CRISPR/Cas9-mediated genome editing via postnatal administration of AAV vector cures haemophilia B mice

Abstract

Haemophilia B, a congenital haemorrhagic disease caused by mutations in coagulation factor IX gene (F9), is considered an appropriate target for genome editing technology. Here, we describe treatment strategies for haemophilia B mice using the clustered regularly interspaced short palindromic repeat (CRISPR)/Cas9 system. Administration of adeno-associated virus (AAV) 8 vector harbouring Staphylococcus aureus Cas9 (SaCas9) and single guide RNA (sgRNA) to wild-type adult mice induced a double-strand break (DSB) at the target site of F9 in hepatocytes, sufficiently developing haemophilia B. Mutation-specific gene editing by simultaneous induction of homology-directed repair (HDR) sufficiently increased FIX levels to correct the disease phenotype. Insertion of F9 cDNA into the intron more efficiently restored haemostasis via both processes of non-homologous end-joining (NHEJ) and HDR following DSB. Notably, these therapies also cured neonate mice with haemophilia, which cannot be achieved with conventional gene therapy with AAV vector. Ongoing haemophilia therapy targeting the antithrombin gene with antisense oligonucleotide could be replaced by SaCas9/sgRNA-expressing AAV8 vector. Our results suggest that CRISPR/Cas9-mediated genome editing using an AAV8 vector provides a flexible approach to induce DSB at target genes in hepatocytes and could be a good strategy for haemophilia gene therapy.

Figure 1

AAV vector-mediated generation of haemophilia B in adult wild-type mice. (a) Schematic diagram of a single AAV vector expressing SaCas9 and sgRNA. HCRhAATp, an enhancer element of the hepatic control region of the Apo E/C1 gene and the human anti-trypsin promoter. (b) AAV vector expressing SaCas9 and each sgRNA targeting F9 was intravenously injected into 7-week-old C57BL/6 J male mice, and plasma levels of FIX:C were measured at indicated times. Solid and dashed lines represent high (1 × 10¹² vector genome/body) and low dose treatments (3 × 10¹¹ vector genome/body), respectively. Values are mean ± SEM (n = 3–6). **P < 0.01, compared with pretreatment (post hoc Bonferroni test). (c) Cas9-mediated cleavage of F9 in the liver was assessed using the Surveyor^® nuclease assay. Control was liver DNA from non-treated C57BL/6 J mice. Red arrows represent a mutation. (d) Distribution of the AAV genome in each tissue at 8 weeks after the injection. Values are mean ± SEM (n = 6).

Figure 2

Phenotypic correction of haemophilia B mice by homologous directed repair (HDR). (a) Schematic representation of the targeting strategy. The AAV8 vector expressing SaCas9 and sgRNA targeting exon 8 of F9 (AAV8-SaCas9 (exon 8)) induces a DSB. Simultaneous administration of the AAV8 vector containing homologous donor sequence (AAV8-HDR Donor) enabled correction of the target sequence by HDR. (b) AAV8-SaCas9 (exon 8) and AAV8-HDR Donor were injected into 0- (intraperitoneal injection, n = 10), 14- (intravenous injection, n = 4) or 26- to 45-week-old (182 days) (intravenous injection, n = 5) haemophilia B (HB) mice. Plasma FIX:C levels at 4–8 weeks following intravenous injections were measured. Vector dose (AAV8-SaCas9 (exon 8)/AAV8-HDR Donor): 6 × 10¹⁰ vg/2 × 10¹¹ in 0-day-old; 2.4 × 10¹¹ vg/6 × 10¹¹ in 14-day-old; 1 × 10¹² vg/3 × 10¹² in 26- to 45-week-old mice. (c,d) Blood coagulation assessed by activated partial thromboplastin time (APTT) (c) and bleeding volume after tail clipping for 10 min (d) were measured in wild-type mice (WT), HB mice, and HB mice treated with HDR mechanism (HDR) (n = 3–8). Values are mean ± SEM. **P < 0.01, compared with HB mice (two-tailed Student’s t-test). (e) To analyse the frequencies of HDR in liver DNA, two-step PCR was performed. Liver DNA was obtained from HB mice treated at P0 (n = 7) or P14 (n = 3). PCR fragments were amplified not to contain the donor vector followed by nested PCR. The frequency of the sequence that underwent HDR in the nested PCR sample was quantified using next-generation sequencing. Values are mean ± SEM.

Figure 3

Efficient increase in plasma FIX:C in haemophilia B mice by both HDR and insertion at DSB. (a) Schematic representation of targeting strategy. AAV8-SaCas9-sgRNA3 for F9 intron (AAV8-SaCas9 (intron 1)), the AAV8 vector to induce a DSB in intron 1; AAV8-Targeting, the gene correction AAV vector; SA, human F9 intron 1 splice acceptor site; F9 exon 2–8, codon-optimized cDNA (exon 2–8) of mouse F9. (b,c) Haemophilia B male (HB) mice (4–8 weeks old) were treated with intravenous injection of AAV8-SaCas9 (intron 1) and AAV8-Targeting. (b) Plasma levels of FIX:C were measured at indicated times after vector administration. Red and blue lines represent high and low dose treatments, respectively. Black line represents treatment with donor vector only. (c) Activated partial thromboplastin time (APTT) was measured in wild-type C57BL/6 mice (WT), HB mice, and HB mice treated with high or low dose vector. Vector dose (AAV8-SaCas9 (intron 1)/AAV8-Targeting): 1 × 10¹² vector genome (vg)/3 × 10¹² vg in high dose, 0.3 × 10¹² vg/1 × 10¹² vg in low dose. Values are mean ± SEM (n = 3–6). **P < 0.01, compared with HB mice (two-tailed Student’s t-test). (d) AAV8-SaCas9 (intron 1) and AAV8-Targeting were injected into 0- (P0: intraperitoneal injection, n = 6), 7- (P7: intravenous injection, n = 6), 28- (P28: intravenous injection, n = 4), or 42-day-old (P42: intravenous injection, n = 4) HB mice. Plasma FIX:C levels were measured at 4–8 weeks following vector injection. Vector dose (AAV8-SaCas9 (intron 1)/AAV8-Targeting): 6 × 10¹⁰ vg/2 × 10¹¹ in 0-day-old; 1.4–2.2 × 10¹¹ vg/4.1–6.6 × 10¹¹ in 7-day-old; 1 × 10¹² vg/3 × 10¹² in 28- and 42-day-old mice. Values are mean ± SEM. **P < 0.01, compared with HB mice (two-tailed Student’s t-test). (e) AAV genome in the liver at 8–16 weeks after vector injection in HB mice treated at 28 days old (P28) or 0 days old (P0). Values are mean ± SEM (n = 3–4). **P < 0.01 (two-tailed Student’s t-test). (f) AAV8-SaCas9 (intron 1) and AAV8-Targeting were intraperitoneally injected into 0-day-old HB mice, and plasma levels of FIX:C were measured at indicated times after vector administration. Values are mean ± SEM (n = 3–6). (g) AAV8-SaCas9 (intron 1) and AAV8-Targeting (HDR + INS) or AAV8-Targeting without arm sequence (INS) were intraperitoneally injected into 0-day-old HB mice, and plasma levels of FIX:C were measured at 4–8 weeks following vector injection. Values are mean ± SEM (n = 5–8). P = 0.096 (two-tailed Student’s t-test).

Figure 4

Disruption of the AT gene, Serpinc1, restores bleeding phenotypes of haemophilia B mice. AAV8 vector expressing SaCas9 and sgRNA targeting exon 8 of Serpinc1 (sgRNA1 for Serpinc1) (AAV8-SaCas9 (AT)) was intravenously injected into 7–8-week-old haemophilia B (HB) male mice (1 × 10¹² vector genome/body). (a) Plasma levels of antithrombin activity (AT:C) were measured at indicated times. Values are mean ± SEM (n = 8). **P < 0.01, compared with pretreatment (post hoc Bonferroni test). (b) Cas9-mediated cleavage of the genome in the liver was confirmed using the Surveyor^® nuclease assay. Red arrows represent a mutation. (c,d) Fibrin formation in vivo after endothelial disruption in wild-type C57BL/6 mice (WT), HB mice, and HB mice treated with AAV8-SaCas9 (AT) was observed by intravital confocal microscopy (Nikon A1RNP; Nikon, Tokyo, Japan) at ×400 magnification. Scale bars, 20 µm. (c) Representative images at 5 min after laser irradiation. Green signal indicates fibrin formation at the site of endothelial disruption (white arrow). (d) Signal intensities of fibrin formation were quantified using NIS-Elements AR 3.2 (Nikon), and are expressed with arbitrary units. Data are representative of at least three experiments. (e) Thrombin generation in plasma obtained from HB mice treated without (HB) or with AAV8-SaCas9 (AT) (AT sgRNA) was assessed by cleavage of the fluorogenic substrate, and is expressed as arbitrary units. Values are mean ± SEM (n = 4). Thrombin generation in mouse plasma containing indicated concentrations of FIX:C (50%, 25%, 10%, 1%, and 0%) was assessed as a reference. **P < 0.01 (two-tailed Student’s t-test).