Oncogene
SEARCH     advanced search my account e-alerts subscribe register
Journal home
Advance online publication
Current issue
Archive
Press releases
For authors
For referees
Contact editorial office
About the journal
For librarians
Subscribe
Advertising
naturereprints
Contact NPG
Customer services
Site features
NPG Subject areas
Access material from all our publications in your subject area:
Biotechnology Biotechnology
Cancer Cancer
Chemistry Chemistry
Dentistry Dentistry
Development Development
Drug Discovery Drug Discovery
Earth Sciences Earth Sciences
Evolution & Ecology Evolution & Ecology
Genetics Genetics
Immunology Immunology
Materials Materials Science
Medical Research Medical Research
Microbiology Microbiology
Molecular Cell Biology Molecular Cell Biology
Neuroscience Neuroscience
Pharmacology Pharmacology
Physics Physics
Browse all publications
 
10 September 2001, Volume 20, Number 40, Pages 5755-5762
Table of contents    Previous  Article  Next   [PDF]
Fusions of the SYT and SSX genes in synovial sarcoma
Marc Ladanyi

Department of Pathology, Memorial Sloan-Kettering Cancer Center, New York, NY, USA

Correspondence to: M Ladanyi, Department of Pathology, Memorial Sloan-Kettering Cancer Center, 1275 York Ave., New York, NY 10021, USA. E-mail: ladanyim@mskcc.org

Abstract

Synovial sarcomas are high grade spindle cell tumors that are divided into two major histologic subtypes, biphasic and monophasic, according to the respective presence or absence of a well-developed glandular epithelial component. They contain in essentially all cases a t(X;18) representing the fusion of SYT (at 18q11) with either SSX1 or SSX2 (both at Xp11). Neither SYT, nor the SSX proteins contain DNA-binding domains. Instead, they appear to be transcriptional regulators whose actions are mediated primarily through protein-protein interactions, with BRM in the case of SYT, and with Polycomb group repressors in the case of SSX. Ongoing work on the SYT-SSX fusion and synovial sarcoma should yield a variety of data of broader biological interest, in areas such as BRM and Polycomb group function and dysfunction, transcriptional targets of SYT-SSX proteins and their native counterparts, differential gene regulation by SYT-SSX1 and SYT-SSX2, control of glandular morphogenesis, among others. Oncogene (2001) 20, 5755-5762.

Keywords

chromosomal translocation; transcription factor; nuclear localization

Introduction

Synovial sarcomas account for almost 10% of all soft tissue sarcomas and typically arise in the para-articular regions in adolescents and young adults, but actually bear no biologic or pathologic relationship to synovium. They are considered high grade sarcomas which kill at least 25% of patients in the first 5 years after diagnosis, in spite of the best currently available management (Lewis et al., 2000). Synovial sarcomas are divided into two major histologic subtypes, biphasic and monophasic. Biphasic tumors contain both epithelial cells arranged in glandular structures and spindle cells, whereas monophasic types are entirely composed of spindle cells, lacking the glandular epithelial component. Expression of epithelial antigens is detected by immunohistochemical methods in the epithelial component in all biphasic tumors and in scattered epithelioid or spindle cells in all biphasic tumors and most but not all monophasic tumors. Ultrastructurally, the epithelial cells lining the glandlike spaces in biphasic tumors show luminal microvilli and are interconnected by various types of cell junctions.

Fusions of SYT with SSX1 or SSX2 in synovial sarcoma: prevalence and structural aspects

A t(X;18) is detected cytogenetically in over 90% of synovial sarcomas, regardless of histologic subtype. In relation to this recurrent X chromosome translocation, it is of incidental interest to note that the male to female ratio for this tumor is approximately even. Cloning of the translocation breakpoints initially showed that the t(X;18) results in the fusion of two novel genes, designated SYT (at 18q11) and SSX (at Xp11) (Figure 1) (Clark et al., 1994). SYT encodes a protein of 387 amino acids, of which all but the C-terminal 8 are fused to SSX. The exon structure of SYT is not available but the intron rearranged in the t(X;18) appears greater than 13 kb (Clark et al., 1994; Geurts van Kessel et al., 1997). Subsequent work showed that the Xp11 breakpoint actually involves either of two closely related genes, SSX1 (Xp11.23) and SSX2 (Xp11.21) (Crew et al., 1995; De Leeuw et al., 1995). These are respectively located in the vicinity of ornithine aminotransferase-like pseudogenes 1 and 2 (OATL1 and OATL2), previously defined by fluorescence in situ hybridization (FISH) as alternative breakpoints in synovial sarcomas (Shipley et al., 1994). The ratio of SYT-SSX1 to SYT-SSX2 cases is close to 2 : 1 (Crew et al., 1995; Kawai et al., 1998; reviewed in Dos Santos et al., 2001). The term SYT-SSX fusion is still used to refer to these fusions as a group.

