Senescence and inflammation are unintended adverse consequences of CRISPR-Cas9/AAV6-mediated gene editing in hematopoietic stem cells

Abstract

Gene editing (GE) using homology-directed repair (HDR) in hematopoietic stem and progenitor cells (HSPCs) offers promise for long-range gene correction of inherited genetic disorders. However, cellular responses induced by CRISPR-Cas9/AAV6 engineering impair the long-term repopulating potential of HDR-edited HSPCs, adversely impacting the safety and efficacy of clinical translation. Our study uncovers a durable senescence-like response in genetically engineered HSPCs triggered by p53 and interleukin (IL)-1/nuclear factor κB (NF-κB) activation, which restricts graft size and clonal diversity in long-term transplantation assays. We show that transient p53 inhibition or blocking inflammatory pathways mitigates senescence-associated responses, improving the repopulating capacity of edited HSPCs. Importantly, we identify treatment with Anakinra, an IL-1 signaling antagonist, as a promising strategy to enhance polyclonal output in HDR-edited cells while minimizing genotoxicity risks associated with the editing procedure. Overall, our findings present strategies to overcome key hurdles in HDR-based HSPC gene therapies, providing a framework for enhancing their efficacy and safety in clinical applications.

Keywords: CRISPR-Cas9; DNA damage; gene editing; gene therapy; genome integrity; hematopoietic stem cells; inflammatory programs; p53; senescence; viral vectors.

Graphical abstract

Figure 1

CRISPR-Cas9/AAV6-mediated gene editing in human HSPCs triggers cellular senescence and inflammatory programs (A) Experimental workflow for cord blood (CB)-derived HSPC GE. (B) CDKN1A expression relative to GUSB housekeeping gene (n = 29, 11, 27, 6, and 6); Kruskal-Wallis test. (C and D) Percentage of SA-β-Galactosidase (SA-β-Gal)⁺ (C: n = 12, 5, 11, 7, and 7 for up to 8 independent donors) or p16⁺ (D: n = 13, 5, 12, 7, and 7 for up to 9 independent donors) HSPCs; Kruskal-Wallis test. (E–G) Inflammatory cytokine expression relative to GUSB housekeeping gene. (E) IL-1α (n = 15, 4, 15, 5, and 5); (F) IL-6 (n = 33, 4, 27, 6, and 6); (G) IL-8 (n = 30, 4, 27, 6, and 6); Kruskal-Wallis test. (H) Number of colonies grown from CB-derived HSPCs treated as indicated (n = 27, 10, 18, 11, and 7); Kruskal-Wallis test; mean ± SEM. (I) Heatmap showing normalized enrichment score (NES) from GSEA against the hallmark gene sets of the Molecular Signatures Database (MSigDB) and senescence-associated gene sets from reported literature. Terms are sorted according to NES, and adjusted p value is reported. p values: ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001. If not indicated, results are not significant (ns). Unless otherwise specified, lines indicate median values. CB HSPCs were collected at 4 days post-GE (B–H) and at 1 day and 4 days post-GE (I).

Figure 2

HSPCs that have undergone CRISPR-Cas9/AAV6-mediated HDR display the highest accumulation of senescence markers (A) Percentage of HDR- and NHEJ-edited alleles in CB-derived HSPCs with increasing doses of AAV6 vector or HS RNP. HDR: 2 × 10³: n = 2, 1 × 10⁴: n = 27, 2 × 10⁴: n = 4. NHEJ: HS: n = 7, 2 × 10³: n = 2, 1 × 10⁴: n = 11, 2 × 10⁴: n = 4; Kruskal-Wallis test; mean ± SEM. (B and C) Percentage of SA-β-Gal⁺ (B) or p16⁺ (C) CB-derived HSPCs (n = 5 technical replicates); Friedman test. (D) Number of colonies grown from CB-derived HSPCs treated as indicated (n = 3 independent CFU replicates); Friedman test; mean ± SEM. (E and F) Percentage of SA-β-Gal⁺ (E, n = 3 independent CFU replicates) or p16⁺ (F, n = 6 independent CFU replicates) cells in the progeny of edited CB-derived HSPCs at indicated conditions; Friedman test. (G) Percentage of hCD45⁺cells in PB of mice transplanted with CB-derived HSPCs edited with increasing AAV6 doses (n = 5). Linear mixed effects models (LME) followed by post hoc analysis; results are shown for the last time point; mean ± SEM. (H) Percentage of BM-derived hCD45⁺ cells in mice in (G) (n = 5); Kruskal-Wallis test. (I–L) Percentage of SA-β-Gal⁺ cells in BM-derived human CD45⁺ (I), CD19⁺ (J), CD13⁺ (K), and CD34⁺ (L) (n = 4, 5, 5, 5, 3, 4, 4, 4, 5, and 5 of mice in H and Figure S2E); Kruskal-Wallis test. p values: ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001. If not indicated, results are not significant (ns). Unless otherwise specified, lines indicate median values. Samples were collected at 4 days post-GE (A–F) and at 15 weeks post-transplantation (H–L).

