Antisense oligonucleotide-induced alternative splicing of the APOB mRNA generates a novel isoform of APOB

Abstract

Background: Apolipoprotein B (APOB) is an integral part of the LDL, VLDL, IDL, Lp(a) and chylomicron lipoprotein particles. The APOB pre-mRNA consists of 29 constitutively-spliced exons. APOB exists as two natural isoforms: the full-length APOB100 isoform, assembled into LDL, VLDL, IDL and Lp(a) and secreted by the liver in humans; and the C-terminally truncated APOB48, assembled into chylomicrons and secreted by the intestine in humans. Down-regulation of APOB100 is a potential therapy to lower circulating LDL and cholesterol levels.

Results: We investigated the ability of 2'O-methyl RNA antisense oligonucleotides (ASOs) to induce the skipping of exon 27 in endogenous APOB mRNA in HepG2 cells. These ASOs are directed towards the 5' and 3' splice-sites of exon 27, the branch-point sequence (BPS) of intron 26-27 and several predicted exonic splicing enhancers within exon 27. ASOs targeting either the 5' or 3' splice-site, in combination with the BPS, are the most effective. The splicing of other alternatively spliced genes are not influenced by these ASOs, suggesting that the effects seen are not due to non-specific changes in alternative splicing. The skip 27 mRNA is translated into a truncated isoform, APOB87SKIP27.

Conclusion: The induction of APOB87SKIP27 expression in vivo should lead to decreased LDL and cholesterol levels, by analogy to patients with hypobetalipoproteinemia. As intestinal APOB mRNA editing and APOB48 expression rely on sequences within exon 26, exon 27 skipping should not affect APOB48 expression unlike other methods of down-regulating APOB100 expression which also down-regulate APOB48.

Figure 1

Design of skip 27 ASOs. The sequences of the intronic regions flanking exon 27 and parts of exon 27 are shown at the top – intronic sequences are in small letters, exonic sequences are in caps, exon 27 is enclosed in the grey-shaded box. The splice-sites (BPS = branch-point sequence; 3'ss = 3' splice-site; 5'ss = 5' splice-site) are outlined with dotted lines. The Shapiro and Senapathy (S&S) scores below each splice-site reflect the degree of similarity of these sequences to the splice-site consensus [9]. The ASO design is indicated below. Arrows run in the 5' to 3' direction. Three classes of ASO tested are distinguished within each box: single splice site ASOs that target individual splice sites; combination splice-site ASOs that target two splice-sites simultaneously; combination splice-site ASOs that target two splice-sites and are linked to an hnRNP A1 binding tail (A1 – the sequence of this tail is presented) or a mutated tail that is unable to bind hnRNP A1 (mut A1).

Figure 2

Alternative splicing of exon 27 induced by APOB exon 27 splice-site ASOs. (a) HepG2 cells were transfected with the indicated ASOs at 250 nM for 48 hours. RT-PCR was carried out on the total RNA extracted from these cells with oligonucleotides annealing to the adjacent exons 26 and 28. The -RT control PCR was performed without reverse transcriptase. The positions of the bands corresponding to the APOB mRNA species with exon 27 (inc 27) and without exon 27 (skip 27) are indicated on the right side. Fragment lengths in bp of pBR322-MspI markers are indicated on the left side. The band corresponding to skip 27 mRNA was cloned and sequenced to confirm that exon 26 is spliced to exon 28. (b) The graph shows the quantitative data from three independent replications of the experiment (error bars show the S.E.M.). Overall one-way repeated measures ANOVA p-value was < 0.0001, indicating statistically significant differences between groups. Dunnett's multiple comparison test was used to compare the no-oligo control versus ASO-transfected cells: statistically significant differences were found with skip 27–53, skip 27-5B, and skip 27-3B.

