Mobilization-based chemotherapy-free engraftment of gene-edited human hematopoietic stem cells

Abstract

Hematopoietic stem/progenitor cell gene therapy (HSPC-GT) is proving successful to treat several genetic diseases. HSPCs are mobilized, harvested, genetically corrected ex vivo, and infused, after the administration of toxic myeloablative conditioning to deplete the bone marrow (BM) for the modified cells. We show that mobilizers create an opportunity for seamless engraftment of exogenous cells, which effectively outcompete those mobilized, to repopulate the depleted BM. The competitive advantage results from the rescue during ex vivo culture of a detrimental impact of mobilization on HSPCs and can be further enhanced by the transient overexpression of engraftment effectors exploiting optimized mRNA-based delivery. We show the therapeutic efficacy in a mouse model of hyper IgM syndrome and further developed it in human hematochimeric mice, showing its applicability and versatility when coupled with gene transfer and editing strategies. Overall, our findings provide a potentially valuable strategy paving the way to broader and safer use of HSPC-GT.

Keywords: CRISPR-Cas gene editing; RNA-based delivery; X-linked hyper-IgM syndrome; autologous stem cell transplantation; conditioning-free; gene transfer; hematopoietic stem cells; human hematochimeric mouse model; mobilization.

Graphical abstract

Figure S1

Long-term donor chimerism is established by M-HSCT, related to Figure 1 (A) Schematic of the M-HSCT strategy, illustrating the proposed exchange between mobilized recipient CD45.2 (in gray) and donor CD45.1 cells (in purple). (B) Schematic of the interaction between mobilization agents and their therapeutic targets within the BM microenvironment. (C) Representative plots showing the gating strategy used to characterize LSK and SLAM HSC circulating in the peripheral blood, stained for lineage markers, SCA1, KIT, CD48, and CD150. (D) Representative plots showing the gating strategy used to characterize the donor cells, extracted from the BM, stained for lineage markers, SCA1, KIT, pre-, and post-purification of lineage negative cells. In all the analyses, p less than 0.05 were considered significant (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. “ns” means non-significance).

Figure 1

Long-term donor chimerism is established by mobilization-based HSCT (M-HSCT) (A) Schematic of the M-HSCT protocol. Recipient CD45.2 mice were mobilized with G-CSF (green dots) and AMD3100 (red triangle), with and without BIO5192 (blue triangle) (G7A; G7AB), and subsequently transplanted with CD45.1 2 × 10⁶ Lin⁻ cells, collected from the BM. (B) Counts of mobilized WBC (left panel), LSK (middle panel), and SLAM HSC (right panel) per mL in the PB in non-mobilized (Sham) and G7A and G7AB mobilized mice. (C) Counts of LSK cells per million of Lin⁻ CD45.1 donor cells, collected from the BM. (D) Long-term follow-up of the donor CD45.1 (blue) and recipient CD45.2 (ocher) chimerism observed within total CD45⁺ cells in PB after transplanting 2 × 10⁶ Lin⁻ cells post-mobilization in recipient CD45.2 mice. (E) The reconstitution of myeloid and lymphoid lineages over time of the recipient CD45.2⁺ cells (left panel) and CD45.1⁺ cells (right panel) in the PB of recipient CD45.2 mice. (F–I) Chimerism of CD45.1 cells was observed within CD19⁺ B cells, CD11b⁺ myeloid cells, CD4⁺ T helper cells, and CD8⁺ T cytotoxic cells in the PB (F) and BM (G), within the Lin⁻, LSK, and SLAM HSC (H) and spleen (I). (J and K) Myeloid and lymphoid lineage composition of CD45.2⁺ cells (left panel) and CD45.1⁺ cells (right panel) in the BM (J) and spleen (K) of CD45.2 mice. The results are mean ± SEM, with n ≧ 10 biological replicates. In all the analyses, p less than 0.05 were considered significant (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. “ns” means non-significance).

