Abstract
Background: Malignant fibrous histiocytoma (MFH) or undifferentiated pleomorphic sarcoma (UPS) is the most common soft-tissue sarcoma of late adult life. Further advances in genetic characterization are warranted. The aim of this study was to search for numerical and structural chromosomal anomalies in UPS. Materials and Methods: We investigated five sarcoma-specific chromosomal translocations, five oncogene amplifications as well as the numerical karyotype of 19 UPS samples and one UPS/MFH cell line (U2197) using FISH probes on interphase nuclei. Results: Our results demonstrate that chromosomal translocations involving CHOP, SYT, EWS, FUS and FKHR genes are absent. Furthermore, amplification of ERBB2 (10.5%) and MDM2 (10.5%) was observed whereas the EGFR, C-MYC and N-MYC genes were not amplified. Interestingly, predominant aneuploidies were found in eight chromosomes. Conclusion: The data demonstrate rarity of sarcoma-specific chromosomal breaks and oncogene amplifications in UPS, yet polysomic chromosomes appear more characteristically in this condition.
Soft tissue sarcomas (STS) are malignant tumors of mesenchymal origin. With approximately 11,280 new cases diagnosed annually in the United States they comprise less than 1% of all malignancies (1, 2). Nevertheless, STS comprise a large heterogeneous group with more than 50 diagnostic entities described, the more common entities include liposarcomas, leiomyosarcomas, synovial sarcomas, malignant peripheral nerve sheath tumors (MPNST) and malignant fibrous histiocytomas (MFH), also referred to as undifferentiated pleomorphic sarcomas (UPS) (3, 4). Since STS often show highly aggressive potential, the overall five-year survival rate is approximately 50% (5). Surgical resection with free margins and adjuvant radiotherapy represents the current gold standard for STS therapy. Yet, morbidity and mortality remain comparatively high and severe late-effects occur quite often (6-8).
Although it is assumed that sarcomas arise due to chromosomal aberrations and/or mutations in mesenchymal progenitor cells, the exact cellular origin of most of these tumors remains elusive (9). With regard to genetic abnormalities sarcomas are divided into two classes: The first class encompasses tumors that have specific genetic mutations, simple karyotypes and translocations potentially resulting in the formation of fusion genes. The second class comprises of sarcomas without specific mutations and chromosomal aberrations, but comprising complex karyotypes with numerous genetic gains and losses with no specific pattern (10). For the sarcoma of the first class, various fusion genes, which represent a characteristic feature for the appropriate entity, have been described over the years. Fusion genes like SS18-SSX1 and EWS-FLI1 are specifically linked to synovial sarcoma and Ewing's sarcoma/Primitive Neuroectodermal Tumor (PNET), respectively (11, 12). Some further sarcoma-specific chromosomal break regions are present in the CHOP gene in myxoid liposarcoma (13, 14) and FKHR in alveolar rhabdomyosarcoma (15). Recently, methods such as fluorescence in situ-hybridization (FISH) have been established to screen for known fusion genes (16). The resulting fusion proteins do not interact with the same interaction partners like the physiological wild-type proteins (17). This can alter the regulation of key proteins or transcription factors in tumor cells. Consequently, an altered expression profile can result, which may contribute to tumorigenesis. Accordingly, it has been found that the SS18-SSX1 fusion protein reduces the tumor-suppressive function of p53 by stabilizing its negative regulator MDM2 (18). This example demonstrates the complexity of genetic alterations in relation to the mechanisms of regulation. The discovery of particular chromosomal breaks and translocations is significant - not only for the development of new therapeutic strategies, but also for the confirmation of diagnosis and pathological findings.
Another important molecular genetic feature of cancer cells concerns copy number gains and amplifications of proto-oncogenes like the transcription factor C-MYC which binds to promoters and regulates up to 15% of all genes (19). C-MYC is amplified in diverse malignancies like in breast and prostate cancer (20, 21). Amplification of further factors as N-MYC and the receptor tyrosine kinases ERBB2 and EGFR have also been described in various malignancies like neuroblastoma, breast cancer and gliomas, respectively (22-24). The oncogene MDM2 is an inhibitor of apoptosis, whose amplification has been shown in sarcomas (25). Amplification or translocation of such genes could result in altered gene expression profile and consequently favor oncogenesis.
Profound knowledge of the origin and pathogenesis of STS is missing, and new insights or diagnostic characteristics are necessary. For a better understanding of the molecular mechanisms, intensive molecular genetic analysis is required, which is difficult due to the rarity and the large number of different histological subtypes of STS. To further investigate the genetics of UPS we examined three types of chromosomal aberrations in 19 UPS samples and one UPS/malignant fibrous histiocytoma cell line (U2197) by the use of specific FISH probes: First, we tried to identify five specific chromosomal breaks, which have already been found in different sarcoma entities (as described above). Secondly, we explored the amplification of five genes that harbor tumorigenic potential in diverse neoplasms. Finally, we determined the numbers of chromosomes in several cell nuclei of UPS cells.