The SSX1 and SSX2 genes encode 188 amino acid proteins which are highly similar. The exon structure of the SSX1 and SSX2 genes has recently been published (Gure et al., 1997). In both genes, the fourth intron is rearranged by the t(X;18). Whether the occurrence of translocations is purely random and their specificity reflects only the selection of functional oncogenic proteins (Barr, 1998), or whether the introns involved are somehow more prone to recombine (Bouffler et al., 1993) is an persistent issue in the pathogenesis of tumors with gene fusions. Because SSX1 is involved in the SYT-SSX fusion nearly twice as often as SSX2, we recently studied the sizes and sequences of these introns for clues to why SSX1 intron 4 is more often rearranged than SSX2 intron 4 (M Ladanyi and MY Lui, GenBank AY034079 and AY034080). These introns are essentially of the same size (SSX1: 1985 bp, SSX2: 1983 bp), arguing against a simple stochastic model of breakpoint occurrence. The sequence analysis shows 88% identity between the two introns (SSX1 and SSX2 are 89.1% identical in their cDNA sequences). The SSX2 genomic sequence is also available in chromosome X working draft sequence segment NT 25348. The high level of identity and its similar level in both intronic and exonic sequences of SSX1 and SSX2 highlights the recency of the duplication event leading to the formation of this gene family. Furthermore, sequence analysis of the SSX1 and SSX2 introns revealed no significant difference in repetitive elements, which have been proposed to be recombinogenic in other translocations. Specifically, both introns contain one ALU-Sc repeat in their 5' portion and 2 MIR repeats at the 5' end. Since there is no obvious structural basis for the differential involvement of SSX1 compared to SSX2, it is tempting to speculate that it may instead reflect small functional differences in the oncogenicity of SYT-SSX1 vs SYT-SSX2 leading to in vivo selection for the former or that it may be due to differences in chromatin structure around SSX1 and SSX2. At least four other SSX family genes have been identified in Xp11 (Chand et al., 1995; De Leeuw et al., 1996; dos Santos et al., 2001). Of these, only SSX4 has been found to be rearranged with SYT, and this only in two otherwise unremarkable cases of synovial sarcoma (Skytting et al., 1999; Mancuso et al., 2000). SSX1 and SSX2 seem to escape X-inactivation (Miller et al., 1995; and M Ladanyi, unpublished results).

Studies of the pattern of developmental and adult expression of SYT and SSX have so far provided few clues to their physiological function. Murine SYT is widely expressed in early embryogenesis and in adult tissues, and somewhat more restricted in fetal development to cartilaginous tissue and certain neuronal and epithelial cells (de Bruijn et al., 1996). The SSX genes appear to belong to a family of human tumor antigens whose expression in adults is restricted to tumors and germ cells ('cancer-testis' antigens) (Tureci et al., 1996; Gure et al., 1997). Most cancer-testis antigens are encoded by genes on chromosome X (Tureci et al., 1998). SSX expression during development has not so far been studied. There are no reported mouse knockouts of the SYT and SSX genes which could further illuminate their normal role in development.

The SYT-SSX chimeric transcript represents a highly sensitive diagnostic marker for synovial sarcoma. The prevalence of this fusion in synovial sarcoma approaches 100% when there is adequate tumor RNA for RT-PCR. Its specificity for synovial sarcoma, not unexpected from the cytogenetic literature, has been confirmed by molecular screening of other morphologically similar sarcomas (Hiraga et al., 1998; van de Rijn et al., 1999; Guillou et al., 2001). The clinical usefulness of its molecular demonstration in cases of synovial sarcoma of unusual histology or location has been reviewed elsewhere (Ladanyi and Bridge, 2000). Molecular variants of the SYT-SSX1 and SYT-SSX2 fusion transcripts are rare and none have been recurrent (reviewed in dos Santos et al., 2001). The presence of the SYT-SSX fusion in both the spindle cells and epithelial components of biphasic tumors has been demonstrated by a variety of methods (Hiraga et al., 1998; Birdsall et al., 1999; Kasai et al., 2000). As expected, the formation of SYT-SSX1 and SYT-SSX2 is mutually exclusive and the fusion type is concordant in primaries and metastases and constant over the course of the disease.

SYT-SSX fusion proteins: functional aspects

SYT and the SSX proteins lack recognizable DNA-binding domains and are presumably transcriptional regulators whose actions are mediated primarily through protein-protein interactions. The SYT-SSX chimeric proteins are likely to be involved in the transcriptional deregulation of specific target genes. The genes normally regulated by SYT or SSX proteins through putative direct or indirect protein-protein interactions with specific transcription factors are presently unknown.

SYT domain structure

In the SYT protein, sequence analysis has identified two distinct domains. On the basis of phylogenetic conservation, Thaete et al. (1999) defined a novel 54 amino acid N-terminal domain (designated the SNH domain). On the basis of the similarity of its amino acid composition to corresponding domains in other transcriptional regulators (such as EWS and TLS), it appeared that a C-terminal domain rich in glutamine, proline, glycine and tyrosine (designated the QPGY domain), might function as a transcriptional activation domain (Brett et al., 1997). Indeed, in standard transcriptional assays, this portion of SYT, mapped to residues 158-387, when fused to the GAL4 DNA-binding domain functions as a transactivation domain (Brett et al., 1997). Deletion of the SNH domain enhances transcriptional activation by SYT mutants, suggesting that this domain acts as an inhibitor of the QPGY activation domain. Thaete et al. (1999) have also demonstrated the interaction in vitro of the SYT (and SYT-SSX) proteins with the SNF protein BRM. Thus, SYT (and SYT-SSX) may interact directly with BRM, a component of the SWI/SNF complex involved in chromatin remodeling. The latter is a multiprotein complex which counteracts repression by chromatin structural proteins such as histones and the polycomb-group of proteins. The SNH domain is dispensable for the BRM interaction, but the rest of the protein, including the portion between the SNH and QPGY domains, appears necessary for the full interaction and colocalization of SYT (or SYT-SSX) and BRM (Thaete et al., 1999). The SYT-binding domain within BRM maps to its amino terminal half.