Figure 2

M-HSCT allows establishing sufficient donor chimerism to rescue the HIGM1 phenotype (A) Schematic of different mobilization protocols tested and the times of analysis in Cd40lg^−/− mice. (B–D) Counts of mobilized WBC (B), LSK (C), and SLAM HSC (D) per mL in the PB of Cd40lg^−/− mice treated with PBS (Sham), G-CSF for 7 days (G7), G-CSF for 7 days and AMD3100 (G7A), G-CSF for 7 days, AMD3100 and BIO5192 (G7AB), half-dose of G-CSF for 7 days, AMD3100 and BIO5192 (G7AB-H), G-CSF for 5 days and AMD3100 (G5A), G-CSF for 5 days, AMD3100 and BIO5192 (G5AB), G-CSF for 3 days, AMD3100 and BIO5192 (G3AB), half-dose of G-CSF for 3 days, AMD3100 and BIO5192 (G3AB-H), and only AMD3100 and BIO5192 (AB) at 0, 1, 3, 6, and 9 h after the last injection of A or AB. (E and F) MMP9 (E) and CXCL12 (F) concentration in the BM extracellular extracts of Cd40lg^−/− mice mobilized with protocols described above. (G) Total counts of SLAM HSC in the lower limbs (left panel) and in the PB per mL (right panel) of Cd40lg^−/− mice treated with PBS or mobilized with G7AB. (H) Long-term follow-up of the donor WT (blue) and recipient Cd40lg^−/− (ocher) chimerism observed within total CD45⁺ cells in PB after transplanting 2 × 10⁶ Lin⁻ cells (collected from the BM) post mobilization in recipient Cd40lg^−/− mice. (I–K) Myeloid and lymphoid lineage composition of Cd40lg^−/− (left panel) and WT (right panel) cells in the PB (I), BM (J), and spleen (K) of Cd40lg^−/− mice after M-HSCT. (L) TNP-KLH-specific IgG concentration in sera collected 7 days before (pre) and after (post) TNP-KLH vaccination of Cd40lg^−/− mice after M-HSCT. (M) The percentage of PNA⁺GL7⁺ splenic germinal centers B cells within the spleen of Cd40lg^−/− mice after TNP-KLH vaccination of Cd40lg^−/− mice treated by M-HSCT. The results are mean ± SEM, with n ≧ 5 biological replicates for the kinetic experiments, except for the AB group (n = 4), and with n ≧ 9 biological replicates for the Cd40lg^−/− M-HSCT experiments. In all the analyses, p less than 0.05 were considered significant (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. “ns” means non-significance).

Figure S2

M-HSCT allows establishing sufficient donor chimerism to rescue the HIGM1 phenotype, related to Figure 2 (A) Counts of mobilized WBC (left panel), LSK (middle panel), and SLAM HSC (right panel) cells per mL in the PB of Cd40lg^−/− mice and CD45.2 mice, after mobilization with PBS (Sham) or mobilized with G7AB. A mixed-effect model (REML), followed by a post hoc analysis with Sidak’s test. (B) The percentage of neutrophils, lymphocytes, monocytes, eosinophils, and basophils in the PB after different mobilization protocols. (C and D) Counts of mobilized monocytes (C) and neutrophils (D) per mL in the PB after different mobilization protocols. (E) Schematic of G-CSF impact on the BM. (F) CXCR4 MFI on circulating LSK after treatments with PBS, G7A, G7AB, and on donor cells purified from untreated bone marrow. A Kruskal-Wallis test, followed by a post hoc analysis with Dunn’s test. (G) Schematic of the M-HSCT protocol applied to Cd40lg^−/− mice. The recipient Cd40lg^−/− mice were mobilized with G-CSF (green dots), AMD3100 (red triangle), and BIO5192 (blue triangle) (G7AB), and subsequently transplanted with 2 × 10⁶ WT CD45.1 Lin⁻ cells, collected from the BM. (H–K) Chimerism of WT cells was observed within CD19⁺, CD11b⁺, CD4⁺, and CD8⁺ cells in the PB (H), spleen (I), and BM (J) and within Lin⁻, LSK, and SLAM HSC in the BM (K). A mixed-effects model (REML), followed by a post hoc analysis with Tukey’s test (between groups) or by post hoc analysis with Dunnett’s test (within groups). (L) Chimerism of CD45.1 cells at 20 weeks, following different M-HSCT protocol, subsequently transplanted with 2 × 10⁶ WT CD45.1 Lin⁻ cells, collected from the BM. (M) Chimerism of CD45.1 cells at 20 weeks, following mobilization with G7AB and subsequently transplanted with different cell doses (WT CD45.1 Lin⁻ cells, collected from the BM). Results are mean ± SEM, with n ≧ 5 for the kinetic experiments, except for the AB group (n − 4), and with n ≧ 9 for the Cd40lg^−/− M-HSCT experiments. In all the analyses, p less than 0.05 were considered significant (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. “ns” means non-significance).

