Genome editing using CRISPR-Cas9 to create the HPFH genotype in HSPCs: An approach for treating sickle cell disease and β-thalassemia

Abstract

Hereditary persistence of fetal hemoglobin (HPFH) is a condition in some individuals who have a high level of fetal hemoglobin throughout life. Individuals with compound heterozygous β-thalassemia or sickle cell disease (SCD) and HPFH have milder clinical manifestations. Using RNA-guided clustered regularly interspaced short palindromic repeats-associated Cas9 (CRISPR-Cas9) genome-editing technology, we deleted, in normal hematopoietic stem and progenitor cells (HSPCs), 13 kb of the β-globin locus to mimic the naturally occurring Sicilian HPFH mutation. The efficiency of targeting deletion reached 31% in cells with the delivery of both upstream and downstream breakpoint guide RNA (gRNA)-guided Staphylococcus aureus Cas9 nuclease (SaCas9). The erythroid colonies differentiated from HSPCs with HPFH deletion showed significantly higher γ-globin gene expression compared with the colonies without deletion. By T7 endonuclease 1 assay, we did not detect any off-target effects in the colonies with deletion. We propose that this strategy of using nonhomologous end joining (NHEJ) to modify the genome may provide an efficient approach toward the development of a safe autologous transplantation for patients with homozygous β-thalassemia and SCD.

Keywords: colony assay; deletion; engineered nucleases; erythroid differentiation; fetal hemoglobin.

Fig. S1.

Deletion type of hereditary persistence of fetal hemoglobin (HPFH). (A) Schematic representation of genomic deletion in common deletion types of HPFH and their corresponding fetal hemoglobin (HbF) level according to published data. The structure of the β-globin cluster. The corresponding genes from γ-globin to the 3′ DNase I hypersensitive site (3′HS-1) are shown on the Top panel. The 3′ break point in HPFH1 to HPFH 4 is beyond the 3′HS-1 site. The corresponding HbF level to each heterozygote deletion is shown on the right. (B) Scheme of normal β-globin locus (Upper) and naturally occurring Sicilian HPFH-5 (Lower). The 12.9-kb deletion from the 5′ δ-globin gene to 3′ of the β-globin gene abolishes the BCL11A binding sites located on the γδ-intergenic region and brings the 3′ β-enhancer and 3′ HS-1 site closer to the γ-globin gene.

Fig. 1.

Targeting clinically relevant HPFH5 deletions by S. aureus CRISPR-Cas9 in human cells. (A) Schematic of gRNAs targeting deletions in the β-globin locus to mimic the naturally occurring Sicilian HPHF deletion. The gRNAs were chosen from 4.5 kb upstream of HBD and 1.13 kb downstream of HBB. Black downward arrowheads, putative BCL11A binding sites. Blue arrowheads and red arrowheads, forward and reverse primers flanking each gRNA-targeting sequence used for amplifying fragments for the T7 endonuclease 1 assay (T7E1 assay). The 5′ break point gRNAs are labeled with “L”; the 3′ break point gRNAs are labeled with “R”; the “13kb” represents the length of the HPFH-5 deletion depicted in Fig. S1A. (B) Representatives of the T7E1 assay of gRNA targeting the β-globin locus in 293T cells. The percentage of indels is indicated under each gRNA. The size of each amplification and the size of cleavage product by T7 assay are listed in Table S3. (C) Dual gRNA-mediated deletion examined by PCR using the primer set of the most left and the most right ones depicted in A. The sizes of each PCR product are listed in Table S3. (D) Representative Sanger-sequencing results of two DSB junction areas from subcloning of the PCR product of L2/R2 transfectants aligned with WT junction sequences. Red arrows indicate the DSB cleavage sites (3 bp from PAM sequences NNGRRT). In between the two gRNA cleavage sites, there is an ∼13-kb fragment that was deleted in all five clones examined. The red sequences in the 5′ junction and green sequences in the 3′ junction are represented as sequences that gRNA, L2, and R2 used for targeting.

Fig. 2.

Dual gRNA-mediated deletion in primary cells. (A) Detection of dual gRNA-mediated deletions in primary erythroblast (Primary Ery), iPSCs, and CD34⁺ HSPCs 72 h after transfection. Red arrowhead, WT; black arrow, deletion. NC, negative control (cells transfected with pX601 without gRNA); NT, nontransfected. (B) Modification of the Cas9 constructs for the enrichment of transfected cells. The 293T, transfected individually with Cas9 conjugated with GFP or mCherry, and HSPCs, cotransfected with both conjugated Cas9s, are shown by the green and red fluorescence signals. (Scale bars, 1 μm.)

Fig. 3.

Dual gRNAs mediated clinical relevant HPFH5 deletion in CD34⁺ HSPCs. (A) Representative images of burst-forming unit-erythroid (BFU-E); colony-forming unit-granulocyte and macrophage (CFU-GM); colony forming unit-granulocytes, erythroid, monocyte/macrophage, megakaryocyte (CFU-GEMM) in the semisolid medium culture. (B) Detection of deletion in colonies from semisolid cultured GFP⁺/mCherry⁺ colonies by PCR using the primer sets shown in the Upper line. The varying sizes of PCR products in the Lower panel indicate NHEJ in those colonies that have only a single double strand break (DSB) site. The clones positive for P1-P4 and P3-P4 amplifications indicate heterozygous deletion. The clone positive for P1-P4, but negative for P3-P4, indicates homozygous deletion. Non-deletion, no 13-kb deletion. (C) Shown is the T7E1 assay for detection of potential off-site targets that are listed in Table S2. L2 and R3 are the on-target controls. The faint bands in GRIK4 are not the expected sizes for corresponding gRNA-targeting cleavage products, indicating that these are derived from nonspecific cleavage of the PCR product.

Fig. S2.

Schematic representation of procedures used for modifying CD34⁺ HSPCs with HPFH deletion.

Fig. S3.

Detection of NHEJ by Sanger sequencing in the clones which shown same size of PCR product. (A) Sequence alignment of representative clones with WT PCR products. PCR products with same-size fragments amplified from the 3′ breakpoint of eight clones showed mixed signal when aligned with WT sequences. The red line indicates the targeting site. The asterisks indicate matched nucleotides. (B) Histogram of Sanger sequencing further confirmed that the mixed signal started from the cleavage site indicated by the red arrow.

Fig. 4.

Elevated γ-globin expression in erythroid cells differentiated from CD34⁺ HSPCs modified with HPFH5 deletion by real-time RT-PCR. (A) Comparison of β-globin mRNA expression normalized to α-globin expression in erythroid colonies with and without 13-kb deletion based on the result from the genomic DNA screen in Fig. 3B. (B) Comparison of γ-globin expression in erythroid colonies with heterozygous and homozygous deletion to the clones without deletion. (C) The ratio of γ-globin expression to combined β- and γ-globin expression in clones with and without deletions. All data represent the mean plus or minus SEM. Statistically significant differences are indicated as follows: **P < 0.01; ***P < 0.0001 as determined by the Student unpaired t test. For all figures: filled circles, clones without deletion (Non-del); filled squares, clones with heterozygous deletion Del (Het); filled triangle, clones with homozygous deletion Del (Hom). HBA, α-globin mRNA; HBB, β-globin mRNA; HBG, γ-globin mRNA.