SSX domain structure

Sequence analysis and functional studies have identified two major domains in SSX. The N-terminal regions of both SSX1 and SSX2 show similarity to a transcription repression domain, the Kruppel-associated box (KRAB) (Witzgall et al., 1994; Margolin et al., 1994). However, the SSX KRAB-like domain is an inefficient or even inactive repressor domain and a stronger novel repressor domain (designated SSX-RD) has been functionally mapped to the C-terminal of both SSX1 and SSX2 (Thaete et al., 1999; Lim et al., 1998). The SSX-RD domain is the most highly conserved portion of the protein among SSX family members, further supporting its critical function. The SSX-RD is also required for the nuclear co-localization with Polycomb group proteins, which function to repress transcription through modification of higher order chromatin structure (dos Santos et al., 2000). Taken together, these data suggest that transcriptional repression by SSX is effected at least in part through association with or recruitment of Polycomb group repressors by the SSX-RD domain.

Domain structure of SYT-SSX

These recent data on the domain structure of SYT and SSX proteins do not lend themselves to any simple models for the oncogenicity of SYT-SSX. The chimeric transcript in synovial sarcoma replaces the 5' portion of SSX1 and SSX2 with all but the eight c-terminal amino acids of SYT. The loss of these eight terminal amino acids of SYT appears to have a negligible effect on transactivation by the SYT QPGY domain. The C-terminal SSX-RD domain present in SYT-SSX contributes a transcriptional repressor domain to the protein. Thus, the fusion protein has transcriptional activating and repressing domains, an observation which complicates models of SYT-SSX-mediated transcriptional deregulation. In transactivation assays, SYT-SSX has a several-fold lower activity than native SYT (Brett et al., 1997). Unlike chimeric transcription factors presumed to function mainly at appropriate enhancers recognized by the intact DNA-binding domain derived from one of the translocation partners, the functional compartment within the nucleus to which SYT-SSX is targeted is more difficult to define and yet will form a key element in understanding its aberrant function. This adds another level of complexity to the oncogenic scenario in synovial sarcoma.

Nuclear distribution of SYT, SSX, and SYT-SSX proteins

Nuclear localization studies with deletion mutants have found that the nuclear targeting domain of SSX coincides with its C-terminal repressor domain (SSX-RD), whereas in SYT, a domain within amino acids 51-90 distinct from canonical nuclear localization signals is required for nuclear entry. Thus, both nuclear targeting signals are included in the fusion protein. Nuclear localization of SYT-SSX has also been confirmed in synovial sarcoma tissues, using specific antibodies (dos Santos et al., 1997; Hashimoto et al., 2000). Both native proteins and the fusion proteins show nucleolar exclusion. The nuclear distribution of the SYT and SSX proteins appears largely determined by their respective interactions with BRM and polycomb-group proteins. The SYT and SSX proteins have speckled and diffuse nuclear patterns of distribution, respectively (Brett et al., 1997). Specifically, SYT localizes to distinct nuclear bodies, separate from PML oncogenic domains and several other previously described nuclear bodies (dos Santos et al., 2000). Thaete et al. (1999) have demonstrated that this nuclear distribution represents co-localization of SYT with BRM. SSX shows a diffuse nuclear distribution with some punctate staining (dos Santos et al., 2000). SSX co-localizes with Polycomb group repressor proteins (e.g. HPC2, RING1, HPH1, ENX, EED, and BMI1) (Soulez et al., 1999; dos Santos et al., 2000), but no direct interaction has so far been identified in co-immunoprecipitation and yeast two-hybrid assays between SSX or SYT-SSX and selected Polycomb group proteins such as RING1 and BMI1 (Soulez et al., 1999). It initially appeared that SYT-SSX proteins followed a speckled distribution similar to native SYT (Brett et al., 1997; dos Santos et al., 1997). However, recent data suggest a more complex picture: in at least some cell types, the nuclear distribution of SYT-SSX seems to follow that of the punctate or non-diffuse component of the staining observed for native SSX proteins, which parallels that of Polycomb group repressor proteins (Soulez et al., 1999). Indeed, it has been suggested that some of the discrepant findings in these nuclear localization studies are due to marked differences in the distribution of Polycomb group repressors (and by association, of SSX proteins) between different cell lines (Soulez et al., 1999).