Figure 3

M-HSCT allows an efficient donor to recipient exchange of HSPCs within the human niche of hematochimeric mice (A) The percentage of human chimerism (CD45⁺; left panel) and lymphoid/myeloid cell composition within the human CD45⁺ population (right panel) over time in NSGW41 mice, following the first transplant of human CD34⁺ cells. (B) Counts of mobilized WBC (left panel), LSK cells (middle panel), and human CD34⁺ CD38⁻ cells (right panel) per mL in the PB of humanized NSGW41 mice non-mobilized or mobilized with G7AB at 0, 1, 3, and 6 h after the last injection of AMD3100 and BIO5192. (C) Total counts of CD34⁺ CD38⁻ cells in the lower limbs (left panel) and the PB per mL (right panel) of humanized NSGW41 mice non-mobilized (Sham) or mobilized (G7AB). (D) Schematic illustration of competitive transplantation, after G7AB mobilization, between human resident cells and newly transplanted CD34⁺ G-CSF mPB GFP-transduced cells in NSGW41 mice. (E) Vector copy number (VCN) in HSPCs population (CD34⁺CD133⁺CD90⁺) (left panel) and the percentage of GFP⁺ cells measured within CD34⁺ cells (right panel), in vitro, after transduction. (F) CXCR4^high MFI over time after thawing, stained with an antibody targeting the N terminus epitope of CXCR4, on the HSPC population (CD34⁺CD133⁺CD90⁺) mobilized with G-CSF. (G and H) Long-term follow-up of human CD45⁺ (G) and GFP⁺/CD45⁺(H) cells chimerism in PB after M-HSCT in NSGW41 mice. (I) Chimerism of GFP⁺ cells was observed within CD19⁺, CD13⁺, and CD3⁺ cells at the end of the experiment in PB after M-HSCT. (J) The reconstitution of myeloid and lymphoid lineages over time of human CD45⁺ (left panel) and CD45⁺/GFP⁺ cells (right panel) in the PB after M-HSCT. (K and L) Human CD45⁺ (K) and GFP⁺/CD45⁺ (L) cells chimerism in the BM, spleen, and thymus after M-HSCT. (M) Myeloid and lymphoid lineage composition of human CD45⁺ cells (left panel) and CD45⁺/GFP⁺ cells (right panel) in the spleen of NSGW41 mice after M-HSCT. (N) Chimerism of GFP⁺ cells was observed within CD19⁺, CD13⁺, and CD3⁺ cells in the spleen after M-HSCT. (O) The percentage of T cells within human CD45⁺ (left panel) and CD45⁺/GFP⁺ cells (right panel) in the thymus after M-HSCT. (P) Chimerism of GFP⁺ cells was observed within T cells in the thymus after M-HSCT. (Q) Myeloid and lymphoid lineage composition of human CD45⁺ (left panel) and CD45⁺/GFP⁺ cells (right panel) in the BM of NSGW41 mice after M-HSCT. (R) Chimerism of GFP⁺ cells was observed within CD19⁺, CD13⁺, and CD3⁺ cells in the BM after M-HSCT. (S) The percentage of HSPCs (CD34⁺ CD38⁻ CD90⁺) in human CD45⁺ (left panel) and CD45⁺/GFP⁺ cells (right panel) in the BM after M-HSCT. (T) Chimerism of GFP⁺ cells was observed within HSPCs in the BM after M-HSCT. The results are mean ± SEM, with n ≧ 9 biological replicates. In all the analyses, p less than 0.05 were considered significant (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. “ns” means non-significance).

