Background
Chondrosarcoma is a malignant bone neoplasm characterized by the deposition of a hyaline cartilaginous extracellular matrix. With an incidence of 1:50,000 it typically occurs in adults in their 3
rd to 6
th decade of life. Chondrosarcoma represents a heterogeneous group of tumors. Primary central chondrosarcoma is defined by the formation of hyaline cartilage with decreasing matrix production in higher grades and constitutes about 80% of all chondrosarcomas [
1]. Dedifferentiated chondrosarcoma is characterized by a low-, or intermediate grade chondrosarcoma juxtaposed to a high grade anaplastic sarcoma and constitutes about 10% of all chondrosarcomas [
2].
Both high grade conventional and dedifferentiated chondrosarcoma respond poorly to conventional chemo- and/or radiotherapy, have a high metastatic rate, and consequently have a very poor prognosis [
3]. It is because of these features that there is an urgent need for model systems in pre-clinical research aimed at evaluating new targeted treatment strategies for chondrosarcoma [
4].
Recently IDH1 and IDH2 mutations were found in conventional central and dedifferentiated chondrosarcomas [
5]. IDH1 and IDH2 mutations are well known in gliomas [
6], but are notoriously difficult to grow in culture [
7]. This is a feature shared by, in particular, grade I chondrosarcomas. Recently, a new cell line derived from a grade II chondrosarcoma was published, CH-3573 [
8]. Over the last years, cell lines derived from dedifferentiated chondrosarcomas have been developed [
9,
10]. In the pursuit of expanding the panel of cell lines we have succeeded in creating three new chondrosarcoma cell lines. L835 is derived from a grade III conventional chondrosarcoma, while L2975 and L3252 originate from dedifferentiated chondrosarcomas of bone. These three new cell lines provide a valuable addition to the current panel of chondrosarcoma cell lines.
Methods
Culture of human chondrosarcoma cells
Tumor-tissue derived from three resected specimens derived from one conventional and two dedifferentiated chondrosarcomas were used for culture. Samples were coded and all procedures were performed according to the ethical guidelines “Code for Proper Secondary Use of Human Tissue in The Netherlands 2002” (Dutch Federation of Medical Scientific Societies
http://www.federa.org/sites/default/files/bijlagen/coreon/codepropersecondaryuseofhumantissue1_0.pdf). Specimens were washed 3x with RPMI1640 (Gibco, Invitrogen Life-Technologies, Scotland, UK) containing 1% penicillin/streptomycin (100U/mL), minced with razor blades and immersed in collagenase dispase overnight. After washing, the cells were transferred into small collagen-coated culture flasks and cultured in RPMI1640 supplemented with 20% heat inactivated Fetal Calf Serum (Gibco, Invitrogen Life-Technologies, Scotland, UK), 1% L-glutamax, and 1% penicillin/streptomycin (100U/mL). Cells were grown in a humidified incubator with 95% air and 5% CO
2 and cultured until stably multiplying.
COBRA-Fluoresence in-situ hybridization
COBRA-FISH on metaphase slides was performed as described previously [
11]. For each cell line several cell culture passages were studied (L835: passage 17 and 35, L2975: passage 20 and 30, L3252: passage 7, 8, and 20) and karyotypes were described for each cell line according to the International System of Human Cytogenetic Nomenclature (ISCN) 2009.
Expression of cartilaginous genes
RNA was isolated from L835 (passage 40), L2975 (passage 58), and L3252 (passage 21). Chondrogenic phenotype was assessed using RT-PCR for collagen I, IIB, III, and X, aggrecan, and SOX9 as described previously [
12].
Assessment of cell line identity
DNA isolation from cell pellets was performed using the wizard genomic DNA purification kit (Promega, Madison, WI) according to manufacturer’s instructions. DNA concentrations were measured using a Nanodrop ND-1000 spectrophotometer and quality was checked on a 1% agarose gel stained with ethidium bromide. Identity of cell lines was confirmed using the PowerPlex® 1.2 system (Promega Benelux BV, Leiden, The Netherlands). For L835 passage 36 was compared to primary tumor tissue, for L2975 passage 37 was used, and for L3252 passage 20 was compared to primary tumor tissue.
