously in the neck-fold 16 weeks after implantation (A). Overview of MM xenograft of the sarcomatoid type in nude mouse (B). Xenograft of the sarcomatoid type with microinvasion of surrounding tissue (C). Xenograft of the epithelioid type (third mouse) widely invading the surrounding muscle tissue (D). Signs of inflammation and fibrosis/hyalinosis in the xenograft of the epithelioid type (fourth mouse; E).

to the manufacturer’s protocol (Abbot Molecular/Vysis). The assay consists of single-stranded, whole chromosome painting probes directly labelled with different sets of fluorophores to assign a specific combination of fluorophores to each of the 24 chromosome homologs. Metaphase images were captured with an epifluorescence microscope (Zeiss, Go¨ttingen, Germany) equipped with a CCD camera (Photometrics, Tucson, AZ) and using single band pass filters (SpectraVysion optical filters; Abbot Molecular/Vysis) corresponding to the fluorophores SpectrumGold, SpectrumFRed, SpectrumAqua, SpectrumRed, and SpectrumGreen, and to fluorescent DAPI counterstain. Further image processing and 24-color karyotyping was performed with the Vysis SpectraVysion software. The complex rearrangements of chromosome 3 were analyzed with the XCyte mBAND kit for chromosome 3 according to the

manufacturer’s Germany).

protocol

(Metasystems,

Altlussheim,

3. Results 3.1. Cell culture characteristics Both cell lines, MM-Z03E and MM-Z03S, were derived from the malignant pleural effusion of an 83-old-male by in vivo selection of epithelioid and sarcomatoid MM cell clones. They were continuously passaged more than 70 times during the past 2 years. Cytomorphologic appearance was relatively uniform in both cell lines from early to late passages. MM-Z03E cells revealed epithelioid cobblestone morphology, while MM-Z03S cells exhibited

H.J. Schulten et al. / Cancer Genetics and Cytogenetics 176 (2007) 35e47

40

Fig. 3. Immunophenotyping in xenografts. Cytokeratin expression in xenograft from MM-Z03S (A) and from MM-Z03E (B). Nuclear MIB-1 expression is lower in the sarcomatoid (C) than in the epithelioid (D) type.

a sarcomatoid storiform growth pattern (Table 2; Fig. 1, A and B). Immunohistochemical analysis on cytospin preparations (Fig. 1, C and D) showed moderate or strong expression of cytokeratin (Fig. 1, E and F), CK 18, and calretinin in both cell lines, whereas EMA was comparably expressed stronger in epithelioid cells of MM-Z03E, and vimentin was comparably expressed stronger in sarcomatoid cells of MM-Z03S. Expression of p53 and of MMnegative/inconstant markers such as CK 5/6, BerEP4,

S-100, SM-actin, CD34, CD117, CEA, TTF-1, D2/40, and E-cadherin was only weak, very weak, or absent in cultured cells. 3.2. Xenotransplantation Subcutaneous injection of cell cultureederived MM cells gave rise to solid tumors within 16 weeks in all four xenotransplanted athymic nude mice (Table 3; Fig. 2A).

