RNA toxicity in myotonic muscular dystrophy induces NKX2-5 expression

Abstract

Myotonic muscular dystrophy (DM1) is the most common inherited neuromuscular disorder in adults and is considered the first example of a disease caused by RNA toxicity. Using a reversible transgenic mouse model of RNA toxicity in DM1, we provide evidence that DM1 is associated with induced NKX2-5 expression. Transgene expression resulted in cardiac conduction defects, increased expression of the cardiac-specific transcription factor NKX2-5 and profound disturbances in connexin 40 and connexin 43. Notably, overexpression of the DMPK 3' UTR mRNA in mouse skeletal muscle also induced transcriptional activation of Nkx2-5 and its targets. In human muscles, these changes were specific to DM1 and were not present in other muscular dystrophies. The effects on NKX2-5 and its downstream targets were reversed by silencing toxic RNA expression. Furthermore, using Nkx2-5+/- mice, we show that NKX2-5 is the first genetic modifier of DM1-associated RNA toxicity in the heart.

Conflict of interest statement

COMPETING INTERESTS STATEMENT

The authors declare competing financial interests: details accompany the full-text HTML version of the paper at http://www.nature.com/naturegenetics/.

Figures

Figure 1

Cardiac conduction abnormalities. (a) ECG abnormalities in induced transgenic mice. P (atrial) and R (ventricular) waves are indicated. * denotes missing R waves in second-degree block. Note the lack of P waves in the complete heart block example. (b) Immuno-histochemistry of a cross-section of mouse hearts, showing CX43 (green), CX40 (red) and CX40 CX43 coexpression (orange). Note the thinning and loss of CX40 staining in the conduction system (Purkinje fibers) of induced mice. (c) CX43 expression is lower in ventricular myocardium of induced mice than in uninduced mice, as detected by immunohistochemistry (IHC), RT-PCR (mean ± s.d.) and protein blotting. Numbers above lanes in protein blot represent the number days of transgene induction. Numbers below lanes represent relative CX43 expression compared to wild-type mice. Numbers below days 6–17 represent the average of those samples. GAPDH was included as a loading control.

Figure 2

NKX2-5 expression is induced by RNA toxicity in the heart. (a) Immunohistochemistry for NKX2-5 in ventricular myocardium demonstrates that induced transgenic mice have more nuclei with detectable NKX2-5, compared to uninduced mice (shown graphically in the lower left panel; mean ± s.d.; P = 0.002). RT-PCR demonstrates that Nkx2-5 mRNA expression in heart is also higher in induced mice (lower right panel; mean ± s.d.; P = 0.03). (b) RNA blotting confirmed RT-PCR results (numbers below the blot refer to expression level relative to uninduced mice). GAPDH was used as a loading control.

Figure 3

NKX2-5 is ectopically induced in skeletal muscle. (a) Transgene expression in induced mice. DAPI staining marks nuclei; GFP expression indicates transgene expression. Immunohistochemistry for NKX2-5 (red) shows that there is no expression in uninduced skeletal myonuclei but markedly increased expression in muscle from induced mice. (b) Protein blotting for NKX2-5 in skeletal muscle. GAPDH was used as a loading control.

Figure 4

NKX2-5 downstream targets are affected by induction of transgene expression. (a) Real-time RT-PCR results for Nppb, Ankrd1 and Gia5 mRNA in hearts of 5-313 transgenic mice, relative to uninduced 5-313 mice (mean ± s.d.). (b) Real-time RT-PCR results for Nppb, Ankrd1 and Adora1 mRNA in skeletal muscle of 5-313 transgenic mice, relative to uninduced 5-313 mice (mean ± s.d.). Note reversion toward normal levels with transgene silencing. (c) Protein blots of downstream targets in skeletal muscle extracts demonstrate that expression of CARP is induced and CX43 is downregulated by RNA toxicity, and normal levels are restored in reverted mice. Lanes 1–2: uninduced mice; 3–5: induced mice; 6–8: reverted mice. GAPDH was used as a loading control.