Doubling time and migration assays
The RTCA xCelligence system (Roche Applied Sciences, Almere, the Netherlands), based on cell-electrode substract impedance detection technology [
13], was used for doubling time and migration assays. Prior to starting experiments cell number curves were run to determine optimal growth curves and for doubling time experiments cell lines were plated at a density of 1,000 cells per well for L2975 and L3252 and 10,000 cells per well for L835 in growth medium (10% FCS in RPMI1640). For migration experiments, 100,000 cells were optimal.
For doubling time assays, 30 minutes after plating, view-plates were loaded into the RTCA station in the cell culture incubator. Cell index (CI) was acquired every hour.
Proliferation was monitored for 400 hrs. Every day plates were taken from the machine and most representative areas were photographed using a Zeiss axiovert 40C light microscope (Rijswijk, the Netherlands).
For migration assays, lower wells of the SIM plates (migration plates) were filled with growth medium (20% FCS in RPMI1640) as a chemoattractant, and cells were plated in the top wells in empty buffer (RPMI1640 only). CIM plates with 8 μm pores were loaded into the RTCA station in the cell culture incubator immediately after plating and cell index (CI) was acquired every 5 minutes. Migration was monitored for 24 hrs. Experiments were performed in triplicate.
Mutation analysis
Mutation analysis was performed for TP53 (exons 5–8) [
14], and IDH1 and −2 exons 4 [
15] using direct sequencing of DNA as described (14;15). Mutation analysis for PIK3CA, KRAS, BRAF, EGFR was performed using hydrolysis probes assay [
16] at L835 (passage 36), L2975 (passage 37), and L3252 (passage 20). Mutation analysis for TP53 was performed at those same passage numbers and IDH mutation analysis was performed at L835 passage 38 and 47, L2975 passage 31 and 46, and L3252 passage 20, as well as on DNA obtained from CH-3573 [
8]. To determine expression of the IDH mutated allele cDNA was generated using 1 μg total RNA as described [
12] for L835 (passage 38), L2975 (passage 31), and L3252 (passage 20). Primers were designed with primer3 software (
http://frodo.wi.mit.edu/primer3/) and ordered from ISOGEN Bioscience BV (Maarssen, the Netherlands). PCR was done with the quantitative PCR core kit for SYBR green I supplemented with fluorescin (Eurogentec, Seraing, Belgium) on 0.2μL cDNA per reaction in an iCycler iQ Real-time Detection system (Bio-Rad Laboratories, Hercules, CA). PCR was done for 40 cycles. PCR products were purified using QIAquick PCR purification Kit (Qiagen, Hilden, Germany) according to manufacturer’s instructions. Purified products were sequenced by Macrogen (Amstelveen, the Netherlands) and resulting sequences were analyzed using MutationSurveyor DNA Variant Analysis software (Softgenetics, UK). Primer sequences and annealing temperatures are listed in Table
1.
IDH1 genomic | Forward | CGGTCTTCAGAGAAGCCATT | 59.4 | 113 |
IDH1 genomic | Reverse | GCCAACATGACTTACTTGATCC | 58.6 | |
IDH2 genomic | Forward | AACATCCACGCCTAGTCC | 56.3 | 90 |
IDH2 genomic | Reverse | CAGTGGATCCCCTCTCCAC | 60.5 | |
IDH1 cDNA | Forward | CGGTCTTCAGAGAAGCCATT | 59.4 | 131 |
IDH1 cDNA | Reverse | AGGCCCAGGAACAACAAAAT | 56.4 | |
IDH2 cDNA | Forward | AGTGTGGCTGCAAGTGTGC | 60.0 | 365 |
IDH2 cDNA | Reverse | GAGATGGACTCGTCGGTGTT | 60.1 | |
Array-CGH analysis
Array-CGH was performed on DNA derived from the primary tumor as well as from cultured cells of all three cases as described [
17]. DNA of L835 passage 36, L2975 passage 37, and L3252 passage 20 was used. In brief, labeling of 1 μg DNA was performed using the BioPrime Total Genomic Labeling System (Invitrogen Corporation, Carlsbad, CA) following the manufacturer’s protocol. As reference, DNA from a commercial source (Promega Corporation, Madison, WI) was used. Labeled test and reference samples were mixed and hybridized as a gender mismatch. Hybridization was performed on an Agilent 105 k oligonucleotide array according to manufacturer’s instructions. Slides were scanned using the Agilent Scanner with 5 μm scan resolution. Scan images were processed with the Feature Extraction Software and the generated raw data files were analyzed using Genomic Workbench (Agilent Technologies, Santa Clara, CA). In short, the mean of the background corrected and Lowess normalized log2 ratios of identical features was calculated. Normalization of ratios was done using the overall values as well as the values of the control reporter probes on the array. Aberrations were calculated using the ADM-2 algorithm with a threshold of 8.6 and displayed with a moving average of 1 Mb.