Table 4 Cytogenetic analysis of cell culture and cell lines Passage

Karyotype

Cell culture Passage 5

37~39,XY,der(1;3)(q10;q10)t(3;3)(q27;p21),-4,der(9;10)(q10;q10),-13,-14,der(16)t(3;16)(q11;q11), der(17)t(3;17)(p21;p12)t(1;3)(p34;p11)t(1;17)(p11;p12),-18,-22[cp8]/38-40,sl, þ5[cp10]/78-81,sdl1x2[cp8]/153-155,sdl2x2[cp2] 38~43,XY,der(1;3)(q10;q10)t(3;3)(q27;p21),-4,þ5,þ5,der(9;10)(q10;q10),-13,-14,der(16)t(3;16)(q11;q11), der(17)t(3;17)(p21;p12)t(1;3)(p34;p11)t(1;17)(p11;p12),-22[cp21]/72,idemx2[1] rev ish dim(1)(p34pter),enh(3)(q11q27),dim(4),amp(5),dim(9)(p),dim(10)(p),dim(13)(q),dim(14)(q), dim(16)(q),dim(22)(q) 38~43,X,-Y,del(1)(q34),der(1;3)(q10;q10)t(3;3)(q27;p21),-4,þ5,þdel(5)(q13),der(9;10)(q10;q10),-13,-14, der(16)t(3;16)(q11;q11),der(17)t(3;17)(p21;p12)t(1;3)(p34;p11)t(1;17)(p11;p12),-22[cp17] 36~37,XY,der(1;3)(q10;q10)t(3;3)(q27;p21),-4,der(9;10)(q10;q10),-11,-13,-14,der(16)t(11;16)(p11;q11), der(17)t(3;17)(p21;p12)t(1;3)(p34;p11)t(1;17)(p11;p12),-18,-22[cp5]/34-38,idem,-21[cp8] rev ish dim(1)(p34pter),dim(3)(q27qter),dim(4),dim(9)(p),enh(9)(q),dim(10)(p),dim(11)(q),dim(13)(q), dim(14)(q),dim(16)(q),dim(18),trenda dim(19),trenda dim(21)(q),dim(22)(q) 36~37,XY,der(1;3)(q10;q10)t(3;3)(q27;p21),-4,der(9;10)(q10;q10),-11,-13,-14,der(16)t(11;16)(p11;q11), der(17)t(3;17)(p21;p12)t(1;3)(p34;p11)t(1;17)(p11;p12),-18,add(19)(p13),-21,-22[cp5]/69~72, idemx2,-2,-17,þ19,-add(19)(p13)x2[cp13]

MM-Z03E Passage 53 MM-Z03E Passage 53 MM-Z03E Passage 85 MM-Z03S Passage 27 MM-Z03S Passage 27 MM-Z03S Passage 74

a

trend, a trend imbalance not reaching the threshold.

H.J. Schulten et al. / Cancer Genetics and Cytogenetics 176 (2007) 35e47

41

Fig. 4. Cytogenetic analysis of MM-Z03E and MM-Z03S and multicolor banding of chromosome 3. Representative karyotypes from MM-Z03E (A and B) and MM-Z03S (C and D) in an inverted DAPI-banded view and M-FISH 24 classification colors. Both cell lines are characterized by a hypodiploid stemline. Numeric losses common to both cell lines include e4, e13, e14, and e22. Structural aberrations include der(1;3)t(3;3), der(9;10), and a der(17)t(3;17)t(1;3)t(1;17). In addition, MM-Z03E is marked by þ5 and a der(16)t(3;16) and MM-Z03E by a der(16)t(11;16). Translocated fragments are assigned by colors corresponding to their chromosomal origin. Multicolor banding of chromosome 3 material with region-specific painting probes (E). The labelling scheme and orientation of segments can be derived from the normal chromosome 3 paint and are designated for der(1:3) and der(17).

42

H.J. Schulten et al. / Cancer Genetics and Cytogenetics 176 (2007) 35e47

Fig. 4. (Continued)

For xenografts from MM-Z03S, the gross shape was either round (first mouse) or oval (second mouse; Fig. 2B), whereas for xenografts from MM-Z03E, the shape was irregular (third and fourth mice). The microscopic appearance of MM cells was predominantly polygonal in xenografts from both cell lines. We could observe differences in the pattern of invasive growth between xenografts from

MM-Z03E and MM-Z03S. Whereas sarcomatoid tumors from MM-Z03S grew only in a minimally invasive fashion (Fig. 2C), a widely invasive growth was determined for one epithelioid tumor from MM-Z03E (Fig. 2D, third mouse). Weak to moderate inflammatory reactions as well as slight fibrosis and hyalinosis were seen in all four tumors (Fig. 2E). Necrosis was not detected at all.