Figure S3

M-HSCT allows an efficient donor to recipient exchange of HSPCs within the human niche of hematochimeric mice, related to Figure 3 (A) Representative plots showing the gating strategy used to characterize the CXCR4^high population in the HSPC (CD34⁺CD133⁺CD90⁺) population. (B) Scheme of CXCR4 cleavage following G-CSF mobilization and related antibodies localization. (C) The percentage of migrating HSPCs, performed at 24 and 72 h post thawing, previously collected with G-CSF. A Kruskal-Wallis test was performed, followed by a post hoc analysis with Dunn’s test. (D) The percentage of CXCR4^high cells (left panel) and MFI (right panel) over time after thawing, stained with an antibody targeting the ECL2 epitope of CXCR4, on the HSPC population mobilized with G-CSF or G-CSF/AMD3100. (E) The percentage of CXCR4^high cells (left panel) and MFI (right panel) over time after thawing, stained with an antibody targeting the N terminus epitope of CXCR4, on the HSPC population mobilized with G-CSF. (F) The percentage of KIT⁺ cells (left panel) and MFI (right panel) over time after thawing, on the HSPC population mobilized with G-CSF or G-CSF/AMD3100. (G) The percentage of ITGA4⁺ cells (left panel) and MFI (right panel) over time after thawing, on the HSPC population mobilized with G-CSF or G-CSF/AMD3100. The longitudinal comparisons, performed by a mixed-effect model (REML), followed by a post hoc analysis with Dunnett’s test. The results are mean ± SEM, with n ≧ 5, with 3 different donors. In all the analyses, p less than 0.05 were considered significant (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. “ns” means non-significance).

Figure S4

Optimization of the mRNA delivery platform, related to Figure 4 (A–D) Fold change of CXCR4^high cells (left panel) and CXCR4^high MFI (right panel) in HSPCs (CD34⁺CD133⁺CD90⁺) electroporated with CXCR4 mRNA differing for the 5′ UTR sequence (A), the 3′ UTR sequence (B), the polyA tail length (C), and the mRNA capping (D), normalized to electroporated-only (EO) cells. (E) Fold change of CXCR4^high cells in HSPCs electroporated with CXCR4 mRNA differing for nucleotides used during mRNA synthesis, normalized to EO cells. (F) The percentage of GFP⁺ cells in HSPCs, electroporated with GFP mRNA differing for nucleotides used for the mRNA synthesis, normalized to EO cells. (G) Schematic of the pVAX CXCR4 mRNA and improved pVAXi CXCR4 mRNA. (H) Fold change of IRF7, OAS1, RIG-I, and ISG15 gene expression, in cells electroporated with CXCR4 mRNA differing for nucleotides used during mRNA synthesis, normalized to EO cells. (I) Fold change of CXCR4^high cells (left panel) and CXCR4^high MFI (right panel) in HSPCs, electroporated with different quantities of CXCR4 mRNA, normalized to EO cells. (J) The percentage of migrating HSPCs, electroporated with different quantities of CXCR4 mRNA. (K) Early and late apoptosis (EA and LA) is induced by the electroporation of different quantities of CXCR4 mRNA, in bulk CD34⁺ cells. (L) The percentage of HSPCs (CD34⁺CD133⁺CD90⁺) in CD34⁺ cells electroporated with different quantities of CXCR4 mRNA. A Kruskal-Wallis test, followed by a post hoc analysis with Dunn’s test (G and K). The results are mean ± SEM, with n ≧ 5 for data collected in vitro (3 different donors). In all the analyses, p less than 0.05 were considered significant (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. “ns” means non-significance).

