Gene correction for SCID-X1 in long-term hematopoietic stem cells

Abstract

Gene correction in human long-term hematopoietic stem cells (LT-HSCs) could be an effective therapy for monogenic diseases of the blood and immune system. Here we describe an approach for X-linked sSevere cCombined iImmunodeficiency (SCID-X1) using targeted integration of a cDNA into the endogenous start codon to functionally correct disease-causing mutations throughout the gene. Using a CRISPR-Cas9/AAV6 based strategy, we achieve up to 20% targeted integration frequencies in LT-HSCs. As measures of the lack of toxicity we observe no evidence of abnormal hematopoiesis following transplantation and no evidence of off-target mutations using a high-fidelity Cas9 as a ribonucleoprotein complex. We achieve high levels of targeting frequencies (median 45%) in CD34⁺ HSPCs from six SCID-X1 patients and demonstrate rescue of lymphopoietic defect in a patient derived HSPC population in vitro and in vivo. In sum, our study provides specificity, toxicity and efficacy data supportive of clinical development of genome editing to treat SCID-Xl.

Fig. 1

In vitro, medium scale genome targeting at IL2RG locus. a Diagram of genomic integration and correction outcomes. b Top: schematic of IL2RG corrective donors containing (+tNGFR) or not (–tNGFR) selectable marker. Bottom: IL2RG cDNA targeting frequencies of frozen mobilized peripheral blood CD34⁺HSPCs (white circles) or freshly purified cord blood male-derived CD34⁺HSCPs (red circles) derived from medium scale (1.0 × 10 ⁶) genome targeting and measured at day 4. Absolute targeting frequencies measured by ddPCR. Median: 23.2% (+tNGFR, n = 11 biological replicates), median 45% (–tNGFR, n = 13 biological replicates). c Single cell-based methylcellulose assay from mock targeted (nucleofected only) or IL2RG cDNA targeted (–tNGFR donor) CD34⁺HSPCs. Absolute number of clones are shown (n = 3 biological replicates). d Fraction of the total for each type of colony scored. e Gene correction outcome of SCID-X1 patient 2 derived CD34⁺HSPCs. Shown is the multi-lineage differentiation using OP9-idll1 in vitro system (n = 23 wells). No growth was derived from uncorrected CD34⁺cells. LT-HSPCs long-term hematopoietic stem cells, ST-HSC short term hematopoietic stem cells, MPP multi-potent progenitor, CMP common myeloid progenitor, LMPP lymphoid multi-potent progenitor, CLP common lymphoid progenitor, HSPCs hematopoietic stem and progenitor cells, ddPCR droplet digital digital droplet PCR. Mean ± s.e.m.; ns not specific (Welch’s t-test). Source data are available in the Source Data file

Fig. 2

Normal hematopoietic reconstitution from IL2RG cDNA targeted CD34⁺ HSPCs. a Timeline of primary (1°) and secondary (2°) human transplants into sub-lethally irradiated NSG mice. CD34⁺ HSPCs are derived from umbilical cord blood of healthy male donors. Adult mice transplanted intra-femoral (IF) with either WT CD34⁺HSPCs (white circles) or mock targeted (yellow circles) or RNP only (black circles) or un-selected IL2RG cDNA targeted (blue-black circles) HSPCs. Three – 4 days old NSG pups transplanted intra-hepatic (IH) with either mock or IL2RG targeted HSPCs. b Combined IF and IH human cells engraftment (hCD45⁺HLA A-B-C⁺) 16 weeks after 1° human transplant into indicated organs. c %IL2RG cDNA targeted HSPCs within human graft in indicated organs,  quantified by ddPCR. BM (n = 24 mice), SP (n = 24 mice), PB (n = 6 mice) (***p = 0.0008, one-way ANOVA). d Percent human engraftment in indicated organs as in (b) 16 weeks post 2° human CD34⁺ HSPCs transplant into adult NSG mice. *p-value SP-IH = 0.025, *p-value BM-IF = 0.043 (Welch’s t-test). e %IL2RG targeted HSPCs quantied by ddPCR 32 weeks after engraftment. Median shown. BM bone marrow, SP spleen, PB peripheral blood. Source data are available in the Source Data file

Fig. 3

Normal multi-lineage development from IL2RG cDNA targeted in the LT-HSC population. a Percent cellular composition of the lymphoid, myeloid and erythroid lineage derived from IH 1° human engraftment, shown in indicated organs and targeting conditions. CD3⁺ BM: **p = 0.0017, CD3+SP: **p = 0.007 (Welch’s t-test). b Same as (a) but IF transplant analysis. CD3⁺ SP: *p = 0.023, CD56⁺ BM: *p = 0.015 (Kruskal–Wallis test). c Percent cellular composition of the lymphoid, myeloid and erythroid lineage derived from secondary transplants. Data shown are combined IH and IF primary transplants. CD3⁺ BM: *p = 0.015, CD56⁺: *p = 0.025, CD19⁺ SP: ***p = 0.0002, CD14⁺ SP: *p = 0.0112, CD11c⁺ SP: ***p = 0.0004. LT long term. Error bars: mean ± s.e.m. Source data are available in the Source Data file

