Cardiac Applications of CRISPR/AAV-Mediated Precise Genome Editing

Abstract

The ability to efficiently make precise genome edits in somatic tissues will have profound implications for gene therapy and basic science. CRISPR/Cas9 mediated homology-directed repair (HDR) is one approach that is commonly used to achieve precise and efficient editing in cultured cells. Previously, we developed a platform capable of delivering CRISPR/Cas9 gRNAs and donor templates via adeno-associated virus to induce HDR (CASAAV-HDR). We demonstrated that CASAAV-HDR is capable of creating precise genome edits in vivo within mouse cardiomyocytes at the neonatal and adult stages. Here, we report several applications of CASAAV-HDR in cardiomyocytes. First, we show the utility of CASAAV-HDR for disease modeling applications by using CASAAV-HDR to create and precisely tag two pathological variants of the titin gene observed in cardiomyopathy patients. We used this approach to monitor the cellular localization of the variants, resulting in mechanistic insights into their pathological functions. Next, we utilized CASAAV-HDR to create another mutation associated with human cardiomyopathy, arginine 14 deletion (R14Del) within the N-terminus of Phospholamban (PLN). We assessed the localization of PLN-R14Del and quantified cardiomyocyte phenotypes associated with cardiomyopathy, including cell morphology, activation of PLN via phosphorylation, and calcium handling. After demonstrating CASAAV-HDR utility for disease modeling we next tested its utility for functional genomics, by targeted genomic insertion of a library of enhancers for a massively parallel reporter assay (MPRA). We show that MPRAs with genomically integrated enhancers are feasible, and can yield superior assay sensitivity compared to tests of the same enhancers in an AAV/episomal context. Collectively, our study showcases multiple applications for in vivo precise editing of cardiomyocyte genomes via CASAAV-HDR.

Keywords: Enhancer; Functional Genomics; Genome editing; Massively Parallel Reporter Assay; PLN-R14Del; Truncated Titin.

Fig. 1. Creation of titin truncating variants via CASAAV-HDR.

A. Genome editing strategy for creation of titin variants or control alleles, with c-terminal insertion of mScarlet tag. B. Experimental timeline. C-E. Exon-mScarlet junction sequences from wildtype control, +AT63117fs, or -T65094fs cDNA, respectively. F-H. In situ imaging of mScarlet fluorescence in cardiomyocytes with Ttn wildtype control, -T65094fs, or +AT63117fs alleles, respectively. Scale bar = 2.5μm. I. Isolated T65094fs cardiomyocyte immunostained with alpha actinin (ACTN2). ACTN2 signal colocalizes with mScarlet (lower panel).

Fig. 2. Creation of a phospholamban point mutation via CASAAV-HDR.

A. Homology directed repair-based strategy for genome editing at the Pln locus. B. Experiment timecourse. Guide RNA and donor template was delivered by AAV9 to newborn mouse pups with cardiomyocyte specific Cas9 activation. C. Wholemount images of 6-week old uninjected control or Pln edited hearts. D. DAPI and WGA stained ventricular sections of control and edited hearts. E. Editing efficiency as the percent of ventricular cardiomyocytes that expressed mScarlet. F. In situ confocal imaging of edited cardiomyocytes. G. mScarlet-expressing cardiomyocytes immunostained with t-tubule marker CAV3. H. Colocalization analysis of CAV3 and mScarlet-PLN in immunostained cardiomyocytes. Student’s t p = 0.052. n = 60, and 55 for WT and R14Del groups, respectively.

Fig. 3. Phenotypes of PLN R14Del cardiomyocytes.

A. Area of edited (mScarlet+) and unedited (mScarlet-) cardiomyocytes. B. Width of cardiomyocytes. C. Organization of the cardiomyocyte t-tubule network. D. Spacing of transverse t-tubule elements. E. Schematic of molecular regulation of calcium release and reuptake. F. Capillary western blot showing that unedited PLN is phosphorylated. G. Capillary western blot showing that R14Del blocks phosphorylation of PLN. H. Quantification of cardiomyocytes with three or more low peaks. I. AnomalyExplorer; beat-to-beat variation in calcium amplitude for a representative SNAP-tag(+) PLN-R14Del cardiomyocyte with multiple low peaks (highlighted in red). J. Quantification of variation in peak height. For each set of calcium traces peak coordinates were fit with a best-fit line. The distance of each peak from the best fit line was normalized to peak height and plotted. n = 429, 430, 395, and 295, for WT SN−, WT SN+, R14Del SN−, and R14Del SN+, respectively. p<0.05*; <0.01**; <0.001***; not significant, ns. A-D) Dunnett’s p-value multiple comparison to R14Del mScarlet(+) group. n = 56, 63, 63, and 62, for WT SN−, WT SN+, R14Del SN−, and R14Del SN+, respectively. J) One-way ANOVA, Tukey multiple comparison correction.

Fig. 4. Precise integration of enhancers for a massively parallel reporter assay.

A. P300 binding intensity at each selected enhancer in either cardiomyocytes or endothelial cells. B. HDR/MPRA strategy. To avoid amplification of AAV-derived DNA or transcripts, at least one primer in each set targets a region outside the homology arms. C. The activity of enhancers inserted into the Tnni1 3’ UTR within cardiomyocytes. D. Comparison of enhancer activities when inserted into the Tnni1 UTR, or when assayed in an episomal AAV STARR-seq vector with an assortment of minimal promoters. Dunnett’s p-value multiple comparison of cardiomyocyte enhancers for each promoter to the cardiomyocyte enhancers of the Tnni1 HDR MPRA. ***p<0.001 E. The activities of enhancers assayed by Tnni1 HDR MPRA poorly correlate with activities obtained by episomal MPRA. Shaded area indicates confidence of fit for best-fit line.