Models of SYT-SSX function

Thus, models of SYT-SSX oncogenesis need to consider which interaction is 'dominant', BRM vs Polycomb group, in terms of subnuclear targeting and which transcriptional effect is dominant, transactivation by the SYT QPGY domain or repression by the SSX-RD. Dos Santos et al. (2000) provide evidence that the targeting signal contained in the SSX RD is dominant over that in the SYT amino terminal portion, and suggest that Polycomb-repression is 'dominant' over BRM. Thus, the prinicipal subnuclear compartment affected by SYT-SSX may be that normally targeted by native SSX proteins in germ cells. Therefore, one hypothesis suggested by these considerations is that the t(X;18) leads to aberrant and constitutive re-expression of the SSX-RD within the SYT-SSX fusion protein, resulting in the Polycomb-mediated repression by the SSX-RD of genes recognized by as yet uncharacterized protein-protein interactions between SSX and putative DNA-binding transcription factors. In the somatic precursor cells of synovial sarcoma, presumably lacking expression of native SSX, these unknown target genes would not be normally repressed. This scenario is complicated by the observation that at least some synovial sarcomas show expression of the remaining unrearranged SSX gene or of other SSX family members, detectable by non-nested RT-PCR. Tureci et al. (1998) found expression of the normal SSX2 gene in two synovial sarcomas with SYT-SSX1 gene fusions. We studied the expression of SSX1 and SSX2 in 30 synovial sarcomas (from 15 men and 15 women) and found expression of SSX1 or SSX2 in eight tumors, six from female patients and two from male patients (M Ladanyi, I Fligman, unpublished data). In the tumors from male patients, the other SSX family gene was expressed, whereas tumors from female patients expressed either the other SSX family gene, or the remaining normal allele of the SSX gene involved in the translocation (as identified by appropriate PCR primers). The expression of normal SSX genes in synovial sarcoma may reflect neoplasia-associated de-repression, as expected for cancer-testis antigens, or may represent residual expression due to low level expression in the precursor cells of synovial sarcoma. Thus, in at least some cases of synovial sarcoma, SYT-SSX could compete directly with native SSX, if the above model of SYT-SSX pathogenesis is correct.

Alternative models of SYT-SSX function, which may not be mutually exclusive, have recently been proposed by Dos Santos et al. (2001). In another model, SYT-SSX may redirect the 'dominant' SSX-RD domain to repress genes recognized and normally activated through putative protein-protein interactions between SYT and DNA-binding transcription factors. It is also possible that the formation of SYT-SSX is associated with the acquisition of novel DNA or protein target specificities, as described for other chimeric transcriptional regulators (Lee et al., 1997; Bertolotti et al., 1998; Petermann et al., 1998). There are so far no reported mouse models transgenic for SYT-SSX1 or SYT-SSX2.

Regarding the significance of the C-terminal SSX domain in vivo, a notable cell line has recently been described, HS-SY-3 (Sonobe et al., 1999). The SYT-SSX1 fusion product in this cell line (and the original tumor material) is truncated at its 3' end, and thereby the protein lacks the entire SSX-RD. Unlike other synovial sarcoma cell lines, HS-SY-3 grows slowly in vitro and is non-tumorigenic in nude mice (Sonobe et al., 1999). The primary tumor was a 10 cm popliteal monophasic synovial sarcoma in an 86 year old woman, an unusually advanced age for this tumor. Assuming that SSX-RD was absent from the time of formation of the gene fusion in this unusual tumor, it is tempting to hypothesize that the SSX-RD is dispensable for synovial sarcoma tumorigenesis, although it may contribute to tumor aggressiveness. This cell line brings into question, albeit with inherent caveats, the model of SYT-SSX oncogenesis which depends on a functional SSX-RD for subnuclear targeting and overall transcriptional effect.

Interaction between SYT and p300

Recently, a cellular role independent of transcriptional activation has been described for SYT, and by extension, for SYT-SSX (Eid et al., 2000). Eid et al. (2000) report a direct interaction of SYT with the nuclear protein p300, implicating SYT in the control of cell adhesion in a transcription activation-independent manner, at least in part by allowing appropriate beta1 integrin/fibronectin receptor function. Interestingly, SYT deletion mutants lacking the eight C-terminal amino acids bound p300 but inhibited cell adhesion, presumably through a dominant negative mechanism. The inability of these SYT mutants to facilitate adhesion may be due to the loss of the SH2 binding motif within the deleted C terminal end. In preliminary assays, SYT-SSX was found to be comparable to this SYT deletion mutant in terms of dominant negative effects on cell adhesion, suggesting that this may another oncogenic effect of SYT-SSX, mediated through the disruption of p300/SYT-dependent maintenance of adhesion responses. Current data on domains of SYT and SSX proteins defined by sequence similarities or functional studies are schematized in Figure 1.

Clinical and phenotypic correlates of SYT-SSX fusion type in synovial sarcoma

Studies in Ewing's sarcoma and alveolar rhabdomyosarcoma have established the concept that alternative forms of the specific fusion products seen in sarcomas with chromosomal translocations may have an impact on one or more features of the associated cancer, such as age at diagnosis, primary site, proliferative rate, or clinical outcome (Zoubek et al., 1996; Kelly et al., 1997; de Alava et al., 1998). These different associations further support the notion that the fusion proteins encoded by these chromosomal translocations are central to the biology of these sarcomas such that the heterogeneity in fusion protein function associated with even minor structural differences can have a detectable and direct impact on either the cellular, histologic, or clinical level (Lin et al., 1999).