Figure 3

Senescence markers are detected in vivo upon transplantation of CRISPR-Cas9/AAV6-edited HSPCs (A) Percentage of hCD45⁺ cells in BM early (6 weeks, 6w) and late (15 weeks, 15w) post-transplantation of mice transplanted at day 1 post-GE with CB-derived HSPCs. (6 weeks: n = 12 and 12; 15 weeks: n = 11 and 12); Mann-Whitney test. (B) Percentage of SA-β-Gal⁺ cells in BM-derived CD34⁺ HSPCs retrieved from mice in (A) (n = 12, 12, 12, 12, 12, and 11); Kruskal-Wallis test. (C and D) Spearman correlation between the percentage of SA-β-Gal⁺ cells within CD34⁺ GFP⁺ cells (y axis) and CD34⁺ GFP⁺ (x axis) at 6 weeks (C) or 15 weeks (D) post-transplantation (n = 12). Each dot represents an individual mouse from (B). (E) Percentage of hCD45⁺ cells in PB over time of mice transplanted at day 1 post-GE with CB-derived HSPCs treated as indicated (n = 8). LME followed by post hoc analysis; results are shown for the last time point; mean ± SEM. (F) Percentage of BM-derived hCD45⁺ cells of mice from (E) (n = 8); Mann-Whitney test. (G) Percentage of hCD45⁺SA-β-Gal⁺ cells in PB of mice from (E) (n = 8); LME followed by post hoc analysis; mean ± SEM. (H) Percentage of SA-β-Gal⁺ cells in BM-derived hCD34+ of mice from (F) (n = 8); Mann-Whitney test. p values: ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001. If not indicated, results are not significant (ns). Unless otherwise specified, lines indicate median values, and samples were collected at 18 weeks post-transplantation (F and H).

Figure 4

Senescence markers are detected in vivo upon transplantation of CRISPR-Cas9/AAV6-edited HSPCs (A and B) Percentage of SA-β-Gal⁺ (A) or p16⁺ (B) mPB HSPCs (n = 10, 5, 5, 10, and 10 for up to 3 independent donors); Kruskal-Wallis test. (C) Number of colonies grown from mPB HSPCs treated as indicated (n = 9, 3, 3, and 9 for up to 3 independent donors); Kruskal-Wallis test; mean ± SEM. (D) Percentage of hCD45⁺ cells in PB of mice transplanted with mPB HSPCs at day 1 post-GE at the indicated conditions (n = 19, 11, 6, and 19). LME followed by post hoc analysis; results are shown for the 6 weeks post-transplantation time point; mean ± SEM. (E) Percentage of BM-derived hCD45⁺ cells in mice from (D) (n = 19, 11, 6, and 19); Kruskal-Wallis test. (F–J) Percentage of SA-β-Gal⁺ (F–I) or p16⁺ (J) cells within human CD45⁺ (F: n = 11, 11, 6, 11, and 11), CD19⁺ (B cells, G: n = 5, 5, 6, 6, and 6), CD13⁺ (myeloid cells, H: n = 5, 5, 6, 6, and 6), and CD34⁺ (HSPCs, I: n = 11, 11, 6, 11, and 11 and J: 5, 5, 6, 6, and 6) retrieved from the BM of mice from either one or two independent experiments from (E) 15 weeks post-transplantation; Kruskal-Wallis test; median. p values: ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001. If not indicated, results are not significant (ns). Unless otherwise specified, lines indicate median values, and mPB HSPCs were collected at 4 days post-GE (A–C) and 18 weeks post-transplantation (D–J).

Figure 5

p53 inhibition or anti-inflammatory treatments at the time of gene editing mitigate senescence response and improve HSPC fitness (A) Experimental workflow. (B) Heatmap showing NES from GSEA against the hallmark gene sets of the Molecular Signatures Database (MSigDB). Terms are sorted according to NES. (C and D) Percentage of SA-β-Gal⁺ (C) or p16⁺ (D) CB HSPCs treated as indicated at day 4 post-GE (n = 5 technical replicates); Friedman test. (E) Number of colonies grown from CB HSPC gene edited and treated as indicated at day 4 post-GE (n = 5 independent CFU replicates); Friedman test; mean ± SEM. (F and G) Percentage of SA-β-Gal⁺ (F) or p16⁺ (G) cells in the progeny of edited CB HSPCs at indicated conditions from (E) (n = 5 independent CFU replicates); Friedman test. (H) Percentage of SA-β-Gal⁺ mPB HSPCs treated as indicated (n = 6, 6, 6, 6, 5, 5, 4, 4, 5, 5, 3, and 3); Kruskal-Wallis test. (I) Number of colonies grown from mPB HSPCs treated as indicated (n = 5, 6, 4, 4, 4, and 3); Kruskal-Wallis test; mean ± SEM. (J) Percentage of SA-β-Gal⁺ cells in the progeny of edited mPB HSPCs at indicated conditions (n = 10, 10, 6, 6, 9, 9, 4, 4, 8, 8, 3, and 3); Kruskal-Wallis test. p values: ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001. If not indicated, results are not significant (ns). Unless otherwise specified, lines indicate median values, and HSPCs were collected at 4 days post-GE (C–J) and at day 1 and 4 post-GE (B). (−): HS/AAV6 condition.

