Bombesin may stimulate proliferation of human pancreatic cancer cells through an autocrine pathwayкод для вставкиСкачать
Int. J. Cancer: 68,5 14-5 19 ( 1996) 0 1996 Wiley-Liss, Inc. Publlcatlon of the lnternatlonal Un on Against Cancer PmI,cat on de I unlon memationale b n t r e le Cancer ESTABLISHMENT, CHARACTERIZATION AND DRUG SENSITIVITY OF FOUR NEW HUMAN SOFT TISSUE SARCOMA CELL LINES Wei-Wei LI', Carlos CORDON-CARDO~, Quanguang cHEN3, Suresh C. J H A N W A R and ~ Joseph R. BERTrN01,4 Laboratories of 'MolecularPharmacology, 'Molecular Immunopathology and 3Solid Tumor Genetics, Memorial Sloan-KetteringCancer Center, New York, M: USA. Four new cell lines were established from patients with soft tissue sarcomas. Drug sensitivity as well as genotypic characterization, which may be related to drug sensitiviky in these cell lines, was determined. Karyotype, H-ras, c-myc and mutant p53 gene expression, Rb, GI- and S-phase cyclins, E2F and major cyclin/CDK inhibitors such as p16 and p21 and p-glycoprotein were analyzed usingcytogenetic, Northern blot and immunological methods. Drug sensitivity was determined using growth inhibition tests.These cell lines differed in their morphologyand growth rates, forming colonies in soft agar with a cloning efficiency of 4.3-13.4%. and 3 of the 4 cell lines grew in nude mice. Cytogenetic analysis of cell lines revealed highly aneuploid karyotypes. Deletion and/or translocation of chromosome 17 was seen in HS-16, HS-18 and HS-30 cells, and both copies of chromosome I 3 were lost or marranged in the HS- I8 cell line. Mutant p53 protein was present in all 4 cell lines. HS- I8 cells showed no expression of the Rb protein and high levels of expression of EZF, cyclin A, cyclin E and CDKZ. HS- I 6 expressed a higher level of cyclin D than the other 3 cell lines. p2Id1 expression was seen in all cell lines, but p16'*' was expressed only in HS-30 and HS-42 cell lines. These cell lines were sensitive to tax01 and relatively resistant to methotrexate, vinblastine and 5-fluorouracil when compared with the fib-coma cell line HT-1080. These new cell lines should provide a useful model for the study of soft tissue sarcomas and for evaluating new drugs or treatments. c 1996 Wiley-Liss,Inc. The soft tissue sarcomas are a heterogeneous group of neoplasms composed of various histological subtypes and remain one of the most chemotherapy-refractory human malignancies (Greenall et al., 1986). A better understanding of the biology and reasons for drug failure in this disease could lead to improved treatment. As detailed investigations are difficult to perform in fresh samples from patients, we sought to establish new cell lines that would reflect the phenotypic and genotypic properties of these tumors. We describe herein the establishment and characterization of 4 new soft tissue sarcoma cell lines, including growth properties, tumorigenicity in nude mice, cytogcnetics, determination of some oncogenes and tumor-suppression genes, cyclins and cyelin-dependent kinase inhibitors. In addition, the sensitivity of these cell lines to different anti-cancer drugs was determined. These cell lines should provide a useful model for further studies in exploring the biology of these tumors and response to chemotherapeutic agents. MATERIAL AND METHODS Tumor samples Tumor specimens of surgically excised soft tissue sarcomas were obtained from the Surgery Department, Memorial SloanKettering Cancer Center. Histopathological analysis o f the 4 tumor samples revealed them to be mesenchymal chondrosarcoma (HS-16), myxoid liposarcoma (HS-18), malignant