An extended inhibitory context causes skipping of exon 7 of SMN2 in spinal muscular atrophy Natalia N. Singh, a Elliot J. Androphy, a,b and Ravindra N. Singh a,b, * a Department of Medicine, University of Massachusetts Medical School, Worcester, MA 01605-2324, USA b Program in Molecular Genetics and Microbiology, University of Massachusetts Medical School, Worcester, MA 01605-2324, USA Received 31 December 2003 Abstract SMN1 and SMN2 represent the two nearly identical copies of the survival of motor neuron gene in humans. The most frequent cause of spinal muscular atrophy (SMA) is loss of SMN1 accompanied by the inability of SMN2 to compensate due to an inhibitory mutation at position 6 in exon 7 (C6U) that causes exon 7 exclusion. How this single exonic nucleotide regulates exon 7 recognition has been of major interest. Based on score matrices and in vitro assays, abrogation of an exonic splicing enhancer (ESE) associated with SF2/ASF has been considered as the cause of exon 7 exclusion. However, a recent report supports the creation of an exonic splicing silencer (ESS) associated with hnRNP A1 as the determining factor for exon 7 exclusion. Here we show that C6U strengthens an inhibitory context that covers a larger sequence than the hnRNP A1 binding site. The inhibitory context can also be strengthened by the addition of a G residue at the first position of exon 7 in SMN1, promoting exon 7 skipping despite the presence of SF2/ASF binding site. Through in vivo selection and a series of mutations we demonstrate that the strengthening of the extended inhibitory context at the 5 0 end of exon 7 is exercised through overlapping sequence motifs that collaborate to regulate exon usage. Ó 2004 Elsevier Inc. All rights reserved. Keywords: Spinal muscular atrophy; Alternative pre-mRNA splicing; SMN genes; Exon 7; Exonic splicing enhancer A fundamental problem in pre-mRNA splicing is ‘exon recognition,’ the process by which exons are dis- tinguished from introns and intron–exon boundaries are accurately defined [1]. In addition to the general splice- osomal machinery, many non-spliceosomal proteins participate in pre-mRNA splicing [2,3]. These include serine–arginine-rich proteins (SR proteins), SR-like proteins, and heterogeneous nuclear ribonucleoproteins (hnRNPs) [1,4–7]. Some of these proteins bind to pre- mRNA sequences called exonic splicing enhancers (ESEs) and silencers (ESSs). Enhancers and silencers promote or suppress splice-site (ss) selection, respec- tively. Two different algorithms have been used to pre- dict ESEs based on small sequence motifs [8,9]. In fact, the most powerful method predicts ESEs with no more than 6 nucleotides, based on the assumption that RNA binding proteins typically interact with 6-nucleotide long motifs [8]. Thus, longer ESEs and/or RNA struc- ture-associated ESEs are not predictable. Furthermore, none of the available algorithms predict the degree of involvement of a given ESE where a series of adjacent or overlapping cis-elements are present. Functional analy- sis of sequences remains the ultimate proof of the presence of cis-elements within an exon [9]. Spinal muscular atrophy (SMA) is the second most common autosomal recessive disorder characterized by the loss of motor neurons in the anterior horn of the spinal cord. In humans, two nearly identical survival of motor neuron genes (SMN1 and SMN2) exist on chro- mosome 5q13. Deletions or mutations within SMN1 but not SMN2 cause all forms of SMA [10]. SMN1 and SMN2 differ by a critical C to T substitution (6U in transcript of SMN2) at position 6 of exon 7 (Fig. 1A). This single nucleotide substitution determines exon 7 usage [11,12]. Exon 7 is known to have a weak 3 0 ss [13], probably due to its suboptimal polypyrimidine tract. An improved polypyrimidine tract promoted SMN2 exon 7 inclusion, however, an AG-rich ESE associated with the SR-like protein Tra2 was still required (Fig. 1A) [14,15]. * Corresponding author. Fax: 1-508-856-6797. E-mail address: Ravindra.Singh@umassmed.edu (R.N. Singh). 0006-291X/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2004.01.067 Biochemical and Biophysical Research Communications 315 (2004) 381–388 BBRC www.elsevier.com/locate/ybbrc