Figure 6

Mitigation of senescence enhances in vivo polyclonal hematopoietic reconstitution by HDR-edited HSPCs (A and B) Percentage of hCD45⁺ (A) or GFP⁺ hCD45⁺ (B) cells in PB of high-cell-dose transplanted mice at day 1 post-GE with CB-derived HSPCs treated as indicated (n = 17, 18, 27, and 14). LME followed by post hoc analysis; results are shown for the 18 weeks post-engraftment time point; mean ± SEM. (C–E) Percentage of hCD45⁺ (C); GFP⁺ hCD45⁺ (D); or GFP⁺ HSPCs (CD34⁺), B cell (CD19⁺), and myeloid cells (CD13⁺) (E) in BM of transplanted mice from (A) (n = 16, 18, 27, and 14); Kruskall-Wallis test. (F) Percentage of SA-β-Gal⁺ cells in BM-derived CD34⁺ retrieved from mice transplanted with high cell doses (n = 5, 5, 6, 6, 12, 12, 4, and 4 from one independent experiment from C); Kruskal-Wallis test. (G) Number of colonies grown from BM-derived CD34⁺ HSPCs from (C) of mice treated as indicated 18 weeks post-transplantation (n = 13, 14, 20, and 14); Kruskal Wallis test; mean ± SEM. (H) Percentage of hCD45⁺ cells in PB of mice transplanted at day 1 post-GE with mPB HSPCs treated as indicated (n = 6, 6, 6, 6, and 5). LME followed by post hoc analysis; results are shown for the 6 weeks post-engraftment time point; mean ± SEM. (I) Percentage of BM-derived hCD45⁺ cells in mice shown in (H) at 15 weeks post-transplantation. (n = 6, 6, 6, 6, and 5); Kruskal-Wallis test. (J) Percentage of hCD45⁺ GFP⁺ cells in PB of mice from (H). LME followed by post hoc analysis; results are shown for the 6- and 15-week post-transplantation time points; mean ± SEM. (K) Percentage of hCD45⁺GFP⁺ cells in BM of mice shown in (I) at 15 weeks post-transplantation; Kruskal-Wallis test. (L) Percentage of SA-β-Gal⁺ cells in PB of mice shown in (H). LME followed by post hoc analysis; results are shown for the 6 and 15 weeks post-transplantation time points; mean ± SEM. (M) Percentage of SA-β-Gal⁺ cells in BM of mice shown in (I) at 15 weeks post-transplantation; Kruskal-Wallis test. (N) Number of colonies grown from BM-derived CD34⁺ HSPCs of mice shown in (I) 15 weeks post-transplantation (n = 6, 6, 6, 6, and 5); Kruskal-Wallis test; mean ± SEM. (O) Number of unique indels in human BM-derived cells of mice transplanted with CB HSPCs from (A) (n = 13, 12, 12, and 12 from 3 independent in vivo experiments); Kruskal-Wallis test. (P) Number of dominant unique BARs in BM-derived cells of mice transplanted with CB HSPCs in (A) (n = 19, 15, 19, and 12 from 4 independent in vivo experiments); Kruskal-Wallis test. p values: ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001. If not indicated, results are not significant (ns). Unless otherwise specified, lines indicate median values. (−): HS/AAV6 condition.

Figure 7

Anakinra treatment reduces the genotoxicity risk associated with the gene-editing procedure (A) Visualization of chromosomal translocation genome wide by CAST-Seq. (B and C) Copies of AAVS1 sequences per human genome 800 bp downstream to the target site in individual colonies generated by edited mPB HSPCs (B: n = 46, 93, 97, 93, and 85) or CB HSPCs (C: n = 48, 91, 92, 98, and 97) from 3 independent donors. LME, linear-mixed model. (D) Mutational rate of total variants in the indicated conditions from BM-derived CD34⁺ HSPCs of mice from one independent experiment from Figure S6I (n = 6, 6, 5, and 4); Kruskal-Wallis test. (E) Percentage of single-nucleotide variants (SNVs), deletions (DELs), and insertions (INSs) in BM-derived CD34⁺ HSPCs retrieved from mice in (D). n = 6, 6, 5, and 4; Kruskal-Wallis test; mean ± SEM. (F) Relative proportion (percentages) of SNV types in mice in (E). n = 6, 6, 5, and 4; Kruskal-Wallis test; mean ± SEM. (G) Percentage of variants with high (HIGH), moderate (MODERATE), or low (LOW) impact found in BM-derived CD34⁺ HSPCs retrieved from mice in (D) (n = 6, 6, 5, and 4); Kruskal-Wallis test; mean ± SEM. (H and I) Quantification and Circos plots representing variants in clonal hematopoiesis (CH)-associated genes (H) or cancer-associated genes (I) (n = 6, 6, 5, and 4); Kruskal-Wallis test. p values: ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001. If not indicated, results are not significant (ns). Unless otherwise specified, lines indicate median values. (−): HS/AAV6 condition.