Figure 4

The transient overexpression of CXCR4 increases chimerism in the humanized context, following M-HSCT (A and B) Fold change of CXCR4^high cells (A) and CXCR4^high MFI (B) in bulk CD34⁺ cells, electroporated with CXCR4 mRNA (pVAX) or optimized CXCR4 mRNA (pVAXi), normalized to electroporated-only (EO) cells. (C and D) Fold change of CXCR4^high cells (C) and CXCR4^high MFI (D) in CD34⁺CD133⁺CD90⁺ HSPCs, electroporated with CXCR4 mRNA or optimized CXCR4 mRNA, normalized to EO cells. (E and F) The percentage of migrating bulk CD34⁺ cells (E) and HSPCs (F), electroporated with GFP, CXCR4, or optimized CXCR4 mRNA. (G and H) Fold change of CXCR4^high cells (G) and CXCR4^high MFI (H) in HSPCs, electroporated with CXCR4 isoform1 or isoform2 mRNA, normalized to EO cells. (I) The percentage of migrating HSPCs, electroporated with GFP, CXCR4 isoform1, or isoform2 mRNA. (J) Long-term follow-up of human CD45⁺ in the PB of NSG mice, following the transplantation of CD34⁺ cells electroporated with GFP or CXCR4 mRNA. (K) The reconstitution of myeloid and lymphoid lineages over time within total CD45⁺ cells in the PB of NSG mice, transplanted with CD34⁺ cells electroporated with GFP or CXCR4 mRNA. (L) Myeloid and lymphoid cell composition of human CD45⁺ cells in the BM of NSG mice, transplanted with CD34⁺ cells electroporated with GFP or CXCR4 mRNA. (M) CD34⁺ subpopulation in the BM of NSG mice, transplanted with CD34⁺ cells electroporated with GFP or CXCR4 mRNA. (N) Schematic illustration of competitive transplantation, after G7AB mobilization, between human resident cells and newly transplanted CD34⁺ G-CSF mPB GFP-transduced and electroporated cells in NSGW41 mice. (O and P) Long-term follow-up of human CD45⁺ (O) and CD45⁺/GFP⁺ (P) cells in the PB after M-HSCT with CD34⁺ cells transiently overexpressing GFP or CXCR4 mRNA, in NSGW41 mice. (Q) Counts of GFP⁺ cells per mL in the PB after M-HSCT with CD34⁺ cells transiently overexpressing GFP or CXCR4 mRNA, in NSGW41 mice. (R) Myeloid and lymphoid lineage composition of human CD45⁺ (left panel) and CD45⁺/GFP⁺ cells (right panel) in the PB after M-HSCT, with CD34⁺ cells transiently overexpressing GFP or CXCR4, in NSGW41 mice. (S) Chimerism of GFP⁺ cells was observed within CD19⁺, CD13⁺, and CD3⁺ cells in the PB after M-HSCT, with CD34⁺ cells transiently overexpressing GFP or CXCR4, in NSGW41 mice. (T) The percentage of human CD45⁺ cells in the PB of NSG mice at 16 weeks, following the secondary transplant of cells collected from groups described in Figure 4O. (U) Percentage of human CD45⁺/GFP⁺ cells in PB of NSG mice at 16 weeks, following the secondary transplant of cells collected from groups described in Figure 4O. The results are mean ± SEM, with n ≧ 10 biological replicates for data collected in vivo, n ≧ 4 biological replicates for secondary transplant experiments, and n ≧ 5 for data collected in vitro (3 biological replicates and 2 technical replicates). In all the analyses, p less than 0.05 were considered significant (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. “ns” means non-significance).

