Zinc-finger nuclease-mediated correction of α-thalassemia in iPS cells

Abstract

Induced pluripotent stem (iPS) cell technology holds vast promises for a cure to the hemoglobinopathies. Constructs and methods to safely insert therapeutic genes to correct the genetic defect need to be developed. Site-specific insertion is a very attractive method for gene therapy because the risks of insertional mutagenesis are eliminated provided that a "safe harbor" is identified, and because a single set of validated constructs can be used to correct a large variety of mutations simplifying eventual clinical use. We report here the correction of α-thalassemia major hydrops fetalis in transgene-free iPS cells using zinc finger-mediated insertion of a globin transgene in the AAVS1 site on human chromosome 19. Homozygous insertion of the best of the 4 constructs tested led to complete correction of globin chain imbalance in erythroid cells differentiated from the corrected iPS cells.

Figure 1

Reprogramming of α-thalassemia fibroblasts and characterization of iPSCs. (A) Phase contrast micrographs illustrating the morphology of fibroblasts (left panel) and iPSCs (middle panel). iPSCs had normal karyotype in culture (right panel). (B) FACS analysis showing that iPSCs express TRA-1-60, TRA-1-80, SSEA-4, and SSEA-3, 4 typical hESCs and iPSCs surface markers. (C) PCR analysis showing loss of the episomal vectors used to reprogram patient-specific fibroblasts after 8 passages in culture. ACTB are control primers that detect a small genomic DNA fragment of the β-actin gene.

Figure 2

Differentiation of reprogrammed α-thalassemia fibroblasts. (A) H&E staining demonstrating that iPSCs derived from α-thalassemia fibroblasts with transgene-free method can form teratoma in NSG mice, and generate cells from the 3 germ layers (bar = 50 μm). (B) HPLC analysis of globin expression and morphology of the erythroblasts obtained after differentiation of α⁰-thal-iPSCs and control H1 hESCs. (Top panels) Basophilic erythroblasts obtained after 14 days of coculture on FH-B-hTERT and 14 days of liquid culture. (Bottom panels) Orthochromatic erythroblasts obtained after an additional 10 days of liquid culture. (Left panels) HPLC profiles and Giemsa staining of cells obtained after differentiation of α⁰-thal-iPSCs. (Right panels) Same as left panels but for control H1 hESCs. α⁰-thal-iPSCs do not express any α-globin chains. Zeta-globin chains are silenced between the 14th and the 24th day of liquid culture.

Figure 3

ZFN-mediated integration of α-globin expression constructs in AAVS1 site. (A) Integration strategy. Four different α-globin constructs were tested. Therapeutic cassettes were inserted between exons 1 and 2 of the PPP1R12C by cotransfection of 2 ZFN-containing plasmids and of a targeting construct containing short homologies to part of intron 1 flanking the therapeutic transgenes. Successful targeting leads to expression of the puromycin gene under the control of the PPP1R12C promoter because of the presence of the splice acceptor (SA). (B) Location of the primer sets for demonstration of ZFN-mediated integration of α-globin constructs after selection of puromycin-resistant colonies. P1 hybridizes to a region just 5′ of the left homology arm. P2 hybridizes to the puromycin-resistance genes. P3 hybridizes to the 3′ part of the α-globin gene; P4 hybridizes to a region just 3′ of the right homology arm. P5 hybridizes to the 5′ region of the LCR. P1/P2, P3/P4, and P5/P6 only yield a PCR product if the transgenes are specifically inserted at AAVS1. P1/P4 detects the unmodified PPP1R12C gene. (C) PCR results demonstrating insertions of α-globin constructs at AAVS1. The PPP1R21C (P1/P4) PCR amplification can be used to identify heterozygous or homozygous insertions. α1-Hom indicates homozygous transgene in same orientation as PPP1R12C, driven by α-globin promoter; β1-Het, heterozygous transgene in same orientation as PPP1R12C, driven by β-globin promoter; β1-Hom, homozygous transgene in same orientation as PPP1R12C, driven by β-globin promoter; and β2-Het and β2-Hom, same as above but transgene is in opposite orientation. (D) Expression of PPP1R12C after insertion of the therapeutic transgenes. Heterozygous clones express PPP1R12C at approximately 50% the level of the unmodified locus. Homozygous clones do not express any detectable levels of PPP1R12C (n = 3). These results confirm the results of the analysis in panel C. (E) Histograms illustrating the number of copies of the corrective α-globin construct inserted in the genome. Q-PCR analyses were performed to compare the number of copies of α and β-globin present in the genome after. Y-axis = 2 × 2^{(Ctα-globin − Ctβ-globin)}. Together with the Southern blot analysis, these results demonstrate that these clones did not harbor any off-target integrations.

Figure 4

Erythroid cell differentiation of corrected α⁰-thal-iPSCs. (A) Chromatograms illustrating the results of HPLC analyses of globin chain expression performed after 14 days (14 FHB 14LC) or 24 days (14FHB 24 LC) of liquid culture of corrected iPSCs differentiated into erythroid cells. The 3 chromatograms on the left illustrate the globin expression pattern in orthochromatic erythroblasts obtained from culture of CD34⁺ cells from cord blood, fetal liver, and peripheral blood from a transfusion dependent adult with α⁰-thalassemia. β1-Het indicates heterozygous transgene integrated in same orientation as PPP1R12C, driven by β-globin promoter; β1-Hom, homozygous transgene in same orientation as PPP1R12C, driven by β-globin promoter; α1-Hom, homozygous transgene in same orientation as PPP1R12C, driven by α-globin promoter. Almost complete correction of chain imbalance in mature erythroid cells was obtained when homozygous transgenes driven by the α-globin promoter were inserted at AAVS1. The β-globin promoter was less effective than the α-globin promoter. Orientation of the transgene had no major effect on expression. (B) Q-RT-PCR analysis of globin expression in hESCs and iPSCs after differentiation into basophilic erythroblasts (14 days of liquid culture). The y-axis indicates the fold-difference compared with GAPDH. (C) IEF electrophoresis on lysates of orthochromatic erythroblasts illustrating the hemoglobin tetramers expressed in controls, in H1 and in corrected iPSCs. Globin content of each tetramer was determined by HPLC after cutting-out each major band from the gel. Corrected iPSCs express Hb F and Hb Gower II in addition to embryonic globins. A vertical line has been inserted to indicate a repositioned gel lane.

Figure 5

Effect of insertion of therapeutic transgenes on neighboring genes at AAVS1. (A) Top panel shows a map of 10 neighboring genes on chromosome 19 around AAVS1 site. (B) Bar graph illustrates Q-RT-PCR analysis of 10 neighbor genes after differentiation of corrected α⁰-thal-iPSCs into basophilic erythroblasts (normalized to parental iPSCs). All iPSCs analyzed carried homozygous insertions at AAVS1. mRNA expression levels were calculated using the Delta (δ Ct) method using differentiated uncorrected α⁰-thal-iPSCs as the controls (n = 3). Four genes were activated. The smallest activation was obtained when the α-globin promoter construct was inserted in the same orientation as the PPP1R12C gene. (C) Same as above but data are expressed relative to GAPDH. Expression of the neighboring genes is low compared with GAPDH.