H.J. Schulten et al. / Cancer Genetics and Cytogenetics 176 (2007) 35e47

43

Fig. 4. (Continued)

Immunophenotypes of xenografts was quite similar to those of corresponding cell cultures. Cytokeratin (Fig. 3, A and B) and vimentin staining was moderate or strong, and expression of epithelial membrane antigen was weak to moderate, and CK18 staining was more pronounced in the epithelioid than in the sarcomatoid phenotype of xenografts. P53 expression was strong in one xenograft from MM-Z03E but weak in the remaining xenografts (Fig. 3, C and D). Negative or inconstant markers for mesotheliomas, including CK5/6, BerEP4, SM-actin, CD34, CD117, and E-cadherin revealed only weak or no expression. In contrast to the xenografts (Fig. 3, C and D), MIB-1 expression of cell cultures was higher in the sarcomatoid than in the epithelioid phenotype (Table 4). This latter finding is in accordance with the proliferation assay conducted on cultured cells of advanced passages, which revealed a significantly higher proliferation rate in MM-Z03S than in MM-Z03E (P50.009). 3.3. Cytogenetic and molecular cytogenetic analyses Both MM-Z03E and MM-Z03S were characterized by stable hypodiploid stem lines in early and late passages, as determined by complementary molecular cytogenetic and M-FISH analysis (Table 4; Fig. 4, AeD). In addition, MM-Z03S was marked by polyploidization in related sidelines of advanced passages leading to chromosomal numbers in the range of 69e72 in the majority of the metaphases. Numerical losses of chromosomes in both cell lines included e4, e13, e14, and e22. Structural aberrations identified in both cell lines included two whole chromosome arm translocations [der(1;3)(q10;q10)t(3;3) and der(9;10)(q10;q10)] and a complex rearrangement involving chromosome 17 with insertion of material from chromosomes 1 and 3 at 17p [der(17)t(3;17)t(1;3)t(1;17)]. The rearrangements of chromosome 3 material could be

clearly identified by using a chromosome 3especific multicolor banding technique (Fig. 4E). Cell lineespecific aberrations included tetrasomy 5 or partial tetrasomy 5p and der(16)t(3;16) in MM-Z03E, and a der(16)t(11;16) in MM-Z03S. All these aberrations contributed to losses of 1p34pter, 4, 9p, 10p, 13q, 14q, 16q, and 22q in both cell lines, as revealed by CGH (Fig. 5, A and B). Cell linee specific CGH imbalances included gains of 3q11q27 and 5 in MM-Z03E, as well as gain of 9q, and losses of 3q27qter, 11q, and 18 in MM-Z03S. CGH ideograms from mouse xenograft tumors of MM-Z03E (Fig. 5C) and MMZ03S (Fig. 5D) exhibited nearly identical chromosomal imbalances compared to the respective cell lines. Cytogenetic analysis of the fifth cell culture passage before both clones were cultured separately revealed the same composition of cytogenetic changes as in later passages of the cell lines, supporting our assumption that the aberrant chromosomal constitution was already present in the primary tumor. By performing aCGH on both cell lines, some minute aberrations and cell lineespecific imbalances that were beyond the resolution of conventional CGH could be detected (Table 5). Remarkably, a homozygous deletion at 9p21.3 spanning approximately 50 megabases included the CDKN2A and CDKN2B gene locus (Fig. 6).

4. Discussion MM is a phenotypically heterogeneous neoplasm with pure epithelioid histology generally having a more favorable prognosis than one with sarcomatoid histology [6]. Several MM cell lines have been described in recent years, reflecting the need to study MM tumor biology by using valid model systems [9e12]. However, only a few MM retaining their biphasic morphology in cell culture have been reported to date [13e15]. The herein presented MM cell

44

H.J. Schulten et al. / Cancer Genetics and Cytogenetics 176 (2007) 35e47

Fig. 5. CGH karyograms of MM-Z03E (A) and MM-Z03S (B). Each chromosome ideogram is shown with its chromosome-specific green/red fluorescent ratio profiles. Bars to the right represent gains and bars to the left indicate losses (n, number of analyzed chromosomes). Both cell lines show losses of 1p34pter, 4, 9p, 10p, 13q, 14q, 16q, and 22q. In addition, MM-Z03E has gains of 3q11q27 and 5 and MM-Z03S gain of 9q and losses of 3q27qter and 18. CGH from mouse xenograft tumors of MM-Z03E (C) and MM-Z03S (D) demonstrated almost identical chromosomal imbalances, as described in the corresponding cell cultures.

