RNA splicing is responsive to MBNL1 dose

Abstract

Myotonic dystrophy (DM1) is a highly variable, multi-system disorder resulting from the expansion of an untranslated CTG tract in DMPK. In DM1 expanded CUG repeat RNAs form hairpin secondary structures that bind and aberrantly sequester the RNA splice regulator, MBNL1. RNA splice defects resulting as a consequence of MBNL1 depletion have been shown to play a key role in the development of DM1 pathology. In patient populations, both the number and severity of DM1 symptoms increase broadly as a function of CTG tract length. However significant variability in the DM1 phenotype is observed in patients encoding similar CTG repeat numbers. Here we demonstrate that a gradual decrease in MBNL1 levels results both in the expansion of the repertoire of splice defects and an increase in the severity of the splice alterations. Thus, MBNL1 loss does not have an all or none outcome but rather shows a graded effect on the number and severity of the ensuing splice defects. Our results suggest that once a critical threshold is reached, relatively small dose variations of free MBNL1 levels, which may reflect modest changes in the size of the CUG tract or the extent of hairpin secondary structure formation, can significantly alter the number and severity of splice abnormalities and thus contribute to the phenotype variability observed in DM1 patients.

Figure 1. Number and severity of splice defects increase when MBNL1 is silenced incrementally from ∼79% to ∼98%.

SkMC were transfected with siRNAs directed against MBNL1 and cell samples on each subsequent day post-siRNA transfection for a period of 5 days, were divided into 4 aliquots where one aliquot was used to measure MBNL1 levels and total RNA was extracted from each of the three other aliquots. Scrambled siRNA transfected samples were harvested on Day 5, the last time point of the experiment. (A) Total protein (10 µg) was analyzed by western blot to measure the silencing achieved for MBNL1 at 24 h intervals for 5 days. Blots were probed for GAPDH as an internal control. (B) Synthesized cDNAs were subjected to PCR analysis to study RNA splicing as indicated with GAPDH RNA as an internal control. In each case the levels of exon inclusion obtained in the experiment shown are indicated. (C) The results of RNA splicing as a function of MBNL1 levels in SkMC are tabulated.

Figure 2. Splice defects observed at ∼67% MBNL1 silencing.

SkMC were transfected with siRNAs directed against MBNL1 and cell samples collected 24 h, 32 h and 40 h post-siRNA transfection were divided into 4 aliquots. One aliquot was used to measure MBNL1 levels and total RNA was extracted from each of the three other aliquots. Scrambled siRNA transfected samples were harvested at ∼48 h. (A) siRNA mediated down-regulation of MBNL1 at 24 h, 32 h and 40 h in SkMC is shown. Blots were probed for GAPDH as an internal control. (B) Synthesized cDNAs were subjected to PCR analysis as indicated with GAPDH RNA as an internal control. In each case the levels of exon inclusion obtained in the experiment shown are indicated. (C) The results of RNA splicing as a function of MBNL1 levels in SkMC are tabulated.

Figure 3. Splice defects in Mbnl1^+/ΔE3 and Mbnl1^ΔE3/ΔE3 skeletal muscle.

Lower limb skeletal muscles from adult wild-type, Mbnl1^+/ΔE3 and Mbnl1^ΔE3/ΔE3 mice were harvested and divided into 2 aliquots. One aliquot was used to measure Mbnl1 levels and the other aliquot was used study RNA splicing. (A) Western blot analysis of steady-state Mbnl1 levels in skeletal muscle of wild-type, Mbnl1^+/ΔE3 and Mbnl1^ΔE3/ΔE3 mice are shown with Gapdh as an internal loading control. (B) cDNAs synthesized from skeletal muscle of wild-type, Mbnl1^+/ΔE3 and Mbnl1^ΔE3/ΔE3 mice were subjected to PCR analysis as indicated with Gapdh RNA as an internal control. In each case the levels of exon inclusion obtained in the experiment shown are indicated. (C) The results of RNA splicing examined as a function of Mbnl1 levels are tabulated.

Figure 4. RNA half-life measurements in SkMC.

Normal myoblasts were treated with a combination of actinomycin-D and α-aminitin to inhibit transcription. Myoblasts were harvested at different time-points after treatment (0, 0.5, 4, 8, 16 and 24 h) and RNA was extracted. Synthesized cDNAs were subjected to RT-PCR analysis to measure RNA half-lives as previously described (17,18). MYC, a short-lived RNA and the long-lived 18S RNA were used as controls. Graphical representation of the average percent of RNA plotted against time from two independent experiments is shown in Figure S2.

Figure 5. Incremental depletion of MBNL1 results in an increase of both the number and severity of RNA splice defects.

RNA splice defects that manifest with the depletion of MBNL1 in SkMC and in Mbnl1^+/ΔE3 and Mbnl1^ΔE3/ΔE3 skeletal muscle are shown. Line thickness represents the severity of the splice defect.