Hybrid splicing minigene and antisense oligonucleotides as efficient tools to determine functional protein/RNA interactions

Abstract

Alternative splicing is a complex process that provides a high diversity of proteins from a limited number of protein-coding genes. It is governed by multiple regulatory factors, including RNA-binding proteins (RBPs), that bind to specific RNA sequences embedded in a specific structure. The ability to predict RNA-binding regions recognized by RBPs using whole-transcriptome approaches can deliver a multitude of data, including false-positive hits. Therefore, validation of the global results is indispensable. Here, we report the development of an efficient and rapid approach based on a modular hybrid minigene combined with antisense oligonucleotides to enable verification of functional RBP-binding sites within intronic and exonic sequences of regulated pre-mRNA. This approach also provides valuable information regarding the regulatory properties of pre-mRNA, including the RNA secondary structure context. We also show that the developed approach can be used to effectively identify or better characterize the inhibitory properties of potential therapeutic agents for myotonic dystrophy, which is caused by sequestration of specific RBPs, known as muscleblind-like proteins, by mutated RNA with expanded CUG repeats.

Figure 1

Alternative splicing of Atp2a1 ex22 is MBNL dependent and efficiently distorted upon AON targeting of MBNL-binding regions. (a) A scheme illustrating the alternative splicing pattern of Atp2a1 ex22 in the presence (green) or absence (red) of MBNLs. Black and white boxes represent constitutive and alternative exons, respectively. The RNA fragment, which was analyzed in vitro, comprising MBNL-binding regions #1 and #2 is marked with the gray frame. (b) Percentage of ex22 inclusion in ATP2A1 endogenous mRNA upon MBNL1,2 silencing in human cells. PSI, percent spliced in index, providing the inclusion level of an alternative exon; n = 3. (c) An experimentally determined secondary structure of Atp2a1-RNA with the YGCY motifs marked with green and in bold if present in humans and mice. Cleavages induced by RNase T2 and nuclease S1 are shown. The AONs bound in vitro to MBNL-binding regions #1 and #2 are accordingly indicated by blue (LNA#1), orange (LNA#2) and red lines (2′OMe/2′OMe-PS). AON, antisense oligonucleotide; 2′OMe-PS, phosphorothioated 2′-O-methyl; LNA, locked nucleic acid. The cleavage sites and intensities of the selected probes are shown using symbols explained in the inset. (d) Quantification of FBA showing the blocking properties of AONs in vitro. On the right, the dissociation constant (Kd) of MBNL1/Atp2a1-RNA complexes without (-AON) or with different AON applications; n = 4. (e) Quantification of FBA showing reduced MBNL1 affinity to Atp2a1-RNA with point mutations, especially YGCYs (marked with red). Regions #1 and #2, significant for MBNL binding, are also indicated; n = 4. (f) MBNL1 dose-dependent percentage of alternative ex22 inclusion in mRNA from the Atp2a1WT minigene in HeLa cells transfected with 200 or 500 ng of the MBNL1 expression vector per well; n = 3. (g) Percentage of alternative ex22 inclusion in Atp2a1WT mRNA upon MBNL1 overexpression and treatment with different AONs at 25–100 nM (2′OMePS) or 100 nM (others). The obtained results were compared with those of the control experiment with MBNL1 overexpression and no AON treatment (green bar); n = 3.

Figure 2

AONs indicate functional MBNL-binding regions within introns but not exons. (a) Schematic representation of the alternative splicing patterns of MBNL-dependent exons from Atp2a1, Pphln1, NASP, Nfix, Ldb3 and Mbnl1 pre-mRNAs, which depend on the localization of the MBNL-binding regions. MBNL1-specific CLIP-seq clusters are shown as green areas for each transcript. The splicing pattern for alternative exons spliced in (exON) upon MBNL binding within a downstream intron is depicted in dark blue, whereas the splicing pattern for alternative exons spliced out (exOFF) upon MBNL binding within an upstream intron or exon is depicted in burgundy. Constitutive and alternative exons are shown as black and white boxes, respectively. Intronic and exonic RNA fragments containing MBNL-binding regions targeted by AONs are marked with grey boxes. (b) RT-PCR analysis showing the percentage of alternative exon inclusion in the tested mRNAs in human cells after silencing of MBNLs and in cells treated with control siRNA (siCtrl). The results for MBNL1 are presented in; n = 3. (c) In vitro MBNL1 binding to RNA fragments derived from the studied transcripts based on FBA assays showing the blocking properties of 2′OMe-PS-AONs and DNA-AONs marked as red and blue curves, respectively; n = 2. (d) RT-PCR analysis showing changes in alternative splicing upon treatment with gene specific AON (AON-Target) or control AON (AON-Ctrl) in human or mouse cell lines; n = 3.