Tumorigenicity in mice
All the experimentation involving laboratory animals was approved by the Institutional Animal Care of Valencia University and the Local Government and was performed in accordance with the national legislation of Spain. Male nude mice were purchased from IFFA-CREDO (Lyon, France), and kept under specific pathogen-free conditions throughout the experiments. For each cell line, 2,000,000 cells were subcutaneously injected in a total of 3 mice (2 months old) under sterile conditions. Tumor was removed when size reached 4 mm in diameter and a fragment of non-necrotic tumor, about 3 to 5 mm3 in size, was used for xenografting into two new male nude mice. The second neoplasm was removed when the size reached 20 mm in diameter. From each tumor a part was snap-frozen in liquid nitrogen, a part fixed in formalin and embedded in paraffin, a part was used for xenografting into 2 new mice, and a part for further culturing of post-xenograft cell lines.
(Immuno)histochemical analysis
L835 (passage 35), L2975 (passage 55), and L3252 (passage 17) cells were fixed in formalin and prepared using the Shandon Cytoblock cell block preparation system (Thermo Scientific, Etten-Leur, the Netherlands). Cells were embedded in paraffin according to standard laboratory procedures for tissue fixation. Sections (4-μm thick) of these paraffin blocks as well as from formalin fixed paraffin embedded original tumor tissue and xenograft passages, were used for H&E, toluidine blue, and immunohistochemistry for ki67, p53, and p16 according to standard procedures [
18]. Details of the antibodies can be found in Table
2.
Table 2
Antibody properties
p16
| g175-405 | 1:800 | Citrate | - | BD Pharmingen (550834) |
p53
| DO-7 | 1:800 | Citrate | - | Dako (M7001) |
Ki67
| MIB-1 | 1:800 | Citrate | - | Dako (M7240) |
Discussion
Chondrosarcoma is the second most common primary sarcoma of bone and to date unresectable chondrosarcomas have a poor outcome [
3]. Grade III and dedifferentiated chondrosarcomas are extremely aggressive in nature and there is an urgent need for model systems facilitating research in order to develop novel therapeutic strategies. Growing chondrosarcoma cells in culture, however, is a challenge and well growing chondrosarcoma cell lines are sparse. We present here the establishment and characterization of three new chondrosarcoma cell lines originating from grade III and dedifferentiated chondrosarcoma.
Recently chondrosarcoma has been found to harbor IDH1 and IDH2 mutations (5;15) and we published that the mutation is retained in a subset of chondrosarcoma cell lines [
15]. In glioma IDH mutations seem to be the earliest event in gliomagenesis even before TP53 mutations occur [
20]. In conventional chondrosarcoma we observe a similar phenomenon, where IDH mutations are present already in a high percentage of low-grade tumors and TP53 mutations are observed to increase with grade (4;5;15). Cell lines created from IDH mutant gliomas have been reported to eliminate their IDH mutation under standard culture conditions [
7]. Recently, however, a glioma cell line carrying an endogenous IDH1 R132H mutation was published, but this cell line showed a slow growth rate in culture [
21]. We here present three chondrosarcoma cell lines, one carrying an IDH1 R132C mutation, one carrying an IDH2 R172W mutation, and one wild type for IDH mutations with stable karyotypes and steady growth patterns. These cell lines show numerical changes and additional mutations. We speculate that in IDH mutant chondrosarcoma the acquisition of additional mutations as we have shown here have facilitated their growth in culture.