Correlation with clinical behavior

This paradigm is also reinforced by the analysis of clinical-molecular correlates in synovial sarcoma. A pilot study of this question showed that, among 39 patients with localized synovial sarcoma, cases with the SYT-SSX2 fusion had a significantly better metastasis-free survival than cases with SYT-SSX1 (Kawai et al., 1998). The finding of a survival advantage for SYT-SSX2 was also reported two subsequent studies with patient groups of similar or smaller size (Nilsson et al., 1999; Inagaki et al., 1999). To confirm these correlations, we recently revisited this question in a retrospective multi-institutional series of 243 patients with synovial sarcoma (Ladanyi et al., submitted). In this more definitive analysis, SYT-SSX2 fusion type is confirmed as a significant positive prognostic factor for overall survival. It appears to exert part of its impact on prognosis prior to presentation, through an association with a lower prevalence of metastatic disease at diagnosis. In patients with localized disease at diagnosis, SYT-SSX fusion type appears to be the single most significant prognostic factor by multivariate analysis.

Correlation with glandular epithelial differentiation

Whether the exact location of the Xp11 breakpoint is related to the morphology of synovial sarcoma was a controversial issue prior to the cloning of the t(X;18) breakpoints. FISH results from three separate groups suggested a relationship between OATL1 (SSX1) vs OATL2 (SSX2) breakpoints and histological subtypes of synovial sarcoma (De Leeuw et al., 1994; Janz et al., 1995; Renwick et al., 1995). However, other results failed to confirm these findings (Shipley et al., 1994, 1996; Crew et al., 1995). The initial molecular analysis of this issue found a statistically significant relationship between histological subtype (monophasic vs biphasic) and SSX1 or SSX2 involvement in the fusion transcript, that could be summarized as follows: tumors containing an SYT-SSX1 fusion transcript were monophasic or biphasic, whereas all (or almost all) tumors with the SYT-SSX2 fusion were monophasic (Kawai et al., 1998). This association of SYT-SSX fusion type and biphasic histology (as defined by the presence of glandular epithelial differentiation with lumen formation) has since been upheld in several independent studies (Nilsson et al., 1999; Antonescu et al., 2000; Mancuso et al., 2000). This is also supported by the results from additional studies, recently tabulated by Dos Santos et al. (2001). In any given individual analysis, the strength of this association with fusion type is obviously affected by the statistical power of a given study group and by differences in diagnostic criteria for biphasic histology. Nonetheless, occasional biphasic synovial sarcomas with the SYT-SSX2 fusion product do exist, but gland formation in these tumors may be less extensive than in SYT-SSX1-positive tumors. The reproducible and statistically significant nature of the association between t(X;18) fusion type and biphasic histology points to a genuine biological phenomenon whose further elucidation may provide insights into the basic histogenetic process of mesenchymal to epithelial conversion and architectural epithelial differentiation (gland formation). Hypothetically, the SYT-SSX1 transcriptional regulator may be more 'permissive' than SYT-SSX2 for this differentiation program, or conversely, may be more repressive of the normal mesenchymal differentiation program of synovial sarcoma precursor cells. In this context, it is interesting to note that epithelial differentiation may arguably represent a 'default' cellular program, reflecting the absence or insufficient activity of transcription factors specific for the hematopoietic or mesenchymal lineages (Frisch, 1997).

Correlation with proliferative rate

Two studies including in aggregate 52 cases, found a statistically signficant relation between SYT-SSX fusion type and tumor cell proliferative activity, with the SYT-SSX1 tumors showing evidence of increased proliferation, as determined by quantitation of immunostaining for Ki-67 (Nilsson et al., 1999; Inagaki et al., 1999). However, in a similar study of 73 cases, we could not confirm this finding (Antonescu et al., 2000). Instead, a comparison of proliferative rate incorporating fusion type, histologic type (monophasic vs biphasic), and sample source (primary vs metastatic) revealed that the main determinants of proliferation rate in this study were the latter two factors. Differences in study results may reflect limitations of the technique used or differences in statistical approaches (i.e. proliferative rate analysed as a continuous variable vs a categorical variable). Further studies using more precise methods of assessing this parameter (e.g. flow cytometry) could clarify this issue.

Possible functional basis for correlates of fusion type

A direct comparison of the transforming or transactivating capabilities of the two different SYT-SSX fusion proteins has not been performed. The C-terminal 78 amino acids of SSX proteins included in SYT-SSX differ at 13 residues between SSX1 and SSX2. These differences are non-conservative, but 12 of the divergent residues lie outside of the C-terminal SSX repressor domain (SSX-RD), arguing against significant differences between the two SSX proteins in terms of their putative interactions with the Polycomb group repressors. The role of the so-called 'divergent domain' of the SSX proteins remains presently unknown (Figure 1).

Expression profiling of synovial sarcoma

High throughput studies of gene expression patterns in synovial sarcoma could begin to address several key issues in the pathobiology of synovial sarcoma, including cell lineage, control of glandular morphogenesis, possible transcriptional targets of SYT-SSX proteins, differential gene regulation by SYT-SSX1 and SYT-SSX2, as well as the establishment of 'diagnostic' expression profiles. Expression profiling studies of synovial sarcoma using cDNA or oligonucleotide microarrays are in progress in several laboratories, and are already yielding interesting leads in some areas (Nielsen et al., 2001; Allander et al., 2001).