H.J. Schulten et al. / Cancer Genetics and Cytogenetics 176 (2007) 35e47

45

Fig. 5. (Continued)

lines are one of the few first established epithelioid and sarcomatoid MM cell lines sharing a clonal origin, although the capacity of MM sub-cell lines to differentiate into either the epithelial or sarcomatoid phenotype, depending solely

on the serum composition, has been reported several years ago [16]. MM-Z03E and MM-Z03S genetically reveal more losses than gains, which is in accordance with other CGH studies

H.J. Schulten et al. / Cancer Genetics and Cytogenetics 176 (2007) 35e47

46

Table 5 Cryptic aberrations identified by aCGH in MM-Z03E and MM-Z03S Cell line MM-Z03E, MM-Z03E, MM-Z03E, MM-Z03S MM-Z03E MM-Z03S MM-Z03E, MM-Z03E MM-Z03E a

MMZ03S MMZ03S MMZ03S

MM-Z03S

Arr cgh resultsa

Aberration size (Mb)

1p13.1p12 (A_16_P15263402/A_16_P00158957)x1 3p21.3p21.2 (A_16_P00696789/A_16_P16225711)x1 9p21.3 (A_16_P18578320/A_14_P118440)x0 10pterp12.1 (A_16_P02196410/A_16_P18873331)x1 10pterp11.22 (A_16_P02196410/A_16_P18879524)x1 10p11.22 (A_16_P02241449/A_16_P02244063)x1 10p11.21 (A_16_P18894209/A_16_P18899328)x1 22q11.1q11.23 (A_16_P41490791/A_16_P03606722)x1 22q11.23q13.33 (A_16_P21293176/A_16_P03641933)x1

6,750 6,760 50 27,240 29,900 2,210 2,350 7,000 26,200

Aberrations considered as uncertain were not included.

on MM [17,18]. The affected chromosomal regions are also known to harbor important tumor-associated genes. The tumor suppressor genes p16(INK4a) and p14(ARF) are located at 9p21, which is commonly targeted in various tumor types, including MM [18e20], and homozygous deletion of the CDKN2A locus has been reported recently in about 70 % of analyzed pleural MM [21]. Gain of 5p has been identified as a common alteration in six of seven cytogenetically characterized MM cell lines of the epithelioid or mixed phenotype [22]. Otherwise, þ5p in combination with e4p, e4q, and e9p was identified as a helpful discriminator between sarcomatoid mesotheliomas from pleural sarcomas and lung carcinomas [23]. In addition, þ5p

has been attributed to the sarcomatoid phenotype in the CGH study of Krismann et al. [18], which compares different MM phenotypes. Tumor-related genes for MM differentiation have not been identified to date, and it should be of interest to further assess the association of þ5p with the histologic phenotype in MM. Mutational analysis of tumor-related genes in MM cell lines has detected p53 mutations in a subset of these tumors [12,24]. In our study, immunohistochemical expression of p53 could be detected in all four xenografts and, remarkably, p53 staining was the highest in an epithelioid xenograft, which was the only tumor exhibiting invasive growth into surrounding muscle tissue. At this time, we do not know if mutated (e.g., overexpressed) p53 is one of the key events triggering invasive behavior in MM. However, we intend to study the mutational status of p53 in our cell lines, considering that the insertion at the derivative chromosome 17 maps to 17p12, which is also the region containing the p53 gene locus. Loss of chromosome 22 is one of the most frequent alterations in MM and may contribute to the inactivation of the NF2 tumor suppressor gene located at 22q12. Loss of wild-type NF2 is seemingly associated with susceptibility to asbestos-induced MM [24,25]. Other genomic imbalances detected in our MM cell lines, including e4, e13q, and e14q, are also common events in MM [17], suggesting that multiple losses at these genomic regions contribute to the dysfunction of known and unknown tumor suppressor genes at these sites. In summary, this study reports on the establishment and characterization of two phenotypically and genetically distinct, immunohistochemically confirmed, independent MM cell lines originating from the same primary MM. Both cell lines represent especially useful model systems to analyze the biologic behavior and differentiation of MM in vivo and in vitro.