The inactivation of tumor suppressor genes is a well-known phenomenon in cancer and p16 mutations have been reported in 20-41% of human chondrosarcomas [
22‐
25]. Interestingly, all studies observed loss of p16 to be correlated with increasing histological grade in conventional chondrosarcoma. Recently, we showed inactivation of p16 in 30/38 (79%) dedifferentiated chondrosarcoma cases [
26]. We previously published three chondrosarcoma cell lines to be negative for p16 using western blot [
18] and upon overexpression of p16 using lentiviral vectors the metabolic activity and cell viability of these cell lines was decreased, indicating loss of p16 to play a role in the proliferative capacity of chondrosarcoma cells. Introduction of p16 in the endogenously TP53 mutant HT-1080 fibrosarcoma cell line, which was recently reported to carry an IDH1 R132C mutation [
5], also led to cell cycle arrest and growth inhibition [
27‐
29]. We report here three new chondrosarcoma cell lines lacking p16 expression based on a homozygous deletion of the CDKN2A locus as shown by aCGH analysis, and confirmed loss of p16 expression using immunohistochemistry. Moreover, aCGH analysis showed a copy number loss around the 17p13.1 locus in L835, whereas a copy number gain was observed in L2975 and L3252. However, mutation analysis for TP53 showed no activating mutations in exons 5–8, and immunohistochemistry showed no p53 overexpression. Together, our data suggest that while IDH mutations are important as early events in a subset of chondrosarcomas, additional inactivation of p16 may be crucial for acquiring a more aggressive phenotype.
The literature presents us with 5 conventional chondrosarcoma cell lines that have been well characterized using pathological, immunohistochemical, and molecular genetic methods [
8,
30‐
32], and we here present L835 as an additional cell line. We previously published L835 to be able to form 3D pellets [
33] and we now show it to be highly stable in culture. L835 cell line showed a slower growth rate compared to the dedifferentiated chondrosarcoma cell lines. In all cases complex genome alterations were observed. Dedifferentiated chondrosarcoma is comprised of two separate components, a high grade anaplastic component and a low to intermediate grade cartilaginous component [
2]. The histogenesis has been under debate but evidence points to a single precursor cell with early separation of the two components as a small number of genetic changes is identical in both components, with additional genetic alterations in the anaplastic component [
26,
34]. Indeed, 3 out of 3 dedifferentiated chondrosarcomas with IDH1 mutations carried the mutation in both components [
26]. Moreover, 79% of the anaplastic and 82% of the cartilaginous components show loss of p16 expression [
26]. L2975 and L3252 were both derived from the recurrence of a dedifferentiated chondrosarcoma; both cell lines exhibited a higher growth rate in vitro, than the L835 cell line, but the cells in culture expressed chondrogenic markers. L2975 proved to be the most aggressive cell line both in culture and in our in vitro migration assay, which may explain why this was the only cell line to be successfully xenografted. We show here the use of L2975 dedifferentiated chondrosarcoma cells with an IDH2 R172W mutation in mouse models, which can be an important asset in the research for new treatment strategies.
Conclusions
We report the establishment and molecular, genetic and functional characterization of one grade III (L835) and two dedifferentiated chondrosarcoma (L2975 and L3252) cell lines. This represents a substantial addition to the already existing panel of chondrosarcoma cell lines, which together may reflect their heterogeneity. In addition to the existing cell lines these cell lines present the field with an extensive model system as heterogeneous in IDH1 and IDH2 and TP53 mutations as the tumors they are derived from. This panel can be implemented in studies ascertaining human chondrosarcoma tumorigenesis, should provide useful tools in the ongoing search for new targeted therapies, and aid in expanding our knowledge on the role of IDH1 and IDH2 mutations in chondrosarcoma formation.
Financial support
Netherlands Organization for Scientific Research (917-76-315: J.G.v.O, and J.V.M.G.B.), Dutch Cancer Society (UL2010-4873: J.G.v.O. and J.V.M.G.B.). Leiden University Medical Centre and University of Valencia Medical School are partners in the context of EuroBoNet, a European Commission granted Network of Excellence to study the pathology and genetics of bone tumors (018814).
Acknowledgements
The authors thank Jaap van den Eendenburg, Annemarie Koornneef, Cathelijn Waaijer and Silvia Calabuig Fariñas for expert technical assistance. The authors also thank Anne-Marie Cleton-Jansen for fruitful discussions.
The cell lines will be made available upon request for non-commercial use only under the restrictions of an MTA.
Competing interests
The authors have no conflict of interest to declare.
Authors’ contributions
JGvO, JVMGB, KS, PCWH, and AL-B conceived of the study. JGvO, DdJ, MAJHvR, IM, and KS performed the experiments. JGvO, JVMG, PDSD, and CSvR participated in analysis and interpretation of patient data and all authors read and approved the final manuscript.