Materials and Methods
Ethics statement. The participants gave their written informed consent, and the study was reviewed and approved by the ethical committee of the BG University Hospital Bergmannsheil, Ruhr-University Bochum, Germany with the registration number 3974-11. Experiments comply with the current laws of Germany.
Cell culture. The human MFH/uUPS cell line U2197 obtained from German Collection of Microorganisms and Cell Cultures (DSMZ), Braunschweig, Germany was authenticated via DNA (STR) profiling by the DSMZ in November 2013. Cells were grown in MEM supplemented with 20% FCS (Thermo Fisher Scientific Inc., Waltham, MA, USA), 0.165% sodium bicarbonate (PAA laboratories, Pasching, Austria) and 1% penicillin/streptomycin (PAA laboratories). The primary human undifferentiated sarcoma samples were harvested from patients' tissues at the BG University Hospital Bergmannsheil (Table 1). This study was approved by the local Ethics Committee and all of the patients gave written informed consent. Cell cultures were obtained from freshly received tumor resections by separation of the cells using Collagenase type II (Worthington Biochemical Corporation, Lakewood, NJ) as previously described (26). Cultures were maintained at 37°C in a humidified 5% CO2 atmosphere.
FISH probes. All probes were obtained from Kreatech (Kreatech, Amsterdam, Netherlands). All used probes are listed in Tables V, VI and VII. 100 nuclei of each UPS sample and the U2197 cell line were analyzed. For comparison and validation each probe was additionally tested on 100 nuclei of fibroblasts from a non-sarcoma patient defined as “healthy control”.
Preparation of samples for FISH. Sample preparation and the following FISH analysis were performed by standard procedures. Briefly, when nearly reaching confluency, cells were trypsinized and harvested in Carnoy's fixative (3:1 methanol to glacial acetic acid). The pellet was dripped onto the cold, wet slides, air-dried and pre-treated in 2x SSC prior to hybridization.
Hybridization of FISH compounds by co-denaturation. FISH probe mixtures were applied according to the manufacturer's protocols (Kreatech, Amsterdam, The Netherlands). Samples were denatured for 5 min at 75°C in a HYBrite hybridization system (Medcompare, South San Francisco, CA, USA), followed by hybridization of 12-16 h at 37°C. After hybridization, slides were immediately incubated in 0.4x SSC/0.3% NP40 at 72°C for 120 sec and afterwards in 2x SSC/0.1% NP40 for 60 sec at room temperature and air dried. After addition of 5-7 μl DAPI solution, slides were evaluated by fluorescence microscopy with the appropriate image processing software (Isis; Metasystems, Altlussheim, Germany).
Results
Absence of special chromosomal breaks. We investigated the distribution of five sarcoma-specific chromosomal breaks (in the CHOP, SYT, EWS, FUS and FKHR genes), in 19 UPS samples, U2197 cells and normal human fibroblasts. One hundred nuclei of each sample were analyzed by break-apart FISH probes. A chromosomal break was assumed when signals were separated by more than the 2-fold of the signal diameter. Based on the experience from routine diagnostics, where among other things the same probes are used, the presence of chromosomal breaks in >10% nuclei was used as a cut-off to categorize the sample as positive for the respective chromosomal break. The analyses revealed that all 19 UPS samples, U2197 cells and normal human fibroblasts were negative for chromosomal breaks in the CHOP, SYT, EWS, FUS and FKHR genes (Figure 1A, B, C, D and E). Among the negative samples, CHOP (4/19), SYT (2/19), EWS (2/19) and FUS (1/19), showed negligible nuclei (1-8%) with putative chromosomal breaks (signals more than 2 diameters apart) where as FKHR exhibited no chromosomal break at all (Figure 1F, G, H and I). The chromosomal breaks in a few nuclei (1-8%) were below the cut-off value, applied to clinical diagnostics of tumor entities (>10% of nuclei). Such chromosomal breaks/translocations were neither found in the U2197 cell line nor in the healthy control. The results are summarized in Table II. Thus, UPS appear characterized by an absence of chromosomal breaks on CHOP, SYT, EWS, FUS and FKHR genes.
Prevalence of ERBB2 and MDM2 amplification. We investigated the prevalence of amplification of five potential oncogenes (EGFR, ERBB2, MDM2, C-MYC and N-MYC). One hundred nuclei of each sample were analyzed using dual-color FISH probes, and the presence of duplication/amplification in >10 % nuclei were used as a cut-off to categorize the sample as positive for the respective gene. ERBB2 and MDM2 amplifications were prevalent in 2 of the 19 samples, each (Table III). No amplification of EGFR, C-MYC and N-MYC was observed. Representative images of amplifications are shown in Figure 2. EGFR gene amplification was observed in U2197 cells. Normal human fibroblasts showed no amplification in any of the 5 genes.
