The alternatively spliced products of LMNA, lamin C and prelamin A (the precursor to lamin A), are produced in similar amounts in most tissues and have largely redundant functions. This redundancy suggests that diseases, such as Hutchinson-Gilford progeria syndrome (HGPS), that are caused by prelamin A–specific mutations could be treated by shifting the output of LMNA more toward lamin C. Here, we investigated mechanisms that regulate LMNA mRNA alternative splicing and assessed the feasibility of reducing prelamin A expression in vivo. We identified an exon 11 antisense oligonucleotide (ASO) that increased lamin C production at the expense of prelamin A when transfected into mouse and human fibroblasts. The same ASO also reduced the expression of progerin, the mutant prelamin A protein in HGPS, in fibroblasts derived from patients with HGPS. Mechanistic studies revealed that the exon 11 sequences contain binding sites for serine/arginine-rich splicing factor 2 (SRSF2), and SRSF2 knockdown lowered lamin A production in cells and in murine tissues. Moreover, administration of the exon 11 ASO reduced lamin A expression in wild-type mice and progerin expression in an HGPS mouse model. Together, these studies identify ASO-mediated reduction of prelamin A as a potential strategy to treat prelamin A–specific diseases.
John M. Lee, Chika Nobumori, Yiping Tu, Catherine Choi, Shao H. Yang, Hea-Jin Jung, Timothy A. Vickers, Frank Rigo, C. Frank Bennett, Stephen G. Young, Loren G. Fong
(A) siRNA knockdown of SRSF2 increases lamin C levels in human cells. Wild-type human fibroblasts were transfected with siRNAs against SRSF1, SRSF2, and SRSF6, and lamin A and C levels measured by Western blotting. Actin levels were measured as a loading control. Transcript levels for SRSF1, SRSF2, and SRSF6 were reduced by ≥80%. The mean ± SEM for 3 independent experiments are shown. *P < 0.05; **P < 0.01, t test. (B) SRSF2 binds to exon 11 RNA sequences. RNA molecules corresponding to nt 22–63 of exon 11 LMNA were biotinylated and bound to streptavidin beads. The binding of HA-tagged SRSF2 was assessed by Western blotting. Binding of SRSF2 to the wild-type RNA was compared with that to a scrambled RNA molecule, with the addition of a 30-fold excess of unlabeled RNA, and RNA molecules in which SRSF2 sites were mutated. (C) Schematic diagram of a LMNA reporter construct. A LMNA fragment (spanning exons 10–12) was cloned behind a glo fragment (colored red) driven by an RSV promoter (black arrowhead). The sequences of the 3 predicted SRSF2 sites and a single predicted SRSF6 (SRp55) site in exon 11 are highlighted. The red asterisk identifies the location of the HGPS mutation (c.1824; C>T). The reporter yields 3 transcripts (glo-prelamin A, glo-lamin C, and glo-progerin). (D) Reporter studies showing that SRSF2 sites 2 and 3 are important for LMNA prelamin A splicing. HeLa cells were cotransfected with the LMNA reporter and a “glo-only” plasmid. After 2 days, glo-prelamin A, glo-progerin, and glo-lamin C transcripts were quantified by qRT-PCR and normalized to levels of the glo-only transcript. Changes in expression, relative to the wild-type reporter (set at a value of 0), for 5 independent experiments (mean ± SEM) are shown.