hemangiopericytoma (HS-30) and malignant mesenchymoma composed of liposarcoma and rhabdomyosarcoma (I IS-42). HS-16, HS-18 and 11s-30 were isolated from primary tumors from the leg of an 18-year-old-male, the retroperitoneum of a 42-yearold female and the pelvis of a 46-year-old malc, respectively. HS-42 was isolated from a metastasis in thc right ovary of a 67-year-old female. None of the patients had received prior chemotherapy or radiotherapy at the time of surgery. Cell culture Tumor samples were trimmed free of connective tissue, washed several times with RPMI-1640 media, finely minced and incubated in RPMI-1640 medium with 10% FBS, 0.6% collagenase I1 and 0.002% DNAse I at 37°C for 2 hr, then passed through a 100 mesh screen, and cell suspensions obtained were placed in RPMI-1640 medium enriched with 10% FBS, 2 mM L-glutamine, 0.2 mM sodium pyruvate, 0.4 mM 1.-serine and 0.3 mM ascorbic acid at a density of 2.5 to 5 x l@ cells/cm2. After 3-7 days of incubation at 37°C in a humidified atmosphere of 5% C o t , the cell monolayer was subcultured after detachment with 0.2% trypsin plus 2 mM EDTA. Cells were maintained in RPMI-1640 medium containing 10% FBS, 2 mM glutamine and 0.2 mM sodium pyruvate. With the use of this procedure, we have been able to establish 12 cell lines from 30 tumors, i.e., a success rate of 40%. In vitro doubling time Triplicate T-25 tissue culture flasks were seeded with 5 X lo4cells in 5 ml of medium/flask. At 24 hr intervals for 10 days, the cell monolayer was trypsinized and cells were counted. These data were plottcd, and population doubling times were calculated. Colony-formingefficiency Colony-forming efficiency was determined both in monolayer culture on plastic and in soft agar. In monolayer cultures, 500-2,000 cells were plated in a 60 mm dish. For soft agar cloning, assays were performed using the method of Hamburger and Salmon (1977). One milliliter of underlayer consisting of RPMI-1640 medium, 10% FBS and 0.5% Noble agar was plated in 24-well plates; 5 to 10 x lo3 cells/ml were suspended in RPMI-1640 with 10% FBS and 0.3% agar, and 1 ml of the suspension was plated on the gelled underlayer. After incubation at 37°C in 5% C 0 2 for 12 days, colonies of greater than 50 cells were counted under a microscope. Growth in nude mice Five-week-old nude mice (Swiss nulnu) were purchased from Charles River (Wilmington, MA) and maintained under pathogen-free conditions. The ability of cultured cells to form tumors was assessed by injecting 3 x lo6cells S.C. in a volume of 0.2 ml in the right flank region. Three to 5 micc were used for each tumor. Tumors were allowed to grow to a volume of approximately 1 cm3 before they were removed a t the time animals were killed. Cytogeneticanalysis Exponentially growing cells were incubated with Colcemid (0.2 kg/ml) for 4-6 hr at 37°C. Cells were then trypsinized, washed and processed for chromosome preparations following 4Towhom correspondence should be addressed, at the Laboratory of Molecular Pharmacology, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA. Fax: (212) 639-2767. Received: Junc lY, 1996 and in revised form August 5. 