A recurrent fusion gene in high-grade endometrial stromal sarcoma : a new tool for diagnosis and therapy ?

Chromosomal translocations that create gene fusions are a leading cause of malignant transformation in hemato­ logical malignancies and sarcomas [1]. Although they are less common in solid tumors, fusion genes are now also recognized as driving events in such tumors ­ for example, in prostate and thyroid cancers. Most of the cancer­causing translocations were detected by chromo­ some banding techniques [2], and they provided reward­ ing entry points for the positional cloning of oncogenes. Combined array comparative genomic hybridi zation and gene expression studies, as well as whole genome or exome sequence analysis, have now been added to the methods available to screen tumor genomes for recurrent aberrations, gene fusions or aberrant transcripts. The prototypic example of a cancer­causing chromo­ somal translocation is the Philadelphia chromosome, first identified in patients with chronic myelogenous leukemia. The Philadelphia chromosome, generated through a reci­ procal translocation between chromosomes 9 and 22 [3], results in a fusion transcript of the genes BCR and ABL1, which encodes a protein with increased tyrosine kinase activity. Patients with this translocation can be effectively treated with tyrosine kinase inhibitors, such as imatinib, which specifically targets the oncogenic fusion protein [4]. Endometrial stromal sarcoma (ESS) belongs to the group of endometrial stromal tumors, which are a category of relatively rare mesenchymal neoplasms of the uterus. The diagnostic spectrum of endometrial stromal tumors ranges from endometrial stromal nodule (ESN) through low­grade ESS (LESS) to high­grade (undifferentiated) ESS. ESNs are benign, well­circumscribed lesions, but LESS and ESS both show malignant behavior. The distinction between LESS and ESS is clinically important because the prognosis of these lesions is drastically different. LESSs are indolent tumors with a low propensity for mainly local recurrences many years after surgical treatment. In contrast, ESSs are highly malignant and often develop extra­uterine metastases. The 5­year survival rate for LESS is close to 100%, compared with 55% for ESS [5]. The histological distinc tion is based on tumor cell necrosis, mitotic index and atypical cell morphology. However, both tumor types can show unusual features, resulting in difficulties for definitive tumor classification. Surgical removal of the fallopian tubes and ovaries on both sides remains the mainstay for the treatment of early stage disease and for LESS, whereas adjuvant or neo­adjuvant chemoradio therapy is used for advanced stages and for ESS. However, a definitive benefit of chemoradiotherapy has yet to be established, highlight­ ing the importance of targeted therapies [6].

The prototypic example of a cancercausing chromo somal translocation is the Philadelphia chromosome, first identified in patients with chronic myelogenous leukemia. The Philadelphia chromosome, generated through a reci procal translocation between chromosomes 9 and 22 [3], results in a fusion transcript of the genes BCR and ABL1, which encodes a protein with increased tyrosine kinase activity. Patients with this translocation can be effectively treated with tyrosine kinase inhibitors, such as imatinib, which specifically targets the oncogenic fusion protein [4].
Endometrial stromal sarcoma (ESS) belongs to the group of endometrial stromal tumors, which are a category of relatively rare mesenchymal neoplasms of the uterus. The diagnostic spectrum of endometrial stromal tumors ranges from endometrial stromal nodule (ESN) through lowgrade ESS (LESS) to highgrade (undifferentiated) ESS. ESNs are benign, wellcircumscribed lesions, but LESS and ESS both show malignant behavior.
The distinction between LESS and ESS is clinically important because the prognosis of these lesions is drastically different. LESSs are indolent tumors with a low propensity for mainly local recurrences many years after surgical treatment. In contrast, ESSs are highly malignant and often develop extrauterine metastases. The 5year survival rate for LESS is close to 100%, compared with 55% for ESS [5]. The histological distinc tion is based on tumor cell necrosis, mitotic index and atypical cell morphology. However, both tumor types can show unusual features, resulting in difficulties for definitive tumor classification. Surgical removal of the fallopian tubes and ovaries on both sides remains the mainstay for the treatment of early stage disease and for LESS, whereas adjuvant or neoadjuvant chemoradio therapy is used for advanced stages and for ESS. However, a definitive benefit of chemoradiotherapy has yet to be established, highlight ing the importance of targeted therapies [6].