For instance, synovial sarcoma, as a monoclonal process recapitulating mesenchymal to epithelial conversion, presents us with naturally occurring setting in which to examine this process at the level of gene expression. The strong association of the type of SYT-SSX transcriptional regulator and morphologic epithelial differentiation (biphasic histology) described above suggests that putative functional differences between SYT-SSX1 and SYT-SSX2 may result in differences in the expression of direct or indirect target genes involved in this process. Yet, there are few data on the biological basis for glandular epithelial differentiation in this tumor. The MET receptor tyrosine kinase, implicated in mesenchymal to epithelial conversion in several in vivo settings, is co-expressed with its ligand, HGF, in the epithelial cells of synovial sarcoma (Kuhnen et al., 1998; Oda et al., 2000). This suggests a possible autocrine basis for this process, analogous to that observed in some in vitro systems (Tsarfaty et al., 1994). Recent cDNA microarray analyses of synovial sarcomas have identified additional genes apparently relevant to this process including Erb-B2/Her2/neu (Allander et al., 2001).

Her2/neu is highly expressed in synovial sarcoma compared to other sarcomas, such as malignant fibrous histiocytoma (Allander et al., 2001). By immunohistochemical analysis, we find it expressed in the epithelial component of biphasic tumors and in solid epithelioid areas of monophasic tumors. FISH analysis demonstrates that this over-expression is not due to gene amplification, and no correlation between SYT-SSX fusion type and Her2/neu expression can be identified so far (Illei et al., 2001). The inter-relationship of the HGF/MET and Erb-B2/Her2/neu signaling systems is potentially interesting. In breast, MET required for tubulogenesis, whereas Her2 required for lobuloalveolar differentiation (Yang et al., 1995). HGF and MET (HGF receptor) can induce the heregulin/Her2 pathway (Castagnino et al., 2000). Aside from its fundamental interest, this finding also provides a biologic basis for evaluating HerceptinTM as a potential therapeutic agent in biphasic synovial sarcoma.

Note added in proof

Nagai et al. (2001) recently examined the transforming activity of SYT-SSX1, its binding to BRM, and its effect on gene expression in a small scale microarray experiment. They found the N-terminal 181 amino acids of SYT-SSX1 to be necessary for its transforming activity. The same portion of SYT-SSX1 was found to interact directly with BRM in both transfected heterologous cells and in a synovial sarcoma cell line. Along with the data of Thaete et al. (1999), this would narrow the core BRM interaction domain of SYT-SSX1 to residues 73 to 181. By co-transfection of competitive BRM deletion constructs with SYT-SSX1, they showed that the interaction with BRM is a major factor in transformation by SYT-SSX1. Finally, using a 1176 gene microarray to analyse gene expression in a cell line stably transfected with an inducible SYT-SSX1 plasmid, they identified DCC as a putative target of transcriptional repression by SYT-SSX1.

Acknowledgements

ML is supported by grant PO1 CA47179 from the National Institute of Health.

References

Allander SV, Chen Y, Illei PB, Antonescu CR, Ladanyi M, Meltzer PS. (2001). Proc. Am. Assoc. Cancer Res. 42, 430.

Antonescu CR, Kawai A, Leung DH, Lonardo F, Woodruff JM, Healey JH, Ladanyi M. (2000). Diagn. Mol. Pathol. 9, 1-8. MEDLINE

Barr FG. (1998). Nat. Genet. 19, 121-124. Article MEDLINE

Bertolotti A, Melot T, Acker J, Vigneron M, Delattre O, Tora L. (1998). Mol. Cell Biol. 18, 1489-1497. MEDLINE

Birdsall S, Osin P, Lu YJ, Fisher C, Shipley J. (1999). Int. J. Cancer 82, 605-608. MEDLINE

Bouffler S, Silver A, Cox R. (1993). Bioessays 15, 409-412. MEDLINE

Brett D, Whitehouse S, Per A, Shipley J, Cooper C, Goodwin G. (1997). Hum. Mol. Genet. 6, 1559-1564. MEDLINE

Castagnino P, Lorenzi MV, Yeh J, Breckenridge D, Sakata H, Munz B, Werner S, Bottaro DP. (2000). Oncogene 19, 640-648. MEDLINE

Chand A, Clark J, Cooper CS, Craig IW. (1995). Genomics 30, 545-552. MEDLINE

Chang CH, Scott GK, Kuo WL, Xiong X, Suzdaltseva Y, Park JW, Sayre P, Erny K, Collins C, Gray JW, Benz CC. (1997). Oncogene 14, 1617-1622. MEDLINE

Clark J, Rocques PJ, Crew AJ, Gill S, Shipley J, Chan AML, Gusterson BA, Cooper CS. (1994). Nature Genet. 7, 502-508. MEDLINE

Crew AJ, Clark J, Fisher C, Gill S, Grimer R, Chand A, Shipley J, Gusterson BA, Cooper CS. (1995). EMBO J. 14, 2333-2340. MEDLINE

de Alava E, Kawai A, Healey JH, Fligman I, Meyers P, Huvos AG, Gerald WL, Jhanwar SC, Argani P, Antonescu CR, Pardo-Mindan FJ, Ginsberg J, Womer R, Lawlor ER, Wunder J, Andrulis I, Sorensen PHB, Barr FG, Ladanyi M. (1998). J. Clin. Oncol. 16, 1248-1255. MEDLINE

de Bruijn DR, Baats E, Zechner U, De Leeuw B, Balemans M, Olde Weghuis D, Hirning-Folz U, Geurts van Kessel A. (1996). Oncogene 13, 643-648. MEDLINE

