Figure 1.
R3hdm2 reacts with anti-NeuN antibody.
A. The depicted GST-fusion proteins were expressed in bacteria and cell lysates analyzed by western. The upper western panel (anti-GST) confirms that proteins of the expected sizes were expressed after IPTG induction. ABS = antibody binding site; FL = full-length R3hdm2; the R3H domain is also shown. The membrane was then stripped and reprobed with anti-NeuN antibody (lower panel), which indicates that the defined ABS region is both necessary and sufficient for anti-NeuN recognition of R3Hdm2. B. Northern of poly(A)+ RNA isolated from adult mouse tissues and probed for R3hdm2. The level of R3hdm2 expression was normalised to β-actin (lower panel) for each tissue and calculated relative to the expression level in thymus.
Figure 2.
Immunoprecipitation and mass spectrometry identify Rbfox3 as the target of anti-NeuN.
A. Silver stain of proteins immunoprecipitated from P20 mouse brain crude nuclear extract using anti-NeuN, and separated by non-reducing SDS-PAGE. No antibody, ands mouse IgG were used as IP controls; the no lysate sample was used to check the integrity of anti-NeuN cross-linking (covalent coupling) to the beads. Arrows point to the bands analyzed by mass spectrometry. B. Western confirms the specificity of the anti-NeuN immunoprecipitation. The “classic” NeuN doublet species below 50 kDa were efficiently immunoprecipitated, and depleted from the supernatant (supe), while the upper immunoreactive bands were not. C. Alignment of the mouse Fox protein family. Accession numbers of the sequences shown are; Rbfox1 (Fox-1, A2bp1, Hrnbp1) NP_067452, Rbfox2 (Fox-2, Rbm9, Fxh) NP_001104299, Rbfox3 (Fox-3, Hrnbp3) NP_001034256. Amino acids that differ in sequence from Rbfox3 are shown in grey. Lower case italics denote regions of Rbfox3 that may be absent due to alternative splicing. The RRM domain (as defined by UniProt) is contained within parentheses. Black lines above the sequence mark peptides identified by MS after in-gel digestion with chymotrypsin (c) or trypsin (t). Boxes surround Rbfox3 residues that match the hPY-NLS consensus. D. Alignment of the anti-NeuN ABS from R3hdm2 with the N-terminus of Rbfox3 and the predicted antibody-binding region of Syn1 from Kim et al. [5]. Amino acids identical to those in Rbfox3 are shaded grey, boxes outline amino acids that are identical across all three proteins.
Figure 3.
The anti-NeuN epitope of Rbfox3 maps to the extreme N-terminus of the protein.
A,B. The proteins depicted in A were expressed in 293T cells and analyzed by western with anti-myc-tag (left panel), or anti-NeuN antibodies on duplicate blots. Anti-NeuN recognises both F20 and R46, and amino acids 5–20 (deleted in Fox3Δ-myc) are necessary for recognition of full-length Rbfox3 by anti-NeuN. C. Three Rbfox3 splice variants were expressed in HeLa cells and compared to NeuN from mouse brain by western with anti-NeuN. Variant 1 (v1) corresponds to NP_001034256, v2 to NP_001034257, and v3 to NP_001020102. Exon numbers are as described in [13]. D. RT-PCR of total RNA from P19 cells prior to neuronal induction (lanes 1), P19 cells 7 days after neuronal induction with RA (lanes 2), and P10 mouse brain (lanes 3). For Rbfox3, 3 reverse primers were used to assess mRNAs that encode the short C-terminus (or -IGTM protein isoforms, 15A-R), the longer, -FTPY isoforms (15-R), or all mRNAs (14-R). A common forward primer to exon 7 (7-F) was used for all. Gapdh was used as a loading control.
Figure 4.
The C-terminus of Rbfox3 specifies sub-cellular location.
Rbfox3 variants 1, 2 and 3 harboring C-terminal myc-tags (A), or N-terminal flag-tags (B) were expressed in HeLa cells and stained with anti-NeuN and DAPI. All C-terminally tagged proteins displayed predominantly cytoplasmic localization, whereas N-terminally tagged Fox3v1 and v2 displayed the expected nuclear localization. Scale bars indicate 50 µm.
Figure 5.
Rbfox3 variants all function to regulate alternative splicing, irrespective of steady-state sub-cellular localization.
N2A or 293T cells were transiently transfected with Rbfox3 variants 1, 2 or 3 harboring N-terminal flag-tags, or pcDNA3 vector control. A. Alternative splicing of endogenous mRNAs encoding Rbfox2, c-src and CaV1.2 (alternative exons 9* and 33) was assayed by RT-PCR. Rbfox2 +e6* represents splicing from exon 5 to a cryptic exon downstream of exon 6. B. Western with anti-flag was used to confirm Rbfox3 overexpression, and anti-actin was used as a loading control. C. Quantification of the results shown in A and 2 additional experiments (i.e. n = 3); results are displayed as average +/− standard deviation. All Rbfox3 variants promoted skipping of Rbfox2 exon 6, but no change was seen in the splicing of c-src exon N or CaV1.2 exons 9* or 33 upon Rbfox3 overexpression.
Figure 6.
Rbfox3 variants enhance inclusion of cryptic exons in Rbfox2, leading to nonsense-mediated decay.
N2A cells were transiently transfected with Rbfox3 variants 1, 2 or 3 harboring N-terminal flag-tags, or pcDNA3 vector control, and treated with emetine to inhibit NMD. A. Alternative splicing of endogenous mRNAs encoding Ptbp2 and Rbfox2 was assayed by RT-PCR. Enhanced exclusion of Ptbp2 exon 10 was used as a positive control for emetine treatment. Rbfox2 spliced products are described in D. B. Western with anti-flag was used to confirm Rbfox3 overexpression; anti-actin was used as a loading control. C. Quantification of the results shown in A and 2 additional experiments (i.e. n = 3); results are displayed as average +/− standard deviation. D. Schematic representation of alternative splicing of the Rbfox2 message around exon 6, indicating the relative positions of cryptic exons 5* and 6*. The functional outcomes for the resulting spliced mRNA products are depicted below; inclusion of exon 5* or exon 6* results in the presence of 2 in-frame nonsense codons in exon 7, and destruction of these messages by nonsense-mediated decay. E. The sequences of mouse Rbfox2 exons 5* and 6* (upper case) and surrounding intronic sequence (lower case) are shown aligned to the corresponding regions from human and chicken.
Figure 7.
NeuN (Rbfox3) nuclear export is not Crm1/exportin 1-dependent.
Rev-GFP fusion protein or Rbfox3 variant 3 with an N-terminal flag-tag were expressed in N2A cells and (A) left untreated or (B) treated with 10 ng/ml leptomycin B for 4 hrs prior to fixation and staining with anti-NeuN and DAPI, or just DAPI in the case of Rev-GFP. Leptomycin B treatment caused Rev-GFP to be retained in the nucleus, but caused no change in the sub-cellular localization of Fox3v3.