1996. ESTAHI.ISHMEN1 OF HUMAN SOFI TISSUE SARCOMA CELL LINES conventional methods using 0.075 M potassium chloride as the hypotonic solution and 3:l methanol acetic acid as the fixative. Air-dried metaphase spreads were stained to reveal G-banding patterns. Clonal chromosome abnormalities wcre defined and karyotypes described according to the international system for human cytogenctic nomenclature (Mitelman, 1991). Southern blot analysis To test c-myc amplification, genomic DNA was isolated from frozen cells. A 10 pg aliquot of each sample was digested by EcoR I and elcctrophoresed on 0.8% agarose gels. After staining with ethidium bromide, DNA was transferred to nylon filters. Pre-hybridization, hybridization and washing were performed. The probe used was a fragment (3rd cxon) of c-myc cDNA (Oncogene Science, Uniondale, NY). Aftcr removal of the probe, filters were rehybridized with a human P-rnicroglobin cDNA as an internal control. Quantification of amplification was performed with a blot analyzer (Betascopc 603). Northern blot analysis Total RNA was extracted from sarcoma cells with RNAzol (Biotecx, Houston, TX); 20 pg RNA were subjected to electrophoresis in a 1.0% agarose gel containing 1.5% formaldehyde. After staining with ethidium bromide, RNA was transfcrred to a nylon membrane. Pre-hybridization, hybridization and wash were performed. Oncogene probes used were C-myc exon 3 (pHSR-1; ATCC, Rockville, MD) and H-ras exon 1 (PCR-generated fragment). Immunological analysis of expression of cell-cycle regulators Mutant p53, Rb, cyclin A, cyclin D1, cyclin E, CDK2, E2F, p16 and p21 were detected following incubation with the appropriate antibodies: monoclonal mouse anti-human p53 (pAb 1801), monoclonal mouse anti-human Rb (IF8), polyclonal rabbit anti-human cyclin A (BF1683), monoclonal mousc anti-human cyclin D1 (HDI l), polyclonal rabbit antihuman cyclin E (HE1 1l), polyclonal rabbit anti-human CDK2 (M2), polyclonal rabbit anti-human E2F (KH95), polyclonal rabbit anti-human p16 (C20) and polyclonal rabbit anti-human p21 (C19) (all from Santa Cruz Biotechnology, Santa Cruz, CA). Immunohistochemical analysis was performed after cytospin using a standard avidin-biotin-peroxidasetechnique (Cordon-Cardo et al., 1987). Expression of the various proteins in tumor cells was classified in 1 of 3 categories by estimating the percentage of cultured cell nuclei staining: (i) negative ( < lo%), (ii) heterogeneous ( 10-70%), (iii) homogeneous ( > 70%). For immunoblotting, 100 pg samples of total cell lysates were size-fractionated by SDS-PAGE and transferred onto nitrocellulose membranes. Protein expression was detected using ECL detection reagents (Amersham, Arlington Heights, IL). Cell growth inhibition Monolayer cells in 6-well plates (10 x 104 cclls/well) were cxposcd to various concentrations o f drugs, including methotrexate (MTX), 10-ethyl-10-dezaaminopterin (10-EDAM), trimetrexate (TMTX), taxol, doxorubicin (DOX), actinomycin D (ACD), vinblastine (VLB) and 5-fluorouracil (5-FUra). After 24 hr, the media were removed and the cell layer was washed twice with cold PBS. Fresh drug-frcc medium was added, and growth was followed for an additional 96 hr. Cells were counted after appropriate dilutions of the cell suspension using a model ZB Coulter counter (Hialeah, FL). Percentage of growth inhibition and EDSOvalues were determined as previously described (Li et al., 1992). Determination of the multidrug resistance gene product (p-glycoprotein) Determination of p-glycoprotein in these cell line was made using a method previously described (Cordon-Cardo et al., 1987). Mouse monoclonal antibodies (MAbs) HYB-241 (a gift 515 from Dr. L Grauer of Hybritech, San Diego, CA) and C-219 (from Centocor, Malvern, PA) wcre used at a concentration of 20 pg/ml. As a negative control, purified mouse MAb directed against the cell surface antigen anthranylate synthase of Escherichia coli was used at a concentration of 20 pg/ml. As a positive control, purified MAbs against cytokeratins and other intermediate