Research ArticleIMMUNOLOGY

CRISPR-based gene editing enables FOXP3 gene repair in IPEX patient cells

See allHide authors and affiliations

Science Advances  06 May 2020:
Vol. 6, no. 19, eaaz0571
DOI: 10.1126/sciadv.aaz0571
  • Fig. 1 The FOXP3 locus is precisely targeted using the CRISPR system in primary HSPCs and T cells.

    (A) Schematic representation of CRISPR-based editing of the FOXP3 gene showing the CRISPR cut site in first coding exon, E1 (exons depicted by gray boxes separated by lines representing introns; the first coding exon, E1, is preceded by the noncoding exon E-1 and the enhancer with TSDR). A zoomed-in view of the sgRNA binding site relative to the start codon, PAM site, and cleavage site is shown. Homology donor depicted below with arms of homology, codon divergent FOXP3 cDNA, BGH polyadenylation (pA) signal included to terminate the FOXP3 transcript, truncated NGFR (tNGFR) marker gene under the Phosphoglycerate Kinase (PGK) promoter to drive marker expression independent of FOXP3 expression, and a second pA. (B) Screening of sgRNAs targeting the first coding exon of the FOXP3 gene. Plasmids encoding WT Cas9 or nickase variant of Cas9 (paired sgRNAs) and FOXP3 sgRNAs nucleofected into K562 cell lines. CRISPR efficiency measured by TIDE analysis to detect insertion deletion (indel) mutations created by nonhomologous end joining (NHEJ)–mediated DNA repair. (C) Experimental method for editing of HSPCs and T cells with functional readouts listed. (D) CRISPR cutting efficiency in CD34+ HSPCs and CD4+ T cells quantified by TIDE analysis for the detection of indel mutations created by the NHEJ repair pathway.

  • Fig. 2 CRISPR combined with a rAAV6 homology donor enables precise HDR-mediated FOXP3 cDNA transgene insertion into the endogenous locus.

    (A) Editing observed at the DNA level by an in-out PCR strategy that uses a primer inside the inserted divergent cDNA construct and a second primer outside of the 5′ arm of homology. Control band represents unmodified region in the FOXP3 gene as a positive control for the presence of genomic DNA. PCR using in-out primers resulted in band only present in samples in which the cDNA was inserted (FL cDNA) and not in FOXP3 knockout (KO) or mock-treated samples. After the ladder, the first four lanes represent CD4+ T cells, and the last three lanes represent HSPCs. (B) Rates of HDR-mediated FOXP3 editing detected by flow cytometry for tNGFR marker gene expression in primary human CD4+CD25++ Tregs, MT-2 Treg cell line, primary CD4+CD25 Teff cells, and CD34+ HSPCs. Editing rates with standard Sp Cas9 (gray) and high-fidelity (HiFi) Cas9 (red) are shown for comparison. Dots represent cells from individual donors. (C) Representative flow cytometry plots showing tNGFR marker gene expression in edited cord blood–derived CD34+ HSPCs. Negative control: Mock-treated cells nucleofected with phosphate-buffered saline in place of CRISPR and transduced with rAAV6-FOXP3 donor. Edited cells enriched using tNGFR selection and purity are shown by flow cytometry. SSC-A, side scatter area. (D) Venn diagram showing overlap in predicted off-target sites identified by COSMID in silico prediction and GUIDE-seq DSB capture. All predicted sites were tested by NGS in edited CD34+ HSPCs derived from cord blood (CB) and validated by NGS in edited bone marrow (BM)–derived HSPCs. (E) The four off-target (OT) sites validated by NGS shown with highlighted mismatches in off-target site versus FOXP3 sgRNA target sequence, closest gene name, gene region, and average frequency of indel mutations detected by NGS (mean, n = 3).

  • Fig. 3 FOXP3 edited Tregs express FOXP3 protein and display characteristic in vitro phenotype and function.

    (A) Quantification of FOXP3 protein expression in the MT-2 Treg cell line by flow cytometry, showing median fluorescence intensity (MFI) values for one representative experiment out of three performed. (B) FOXP3 homology donor constructs designed to improve FOXP3 protein expression, including further codon optimization (FOXP3FLco) and addition of a WPRE element (FOXP3FLcoW). (C) Flow cytometry for FOXP3 in MT-2 cells comparing the different donor constructs. (D) FOXP3 expression in primary Tregs by flow cytometry and quantification represented as ratio of MFI in treated cells versus WT unmodified Tregs [mean ± SD, one-way analysis of variance (ANOVA), Tukey’s multiple comparisons]. Significance comparing FOXP3FL (n = 2, **P < 0.01) or FOXP3FLcoW (n = 6, **P < 0.01) to WT mock treated (n = 2). (E) Expression of signature Treg marker proteins on primary Tregs by flow cytometry and corresponding quantification (mean ± SD, n = 2 to 4). (F) Suppression assay showing unstimulated responders (R), stimulated responders (R*) alone or in coculture with Tregs, and calculated percent suppression. Responders stained with proliferation dye, carboxyfluorescein diacetate succinimidyl ester (CFSE). Average suppressive potential over several independent experiments is quantified (right). Significance comparing the WT Tregs with following Tregs: IPEX, FOXP3 KO, FOXP3FL edited, FOXP3FLco edited, and FOXP3FLcoW edited (mean ± SD, one-way ANOVA, Tukey’s multiple comparisons test, *P < 0.05; ****P < 0.0001; ns, not significant).