De Leeuw B, Balemans M, Geurts van Kessel A. (1996). Cytogenet. Cell Genet. 73, 179-183. MEDLINE

De Leeuw B, Balemans M, Olde Weghuis D, Geurts van Kessel A. (1995). Hum. Mol. Genet. 4, 1097-1099. MEDLINE

De Leeuw B, Suijkerbuijk RF, Olde Weghuis D, Meloni AM, Stenman G, Kindblom LG, Balemans M, van den Berg E, Molenaar WM, Sandberg AA, Geurts van Kessel A. (1994). Cancer Genet. Cytogenet. 73, 89-94. MEDLINE

dos Santos NR, de Bruijn DR, Kater-Baats E, Otte AP, Geurts van Kessel A. (2000). Exp. Cell Res. 256, 192-202. MEDLINE

dos Santos NR, de Bruijn DR, Guerts van Kessel A. (2001). Genes Chromosom. Cancer 30, 1-14. MEDLINE

dos Santos NR, de Bruijn DRH, Balemans M, Janssen B, Gartner F, Lopes JM, De Leeuw B, van Kessel AG. (1997). Hum. Mol. Genet. 6, 1549-1558. MEDLINE

Eid JE, Kung AL, Scully R, Livingston DM. (2000). Cell 102, 839-848. MEDLINE

Frisch SM. (1997). Bioessays 19, 705-709. MEDLINE

Geurts van Kessel AG, dos Santos NR, Simons A, de Bruijn D, Forus A, Fodstad O, Myklebost O, Balemans M, Baats E, Weghuis DO, Suijkerbuijk RF, van den Berg E, Molenaar WM, De Leeuw B. (1997). Cancer Genet. Cytogenet. 95, 67-73. MEDLINE

Guillou L, Coindre J, Gallagher G, Terrier P, Gebhard S, de Saint A, Michels J, Jundt G, Vince DR, Collin F, Trassard M, Le Doussal V, Benhattar J. (2001). Hum. Pathol. 32, 105-112. MEDLINE

Gure AO, Tureci O, Sahin U, Tsang S, Scanlan MJ, Jager E, Knuth A, Pfreundschuh M, Old LJ, Chen YT. (1997). Int. J. Cancer 72, 965-971. Article MEDLINE

Hashimoto N, Araki N, Yoshikawa H, Myoui A, Matsumine A, Kaneko M, Sonobe H, Ochi T. (2000). Cancer Lett. 149, 31-36. MEDLINE

Hiraga H, Nojima T, Abe S, Sawa H, Yamashiro K, Yamawaki S, Kaneda K, Nagashima K. (1998). Diag. Mol. Pathol. 7, 102-110.

Illei PB, Allander SV, Chen Y, Meltzer PS, Antonescu CR, Ladanyi M. (2001). Mod. Pathol. 14, 13A.

Inagaki H, Murase T, Otsuka T, Eimoto T. (1999). Am. J. Clin. Pathol. 111, 528-533. MEDLINE

Janz M, De Leeuw B, Olde Weghuis D, Werner M, Nolte M, Geurts van Kessel A, Nordheim A, Hipskind RA. (1995). J. Pathol. 175, 391-396. MEDLINE

Kas K, Finger E, Grall F, Gu X, Akbarali Y, Boltax J, Weiss A, Oettgen P, Kapeller R, Libermann TA. (2000). J. Biol. Chem. 275, 2986-2998. MEDLINE

Kasai T, Shimajiri S, Hashimoto H. (2000). Mol. Pathol. 53, 107-110. MEDLINE

Kawai A, Woodruff J, Healey JH, Brennan MF, Antonescu CR, Ladanyi M. (1998). New Engl. J. Med. 338, 153-160. MEDLINE

Kelly KM, Womer RB, Sorensen PHB, Xiong Q-B, Barr FG. (1997). J. Clin. Oncol. 15, 1831-1836. MEDLINE

Kuhnen C, Tolnay E, Steinau HU, Voss B, Muller KM. (1998). Virchows Arch. 432, 337-342. MEDLINE

Ladanyi M, Bridge JA. (2000). Hum. Pathol. 31, 532-538. MEDLINE

Lee SB, Kolquist KA, Nichols K, Engler C, Shyamala M, Ladanyi M, Gerald WL, Haber DA. (1997). Nature Genet. 17, 309-313. MEDLINE

Lewis JJ, Antonescu CR, Leung DH, Blumberg D, Healey JH, Woodruff JM, Brennan MF. (2000). J. Clin. Oncol. 18, 2087-2094. MEDLINE

Lim FL, Soulez M, Koczan D, Thiesen HJ, Knight JC. (1998). Oncogene 17, 2013-2018. MEDLINE

Lin PP, Brody RI, Hamelin A, Bradner JE, Healey JH, Ladanyi M. (1999). Cancer Res. 59, 1428-1432. MEDLINE

Mancuso T, Mezzelani A, Riva C, Fabbri A, Dal Bo L, Sampietro G, Perego P, Casali P, Zunino F, Sozzi G, Pierotti MA, Pilotti S. (2000). Lab. Invest. 80, 805-813. MEDLINE

