Molecular studies of genetic changes in myxoid and round cell liposarcoma

Abstract: Chromosomal translocations commonly result in the production of fusion genes and the fusion genes are often tumor-type specific. In myxoid and round cell liposarcomas (MLS/RCLS), almost 95% of the cases carry a t(12;16)(q13;p11). In the remaining 5% of the MLS/RCLS tumors, another translocation and fusion gene can be found, i.e. the t(12;22)(q13;q12). These translocations fuse CHOP on chromosome 12 to either TLS on chromosome 16 or EWSR1 on chromosome 22. The result is a TLS-CHOP or EWS-CHOP fusion protein that is believed to be a key player in the development of MLS/RCLS. Normally, CHOP plays a fundamental role in adipogenesis by forming heterodimers with members of the C/EBP family of transcription factors. TLS-CHOP is capable of dimerizing with the same proteins as CHOP, but the outcome is radically different. This thesis focuses on TLS-CHOP and the hypothesis that TLS-CHOP may function as an aberrant transcription factor. As such TLS-CHOP is believed to regulate the expression of a cascade of downstream genes that contribute to the phenotype and malignancy of MLS/RCLS. We have found that the TLS-CHOP fusion protein localizes to distinct nuclear structures which differs largely from the normal TLS and CHOP patterns. The nuclear structures do not overlap with PML nuclear bodies and furthermore, PML nuclear bodies are dislocated in TLS-CHOP-expressing cells. The abnormal localization of the PML tumor suppressor may be an important part of the transforming process in these tumors. Using a subtraction hybridization technique, we have found that several genes are aberrantly expressed in MLS/RCLS compared to normal adipose tissue and may have a function in the development of these tumors. This work resulted in the identification of a novel gene which seems to be associated with a low degree of differentiation. At present the normal function of this protein, LEPREL1, is not known. However, its different domains and intracellular distribution points in the direction that this protein functions as an enzyme in the ER and Golgi pathway. One of the genes identified as over-expressed in MLS/RCLS was the RET proto-oncogene. Further studies showed that the tumor cells produce the activating ligand persephin and that the receptor is phosphorylated. This indicates that it is activated and capable of signaling proliferation and anti-apoptosis into the cell nucleus. Thus, there seems to be an autocrine growth stimulatory loop running in MLS/RCLS tumors and disrupting the RET signaling pathway may be an attractive target for therapy in MLS/RCLS.