Figure 3

Dose titration of selected APOB skip 27 ASOs. (a) HepG2 cells were transfected with the indicated ASOs at increasing concentrations from 25 to 250 nM for 48 hours. RT-PCR was carried out on the total RNA extracted from these cells with oligonucleotides annealing to the adjacent exons 26 and 28. The -RT control PCR was performed without reverse transcriptase. The positions of the bands corresponding to the APOB mRNA species with exon 27 (inc 27) and without exon 27 (skip 27) are indicated on the right side. Fragment lengths in bp of pBR322-MspI markers are indicated on the left side. (b) The graph shows the quantitative data from three independent replications of the experiment (error bars show the S.E.M.). The data was analyzed using a repeated measures two-way ANOVA, using Bonferroni's test to compare exon 27 skipping between skip 27-3B and the other ASOs at 250 nM: statistically significant differences are seen with skip 27–53, skip 27-3B scram and skip 27-3B A1. No statistically significant difference is seen with skip 27-5B and skip 27-3B mutA1.

Figure 4

Effect of ASOs targeting APOB exon 27 sequences. (a) ESEfinder output for APOB exon 27, using version 2.0 matrices and thresholds [21]. Y-axis shows ESEfinder above-threshold scores for identified matches; X-axis corresponds to the exon 27 sequence. Red = SF2/ASF, blue = SC35, green = SRp40, yellow = SRp55. The light grey horizontal lines indicate the matches that were found by RESCUE-ESE [22]. The dark grey lines indicate the ESE candidates identified by PESXs server [23, 24]. ASOs were designed to bind to locations corresponding to ESE clusters, and these locations are represented below the X-axis by the arrows, which run from the 5' to 3' direction. (b) RT-PCR assay for exon 27 skipping with RNA from cells transfected with ASOs at 250 nM. Exonic ASOs 6–25, 31–50, 61–80, 88–107 weakly promote exon 27 skipping, whereas scrambled versions of these ASOs had no effect on exon 27 skipping. (c) Quantitative data from three independent experiments. Error bars show S.E.M. Overall repeated measures one-way ANOVA showed p < 0.0001, indicating significant differences between compared groups. Dunnett's multiple comparison test showed a significant difference between skip 27-3B and 3B scram (p < 0.01), but no statistically significant differences were seen between skip 27-3B scram and other skip 27 ASOs.

Figure 5

APOB skip 27 ASOs have no effect on the alternative splicing of other exons. (a) The indicated ASOs were transfected at 250 nM into HepG2 cells, and total RNA from these was analyzed by RT-PCR for inclusion/skipping of APOB exon 27; GAPDH, as loading control; inclusion/skipping of insulin receptor (INSR) exon 11; inclusion/skipping of TSC2 exon 25; and inclusion/skipping of Caspase-9 exons 3–6. The identity of each band is listed below each set of 6 lanes from top band to bottom band: APOB inc 27 = exon 27 included; APOB skip 27 = exon 27 skipped, GAPDH = amplified region of GAPDH mRNA 7–261 nt; INSR B = exon 11 included, INSR A = exon 11 skipped; TSC2 +25 = exon 25 included, TSC2 -25 = exon 25 skipped; Caspase-9 isoform 9 = exons 3–6 included, Caspase-9 isoform 9b = exons 3–6 skipped. Transfection of ASOs does not affect the alternative splicing of INSR, TSC2 and Caspase-9. (b) Quantitative data for alternative splicing of each gene from three independent replicates. Error bars represent S.E.M. One-way repeated measures ANOVA of the APOB skip 27 dataset showed p = 0.0007, representing significant statistical differences between groups: Dunnett's multiple comparison test showed statistically significant differences between the no-oligo control and skip 27–53, 5B and 3B, as indicated. One-way ANOVA p-values for the INSR, TSC2 and Caspase-9 datasets were 0.3716, 0.5887, and 0.1674 respectively, indicating no statistically significant differences between groups.

Figure 6

Transfection of APOB skip 27 ASOs causes expression of a short isoform of APOB, APOB87_SKIP27. HepG2 cells were transfected with the indicated ASOs at 250 nM, then pulse-labelled with Trans-³⁵S-Label for 1 hour in the presence of oleic acid-BSA, which stimulates APOB100 expression [62]. Cell lysates or cell culture media were immunoprecipitated with a polyclonal antibody for APOB. The immunoprecipitates were subjected to SDS-PAGE (4% Tris-glycine), followed by autoradiography. The gel shown is representative of three experiments. A short isoform of APOB, APOB87_SKIP27, is seen with the skip 27-3B ASO. This short isoform is not seen with the 3B scram ASO, nor with the no-oligo control. This matches the presence and absence of skip 27 mRNA seen with these ASOs.