A novel mechanism in high-grade endometrial stromal sarcoma
In a recently published study in the Proceedings of the National Academy of Sciences USA, Lee and colleagues

Abstract
High-grade endometrial stromal sarcomas (ESSs) are an aggressive group of endometrial stromal tumors. A recent study described a recurrent chromosomal translocation (t(10;17)) occurring in ESS, which joins the gene 14-3-3ε with either FAM22A or FAM22B. Expression of the resulting fusion gene leads to malignant transformation, and silencing of its expression reverses the malignant phenotype. Because the fusion can be readily detected in diagnostic samples using fluorescent in situ hybridization, this chromosomal aberration could be used to differentiate high-grade ESS from the low-grade, less aggressive form. Discovery of the new oncoprotein could also provide entry points for targeted therapies.

© 2010 BioMed Central Ltd
A recurrent fusion gene in high-grade endometrial stromal sarcoma: a new tool for diagnosis and therapy? [7] report a novel chromosomal aberration in this particu larly aggressive form of uterine sarcoma. The chromo somal rearrangement identified is a translocation t(10;17), which fuses two genes: 14-3-3ε, which encodes a member of the 1433 family (on chromosome 17), and either FAM22A or FAM22B, both encoded on chromosome 10. The 1433 family of proteins comprises seven members with numerous functions, including signal transduction, cytoskeletal configuration, metabo lism, differentiation, survival and transcription [8]. Some of the members of this family are involved in tumori genesis, either as suppressors or as oncogenes [9]. Very little is known about FAM22 proteins, and an involve ment in malignant transformation has not been described before.
Lee et al. [7] performed a series of experiments to verify and characterize the chromosomal rearrangement. Wholetranscriptome sequencing showed that the fusion transcript was formed through joining of exon 5 of the 1433ε gene with exon 2 of either of the FAM22 genes. Fluorescent in situ hybridization analysis showed that the translocation is specific for the aggressive form of the uterine sarcomas and was not found in other mesen chymal tumors. The expression of the fusion protein was demonstrated by westernblot analysis in all trans locationpositive tumors. Subsequently, the oncogenic properties of the fusion protein were explored by silencing the expression using RNA interference in a cell line derived from one of the tumors. This manipulation reduced the viability of the cell line, and overexpression of the fusion construct increased the migratory capacity of fibroblasts.
The authors [7] then sought to elucidate the functional properties of the fusion protein. Wildtype 1433ε is localized mainly in the cytoplasm, where it dimerizes. As a consequence of the translocation, abnormal nuclear localization of 1433ε protein was observed. This locali zation is possibly triggered by a nuclear localization sequence encoded by exon 7 of both FAM22A and FAM22B. The aberrant localization of a functional 1433 protein to the nucleus most likely explains the trans forming properties of the fusion protein.
It was already assumed that the inactivation of promi nent tumor suppressors, including TP53, could explain the differences in tumor aggressiveness in ESS [10]. This view is now supported by comparing the gene expression of the translocationpositive highergrade tumors with their lowgrade counterparts. It became evident that genes involved in proliferation, such as CCND1 (encoding cyclin D1) and CEBPA (encoding CCAAT/enhancer binding protein alpha), in addition to genes encoding factors that promote invasion, such as tissue inhibitor of metalloproteinase 1 (TIMP1) and fascin 1 (FSCN1), are more strongly expressed in the translocation positive highgrade tumors. CCND1, in particular, is a notorious oncoprotein that promotes cell cycle progression through cyclindependentkinaseindependent mechanisms.

Clinical application in endometrial stromal sarcoma
The immediate benefit of this study for the clinical management of patients with ESS is that the detection of the fusion protein greatly aids in diagnosing highgrade variants that may benefit from adjuvant chemo radio therapy. Ideally, the identification of aberrant 1433 signal ing could be developed as a molecular therapeutic target, with the hope that evidencebased therapies will improve the outcome of patients with this aggressive cancer. Additional studies aimed at dissecting the precise role of the aberrant nuclear interaction that results from this translocation will undoubtedly identify downstream targets, which might also offer new therapeutic oppor tunities. The recent report by O'Meara and colleagues [11] that describes a 1433FAM22 fusion in a clear cell carcinoma of the kidney highlights the relevance of these particular cancerinducing genes.
In summary, the paper by Lee et al. [7] presents an elegant example of a topdown strategy, where the identification of a specific and recurrent chromosomal translocation leads to the identification of a novel mecha nism of deregulation of 1433 family members, the identification of the genetic basis of a subgroup of tumors, the development of a novel, specific tool for differential diagnosis, and hopefully an entry point for targeted therapeutic strategies.