  • Fig. 4 FOXP3 editing of Teff cells preserves physiological regulation of FOXP3 expression and in vitro function.

    (A) Flow cytometry time course showing kinetics of FOXP3 expression in nonactivated Teff cells and activated Teff cells on subsequent days after activation [day 3 (d3), d6, and d14], comparing WT unmodified, WT mock-treated, and FOXP3FLcoW edited Teff cells. FOXP3 expression quantified in Teff cells over time course of activation, showing average MFI (mean ± SD, n = 3). (B) Cytokine production in WT and FOXP3 gene edited Teff cells determined by enzyme-linked immunosorbent assay (ELISA). Supernatants collected at 24 hours (IL-2) and 48 hours (IFN-γ and IL-17) after activation with anti-CD3/28 (mean ± SD, n = 3). (C) Teff cell proliferation in response to activation measured by the proliferation assay. Flow cytometry plots of carboxyfluorescein diacetate succinimidyl ester (CFSE) dye–stained Teff cells with progressive dilution of dye as proliferation progresses from nonactivated to day 2 and day 3 after activation with anti-CD3/28 Dynabeads are shown. Comparison of proliferation rates in response to activation with a bead:cell ratio of 1:100 and 1:25. Quantification of average proliferative response of Teff cells from proliferation assay at day 3 is shown to the right, comparing different doses of activation beads (mean ± SD, n = 3).

  • Fig. 5 CRISPR-based editing enables FOXP3 gene correction in IPEX patient cells.

    (A) Schematic of FOXP3 gene highlighting mutations of patients involved in this study. (B) Editing of IPEX and HD T cells observed at the DNA level by in-out PCR strategy. Forward primer is in the tNGFR cassette, and the reverse primer is in the FOXP3 gene locus outside the 3′ arm of homology. Positive and negative fractions after tNGFR enrichment (+/−) analyzed by PCR. (C) Flow cytometry plots of tNGFR staining. (D) Expression of FOXP3 mRNA demonstrated by RT-PCR in activated IPEX and HD T cells (schematic mRNA isoforms are shown to the right of corresponding PCR product bands). After ladder, first lane represents WT cells with two naturally occurring alternatively spliced isoforms, FL and dE2. The second lane shows aberrant skipping of E1 in IPEX pt24 mRNA that results in truncated transcripts. The third lane shows edited IPEX cells with restoration of the FL mRNA. (E) Sanger sequencing showing c.1150G>A mutation in the mRNA of pt37, resulting in an Ala>Thr change, and CRISPR-based insertion of divergent FOXP3 cDNA restoring correct amino acid sequence. (F) Proliferation of Teff cells in response to activation measured by the proliferation assay, comparing HD WT cells and IPEX pt78 cells. Quantification of average proliferation response of Teff cells from proliferation assay at day 3. (G) Functional testing of gene edited IPEX Tregs using the in vitro suppression assay. Flow cytometry plots of CFSE-stained Teff responder cells (R) cultured with or without Tregs. Calculated suppressive potential shows diminished suppressive function of IPEX pt. 64 Tregs, which was partially restored by FOXP3 gene editing.

  • Fig. 6 FOXP3 edited HSPCs undergo multilineage hematopoietic differentiation and engraftment in vitro and in vivo.

    (A) Differentiation potential of edited HSPCs tested by the in vitro CFU assay. Four resulting hematopoietic progenitor colony types: CFU-E (mature erythroid progenitors), CFU-GEMM (granulocyte, erythrocyte, macrophage, and megakaryocyte), BFU-E (primitive erythroid progenitors), and CFU-GM (granulocyte and macrophage progenitors). Representative images of colonies from the CFU assay, showing similar morphology (×10 magnification). (B) Experimental timeline of hu-mouse study using NSG-SGM3 mice. (C) Human engraftment kinetics in the peripheral blood of hu-mice at corresponding weeks after injection. Engraftment was measured by flow cytometry for hCD45 marker on human cells, and frequency was quantified relative to the total of human (hCD45+) and mouse (mCD45+) cells (mean ± SD). (D) Representative flow cytometry plots of engrafted human hematopoietic subsets in the bone marrow (left) and spleen (right) of hu-mice at 14 weeks after injection. Populations gated out of human cells (hCD45+). FOXP3 edited samples were divided into tNGFR+ and tNGFR gates for comparability. (E) Quantification of human hematopoietic lineages by flow cytometry with each symbol representing a single mouse (mean ± SD). In spleen, the CD8+, CD4+, and CD4+CD8+ double-positive (DP) populations were gated out of CD3+ T cells. The CD25+FOXP3+, naïve CD45RA+, and memory CD45RA populations were gated out of CD4+ single-positive T cell subset (*P < 0.5, **P < 0.01, ***P < 0.001).

Supplementary Materials

  • Supplementary Materials

    CRISPR-based gene editing enables FOXP3 gene repair in IPEX patient cells

    M. Goodwin, E. Lee, U. Lakshmanan, S. Shipp, L. Froessl, F. Barzaghi, L. Passerini, M. Narula, A. Sheikali, C. M. Lee, G. Bao, C. S. Bauer, H. K. Miller, M. Garcia-Lloret, M. J. Butte, A. Bertaina, A. Shah, M. Pavel-Dinu, A. Hendel, M. Porteus, M. G. Roncarolo, R. Bacchetta

    Download Supplement

    The PDF file includes:

    • Figs. S1 to S7
    • Tables S1 to S4

    Other Supplementary Material for this manuscript includes the following:

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article