“…-745; data not shown, Senapathy et al+, 1990) or in terms of the number of base-pairs formed with U1 snRNA (IVS2-705 . -654 ϭ -745; Table 2)+ Since the variable part of the internal exons does not seem to contain typical splicing enhancers or silencers (Fig+ 1B), it seems likely that it is the length of the aberrant exon that modulates its recognition+ The effects of exon length on splicing efficiency and exon inclusion both in vitro and in vivo have been recognized previously (Reed & Maniatis, 1986;Furdon & Kole, 1988;Dominski & Kole, 1991;Sterner & Berget, 1993)+ Nevertheless, the fact that aberrant exons IVS2-654 and IVS2-705, which in vivo are both fully included into the spliced product, differ in their respective sensitivity to the antisense oligonucleotides at the 59 splice site by 29-fold was surprising+ Note that a trace of correctly spliced product is detectable in untreated IVS2-705 but not in IVS2-654 HeLa cells (cf+ lanes 1 in Figs+ 3 and 2, respectively)+ Thus the efficiency of exon inclusion and its sensitivity to oligonucleotides are in qualitative if not quantitative agreement+ A similar correlation was observed in vitro in cell-free extracts (Dominski & Kole, 1993, 1994a) Accordingly, the differences in sensitivity to the oligonucleotides are even more striking for IVS2-654 and -745 exons, the latter being partly included in the aberrant b-globin mRNA in the absence of the oligonucleotide (Fig+ 4, lane 1 and Table 2)+ The fact that consensus mutations at the 59 splice sites in IVS2-654 and IVS2-705 mutants affect sensitivity of the 39 splice site to ON-39cr oligonucleotide are consistent with the concept of exon bridging (Kuo et al+, 1991;Hertel et al+, 1997) inherent in the exon definition model (Robberson et al+, 1990)+ In this interpretation, the ability of ON-39cr oligonucleotide to inhibit aberrant splicing of IVS2-705con but not of IVS2-654con premRNA points to differences in bridging and recognition of the two exons flanked by the same 39 splice site upstream and almost identical consensus 59 splice site downstream (Table 2)+ Whether the kinetic or steric hindrance models are responsible for the differences in the ability of the oligonucleotides to correct splicing of mutant pre-mRNAs, it appears that the oligonucleotides compete with the splicing factors for the target sequences and that the interactions of the latter with pre-mRNAs vary significantly+ Although the experiments did not identify the competing factors, they did indicate that antisense oligonucleotides may provide useful tools for modifying and probing the interactions of the spliceosomes with the pre-mRNAs+…”