filaments wcre also used at the same coneentration. Secondary antibodies used were biotinylated horse antimouse IgG affinity-purified antibodies (1:lOO dilution in PBS). Immunocytochemical analysis was performed as described above. RESULTS In vitro growth characteristics All cell lines grew as monolayer cultures with loss of contact inhibition at confluence. Each cell line showed different morphological features with considerable variation in the size and shape of cells (Fig. 1). Doubling times varied from 33 to 40 hr HS-18 cells proliferated more rapidly than the other cell lines (Table I). Plating efficiencies (PE) were measured at passages 80,66,68 and 58 for HS-16, HS-18, HS-30 and HS-42, respectively, and ranged from 16.4% for HS-16 to 6.4% for HS-18. Cloning efficiencies in soft agar were similar for HS-18 and HS-30 but lower for HS-16 and HS-42 cell lines (Table I). Tumorigeniciv of cultured cells in nude mice Cultured HS-16, HS-18 and HS-30 cells (3 x lo6) injected S.C. formed detectable tumors at the injection site within 5, 9 and 5 weeks, rcspectively, and grew to reach 1 cm in size in about 12 wccks for the HS-16 and HS-30 cell lines and in about 20 weeks for the HS-18 cell line. Histopathological characteristics of tumors from nude mice were similar to that from the patients' initial tumors (not shown). The cells obtained from nude mousc tumors were plated in vitro and immediately showed outgrowth. No tumor growth was observed after inoculation of HS-42 cells, even at 26 weeks (Table I). Karyotype Cytogenetic analysis of the 4 cell lines showed highly aneuploid karyotypes (Fig. 2; Table 11). Chromosome numbers ranged 72-77, 67-72, 69-79 and 82-85 for HS-16, HS-18, HS-30 and HS-42 cell lines, respectively. Chromosomal alterations included trisomics, monosomies, translocations, deletions, derivative chromosomes, additional or missing whole chromosomes and marker chromosomes, but no double minutes or homogeneously staining regions were seen in any of the cell lines (Table 11). Deletion and/or translocation of chromosome 17 was seen in HS-16, HS-18 and HS-30 cells, and both copies of chromosome 13 were lost or re-arranged in the HS-18 cell line. A representative karyotype from the cell line HS-18 is presented in Figure 2 to dcmonstratc the complexity of chromosomal abnormalities seen. Oncogene amplification and expression DNA and RNA from normal human lymphocytes and from a non-malignant lymphoblastoid cell line (RPMI-1788) were used for comparison with the 4 sarcoma cell lines. Southern blot analysis showed no significant increase of c-myc gene copy number in any of the cell lines. Over-expression of c-myc and H-ras mRNA was also not observed in these cells (data not shown). Expression of cell-cycleregulatoryproteins Mutated p53 and Rb protein expressions were analyzed immunohistochemically using MAbs. PAB-1801, an antibody that recognizes p53, showed intense diffuse nuclear staining in >70% of cells in all 4 sarcoma cell lines. In HT-1080, a fibrosarcoma obtained from the ATCC, nuclear staining was observed in about 20% of cells. HS-16, IIS-30 and HS-42 cell LI ETAL. 516 FIGURE 1 - Morphology of the soft tissue sarcoma cell lines by phase contrast microscopy (original magnification ~ 2 0 0 )(n) . HS-16. (b) HS-18, (c) HS-30, (d) HS-42. TABLE I - GROWTIi PROPERTIES OF4 KEWLY ESTABLISHED IiUMAN SOFT TISSUE SARCOMA CELL LlXES Cell line HS-16 HS-18 HS-30 HS-42 Pacsage number - 198 187 186 165 Doubling (hr) % Plating o/, Cloning efficiency efficiency 40 33 39 34 16.4 6.4 11.6 5.9 5.2 12.4 2.3 7.0 Growth in nude mice 1 + + 2 ' 3 4 5 -I- - 6 7 \ 9 10 11 12 lines expressed pRb in the nuclei of all cells. However, there was no dctectable Rb protein in the HS-18 cell line. All cell lines expressed E2F-1, cyclin A, cyclin E and CDK2, with the highest expression of these proteins observed in HS-18 cells. Immunoblotting was used to determine expression of cyclin D1, p16Ink4and p2lWaf1and to confirm the absence of Rb protein in HS-18 cells. HS-16 cells expressed a higher levcl of cyclin D1 than the other 4 cell lines. There was no significant difference of p2Iwaf*expression observed in thc cell lines, while ~ 1 expression 6 ~ was~not detected ~ ~ in HT-1080, HS-16 and HS-18 cell lines. Higher cyclin A expression and absence of Rb protein were confirmed in HS-18 cells. These results are summarized in Figure 3 and Table 111. FIGURE2 - A representative karyotype of HS-18 cells. Structural chromosomal alteration such as translocations and deletions (arrows). Note loss of both copies of chromosome 13 and the presence of marker chromosomes. Drug sensithiy The sensitivity of the 4 new sarcoma cell lines and the HT-1080 cell line to several chemotherapeutic agents was determined by exposure to different concentrations of drugs for 24 hr (Table IV). All cell lines were very sensitive to taxol (ED5"values were 3.0 nM for HS-18 cells and less than 1.0 nM for the other 3 cell lines). Sensitivity to ACD and DOX was also observed (EDso values ranged 1-9 and 6 1 6 nM, respectively, for different cell lines). Sensitivity of cells to VLB varied more markedly. The EDSOvalues for the 4 new cell lines were 3- to 17-fold higher than that for HT-1080 eells I1 (6.0 nM). - !? 2 1 ' 2 2 ic x x ESTABLISHMENT OF HUMAN SOFT TISSUE SARCOMA CEI.1. LINES TABLE II - cmoGENmc ANALYSIS 517 lines are composed of malignant cells: ( i ) the cells grew rapidly to a high saturation density in vitro and showed a lack of Cell Numberof Chromoxmal contact inhibition after slow proliferation during the first few line chromosomes alterations passages (generally passages 1-5). (ii) all 4 cell lines have been cultured in vitro for at least 160 passages, (iii) all 4 cell lines HS-16 50 72-77 XX,-X(3n? , +1 +11 +21 form colonies in soft agar, (iv) cytogcnetic analysis displayed +22 2X deI)9)(q22 or 3L), +2X de (22Ma12). marked aneuploidy and (v) the cell lines were found to be +del(5)($ljj,bef(l7)(pll) tumorigenic in nude mice (with the exception of HS-42). random monosomies and triMutant p53 protein, as measured immunohistochemically, somies, random mars was observed in all 4 new cell lines. As the presence of a XX, -X (3nc), -2, +3, -4, HS-18 50 67-72 mutant p53 was not determined in the fresh tumors from which -7, -9, -10, -13, -13, -14, -16, -17, -18, +19, -21, the lines were derived, it is not clear whether the p53 -22, -22, del 1)( 13 32), mutations were present in the fresh tumor from patients del(6)(q16 2!9, Jel(~(plS), before these cell lines were established or grew in culture 2Xdel 11)( \3), del(l7)(pll), selected for cells with p53 mutations. p53 mutations have been +der( ), + Xder (22)t(9; found frequently in a wide variety of tumors, including soft 22)(qll; 13), +der(17)t(17; tissue sarcomas (Toguchida et af., 1992) and are associated ?)(pll;?p, +mar 1, +mar with tumorigenesis, cellular transformation, proliferation and 2 + random mars 69-79 protection of cells from apoptosis and drug cytotoxicity HS-30 50 XX, -Y, (3nz), de1(2)(p21), der(2)t(2;17) q l l ; (Zambetti and Levine, 1993). Abnormalities of thep53 and the pW,der( 13) tq32) Rb genes may occur together in primary human tumors der(19)t(7;19)(qll~pll). (Stratton et al., 1990), as found in the HS-18 cell line. Thep53 +mar 1, +mar 2 andpRb genes reside on the short arm of chromosome 17 and 82-85 XX, -X, (3nz), der(l)t(l; HS-42 50 the long arm of chromosome 13, respectively. Abnormalities of l)(p22;q32), de1(7)( 13, these genes may result from the structural alteration of these der(7)t( 7;?)