Fig. 4

In vivo rescue of SCID-X1 mutation. a Genomic mapping and description of SCID-X1 mutations. b Percent viability determined at indicated days pre- and post- targeting. Mock (nucleofected only), RNP (nucleofected with RNP only), RNP+AAV6 (nucleofected with RNP and transduced with AAV6-based IL2RG corrective donor). Shown is data for mobilized peripheral blood CD34⁺ HSPCs (n = 5). c Medium scale (1.0 × 10⁶ cells) ex vivo genome targeting frequencies of frozen mobilized peripheral blood SCID-X1, at day 2 (blue-black circles, n = 6). Arrow shows 45% genome targeting of SCID-X1 patient 2 derived CD34⁺ HSPCs. d Human cells engraftment analysis at week 17 after intra-hepatic (IH) delivery of IL2RG cDNA targeted (blue-black circles, n = 15) or mutant CD34⁺ HSPCs (gray circles, n = 4). e Percent cellular composition of the lymphoid, myeloid, and erythroid lineage derived from IL2RG corrected or mutant CD34⁺ HSPCs. CD3⁺: ****p < 0.0001, CD56⁺: *p = 0.0146, CD16⁺: **p = 0.0013, CD19⁺: **p = 0.0015, CD235a⁺: **p = 0.0022 (Welch’s t-test). RNP ribonuclearprotein. f Absolute numbers derived from (e). Source data are available in the Source Data file

Fig. 5

Evaluation of IL-2 receptor function in IL2RG cDNA targeted T cells. a Schematic of signaling (pSTAT5—bottom) and proliferation (CFSE—top) in vitro assays. b pSTAT5 assay derived FACS plots. Top: healthy male-derived T cells genome targeted with IL2RG cDNA tNGFR (KI) or with tNGFR+ only cassette integrated at the IL2RG endogenous locus (KO). In red are the percent of double positive IL2RG-tNGFR⁺pSTAT5⁺[4.42%/(4/42% + 3.18%)]×100. We compare 58.2% cells (IL2RG targeted T cells) with 58.7% (IL2RG from WT T cells), (n = 3 biological replicates). c Quantification of IL-2R signaling through phosphoSTAT5 pathway. d pSTAT5 MFI for WT, KI, and KO experiments from (b) p = 0.02, Welch’s t-test. WT T cells (gray circles, n = 6), IL2RG KI (blue circles, n = 3) and IL2RG KO (orange circles, n = 3). e Proliferation profile of CFSE labeled, TCR stimulated IL2RG cDNA tNGFR+ sorted or mock-targeted T cells. Mock-targeted T cells are WT T cells cultured for the same amount of time as the tNGFR+ targeted cells and have been nucleofected in the absence of RNP or absence of transduction with AAV6. Shown FACS analysis at days 2, 4, 6, and 8. pSTAT5 phosphorylated STAT5, CFSE carboxyfluorescein succinimidyl ester, KI knocked in, KO knocked out, tNGFR truncated nerve growth factor receptor, IL-2 interleukin 2. Source data are available in the Source Data file

Fig. 6

Genome specificity of IL2RG sgRNA guide. a Heat map of on-target INDEL frequencies quantied by NexGen-Seq at COSMID identified putative on-target locations from healthy CD34⁺ HSPCs. Levels of NHEJ induced by 20 nt IL2RG sgRNA and truncated 19 nt, 18 nt and 17 nt pre-complexed with WT Cas9 protein at 5:1 molar ratio. b Heat map as in (a) of on-target INDEL frequencies derived from 19 nt IL2RG sg-1 in the genome of CD34⁺ HSPCs SCID-X1 patient 1 derived cells. c Percent viability at day 4 of SCID-X1 patient-derived CD34⁺ HSPCs nucleofected with either wild-type (WT) or high-fidelity (HiFi) SpCas9 protein pre-complexed with either the 20 nt or the 19 nt IL2RG sg-1 (n = 1). d Percent INDELs measured by TIDE at day 4 in cells as in (c) using WT or HiFi Cas9 protein pre-complexed with the 20 nt IL2RG sg-1 (green bars) or 19 nt IL2RG sg-1 (blue bars). e Percent IL2RG cDNA targeting (% HR) as measured by ddPCR at day 4 in cells as in (c) generated by either WT or HiFi Cas9 protein pre-complexed with the 20 nt IL2RG sg-1 or (f) 19 nt IL2RG sg-1. Source data are available in the Source Data file