Presence of chromosome-specific aneuploidy in UPS. In order to address whether UPS are characterized by aneuploidy or specific chromosomal gains, numerical karyotype of all samples were established. In normal human fibroblasts, UPS and U2197 disomy and tetrasomy were observed (Figure 2A). As normal cells are diploid and natural DNA replication in the S phase can exhibit tetrasomy, i.e. 2 chromatids of each chromosome (2n4c), it is indistinguishable from disomy with pathological rearrangements and tetraploidy (4n4c). Hence we ignored chromosomes with 2 and 4 copies from the analysis. In normal human fibroblasts, the presence of DNA ploidies other than disomy and tetrasomy were negligible (see below). In sharp contrast, UPS and U2197 conspicuously exhibited abnormal ploidies (Figure 3). In UPS cells and U2197 cells, the most noticeable increase in aneuploidy was trisomy in 5.1% and 20.9%, respectively, in comparison to a negligible 0.6% in normal human fibroblasts. Also, increased rates of pentasomy (p) and hexasomy (h) were observed in UPS cells (p=1%, h=1.7%) and U2197 cells (p=8%, h=4.1%) in comparison to a negligible percentage in normal human fibroblasts (p=0%, h=0.1%). Furthermore, 13/19 UPS samples displayed highly complex karyotypes whereby each chromosome was present in five to ten copies in more than 4% of the nuclei. In U2197 cells each chromosome was present in five to ten copies in almost 16% of the nuclei characteristic for a deteriorated karyotype.
We also investigated individual chromosome-specific aneuploidy in fibroblasts, UPS and U2197 cells. The fibroblasts showed a constant karyotype. None of the chromosomes existed in more than three copies in excess of 10% of cell nuclei. For each chromosome only 1-2% of nuclei showed a chromosomal copy number gain. In sharp contrast, UPS exhibited copy number changes in different chromosomes, namely 1, 3, 4, 7, 8, 11, 12, 17 (Figure 4, Table IV) in excess of 10% of cell nuclei. Chromosomes 1, 3, 4, 7, 8, 11, 12 and 17 were found trisomic in 5.3% to 21.1% (4/19) UPS, respectively, in >10% of nuclei. Chromosomes 4, 7, 8, 12 and 17 exhibited hexasomy in 5.3% (1/19) UPS in >10% of nuclei. Chromosomes 1, 7, and 17 demonstrated pentasomy in 10.5% (2/19), 5.3% (1/19) and 5.3% (1/19) UPS in >10% of nuclei. The most frequent aneuploidies in UPS patients are summarized in Table IV. In U2197 cells, twelve different chromosomes appeared in more than three copies with nine of these chromosomes existing in excess of 90% of nuclei.
Discussion
We investigated numerical and structural chromosome anomalies in undifferentiated pleomorphic sarcoma and examined five special chromosomal breaks, five potential gene amplifications as well as performed numerical interphase karyotyping in UPS and the UPS/MFH cell line (U2197). Chromosomal breaks/translocations in the genes CHOP, SYT, EWS and FKHR are characteristic for myxoid liposarcoma (13, 14), synovial sarcomas (11), Ewing's sarcoma/PNET(12) and alveolar rhabdomyosarcoma(15), respectively. Our study demonstrates that UPS exhibit no chromosomal breaks of the CHOP, SYT, EWS, FUS and FKHR genes detectable by standard FISH probes. Absence of CHOP rearrangements has been independently reported in MFH (27). We propose that although the above-mentioned five different breaks/translocations may be relevant in the pathogenesis of other sarcoma entities, they play no critical role in the pathogenesis of UPS. To date, 41 gene fusions have been described in 17 different sarcoma types (28), but no specific chromosomal translocation has been reported in UPS. The present study reinforces this observation (29).