Fig. 6. Scatter plot of aCGH results depicting a homozygous deletion in both cell lines of approximately 50 megabases at 9p21.3 covering the CDKN2A and CDKN2B gene loci. The homozygous deletion is represented by an extended shift of features (green spots) to the left.

Acknowledgments We thank Mrs. Wolf-Salgo´ and Mrs. Hottenrott for excellent technical assistance.

H.J. Schulten et al. / Cancer Genetics and Cytogenetics 176 (2007) 35e47

References [1] Travis WD, Brambilla E, Mu¨ller-Hermelink HK, Harris CC, editors. Pathology and Genetics of Tumours of the Lung, Pleura, Thymus and Heart. Lyon: IARC Press, 2004. [2] Simon F, Johnen G, Krismann M, Muller KM. Chromosomal alterations in early stages of malignant mesotheliomas. Virchows Arch 2005;447:762e7. [3] Ceresoli GL, Zucali PA, Favaretto AG, Grossi F, Bidoli P, Del Conte G, Ceribelli A, Bearz A, Morenghi E, Cavina R, Marangolo M, Parra HJ, Santoro A. Phase II study of pemetrexed plus carboplatin in malignant pleural mesothelioma. J Clin Oncol 2006;24:1443e8. [4] Gordon GJ, Rockwell GN, Godfrey PA, Jensen RV, Glickman JN, Yeap BY, Richards WG, Sugarbaker DJ, Bueno R. Validation of genomics-based prognostic tests in malignant pleural mesothelioma. Clin Cancer Res 2005;11:4406e14. [5] Kindler HL, Vogelzang N. Mesothelioma. In: Vokes EE, Golomb H, editors. Oncologic therapies. Berlin: Springer-Verlag, 2003. pp. 415e23. [6] Johansson L, Linden CJ. Aspects of histopathologic subtype as a prognostic factor in 85 pleural mesotheliomas. Chest 1996;109: 109e14. [7] Shaffer LG, Tommerup N. ISCN 2005: An International System for Human Cytogenetic Nomenclature. Basel: S. Karger, 2005. [8] Gunawan B, Schulten HJ, von Heydebreck A, Schmidt B, Enders C, Hoer J, Langer C, Schuler P, Schindler CG, Kuhlgatz J, Fu¨zesi L. Site-independent prognostic value of chromosome 9q loss in primary gastrointestinal stromal tumours. J Pathol 2004;202:421e9. [9] Kivipensas P, Bjorkqvist AM, Karhu R, Pelin K, Linnainmaa K, Tammilehto L, Mattson K, Kallioniemi QP, Knuutila S. Gains and losses of DNA sequences in malignant mesothelioma by comparative genomic hybridization. Cancer Genet Cytogenet 1996;89:7e13. [10] Ruco LP, de Laat PA, Matteucci C, Bernasconi S, Sciacca FM, van der Kwast TH, Hoogsteden HC, Uccini S, Mantovani A, Versnel MA. Expression of ICAM-1 and VCAM-1 in human malignant mesothelioma. J Pathol 1996;179:266e71. [11] Philippeaux MM, Pache JC, Dahoun S, Barnet M, Robert JH, Mauel J, Spiliopoulos A. Establishment of permanent cell lines purified from human mesothelioma: morphological aspects, new marker expression and karyotypic analysis. Histochem Cell Biol 2004;122: 249e60. [12] Kumar K, Rahman Q, Schipper H, Matschegewski C, Schiffmann D, Papp T. Mutational analysis of 9 different tumour-associated genes in human malignant mesothelioma cell lines. Oncol Rep 2005;14: 743e50. [13] Reale FR, Griffin TW, Compton JM, Graham S, Townes PL, Bogden A. Characterization of a human malignant mesothelioma cell

