To analyze the function of SPF45, HeLa nuclear extracts were immunodepleted of this factor and their splicing activity was analyzed. We conclude that interference with SPF45 expression compromises the ability of SXL to promote exon 3 skipping. The effects observed were not due to changes in the expression of SXL, because neither dsRNA-mediated interference nor overexpression of SPF45 changed the levels of SXL protein ( Figure 3B). The nucleotide sequence of human and Drosophila SPF45 is sufficiently different so that Drosophila SPF45 dsRNA does not interfere with human SPF45 expression. The effect could be specifically attributed to SPF45 expression because efficient regulation by SXL was restored by overexpression of human SPF45 in these cells (lanes 9–12). For example, while transfection with 0.004 μg/ml of SXL-encoding plasmid resulted in quantitative skipping of exon 3 in cells cotransfected with control dsRNA (lane 2), transfection with 20-fold higher concentrations of SXL-encoding plasmid resulted in less than 50% exon skipping in the presence of SPF45 dsRNA (lane 8). While cotransfection with limited amounts of SXL-encoding plasmid induced complete skipping of exon 3 in cells that had been treated with control dsRNA ( Figure 3A, lanes 1–4), SXL-induced exon skipping was reduced in cells treated with SPF45 dsRNA (lanes 5–8). A reporter minigene containing Sxl exons 2–4 was cotransfected with increasing concentrations of a SXL-encoding plasmid or with empty vector.
The levels of SPF45 mRNA were reduced at least 20-fold (data not shown). Schneider cells were transfected with dsRNA corresponding to a Drosophila SPF45 cDNA obtained from an embryo library, or transfected with various control dsRNAs. To test this prediction, we used double-stranded RNA (dsRNA) to interfere with endogenous SPF45 function in Drosophila cells. If this recognition involves SPF45, interference with SPF45 activity should compromise SXL function. ), suggesting that recognition of the proximal site facilitates Sxl autoregulation. This suggests that the distal 3′ ss belongs to the category of AG-dependent 3′ ss, which are reliant upon interaction between U2AF 35 and the downstream 3′ ss AG to stabilize U2AF 65 binding to a Py-tract ( Mutation of the distal 3′ ss AG reduced U2AF 65 crosslinking (lane 3), while mutation of the proximal AG did not (lane 2). This suggests that U2AF 65 interacts with the Py-tract immediately upstream from the distal 3′ ss. U2AF 65 was crosslinked to the wild-type RNA but not to a mutant RNA in which a stretch of eight uridines at the Py-tract was substituted by cytidines ( Figure 2A, compare lanes 1 and 4). To test whether U2AF 65 was associated with the Py-tract of A3S RNA, the transcript was uniformly labeled with uridine, incubated with nuclear extract under splicing conditions, and irradiated with UV light, and after treatment with RNase A, immunoprecipitation was carried out using anti-U2AF 65 antibodies. To identify factors associated with the proximal and distal splice sites, ultraviolet (UV) light-mediated crosslinking assays were carried out.