Induction of a local muscular dystrophy using electroporation in vivo: an easy tool for screening therapeutics

Abstract

Intramuscular injection and electroporation of naked plasmid DNA (IMEP) has emerged as a potential alternative to viral vector injection for transgene expression into skeletal muscles. In this study, IMEP was used to express the DUX4 gene into mouse tibialis anterior muscle. DUX4 is normally expressed in germ cells and early embryo, and silenced in adult muscle cells where its pathological reactivation leads to Facioscapulohumeral muscular dystrophy. DUX4 encodes a potent transcription factor causing a large deregulation cascade. Its high toxicity but sporadic expression constitutes major issues for testing emerging therapeutics. The IMEP method appeared as a convenient technique to locally express DUX4 in mouse muscles. Histological analyses revealed well delineated muscle lesions 1-week after DUX4 IMEP. We have therefore developed a convenient outcome measure by quantification of the damaged muscle area using color thresholding. This method was used to characterize lesion distribution and to assess plasmid recirculation and dose-response. DUX4 expression and activity were confirmed at the mRNA and protein levels and through a quantification of target gene expression. Finally, this study gives a proof of concept of IMEP model usefulness for the rapid screening of therapeutic strategies, as demonstrated using antisense oligonucleotides against DUX4 mRNA.

Figure 1

Hyaluronidase pre-treatment improves β-galactosidase expression in mouse Tibialis Anterior muscle (TA). (A,B) Representative sections of TA electroporated (A) without hyaluronidase pre-treatment (IMEP) or (B) with hyaluronidase pre-treatment 2 h before the electroporation procedure (hIMEP). TA muscles were injected by IMEP or hIMEP with 40 µg of pCMV-lacZ reporter plasmid. TA were harvested 1-week post-injection and cryosections stained with X-gal (blue) and counterstained with Eosin (pink). Scale 500 µm. (C) Percentage of surface area expressing β-galactosidase (β-gal⁺) quantified by color thresholding using ImageJ. Data are represented as boxplots, ****p < 0.0001 Mann–Whitney Rank Sum Test; n = 4 for each group. The graph was generated using GraphPad Prism 6.01.

Figure 2

hIMEP of DUX4-expression plasmid induces muscle lesions. (A,B) Representative sections of TA electroporated with 10 µg of pCMV-lacZ and 40 µg of (A) pCIneo or (B) pCIneo-DUX4 plasmids. TA muscles were harvested 1-week post injection and cryosections from distal, medial and proximal regions stained with X-gal to assess β-gal⁺ areas (small pictures). Adjacent sections were stained with HEB coloration for muscle damages evaluation (large pictures). Scales 50 µm (red) and 100 µm (black). (C,D) Percentage of surface area (C) expressing β-galactosidase (β-gal⁺) and (D) damaged in mouse TA 1-week post hIMEP using 10 µg of pCMV-lacZ concomitant with 40 µg of pCIneo (Ctrl) or pCIneo-DUX4 (DUX4) plasmids. β-gal⁺ and lesion area percentage were evaluated on total section stained with X-gal or HEB respectively, and quantified by color thresholding using ImageJ. (E) Recirculation test. Both TA from same mouse were electroporated, one using 40 µg of pCIneo-DUX4, the other saline solution. One week after hIMEP, lesion area was evaluated on HEB stained TA cryosections by color thresholding using ImageJ. All results are presented as boxplots, **p < 0.01 and ****p < 0.0001 Mann–Whitney Rank Sum Test; n = 5 for each group except in (E) n = 2. Graphs were generated using GraphPad Prism 6.01.

Figure 3

Dose–response of muscle lesion area in mouse TA 1-week post hIMEP procedure. (A,B) Lesion area percentage was evaluated (A) on total cryosection or (B) in the injected area (defined by β-gal⁺ region) from distal, medial and proximal part of TA electroporated with different doses of control or DUX4-expression plasmid and 10 µg of pCMV-lacZ. Sections were stained with HEB and lesions were quantified by color thresholding using ImageJ. Control groups (Ctrl) include 1-, 20-, 40-µg pCIneo-injected groups as there was no statistical difference among them (one-way ANOVA followed by Dunn’s post hoc test, p = 0.959 and 0.153 for total and injected region quantifications respectively, NS). (C) Comparison of injured area percentages quantified on total section (black) or in injected area (defined by the β-gal⁺ region) (grey). DUX4 groups include 1-, 5-, 20-, 40-µg pCIneo-DUX4-injected groups as there was no statistical difference among them (one-way ANOVA followed by Dunn’s post hoc test, p = 0.894 and 0.957 for total and injected regions respectively, NS). All results are presented as boxplots, ****p < 0.0001, **p < 0.01, (vs ctrl (A,B) or as indicated (C)). (A,B) One-way ANOVA on the ranks followed by Dunn’s post hoc test; n = 3, except 5 µg DUX4 (n = 4), 40 µg DUX4 (n = 5) and Ctrl (n = 11). (C) Wilcoxon signed rank test; n = 11 (ctrl) or 15 (DUX4). Graphs were generated using GraphPad Prism 6.01.

Figure 4

Lesion distribution through TA regions. Mouse TAs were electroporated with different doses of pCIneo-DUX4 (1, 5,10, 20 and 40 µg) and 10 µg of pCMV-lacZ. Muscle damaged area 1 week following TA hIMEP was evaluated in the injected area (β-gal⁺) of HEB stained cryosections from distal, medial and proximal regions of TA (color thresholding, ImageJ). Since no statistical difference was observed between tested doses considering each region individually, data were respectively pooled to form single proximal, medial and distal groups (data not shown, One-way Anova, p = 0.871, 0.821 and 0.395 in proximal, medial and distal region respectively). All data are presented as boxplots, *p < 0.05 and **p < 0.01. Friedman Repeated Measures Analysis of Variance on Ranks; n = 15 for each group. Graphs were generated using GraphPad Prism 6.01.

Figure 5

Time-course analysis. TA muscles were electroporated with 1 µg of pCIneo-DUX4 and 10 µg of pCMV-LacZ. Lesion area was evaluated 1-, 3- or 7-days after hIMEP in the injected area (β-gal⁺) on HEB stained cryosections of medial and proximal TA regions (color thresholding, ImageJ). Results are presented as boxplots, *p < 0.05 and **p < 0.01. One-way ANOVA on the ranks followed by Dunn’s post hoc test; n = 3 for each group. Scale 1 mm. The graph was generated using GraphPad Prism 6.01.

Figure 6

Confirmation of DUX4 mRNA expression by 3′RACE. (A) Nested PCR was used to amplify the 3′ end of DUX4 transcripts from total RNA extracted from TA muscles electroporated with 1 µg of control (left panel) or DUX4-expression pDNA (right panel). Representative cropped gels. Lanes 1 and 2 come from one gel; lanes 3, 4 and 5 come from a second gel ran in parallel. Corresponding full-length gels are available in supplementary information (SI Fig. S2 online). 3′RACE were performed 1-, 3- and 7-days post hIMEP. Wells (−) show negative controls without retro-transcription. Several fragments were detected (~ 1,350 bp, ~ 550 bp and ~ 350 bp; colored arrows) in each time point (not shown), cloned, sequenced and analyzed to confirm DUX4 specificity. (B) Schematic representation (not to scale) of 3′RACE products analysis. In silico alignment confirmed conservation of vPMO [pLAM3A (− 12 + 13)] target sequence (black line).

Figure 7

Confirmation of DUX4 protein expression by Immunofluorescence. Mouse TAs were injected by hIMEP with 10 µg pCMV-lacZ plasmid concomitantly with 40 µg either pCIneo-DUX4 or pCIneo, or with a saline solution (negative control), as indicated. Cryosections were analyzed 1, 3 and 7 days post-injection by immunofluorescence with antibodies directed against either DUX4 (E5-5) (red) or laminin α2 (green) to stain the myofibre basal lamina (for details, see “Methods”). DAPI was used to visualize nuclei (blue). Pictures were taken with a Nikon Eclipse 80i microscope and merged using NIS-Elements software. Scale bar 50 µm.

Figure 8

RTqPCR analyses of DUX4 target genes (Wfdc3, Zscan4c) and Myog expression level in TA muscle. TAs were injected by hIMEP with either 1 µg of control pCIneo DNA, 1 µg of pCIneo-DUX4- or with 50 µl of saline solution and harvested 1-, 3- or 7-days post injection. Total RNA was extracted with Trizol and qPCRs were performed in duplicates using SYBR Green FastStart Essential DNA Green Master. Analyzes were performed with LightCycler 96 software. At each time, results obtained from control plasmid and saline solution injected groups were pooled to form a single control group, as no statistical difference could be highlighted between them for all tested genes (one-way ANOVA on the ranks followed by Dunn’s post hoc test). Results are presented as fold difference to Rplp0 for Wfdc3 and Zscan4c and as relative to control for Myog. All data are presented as boxplots, *p < 0.05, **p < 0.01, ***p < 0.001 and ^#p < 0.05 vs DUX4 7 days, One-way ANOVA on the ranks followed by Dunn’s post hoc test n = 14 (Ctrl) or 7 (DUX4) (Wfdc3 and Zscan4c) and n = 8 (ctrl) or 4 (DUX4) (Myog). Graphs were generated using GraphPad Prism 6.01.

Figure 9

vPMO treatment of mice electroporated with DUX4-expression plasmid decreases TA muscle lesions. TA muscles were injected by hIMEP with 20 µg of pCineo-DUX4 plasmid. Six hours later, mice received either an intraperitoneal injection of 250 µg of vPMO pLAM3A (− 12 + 13) targeting DUX4 mRNA (vPMO group) or no supplemental treatment (DUX4 group). Percentage of lesion area was evaluated 7 days post IMEP by color thresholding using ImageJ on HEB colored cryosections of TA proximal and medial parts. Results are presented as Box plots, ***p < 0,001 by Mann–Whitney Rank Sum Test; n = 4 for each group. The graph was generated using GraphPad Prism 6.01. Black arrows show fibrosis, white arrows atrophic fibres, black circles inflammatory infiltrate. Scales 1 mm (black) 250 µm (green), 100 µm (yellow) and 50 µm (red).