Precision and efficacy of RNA-guided DNA integration in high-expressing muscle loci

Abstract

Gene replacement therapies primarily rely on adeno-associated virus (AAV) vectors for transgene expression. However, episomal expression can decline over time due to vector loss or epigenetic silencing. CRISPR-based integration methods offer promise for long-term transgene insertion. While the development of transgene integration methods has made substantial progress, identifying optimal insertion loci remains challenging. Skeletal muscle is a promising tissue for gene replacement owing to low invasiveness of intramuscular injections, relative proportion of body mass, the multinucleated nature of muscle, and the potential for reduced adverse effects. Leveraging endogenous promoters in skeletal muscle, we evaluated two highly expressing loci using homology-independent targeted integration (HITI) to integrate reporter or therapeutic genes in mouse myoblasts and skeletal muscle tissue. We hijacked the muscle creatine kinase (Ckm) and myoglobin (Mb) promoters by co-delivering CRISPR-Cas9 and a donor plasmid with promoterless constructs encoding green fluorescent protein (GFP) or human Factor IX (hFIX). Additionally, we deeply profiled our genome and transcriptome outcomes from targeted integration and evaluated the safety of the proposed sites. This study introduces a proof-of-concept technology for achieving high-level therapeutic gene expression in skeletal muscle, with potential applications in targeted integration-based medicine and synthetic biology.

Keywords: CRISPR; MT: RNA/DNA Editing; endogenous-promoter; gene editing; homology-independent targeted integration (HITI); integration; muscle; muscle-specific promoters; overexpression; sequencing.

Graphical abstract

Figure 1

Schematic of ideal safe-harbor sites The red bar in the middle represents the optimal location of safe-harbor sites. The site should be unique and located in open chromatin.^, Within 300 kb, the site should be free from any cancer-related genes, microRNAs, or other functional small RNAs. Within 50 kb, the site should be free from the 5′ end of a gene, replication origin, and ultra-conserved element.^,, Additionally, the site should be in a region of low transcriptional activity or have no mRNA within 25 kb.

Figure 2

Targeted integration of a promoterless GFP at Mb and Ckm leads to GFP expression in myotubes (A) Top: overview of transgene integration to hijack endogenous Ckm or Mb promoters. The HITI-based gene editing system is delivered via plasmid transfection, which encodes the CMV-driven S. pyogenes Cas9 protein, U6 promoter-gRNA expression cassette, and a donor plasmid containing a transgene fragment flanked by Cas9 target sites in the reverse orientation relative to genomic DNA. Bottom: the mechanism at the molecular level shows that, by integrating the transgene in the intronic region, only correct integration is expected to produce the protein, while partial or reverse integration is expected to not affect endogenous genes. The figure is not drawn to scale. SSA, strong splice acceptor; PAM, protospacer adjacent motif; gRNA, guide RNA; SVpoly(A), simian virus polyadenylation signal; pA, polyadenylation signal. (B) The gRNA target site map in Ckm and Mb locus. Five gRNA target sites were selected within 250 bp in the splice acceptor and 5′ UTR of the gene. (C) The quantification of gRNA efficiency in the C2C12 cell line with short-read next-generation sequencing (NGS) based on indel percentage. The error bars were computed as the standard deviation obtained through the bootstrap resampling technique. (D) Validation of correct GFP integration in genomic DNA in both 5′ and 3′ integration regions in the treated group. The schematic shows that the forward primer in the GFP transgene overlaps with the reverse primer (primer list). (E) Validation of correct mRNA splicing from GFP integration by cDNA PCR in both loci. At the transcriptome level, only 5′ integration was assessed. (F) Fluorescence microscopy images assessing GFP expression in edited myotubes 10 days after editing; the scale bar represents 50 μm. (G) Flow cytometry analysis of GFP-positive cells 48 h after treatment in the C2C12 cell line. Three independent biological replicates; means ± SEM.

Figure 3

Targeted integration of a promoterless hF9 gene leads to sustained expression (A) Schematic of the structure of the HITI insert plasmid for human F9 expression. The full-length cDNA region from exon 1 to exon 8 is cloned to replace the GFP transgene in the previous construct. (B and C) Genotyping of the F9 integration in the C2C12 cell line using primers spanning the junction between the integration site and the transgene in genomic DNA and cDNA. (D) Relative hF9 expression following plasmid transfection. Five biological independent samples, mean ± SEM. All samples were processed 10 days post transfection, including 7 days of myotube differentiation. (E) Time course cell culture to assess the expression of hF9 in treated C2C12 cells for 30 days without selection compared to the scrambled group. Two independents biological replicates; means ± SEM.

Figure 4

Deep sequencing reveals precise and imprecise outcomes of targeted integration (A and B) Deep sequencing shows a varied range of precise integration in 5′ and 3′ junctions. A similar pattern was observed in 5′ and 3′ junctions in both loci—vector chewback that predominantly happened on the insert side. (C) Deep sequencing at the mRNA level showed a higher percentage of precise integration in three biological replicates in the treated Ckm locus. (D) Modification frequency trace from the fusion of Ckm and the GFP transgene. (E) Top: Schematic Tn5 tagmentation-based sequencing to quantify integration efficiency. Bottom: The graph shows the percentage of reads aligned to the Ckm gene and reference amplicon. (F) Characterization of the fusion of Ckm and hF9 cDNA using 5′ RACE with the GSP reverse transcriptase primer.

Figure 5

RNA-seq reveals significant increases in transgene expression and relative precision of integration at Ckm over Mb (A) Pipeline for the bulk RNA-seq experiment on Ckm- and Mb-integrated and scrambled non-integrated C2C12 myotube cells. (B) PCA of two biological replicates of C2C12 myotubes with Ckm, Mb, or scrambled treatment. (C) Chromosome distribution of differentially expressed genes. (D) Differential expression of genes following hF9 integration in the Ckm locus; highlighted in red are Ckm and hF9. (E) Differential expression of genes following hF9 integration in the Mb locus; highlighted in red are Mb and hF9.

Figure 6

In vivo validation of hF9-targeted integration at the Ckm locus (A) Schematic of the AAV-CRISPR mediated integration in vivo experiment. (B) Relative hF9 expression in primary skeletal muscle cells following myoAAV-CRISPR transduction; mean ± SEM. Ctrl Nuc (−), control without the nuclease group; Ckm Nuc (+), Ckm target with the nuclease group. (C) Relative hF9 expression in mice following myoAAV CRISPR injections at two time points with respective control groups. Mean ± SEM. 3wk Nuc(−) and 8wk Nuc(−), no nuclease group at the 3- and 8-week time point, respectively; 3wk Ckm, Ckm target with the nuclease group at the 3- and 8-week time point, respectively. (D) Integration results from 5′ and 3′ gene-specific primer direction in gDNA. (E) PCA plot of treated, control without nuclease, and untreated mice (NT). (F) Differential expression of genes following hF9 integration in the Ckm locus at the 3-week time point; highlighted in red are Ckm and hF9. (G) Differential expression of genes following hF9 integration in Ckm locus at the 8-week time point; highlighted in red are Ckm and hF9.