The TLS/FUS gene is involved in a recurrent chromosomal translocation in human myxoid liposarcomas. We previously reported that TLS is a potential splicing regulator able to modulate the 5-splice site selection in an E1A pre-mRNA. Using an in vitro selection procedure, we investigated whether TLS exhibits a specificity with regard to RNA recognition. The RNAs selected by TLS share a common GGUG motif. Mutation of a G or U residue within this motif abolishes the interaction of TLS with the selected RNAs. We showed that TLS can bind GGUG-containing RNAs with a 250 nM affinity. By UV cross-linking/competition and immunoprecipitation experiments, we demonstrated that TLS recognizes a GGUG-containing RNA in nuclear extracts. Each one of the RNA binding domains (the three RGG boxes and the RNA recognition motif) contributes to the specificity of the TLS⅐RNA interaction, whereas only RRM and RGG2-3 participate to the E1A alternative splicing in vivo. The specificity of the TLS⅐RNA interaction was also observed using as natural pre-mRNA, the G-rich IVSB7 intron of the ␤-tropomyosin pre-mRNA. Moreover, we determined that RNA binding specificities of TLS and high nuclear ribonucleoprotein A1 were different. Hence, our results help define the role of the specific interaction of TLS with RNA during the splicing process of a pre-mRNA. TLS (Translocated in LipoSarcoma)1 or FUS has been first characterized as a rearranged gene in chromosomal translocations specific of human myxoid liposarcoma (1, 2). The resulting fusion protein contains the N-terminal part of TLS fused to a transcription factor of the CAAT/enhancer-binding protein family of proteins: CHOP. In an acute myeloid leukemia, TLS is also involved in a chromosomal breakpoint that juxtaposes the same N-terminal region of TLS to a transcription factor of the ETS proteins family: ERG-1 (3, 4). TLS is highly similar to EWS, a gene implicated in chromosomal translocations that are specific of the tumors of the Ewing family (5, 6). In most of these sarcoma, the N-terminal region of EWS is fused to the DNA binding domain of either ERG-1 or FLI-1, which are two closely related ETS proteins.Both TLS and EWS have in common a similar structural organization. Their C terminus part contains multiple domains that are involved in RNA⅐protein interactions: an RNA recognition motif (RRM) flanked by two regions rich in Arg-Gly-Gly repeats (RGG domains) and a C 2 C 2 zinc finger. In their N terminus domain, they contain a glutamine-, serine-, and tyrosine-rich region that functions as a transcriptional activation domain when fused to a heterologous DNA binding domain (7-9). In oncogenic chimera, the adjunction of this N-terminal region of TLS or EWS to the transcriptional regulators CHOP, FLI-1, or ERG-1 generates proteins with transcriptional activities that differ from those of the wild-type counterpart (8,10). From these data it has been proposed that the oncogenic fusion proteins disturb the expression of genes that are regulated by CHOP, FLI-1, or ERG-1. However, no cellular targe...

Computer analysis of human intron sequences have revealed a 50 nucleotide (nt) GC-rich region downstream of the 5' splice site; the trinucleotide GGG occurs almost four times as frequently as it would in a random sequence. The 5' part of a beta-tropomyosin intron exhibits six repetitions of the motif (A/U)GGG. In order to test whether these motifs play a role in the splicing process we have mutated some or all of them. Mutated RNAs show a lower in vitro splicing efficiency when compared with the wild-type, especially when all six motifs are mutated (> 70% inhibition). Assembly of the spliceosome complex B and, to a lesser extent, of the pre-spliceosome complex A also appears to be strongly affected by this mutation. A 55 kDa protein within HeLa cell nuclear extract is efficiently cross-linked to the G-rich region. This protein is present in the splicing complexes and its cross-linking to the pre-mRNA requires the presence of one or several snRNP. Altogether our results suggest that the G-rich sequences present in the 5' part of introns may act as an enhancer of the splicing reaction at the level of spliceosome assembly.

The chicken beta‐tropomyosin gene contains 11 exons, two of which are spliced into mRNA only in skeletal muscle. One pair of alternative exons, 6A and 6B, is found in the middle of the gene; they are spliced in a mutually exclusive manner. The non‐muscle splice 6A‐7 is by far the predominant in vitro reaction in a HeLa cell nuclear extract. A minor product is the 6A‐6B splice, which is excluded in all tissues. This minor product results from the use of a branch point located 105 nt upstream of the 3′ end of the intron separating exons 6A and 6B. The region between the branch point sequence and the final AG contains a stretch of approximately 80 pyrimidines. We have examined the role of the distance of the branchpoint to the 3′ splice site and of the sequences between these two elements. Our results suggest that at least two cis‐acting elements contribute to the mutual exclusivity of exons 6A and 6B. The intron between exons 6A and 6B is intrinsically poorly ‘spliceable’ both because the branch point is too far upstream of the 3′ end of the intron to give efficient splicing and because of the particular sequence lying between this branch point and the 3′ splice site.