(q11.2;?7 chromosomes. Structural changes of chromosome 17 were present in HS-16 (del(l7)(pl l)), HS-18 (del(l7)(pll), dcr(17)t(17;?)(pll;?)) and HS-30 (der(2)t(2;17)(qll;p13)), 1 2 3 4 5 respectively. In addition, we observed loss of both copies of chromosome 13 in the HS-18 cell line. H-ras expression and amplification of the c-rnyc gene are frequently present in a wide spectrum of human tumors and are involved in tumorigenesis, transformation, cellular growth ---mmwcyclin D and signal transduction, but information is limited for soft tissue sarcomas. Activated c-myc or H-ras could co-operate with mutant p53 protein in the oncogenic development of cells (Toguchida et af., 1992), and alteration of these oncogenes may -pRb result in anti-cancer drug resistance in tumor cells (Denis et al., 1991; Niimi et al., 1991). However, significant amplification or over-expression of these oncogenes was not observed in these cell lines. p16 Cyclin D1, cyclin A and cyclin E and associated CDKs are regarded for GI- to S-phase progression and are linked to tumor development (Grana and Rcddy, 1995). E2F-1, and its related family mcmbers, is a crucial transcription factor in cellular proliferation and activates the transcription of growthFIGURE 3 - Protein expression of cyclin A, cyclin D1, Rb, plfPkk4 associated genes, including DHFR, TS and DNA polymerase and p21”afI in soft tissue sarcoma cell lines as shown by Western (Johnson et al., 1994). E2F over-expression may induce cellular blotting. One hundred micrograms of protein extract o f each cell transformation, tumor formation in nude mice and even drug line were separated on SDS/PAGE and clectroblotted to a resistance (Johnson et al., 1994; Li ef al., 1995). However, nitrocellulose membrane. Proteins were detected with the various E2F-mediated transactivation can be normally inhibited by antibodies as described in “Material and Methods”. Lane 1, pRb (Helin et af., 1993). ~ 1 6 ’and ” ~ p2lWaf1, ~ as major cyclin/ HT-1080; lane 2, 11s-16; lane 3, HS-18; lane 4, HS-30; lane 5, CDK inhibitors, play an important role in G I arrest and HS-42. suppression of cellular transformation (Scrrano et al., 1995; Michieli et al., 1996). These proteins interact with p53 and pRb to control cell-cycle progression (Slebos et af., 1994). Intensive TMTX and 10-EDAM were more inhibitory than MTX in the expression of E2F was observed in all new cell lines; cyclin A new cell lines compared with HT-I080 cells. 5-FUra was 2-5 was highly expresscd in HS-18 cells, cyclin D levels were higher time less inhibitory to growth in the new cell lines cornpared in HS-16 cells and p16 was not detected in the HS-16 and with HT-1080. HS-18 cells. These results suggest that oncogenic transformation of some soft tissue sarcomas may be due to abnormalities Expression ofp-glycoprotein of cell-cycle regulators rather than changes in expression of ras None of the cell lines had detectable p-glycoprotein. or c-rnyc genes. An important aspect of our results is the use of these cell lines to determine drug sensitivity. The success of chemothcrDISCIJSSION apy is limited in soft tissue sarcomas, and in vitro drug Four new cell lines were established from primary soft tissue sensitivity or resistance information may bc of value to guide sarcomas. The following evidence establishes that these cell individual treatment or new drug development. Taxol, a potent $ 1 - -- LI E T A I . 