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

47

line (H-MESO-1): a biphasic solid and ascitic tumor model. Cancer Res 1987;47:3199e205. Sun X, Dobra K, Bjornstedt M, Hjerpe A. Upregulation of 9 genes, including that for thioredoxin, during epithelial differentiation of mesothelioma cells. Differentiation 2000;66:181e8. Usami N, Fukui T, Kondo M, Taniguchi T, Yokoyama T, Mori S, Yokoi K, Horio Y, Shimokata K, Sekido Y, Hida T. Establishment and characterization of four malignant pleural mesothelioma cell lines from Japanese patients. Cancer Sci 2006;97:387e94. Klominek J, Robert KH, Hjerpe A, Wickstrom B, Gahrton G. Serumdependent growth patterns of two, newly established human mesothelioma cell lines. Cancer Res 1989;49:6118e22. Bjorkqvist AM, Tammilehto L, Anttila S, Mattson K, Knuutila S. Recurrent DNA copy number changes in 1q, 4q, 6q, 9p, 13q, 14q and 22q detected by comparative genomic hybridization in malignant mesothelioma. Br J Cancer 1997;75:523e7. Krismann M, Muller KM, Jaworska M, Johnen G. Molecular cytogenetic differences between histological subtypes of malignant mesotheliomas: DNA cytometry and comparative genomic hybridization of 90 cases. J Pathol 2002;197:363e71. Cheng JQ, Jhanwar SC, Lu YY, Testa JR. Homozygous deletions within 9p21-p22 identify a small critical region of chromosomal loss in human malignant mesotheliomas. Cancer Res 1993;53:4761e3. Nagel H, Schulten HJ, Gunawan B, Brinck U, Fu¨zesi L. The potential value of comparative genomic hybridization analysis in effusion-and fine needle aspiration cytology. Mod Pathol 2002;15:818e25. Illei PB, Rusch VW, Zakowski MF, Ladanyi M. Homozygous deletion of CDKN2A and codeletion of the methylthioadenosine phosphorylase gene in the majority of pleural mesotheliomas. Clin Cancer Res 2003;9:2108e13. Pelin-Enlund K, Husgafvel-Pursiainen K, Tammilehto L, Klockars M, Jantunen K, Gerwin BI, Harris CC, Tuomi T, Vanhala E, Mattson K, Linnainmaa K. Asbestos-related malignant mesothelioma: growth, cytology, tumorigenicity and consistent chromosome findings in cell lines from five patients. Carcinogenesis 1990;11:673e81. Knuuttila A, Jee KJ, Taskinen E, Wolff H, Knuutila S, Anttila S. Spindle cell tumours of the pleura: a clinical, histological and comparative genomic hybridization analysis of 14 cases. Virchows Arch 2006;448:135e41. Metcalf RA, Welsh JA, Bennett WP, Seddon MB, Lehman TA, Pelin K, Linnainmaa K, Tammilehto L, Mattson K, Gerwin BI, Harris CC. p53 and Kirsten-ras mutations in human mesothelioma cell lines. Cancer Res 1992;52:2610e5. Cheng JQ, Lee WC, Klein MA, Cheng GZ, Jhanwar SC, Testa JR. Frequent mutations of NF2 and allelic loss from chromosome band 22q12 in malignant mesothelioma: evidence for a two-hit mechanism of NF2 inactivation. Genes Chromosomes Cancer 1999;24:238e42.