Figure 3

Hybrid Atp2a1 minigenes confirm the functional intronic and exonic MBNL-binding regions. (a) Percentage of alternative ex22 inclusion in Atp2a1WT and Atp2a1Δ mRNA, lacking a fragment containing MBNL-binding regions, upon MBNL1 overexpression with or without 2′OMe-PS (AON) treatment; n = 3. Atp2a1Δ-Ctrl, a hybrid minigene with an inserted control sequence cassette lacking MBNL-binding regions; n = 2. (b) The schemes illustrate a general organization of hybrid Atp2a1 minigenes containing an MBNL-responsive cassette derived from analyzed transcripts in the place of a natural Atp2a1 cassette within intron 22. We expect to observe the promotion of ex22 through MBNLs binding to YGCY sequence motifs within inserted cassettes (green line) and ex22 exclusion upon AONs blocking MBNL-binding regions (red line). A thermodynamically stable structure having 14-nt-long complementary sequences derived from intron 22, restriction sites and 5-bp artificial helix, is marked with black. It is distant by 32–164 nucleotides from the MBNL-binding sites. (c) Experimentally determined secondary structures of selected areas of analyzed short intronic and exonic RNA fragments containing significant MBNL-binding regions complementary to specific AONs and marked with a red line. The entire structures of the analyzed in vitro RNA fragments are presented in Supplementary Fig. S4b. The number of not shown nucleotides is depicted on the 5′ and/or 3′ end of each structure. YGCY motifs are marked with green and in bold if present in humans and mice. Mbnl1-RNA was previously described. (d) Percentage of alternative ex22 inclusion in the mRNA of a set of hybrid Atp2a1 minigenes upon MBNL1 overexpression with or without treatment with specific 2′OMe-PS AONs. The asterisk indicates an artificial splicing isoform of the Atp2a1Δ-Nfix minigene; n = 3.

Figure 4

Potential crosstalk between SRSF1 and MBNLs. (a) Experimentally determined secondary structures of Mbnl1-RNA and Nfix-RNA with in silico-predicted SRSF1 binding regions marked with blue boxes^,. 2′OMe-PS AONs blocking MBNL-binding regions are marked with a red line. Point mutations are marked with red. (b) Results of EMSA showing the interaction between SRSF1 at the indicated concentrations and 5′-³²P-labeled Mbnl1-RNA (upper panel) or Nfix-RNA (lower panel) with either intact MBNL-binding regions or the regions blocked by individual AONs. The calculated Kd values are indicated below each electrophoretogram; n = 4. On the right is the quantification of EMSA results based on the decline of the free RNA signal in favor of forming SRSF1/RNA complexes. Antagonistic role of SRSF1 and MBNL1 in the alternative splicing regulation of (c) ex1 of Mbnl1 and (d) ex7 of Nfix. Note the SRSF1 dose-dependent promotion of both exon inclusion into the mRNA of wild-type Mbnl1 or Nfix (left panels) or Mbnl1Mut or NfixMut minigenes (right panels). The amount of MBNL1 or SRSF1 expressing vector constituted 500 ng or ranged from 250 to 1000 ng per well, respectively; n = 2 (Mbnl1), n = 3 (Nfix).

Figure 5

The hybrid minigene is useful to screen potential RNA-binding therapeutics. (a) A scheme illustrating MBNL-dependent ex22 inclusion to mRNA derived from the Atp2a1Δ-(CUG) ₁₇ minigene by extensive MBNL1 binding to the hairpin structure formed by (CUG)₁₇. Constitutive and alternative exons are shown as black and white boxes, respectively. (b–c) Top; Schemes illustrating the inhibition of MBNL-dependent ex22 inclusion to mRNA derived from the Atp2a1Δ-(CUG) ₁₇ minigene by antisense oligonucleotide LNA-PS-CAG-8 (left panel) and erythromycin (EM) by their binding to the hairpin structure formed by (CUG) ₁₇ repeats. Bottom; reduction of alternative ex22 inclusion to Atp2a1Δ-(CUG) ₁₇ mRNA in cells treated with (b) LNA-PS-CAG-8 or (c) EM. The results are normalized to the splicing response of ex22 in Atp2a1Δ-(CUG) ₁₇ pre-mRNA without (0%) or with (100%) MBNL1 overexpression (Ctrl, green bar). The LNA-PS-CAG-8 and EM were used in a range of 25–125 nM or 50–500 µM concentrations, respectively; n = 2.