EGFR, ERBB2, MDM2, C-MYC and N-MYC gene amplifications have been reported in various cancers. Our study demonstrates that UPS exhibit no EGFR, C-MYC and N-MYC amplifications. EGFR has been reported to be over-expressed in MFH (30), and the EGFR gene amplification in UPS was analyzed to explain high EGFR expression levels. We observed no EGFR amplification in UPS (yet exclusively in U2197 cell line). Therefore, we conclude that the reported over-expression of EGFR in UPS/MFH cannot be substantiated via gene amplification/dosage events. Hence, other regulatory mechanisms, like elevated transcriptional activity or inactivation of transcriptional suppressors, need to be considered as plausible mechanisms of EGFR overexpression in MFH. Interestingly, ERBB2 and MDM2 genes demonstrate amplification in UPS. In 10.5% (2/19) UPS studied, the number of signals for ERBB2 (17q12) and MDM2 (12q15) was greater than the number of signals for centromeres of chromosome 17 and chromosome 12, respectively. In addition, chromosome 12 and 17 also exhibited conspicuous gains of copy numbers. The elevated dosage of ERBB2 and MDM2 in UPS may result from complex chromosomal rearrangements or, both, individual gene amplification as well as copy number gain of chromosome 17 and 12. Such increases in gene dosage are consistent with the hypothesis that gene amplifications confer growth advantages (31, 32). Momand et al. investigated MDM2 amplification in 3889 samples of 28 tumor types from previously published sources and reported a 7% overall frequency of MDM2 amplification. MDM2 amplification was observed in 19 tumor types with the highest frequency observed in soft tissue tumors (20%) (33). Interestingly, 21% among the 163 MFH exhibited MDM2 amplification (33). Recently, a small-molecule MDM2 antagonist has been described as a possible therapy, and MDM2 overexpressing sarcomas yielded good responses to the respective inhibitor (34, 35). Despite these advances, further studies on amplification mechanisms, and in particular, about the initiating processes of gene amplification are certainly warranted.
Sarcomas display multiple, complex karyotypic abnormalities, and these genetic alterations are an important adjunct to standard morphological and immunohistochemical diagnoses (36-39). Consistent with our observation of aneuploidy in UPS, independent cytogenetic studies in MFH mention also aneuploidy (36-39). In the current analysis, we omitted to include disomy and tetrasomy, as normal cells are diploid and natural DNA replication in the S phase can pretend tetraploidy, making pathological diploidy and tetraploidy indistinguishable from normal disome and tetrasome phases of the cell. Thereafter, the most conspicuous copy number gains that emerged were in specific chromosomes namely trisomy (1, 3, 4, 7, 8, 11, 12, 17), pentasomy (1, 7, 17) and hexasomy (4, 7, 8, 12, 17). Of these, the copy number gains in chromosomes 1, 7, 8 in MFH has been independently reported by Tarkkanen et al. using comparative genomic hybridizations (CGH) (40). In our study, on average each chromosome exhibited 3-10 copy gains in ~10% of UPS nuclei. Chromosomal mis-segregation in normal diploid cells occurs generally below 1% (reviewed in (41)), as observed in our results from normal human fibroblasts. Not surprisingly, the numerical aberrations in the UPS/MFH cell line U2197 was larger than in the UPS samples. In U2197 cells, each chromosome was on average present in 3-10 copies in almost 36.8% of the nuclei. During a long period of cell cultivation, cell lines may progressively “evolve” in several aspects including morphology, vitality, genetics and/or epigenetics (42).
Among the chromosomal gains reported in our study, gain of chromosome 8 has been earlier reported in prostate cancer (21) and gastric cancer (43), gains of chromosomes 8 and 12 in Ewing's sarcoma (44) and gains of chromosomes 8, 12 and 17 in stage c colon cancer (45). Additionally, gain of the q-arm of chromosome 8 has also been reported in a wide variety of cancers like breast, head and neck, gastric and pancreatic cancer (43, 46-48). Such numerical abnormalities of chromosome 8, on which C-MYC is located, has been suggested as an important mechanism in the increase of the C-MYC copy number (49) and its availability to bind and regulate up to 15% of all genes (19). Although chromosome 8 exhibited trisomy and hexasomy in UPS, the number of signals for 8q24 (the locus of C-MYC) was similar to the number of signals for centromeres of chromosome 8 ploidy. Therefore, our data may indicate that elevation of C-MYC protein in MFH may be achieved via copy number gain of the entire chromosome 8 and not through individual C-MYC amplifications (40). Other genes involved in cell growth, motility and survival located on chromosome 8q and reported in various cancer entities are SNAI2, PLAG1, Sulf1, CTHRC1, ENPP2 and ASAP1. Therefore their role in UPS pathogenesis needs further evaluation.
UPS cells lack specific chromosomal breaks but are characterized by ERBB2 and MDM2 amplification and specific polysomies. In the current study we observed no correlation between these structural and numerical aberrations and the histological subtype or tumor grading. This observation is consistent with the report of the chromosomes and morphology (20) study group, that a differential diagnostic sub-classification of pleomorphic sarcomas by means of cytogenetic analysis is implausible and that the karyotype could not be used to predict clinical outcome (50). Metaphase analyses and array CGH may be useful tools to further elucidate clonal chromosomal changes in UPS.
Acknowledgements
The Authors would like to thank Viktoria Albrecht for her expert technical assistance. This work was supported by the Medical Faculty of the Ruhr-University Bochum (FORUM: F667N-2010).
Footnotes
-
↵* These Authors contributed equally and are named in alphabetical order.
- Received April 10, 2014.
- Accepted May 22, 2014.
- Copyright© 2014 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved