CRISPR-Cas9 base editing of pathogenic CaMKIIδ improves cardiac function in a humanized mouse model

Figures

• 

  Figure 1. Analysis of on- and off-target editing in human iPSCs. (A) Schematic showing the structure of CaMKIIδ with its 3 domains (blue, catalytic domain; green, regulatory domain; brown, association domain). Upon oxidative stress, 2 critical methionine residues at position 281 and 282 become oxidized, thereby preventing association of the regulatory with the catalytic domain, and resulting in cardiac disease. We deployed 2 ABE strategies to ablate either 1 (sgRNA1) or both (sgRNA2) oxidative activation sites of CaMKIIδ. (B) Human iPSCs were nucleofected with ABE8e(V106W) and either sgRNA1 or sgRNA2. (C) Sequence of sgRNA1 (CAMK2D on-target, ON) and the corresponding DNA and PAM sequences of the top 8 predicted potential off-target sites (OT), as predicted by CRISPOR. Nucleotides highlighted in yellow are different from sgRNA1. (D) Percentage of adenine (A) to guanine (G) editing for all adenines within the on- and off-target sites (ordered from 5′ to 3′) following adenine base editing with ABE8e(V106W)-SpCas9 + sgRNA1 (n = 3). (E) Sequence of sgRNA2 (CAMK2D on-target, ON) and the corresponding DNA and PAM sequences of the top 8 predicted potential off-target sites, as predicted by…

• 

  Figure 2. Editing CAMK2D in a humanized mouse model upon IR injury. (A) Schematic showing humanization of the mouse sequence of the regulatory domain of CaMKIIδ, which is encoded by exons 11 and 12. (B) Flowchart showing the experimental design for subjecting 12-week-old humanized CAMK2D female and male mice to IR. Cardiac function was assessed 1 week before as well as 24 hours and 3 weeks after IR by echocardiography. At 4 weeks after IR, mice were subjected to a treadmill exhaustion test. At 5 weeks after IR, mice were sacrificed, and tissue was collected for further analyses. (C) Illustration of the split-MyoAAV 2A that was used to deliver ABE8e(V106W)-SpCas9 + sgRNA1 in vivo. (D) Illustration of the split-MyoAAV 2A that was used to deliver ABE8e(V106W)-SpRY + sgRNA2 in vivo. (E) Representative M-mode traces of hearts from mice subjected to either sham, IR+control virus, IR+sgRNA1 or IR+sgRNA2 (echocardiography; 3 weeks after IR; in total n = 8 per group). (F) Mean fractional shortening 3 weeks after IR (n = 8 per group). (G) Mean LVIDd 3 weeks after IR (n = 8 per group). (H) Mean LVIVd 3 weeks after IR (n = 8 per group). All data are individual data points with mean ± SEM and all…

• 

  Figure 3. CAMK2D-edited mice show improved exercise performance after IR. (A) Protocol used for the treadmill exhaustion test. (B) Mean maximal velocity achieved on the treadmill prior to exhaustion (n = 8 per group). (C) Mean total distance achieved on the treadmill prior to exhaustion (n = 8 per group). (D) Linear regression analysis of fractional shortening and the corresponding maximal velocity achieved on the treadmill (n = 8 per group, n = 32 in total). (E) Linear regression analysis of fractional shortening and the corresponding total distance achieved on the treadmill (n = 8 per group, n = 32 in total). All data are individual data points with mean ± SEM and all replicates are individual mice. Statistical comparisons are based on 1-way ANOVA posthoc corrected by Holm-Šidák (B and C) and linear regression analysis (D and E); **P < 0.01, ****P < 0.0001.

• 

  Figure 4. Normal organ weights and prevention of fibrosis in sgRNA2-treated mice after IR. (A) Mean body weight for mice subjected to sham, IR+control virus, IR+sgRNA1 or IR+sgRNA2 (n = 8 per group). (B) Mean heart weight normalized to tibia length (n = 8 per group). (C) Mean lung weight normalized to tibia length (n = 8 per group). (D) Mean liver weight normalized to tibia length (n = 8 per group). (E) Representative picrosirius red staining of transverse cardiac sections from all groups (scale bar: 1,000 μm). (F) Mean percentage of fibrotic tissue (n = 3 per group). All data are individual data points with mean ± SEM and all replicates are individual mice. Statistical comparisons are based on 1-way ANOVA post-hoc corrected by Holm-Šidák (A–D and F); **P < 0.01, ***P < 0.001, ****P < 0.0001.

• 

  Figure 5. Analysis of editing efficiency. (A) Percentage of adenine (A) to guanine (G) editing at the DNA and cDNA level in the anterior cardiac wall of mice following treatment with ABE8e(V106W)-SpCas9 + sgRNA1 (n = 5 per group). (B) Percentage of adenine (A) to guanine (G) editing at the DNA and cDNA level in the anterior cardiac wall of mice following treatment with ABE8e(V106W)-SpRY + sgRNA2 (n = 5 per group). (C) Western blot analysis of oxidized CaMKII, total CaMKII, and GAPDH in mice subjected to either sham, IR+control virus, IR+sgRNA1 or IR+sgRNA2 (in total n = 5 per group). (D) Mean densitometric analysis for oxidized CaMKII normalized to GAPDH (n = 5 per group). (E) Mean densitometric analysis for total CaMKII normalized to GAPDH (n = 5 per group). (F) Mean densitometric analysis for oxidized CaMKII normalized to total CaMKII (n = 5 per group). (G) Mean CaMKII activity for all groups (n = 5 per group). All data are individual data points with mean ± SEM and all replicates are individual mice. Statistical comparisons are based on 1-way ANOVA post-hoc corrected by Holm-Šidák (D–G); **P < 0.01, ***P < 0.001, ****P < 0.0001.