Margolin JF, Friedman JR, Meyer WK, Vissing H, Thiesen HJ, Rauscher FJ. (1994). Proc. Natl. Acad. Sci. USA 91, 4509-4513. MEDLINE

Miller AP, Gustashaw K, Wolff DJ, Rider SH, Monaco AP, Eble B, Schlessinger D, Gorski JL, van Ommen GJ, Weissenbach J, Willard HF. (1995). Hum. Mol. Genet. 4, 731-739. MEDLINE

Nagai M, Tanaka S, Tsuda M, Endo S, Kato H, Sonobe H, Minami A, Hiraga H, Nishihara H, Sawa H, NAgashima K. (2001). Proc. Natl. Acad. Sci. U.S.A. 98, 3843-3848. MEDLINE

Nielsen TO, Brown PO, Botstein D, van de Rijn M. (2001). Mod. Pathol. 14, 16A.

Nilsson G, Skytting B, Xie Y, Brodin B, Perfekt R, Mandahl N, Lundeberg J, Uhlen M, Larsson O. (1999). Cancer Res. 59, 3180-3184. MEDLINE

Oda Y, Sakamoto A, Saito T, Kinukawa N, Iwamoto Y, Tsuneyoshi M. (2000). Hum. Pathol. 31, 185-192. MEDLINE

Oettgen P, Alani RM, Barcinski MA, Brown L, Akbarali Y, Boltax J, Kunsch C, Munger K, Libermann TA. (1997). Mol. Cell Biol. 17, 4419-4433. MEDLINE

Petermann R, Mossier B, Aryee D, Khazak V, Golemis E, Kovar H. (1998). Oncogene 17, 603-610. MEDLINE

Renwick PJ, Reeves BR, Dal Cin P, Fletcher CD, Kempski H, Sciot R, Kazmierczak B, Jani K, Sonobe H, Knight JC. (1995). Cytogenet. Cell Genet. 70, 58-63. MEDLINE

Shipley J, Crew J, Birdsall S, Gill S, Clark J, Fisher C, Kelsey A, Nojima T, Sonobe H, Cooper C, Gusterson B. (1996). Am. J. Pathol. 148, 559-567. MEDLINE

Shipley JM, Clark J, Crew AJ, Birdsall S, Rocques PJ, Gill S, Chelly J, Monaco AP, Abe S, Gusterson BA, Cooper CS. (1994). Oncogene 9, 1447-1453. MEDLINE

Skytting B, Nilsson G, Brodin B, Xie YT, Lundeberg J, Uhlen M, Larsson O. (1999). J. Natl. Cancer Inst. 91, 974-975. MEDLINE

Sonobe H, Takeuchi T, Liag SB, Taguchi T, Yuri K, Shimizu K, Iwata J, Furihata M, Ohtsuki Y, Testa JR. (1999). Int. J. Cancer 82, 459-464. MEDLINE

Soulez M, Saurin AJ, Freemont PS, Knight JC. (1999). Oncogene 18, 2739-2746. MEDLINE

Thaete C, Brett D, Monaghan P, Whitehouse S, Rennie G, Rayner E, Cooper CS, Goodwin G. (1999). Hum. Mol. Genet. 8, 585-591. MEDLINE

Tsarfaty I, Rong S, Resau JH, Rulong S, da Silva PP, Vande Woude GF. (1994). Science 263, 98-101. MEDLINE

Tureci O, Chen YT, Sahin U, Gure AO, Zwick C, Villena C, Tsang S, Seitz G, Old LJ, Pfreundschuh M. (1998). Int. J. Cancer 77, 19-23. MEDLINE

Tureci O, Sahin U, Schobert I, Koslowski M, Scmitt H, Schild HJ, Stenner F, Seitz G, Rammensee HG, Pfreundschuh M. (1996). Cancer Res. 56, 4766-4772. MEDLINE

van de Rijn M, Barr FG, Collins MH, Xiong Q-B, Fisher C. (1999). Am. J. Clin. Pathol. 112, 43-49. MEDLINE

Witzgall R, O'Leary E, Leaf A, Onaldi D, Bonventre JV. (1994). Proc. Natl. Acad. Sci. USA 91, 4514-4518. MEDLINE

Yang Y, Spitzer E, Meyer D, Sachs M, Niemann C, Hartmann G, Weidner KM, Birchmeier C, Birchmeier W. (1995). J. Cell Biol. 131, 215-226. MEDLINE

Zoubek A, Dockhorn-Dworniczak B, Delattre O, Christiansen H, Niggli F, Gatterer-Menz I, Smith TL, Jurgens H, Gadner H, Kovar H. (1996). J. Clin. Oncol. 14, 1245-1251. MEDLINE

Figures

Figure 1 Schematic diagram of domain structure of the SYT, SSX, and SYT-SSX proteins. The amino acid residues representing the boundaries of selected domains are indicated. The scale is approximate. Abbreviations: SNH: SYT amino terminal domain; QPGY: SYT glutamine-, proline-, glycine- and tyrosine-rich domain; KRAB: Krüppel-associated box; DD: SSX divergent domain; SSXRD: SSX repressor domain

10 September 2001, Volume 20, Number 40, Pages 5755-5762
Table of contents    Previous  Article  Next    [PDF]