A novel FOXP3 knockout-humanized mouse model for pre-clinical safety and efficacy evaluation of Treg-like cell products

Abstract

Forkhead box P3 (FOXP3) is an essential transcription factor for regulatory T cell (Treg) function. Defects in Tregs mediate many immune diseases including the monogenic autoimmune disease immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome (IPEX), which is caused by FOXP3 mutations. Treg cell products are a promising modality to induce allograft tolerance or reduce the use of immunosuppressive drugs to prevent rejection, as well as in the treatment of acquired autoimmune diseases. We have recently opened a phase I clinical trial for IPEX patients using autologous engineered Treg-like cells, CD4^LVFOXP3. To facilitate the pre-clinical studies, a novel humanized-mouse (hu-mouse) model was developed whereby immune-deficient mice were transplanted with human hematopoietic stem progenitor cells (HSPCs) in which the FOXP3 gene was knocked out (FOXP3KO) using CRISPR-Cas9. Mice transplanted with FOXP3KO HSPCs had impaired survival, developed lymphoproliferation 10-12 weeks post-transplant and T cell infiltration of the gut, resembling human IPEX. Strikingly, injection of CD4^LVFOXP3 into the FOXP3KO hu-mice restored in vivo regulatory functions, including control of lymphoproliferation and inhibition of T cell infiltration in the colon. This hu-mouse disease model can be reproducibly established and constitutes an ideal model to assess pre-clinical efficacy of human Treg cell investigational products.

Keywords: CRISPR-Cas9; FOXP3; Humanized mouse model; IPEX syndrome; Regulatory T cells.

Graphical abstract

Figure 1

FOXP3 locus can be efficiently targeted by multiple sgRNAs (A) sgRNAs against FOXP3 exon1 and exon3. TSS, transcription start site. (B) Screening of the sgRNAs against FOXP3 locus. (C) Representative image of multiple sgRNA FOXP3 KO (sgRNA4, 5, and 6) analyzed by ICE software. Significance determined by one-way ANOVA followed by Tukey’s multiple comparison tests. ∗p < 0.05. Data are representative of three independently repeated experiments.

Figure 2

In FOXP3 KO hu-mice morbidity and mortality (A) Frequency of INDEL were measured by ICE analysis of HSPCs (n = 6–7/conditions, mean $\pm$ SEM). (B) Schematic representation of the experimental setup. NSG, NOD.Cg-Prkdc scid Il2rg tm1Wjl/SzJ. (C) Survival curve and (D) body weight of FOXP3 KO hu-mice model comparing mice transplanted with FOXP3 WT HSPC (FOXP3WT), FOXP3 KO HSPC alone (FOXP3KO) (n = 16/group). Significance determined by one-way ANOVA followed by Tukey’s multiple comparison tests. ∗p < 0.05, ∗∗∗∗p < 0.0001. Data are representative of four independently repeated experiments.

Figure 3

CD4^LVFOXP3 cells can ameliorate CD4 dominant lymphoproliferation in FOXP3 KO hu-mice (A) Schematic representation of the experimental setup. (B) Survival curve and (C) body weight of FOXP3 KO hu-mice model comparing mice transplanted with FOXP3 KO HSPC alone (FOXP3KO) and FOXP3 KO HSPC treated by CD4^LVFOXP3 (FOXP3KO + CD4^LVFOXP3) (n = 16/group). (D) Representative flow cytometry dot plots. (E) Percentage of hCD45⁺/CD3⁺/CD4⁺/CD8⁺ cells in PB between weeks 8 and 16. (n = 12–16, mean $\pm$ SEM). Data are representative of four independently repeated experiments. All phenotypes were performed by FACS. Significance determined by one-way ANOVA followed by Tukey’s multiple comparison tests. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

Figure 4

CD4^LVFOXP3 cells can ameliorate imbalance between naive/memory CD4⁺ T cell populations in FOXP3 KO hu-mice (A) Representative flow cytometry dot plots. (B) Percentage of hCD45⁺/CD3⁺/CD4⁺/CD8⁺ cells in spleen in week 16 (n = 12–16, Median). (C) Naive/memory phenotype of CD4⁺ T cells in spleen in week 16 (n = 12–16, median). Data are representative of four independently repeated experiments. All phenotypes were performed by FACS. Significance determined by one-way ANOVA followed by Tukey’s multiple comparison tests. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

Figure 5

FOXP3 expression was abrogated in the Treg population in FOXP3 KO hu-mice (A) Representative flow cytometry dot plots. Frequency of CD25⁺CD127^– cells out of total CD4⁺ T cells (top columns) and FOXP3⁺ cells out of CD25⁺CD127^– cells (bottom columns). (B) Percentage of Tregs (CD4⁺CD25⁺CD127^–), Percentage and MFI of FOXP3⁺ Tregs (FOXP3⁺/CD4⁺CD25⁺CD127^–) in spleen in week 16 (n = 12–16, median). Data are representative of four independently repeated experiments. All phenotypes were performed by FACS. Significance determined by one-way ANOVA followed by Tukey’s multiple comparison tests. ∗∗p < 0.01, ∗∗∗∗p < 0.0001.

Figure 6

CD4^LVFOXP3 cells can ameliorate gut infiltration of CD3⁺/CD4⁺ T cell populations in FOXP3 KO hu-mice (A) Representative immunohistochemistry (IHC) analysis of colon histology (CD3⁺/CD4⁺). (B) The quantification of CD3⁺/CD4⁺ T cells detected in colon in week 16 (n = 12–16, median). Data are representative of four independently repeated experiments. Significance determined by one-way ANOVA followed by Tukey’s multiple comparison tests. ∗∗∗p < 0.001.

Figure 7

CD4^LVFOXP3 cells can be reduced by basiliximab in vivo (A) Schematic representation of the experimental setup. Percentage of human (h)CD45⁺/CD3⁺/CD4⁺/CD8⁺/NGFR⁺ cells in (B) PB and (C) spleen (n = 6, mean + SEM). (D) Multilineage engraftment of hu-mice injected with CD4^LVFOXP3 cells with basiliximab (n = 6, mean + SEM). Data are representative of two independently repeated experiments. All phenotypes were performed by FACS. Significance determined by one-way ANOVA followed by Tukey’s multiple comparison tests. ∗p < 0.05, ∗∗p < 0.01. Bacchetta and colleagues generated FOXP3 KO hu-mice transplanted with human HSPCs in which FOXP3 gene was knocked out using CRISPR-Cas9. Mice transplanted with FOXP3KO-HSPCs developed lymphoproliferation and T cell infiltration of the gut. Injection of CD4^LVFOXP3 engineered Tregs into FOXP3KO hu-mice restored in vivo regulatory functions.