Figure 5

NKX2-5 is sufficient to transcriptionally activate downstream targets in skeletal myoblasts. RT-PCR of transient transfection assays demonstrates that Nkx2-5 expression in C2C12 skeletal myoblasts is sufficient to drive dose-dependent expression of cardiac-specific downstream targets Nppa and Nppb. β-actin is a loading control. Leftmost lane is a negative control (no plasmid).

Figure 6

NKX2-5 is expressed in muscle tissue from individuals with DM1. (a) DAPI staining for myonuclei and immunohistochemistry for NKX2-5 in skeletal muscle from individuals with or without DM1. NKX2-5 expression is present in muscle from individuals with DM1 (note green nuclear staining). (b) Protein blotting of muscle protein extracts shows NKX2-5 in tissue from individuals with DM1 (lanes 1–4) but not in normal muscle (lanes 5, 6), in muscle from individuals with Duchenne muscular dystrophy (lane 7) or in muscle from individuals with X-linked myopathy (lane 8). Lane 9: positive control (normal human heart). GAPDH was used as a loading control. NKX2-5 expression seems higher in hearts from individuals with DM1 than in hearts from individuals without DM1. (c) RT-PCR shows that expression of NKX2-5 and its downstream targets NPPA and NPPB is specific to muscles from individuals with DM1. GAPDH was used as a loading control. MD, muscular dystrophy.

Figure 7

Reversal of RNA toxicity and molecular changes. (a) CX40 expression (red) is reduced in the cardiac conduction system of induced mice but restored to more normal levels in reverted mice. Insets show differences in CX40 (red) and CX43 (green) expression in uninduced and induced ventricular myocardium. (b) Upper panel: protein blot shows that CX43 expression is restored in hearts of reverted mice. Lanes 1–3: uninduced mice; 4–6: induced mice; 7–9: reverted mice. Numbers below lanes indicate relative expression. Lower panel: RT-PCR shows that Gia1 (Cx43) mRNA expression is restored in reverted mice. (c) Reverted mice have normal Nkx2-5 expression in cardiac muscles, as shown by a count of nuclei and by RT-PCR. (d) NKX2-5 in skeletal muscle is undetectable by immunohistochemistry (upper panel) or by protein blotting (lower panel) in reverted mice. Lanes 1–2: uninduced mice; 3–5: induced mice; 6–8: reverted mice. GAPDH was used as a loading control. Graphs in b and c show mean ± s.d.

Figure 8

Nkx2-5 haploinsufficiency protects against cardiac conduction defects induced by DMPK 3′ UTR mRNA. (a) PR intervals in Nkx2-5+/+ 5-313 heterozygotes increase much more than in Nkx2-5LacZ/+ 5-313 heterozygotes after induction of DMPK 3′ UTR mRNA toxicity (mean ± s.d.). * indicates that PR intervals are not significantly different before doxycycline induction (P = 0.08). ** indicates that PR intervals are significantly different after doxycycline induction (P = 0.00003). (b) RT-PCR confirms that Nkx2-5 mRNA expression is much lower in Nkx2-5LacZ/+ 5-313 heterozygote hearts than in Nkx2-5+/+ 5-313 heterozygote hearts (mean ± s.d.). (c) Representative protein blot demonstrating that NKX2-5 expression is lower in Nkx2-5LacZ/+ 5-313 heterozygote hearts than in Nkx2-5+/+ 5-313 heterozygote hearts (+/− indicates Nkx2-5LacZ/+ mice). Numbers below blot show relative NKX2-5 expression compared to uninduced Nkx2-5+/+ 5-313 heterozygotes. Uninduced skeletal muscle (sk.m.) extract was used as a negative control for NKX2-5. At least five mice were analyzed for each group for RT-PCR and protein blotting. Dox, doxycycline.