Therapeutic gene editing of T cells to correct CTLA-4 insufficiency

Abstract

Heterozygous mutations in CTLA-4 result in an inborn error of immunity with an autoimmune and frequently severe clinical phenotype. Autologous T cell gene therapy may offer a cure without the immunological complications of allogeneic hematopoietic stem cell transplantation. Here, we designed a homology-directed repair (HDR) gene editing strategy that inserts the CTLA-4 cDNA into the first intron of the CTLA-4 genomic locus in primary human T cells. This resulted in regulated expression of CTLA-4 in CD4⁺ T cells, and functional studies demonstrated CD80 and CD86 transendocytosis. Gene editing of T cells isolated from three patients with CTLA-4 insufficiency also restored CTLA-4 protein expression and rescued transendocytosis of CD80 and CD86 in vitro. Last, gene-corrected T cells from CTLA-4^-/- mice engrafted and prevented lymphoproliferation in an in vivo murine model of CTLA-4 insufficiency. These results demonstrate the feasibility of a therapeutic approach using T cell gene therapy for CTLA-4 insufficiency.

Figure 1. Targeting the CTLA4 locus with CRISPR/Cas9 and repair of a point mutation

(A) Schematic representation of the mutational landscape of CTLA4 insufficiency. Mutations are colour coded by citation (key, bottom right). (B) Schematic representation of HDR donor 1 (P2A-GFP-WPRE-PolyA). (C) Average HDR rate (n=3, percentage GFP+ in cells from separate healthy donors) (Mean 55.83 SD 1.626). (D) Median fluorescent intensity (MFI) of CTLA4 from 5 separate healthy controls and 3 separate samples from a single patient with p.T124P c.370A>C unedited or edited. A significant difference was seen in CTLA4 MFO between WT and p. T124P het mutant cells (p=0.036, Man Whitney test). After editing the difference in MFI was no longer significant. (E) Flow cytometry plot demonstrating surface CTLA4 expression in cells from a healthy individual in an unedited control (left), edited with a gRNA specific for the wild type CTLA4 sequence (gRNA 1) with resulting knock down of CTLA4 protein (centre) and a on edited with a gRNA specific for the p.T124P c.371A>C (gRNA 2) (right) demonstrating minimal activity on the wild-type sequence.

Figure 2. Universal editing strategies

(A) Schematic representation of the editing strategy using gRNA 3 (exon 1) and donor 3 (HA-CTLA4-P2A-GFP-WPRE-HA). (B) Representative flow cytometry plots of the editing strategy shown in (A) demonstrating a non-edited control (left), gRNA only control with resulting knock down of CTLA4 (centre) and HDR mediated by the CTLA4 cDNA-P2A-GFP-WPRE AAV6 donor (48.8% GFP positive cells). (C) Schematic representation of the the intronic editing strategy (donor 4 HA-splice acceptor-CTLA4 exons 2, 3, 4-P2A-GFP-WPRE-HA) (D) Representative flow cytometry plots plots showing CTLA4 expression and GFP expression (HDR) in cells edited with the gRNA 3/Cas9 RNP alone (intron 1) (centre left), and then with transduction of donor 4 (WPRE) and donor 5 (3’UTR) (far right). (E) Mean HDR rate (n=3, percentage GFP+ in cells from separate healthy donors). Exon 1 approach (gRNA 3 + donor 3) mean=42.47% GFP+, SD 8.13, intronic WPRE donor (gRNA 4 + donor 4A) mean=64.63% GFP+, SD 3.06, Intron 3’UTR donor (gRNA 4, donor 4B) mean=38.13% GFP+, SD 2.70.

Figure 3. Functional characteristics of edited T cells

(A) Schematic representation of transendocytosis assay. Cells (edited or unedited controls) are incubated in a 5:1 ratio with DG75 cells expressing either fluorescent labelled (mCherry) CD80 or CD86. Uptake of ligand can then be assessed by flow cytometry (mCherry uptake into T cells). (B) Representative FACS plots demonstrating TE of mCherry-bound CD80 and CD86 (top right quadrant each plot) in healthy control CD4+ T cells (top row), CD4+ cells that have undergone knock out of CTLA4 (upper middle row) and CD4+ cells that have undergone repair using the different editing strategies (gated on edited GFP+ cells). DG75 cells that do not express either ligand are used as a negative control (left column). (C) mCherry uptake relative to the unedited control with the different universal editing strategies in healthy CD4+ cells (CTLA4 KO mean = 0.69, SD = 0.155, N=3, gRNA 3 + donor 3 mean = 0.79, SD = 0.13, n=3, gRNA 4 +donor 4 mean = 0.94, SD = 0.08, n=3). (D) Graphs showing increase in CTLA4 and FOXP3 MFI in unedited cells (blue) and edited cells (red) when costimulated with CD80 and CD86.

Figure 4. Restoration of CTLA4 expression and function in CD4+ cells from patients with CTLA4 haploinsufficiency

(A) HDR rates (% cells GFP+) in edited CD4+ cells from patients with CTLA4 haploinsufficiency resulting from three different mutations. (B) Graph showing restoration of surface CTLA4 in heterozygote mutant CD4+ cells following editing with gRNA 4 and HDR donor 4. %CTLA4 positive relative to a healthy control assessed at the same time are shown. GFP+ edited cells are compared to mock edited cells. (C, D, E) Overnight TE assays gated on CD3+ CD4+ FOXP3+ cells in healthy control (top rows), patient cells with three different mutations (middle rows) and patient cells corrected with the intronic editing strategy (bottom rows). (F) Graph showing the increase in ligand acquisition (% CD4+ FOXP3+ T cells mCherry positive) in patient cells after editing relative to healthy control.

Figure 5. Kinetics of CTLA4 surface expression in resting and activated states

(A) Representative time course of CTLA4 expression (MFI histogram) on healthy control CD4+ cells (top), c.371A>C heterozygous mutant cells (middle) and c.371A>C heterozygous mutant cells edited with gRNA 4/Cas9 RNP and donor 4. (B) Percentage and MFI (C) of CTLA4 surface expression over time (n=3 for all conditions).

Figure 6. Assessment of T cell GT for CTLA4 insufficiency using an in vivo murine model

(A) FACS plots demonstrating typical editing efficiencies achieved in murine CTLA4^-/- T cells (GFP+ cells, upper plots) with restoration of CTLA4 expression in both the FOXP3+ and FOXP3- compartments (lower plots). (B) FACS plots post sort demonstrating % GFP+ in the sorted edited cells (upper plots) and CTLA4 expression in FOXP3- and FOXP3+ cells in the two sorted populations (GFP- left lower plot, GFP+ left upper plot).(C) Serial tail vein bleeds demonstrating persistence and stability of the GFP+ population after adoptive transfer. (D) Lymph node and spleen size in mice that received wild type T cells (left) mock edited and GFP-cells (middle) and GFP+ enriched edited cells (right). (E) Lymph node (left) and spleen (right) cell counts in mice that received wild-type T cells, mock edited, GFP-and GFP+ T cells. (F) Number of Tconv per μg of heart tissue in mice that received WT, mock edited, GFP- and GFP+ T cells. (G) Representative FACS plots (left) and collated data (right) showing CTLA4 expression in lymph node Treg and Tconv cells. (H) Representative FACS plots (left) and collated data (right) showing CTLA4 expression in spleen Treg and Tconv cells. Data collated from 2 independent experiments; n=4-5. One-way ANOVA; mean ± SD are shown; ****, p < 0.0001; ***, p < 0.001; **, p < 0.01; *, p < 0.05; ns, not significant.