518 TABLE 111 -MUTANT p53, pRb, CYCLIN A CYCLIN E,CDKZ, E2F AND OLYCOPROTEIN EXPRESSION I N SOFT +ISSUE SARCOMA CELL LiNiH Cell line HS-16 HS-18 HS-30 HS-42 HT-1080 Mutant p53 Rh Cyclin A loo++ 1(K)+++ 70++ 100+++ 20++ 100+ loo+++ 20++ 60+++ 100+++ 20+ ND 100+ Staining (70)’ .~ Cyclin E CDKZ 10+ 40+ 70+++ 50++ 70++ ND 70++ 80+++ 80++ 50++ ND - E2F P-glycoprotein loo+++ lo()+++ - 70+++ 70+++ - ND -2 I % indicates the number of positive cells; + indicates the degree of intensity.-2From Slovak et al. (1991). See “Material and Methods” for details. ND, not done. TABLE IV - GROWTH-INHIBITORY LFFEClS OF 8 ANTI-CANCER SARCOMA CEI.1. LINES Cell line ~~ HS-16 HS- 18 HS-30 HS-42 HT-1080 DRUGS O N SOFT TISSUE ICCO (nM) MTX - TMTX 10-EDAM DOX ACD VLB Taxol 5-FUra 1,990 1,740 600 2,830 170 100 2,100 150 220 230 250 80 13.0 12.0 16.0 7.0 6.0 1.0 2.0 9.0 2.0 2.0 18.0 100.0 70.0 30.0 6.0 0.1 3.0 0.2 0.3 0.1 9,750 9,330 22,230 7,620 4,900 170 30 180 Cells were exposed to drugs for 24 hr, then washed for an additional 96 hr and counted. Values are the averagc of 2-4 different experiments. inhibitor of tumor cell replication isolated from the plant Tarus brevifolia, has been used in the clinic to treat solid tumors, including breast, ovary and colon cancer. Phase I1 trials showed that the response rate to this drug in a small group of soft tissue sarcomas was only 18% (5/28) (Kaye, 1995). However, this compound is a potent inhibitor of growth of these sarcoma cell lines. Based on these studies, this drug should be further evaluated in this disease. DOX and ACD have some activity against soft tissue sarcomas clinically (Greenall et al., 1986), and resistance to these drugs may be mediated by increased cxpression of p-glycoprotein. Overexpression of p-glycoprotein is present in approximately 20% of some soft tissue sarcomas (Gerlach et al., 1987). The sarcoma cell lines we studied were sensitive to DOX and ACD, consistent with the lack of expression of p-glycoprotein. The 4 new cell lines were relatively resistant to MTX and VLB compared with HT-1080, a fibrosarcoma ccll line. Resistance to MTX in HS-16, HS-30 and HS-42 cells was partially explained by the inability of these cell lines to accumulate long-chain polyglutamates of MTX (Li et al., 1992). The mechanism of resistance of the HS-18 cells to MTX was associated with increased expression of DHFR, linked to increased free E2F due to the absence of pRb (Li et al., 1995). Alteration of the other cell-cycle regulators in these cell lines may also contribute to drug resistance. The establishment and phenotypic and genotypic characterization of these cell lines should provide a useful model for the study of soft tissue sarcomas and for evaluating new drugs or treatments. ACKNOWLEDGEMENTS The authors thank Drs. M. Brennan, L. Freedman, E. Casper and J. Woodruff for their continued support and co-operation. J.R.B. is an American Cancer Society Professor of Medicine and Pharmacology. This work was supported by NIH grant CA 47179. REFERENCES CORDOS-CARDO. C., FINSTAD, C.L., B A N D ~ K N.H. , and MELAMED, M.R., Immunoanatomical distribution of cytostructural and tissueassociated antigens in the human urinary tract. Amer. J. Parhol., 126, 269-284 (1 987). DENIS,N., Iorm, A,, KRUH, J., DAUTRY, F. and CORCOS,D., Stimulation of methotrexate resistance and dihydrofolate reductase gene amplification by c-myc. Oncogene, 6, 1453-1457 (1991). GERLACH, J.II., BELL,D.R., KARAKOUSIS, C., SLOCUM, H.K., KARTNER, N., RUSTUM, Y.M., LING,V. and BAKER,R.M., P-gl coprotein in human sarcoma: evidence for multidrug resistance. J. c b . Oncol., 9, 1452-1460 (1987). GRANA,X. and REDDY,E.P., Cell cycle control in mammalian cells: role of cyclins, cyclin dependent kinase CCDKs, rowth suppressor clin-dependent kinase inhibitors CCl?l’s. Oncogene, 11, GREENALL, M.J., MAGILI., G.B., DECOSSE,J.J. and BRENNAS,M.F., Chemotherapy for soft tissue sarcoma. Sue. Gynecol. Obsrer., 62, 193-198 (1986). IIAMBURGER, A.W. and SALMON, S.E., Primary bioassay of human tumors stem cells. Science, 197,461463 (1977). HELIN,K., HARLOW,E. and FAT~AEY, A., Inhibition of E2F-1 transactivation by direct binding of the retinoblastoma protein. Mol. cell. Biol., 13,6501-6508 (1993). JOHNSON, D.G., CRESS, W.D., JAKOI, I.. and NEVENS, J.R., Oncogenic capacit of the E2F1 gene. Roc. nut. Acad. Sci. (Wash.), 91, 1282312827 4994). KAYE, S.B., Docetaxel (taxotere) in the treatment of solid tumors other than breast and lung cancer. Semin. Oncol., 22 (Suppl. 4), 30-33 (1995). 1.1, W.W., FAN,J., HOCHHAUSER, D., BANERJEE, D., ZIELISSKI, Z., ALMASAN, A., YIS, Y., KELLY,R., WAHL,G.M. and BERTINO, J.R., Lack of functional retinoblastoma protein mediates increased resis- tance to antimetabolites in human sarcoma cell lines. Proc. nut. Acad. Sci. (Wash.),92,10400-10436 (1995). LI, W.W., LIN,J.T.. SCHWEI,IZER, B.I., TONG,W.P., NIEDZWIECKI, D. and BERTINO,J.R., Intrinsic resistance to methotrexate in human tissue sarcoma ccll lines. Cuncer Res., 52,3908-3913 (1992). MICHILLI, P., LI, W., LORENZI, M.V., MIKI,T., ZAKUT,R., GIVOL, D. and PIERCE, J.H., Inhibition of onco ene mediated transformation by ectopic expression of p2lWa“in Nl&T3 cells. Oncogene, 12, 775-784 (1996). MITELMAN, F. (ed.), Guidelinesfor cancer cytogenetics. Supplement to an intemarional vstem for cyrogenctic nomencluture, Karger, Basel (1991). ESTABLISHMENTOF HUMAN SOFT TISSUE SARCOMA CELL LINES 519 NIIMI,S., NAKAGAWA, K., YOKUTA, J.. TSUNOKAWA, Y.,NISHIO, K., do not overexpress receptors for epidermal growth factor. Brit. J. Cancer, 64,296-298 (1991). TERASHIMA. Y., SHIBURA, M., TERUDA, M. and SAIJO,N., Rcsistancc to anticancer drugs in NIH3T3 cclls transfected with c-rnyc and/or SI'RA-ITON, M.R., Moss, S., WARRES,W., PATTERSON, H., CLAH,J., c-II-rus genes. Brit. J. Cancer, 63,237-241 (1991). FISHER,C., BALL,A., THOMAS,M., GUSTERSON, B.A. and COOPER, SERRANO, M., GOMEZ-LAtlWL, E., DEPINHO,R.A., BEACH,D. and C.S., Mutation of the p53 gene in human soft tissue sarcomas: BAR-SAGI,D., Inhibition of ras-induced proliferation and cellular association with abnormalities of the RBI gene. Oncogene, 5, 12971301 (1990). transformation by ~ 1 6 ' "Science, ~ ~ . 267,249-252 (1995). R.L., SLEBOS,R.T.C., LEE, M.H., PLCNUTT, B.S., KESSIS,T.. WILLIAMS, TOGUCHIDA, J., YAMAFCCHI,T., RITCHIE, B., BEAUCHAMP, s.11..HERRERA,G.E., YAMAMURO,T., KOTOURA,Y., SASAKI, B.O., JACKS,T., HEDRICK, L., KASTAN, M.B.and CHO, K.R., p53- DAYTON, R.R., ISHIZAKI, K. and YAVDELL, dependent G I arrest involves Rb related proteins and is disrupted by M.S., LITTLE,J.B., WEICHSEI.BAUM, the human papillomavirus l x E; oncoprotein. Proc. nut. Acud. Sci. D.W., Mutation spectrum of the p53 gene in bone and soft tissue sarcomas. Cancer Res., 52,6194-6199 (1992). (Wash.), 91,5320-5324 (1994). SLOVAK, M.L., MICSKI, S.E.L., COLE,S.P.C., GERLACH, J.H., YOHEM, ZAMBETII, G.P. and LEVINE,A.J., A comparison of the biological K.H. and TRENT, J.M., Tumorigenic multidrug-resistant HT 1080 cells activities of wild-type and mutant p53. FASEBJ., 7,855-865 (1993).