Figure S5

The transient overexpression of CXCR4 increases chimerism in the humanized context, following M-HSCT, related to Figure 4 (A and B) Myeloid and lymphoid cell composition of human CD45⁺ in the spleen (A) and thymus (B) of NSG mice, transplanted with CD34⁺ cells electroporated with GFP or CXCR4 mRNA. The comparison of lineages between human CD45⁺ and GFP⁺ cells by a mixed-effects model (REML), followed by a post hoc analysis with Dunnett’s. (C, E, and G) Myeloid and lymphoid lineage composition of human CD45⁺ (left panel) and CD45⁺/GFP⁺ cells (right panel) in the spleen (C), BM (E), and thymus (G) after M-HSCT (as in Figure 3D) with CD34⁺ cells transiently overexpressing GFP or CXCR4 in NSGW41 mice. The comparison of lineages between human CD45⁺ and CD45⁺/GFP⁺ cells was performed by a mixed-effect model (REML), followed by a post hoc analysis with Dunnett’s. (D, F, and H) Chimerism of GFP⁺ cells was observed within CD19⁺, CD13⁺, and CD3⁺ cells in the spleen (D), BM (F), and within CD3⁺, CD4⁺, and CD8⁺ T cells in the thymus (H) after M-HSCT with CD34⁺ cells transiently overexpressing GFP or CXCR4 in NSGW41 mice. A mixed-effect model (REML), followed by a post hoc analysis with Tukey’s test (between groups) or by a post hoc analysis with Dunnett’s test (within groups). The results are mean ± SEM, with n ≧ 10. In all the analyses, p less than 0.05 were considered significant (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. “ns” means non-significance).

Figure S6

M-HSCT confers a significant advantage to gene-edited cells when paired with an engraftment enhancer, related to Figure 5 (A) Schematic illustration of competitive transplantation, after G7AB mobilization, between human resident cells (initially transplanted with 3 × 10⁵ CD34⁺ G-CSF mPB cells, counted on d1 p.t.) and newly transplanted CD34⁺ G-CSF mPB gene-edited cells (an outgrowth of 3 × 10⁵ CD34⁺ cells, counted on d1 p.t., and transplanted on d4 p.t.) in NSGW41 mice. (B) HDR efficiency in edited cells (GFP⁺) on the bulk CD34⁺ population, assessed in vitro, 15 days post electroporation. (C) The reconstitution of myeloid and lymphoid lineages over time of human CD45⁺ (left panel) and CD45⁺/GFP⁺ cells (right panel) in PB after M-HSCT with CD34⁺ cells gene edited as in main Figure 5A, in NSGW41 mice. The comparison of lineages between human CD45⁺ and CD45⁺/GFP⁺ cells was performed at the last time point by a mixed-effects model (REML), followed by a post hoc analysis with Dunnett’s test (within groups). (D and E) Myeloid and lymphoid lineage composition of human CD45⁺ (left panel) and CD45⁺/GFP⁺ cells (right panel) in the spleen (D) and BM (E) after M-HSCT with CD34⁺ cells gene edited as in main Figure 5A, in NSGW41 mice. The comparison of lineages between human CD45⁺ and CD45⁺/GFP⁺ cells was performed by a mixed-effect model (REML), followed by a post hoc analysis with Dunnett’s test (within groups). (F–H) Chimerism of GFP⁺ cells was observed within CD19⁺ B cells, CD13⁺ myeloid cells, and CD3⁺ T cells at the end of the experiment in the PB (F), spleen (G), and BM (H) after M-HSCT with CD34⁺ cells gene edited as in main Figure 5A, in NSGW41 mice. A mixed-effect model (REML), followed by a post hoc analysis with Tukey’s test (between groups) or by post hoc analysis with Dunnett’s test (within groups). (I) The percentage of CXCR4^high cells in HSPCs (CD34⁺CD133⁺CD90⁺), electroporated with CXCR4 variants mRNA. A Kruskal-Wallis test, followed by a post hoc analysis with Dunn’s test. (J) CXCR4^high MFI in HSPCs (CD34⁺CD133⁺CD90⁺), electroporated with CXCR4 variants mRNA. A Kruskal-Wallis test, followed by a post hoc analysis with Dunn’s test. The results are mean ± SEM, with n ≧ 8 for data collected in vivo on NSGW41 mice, with n ≧ 5 for the data collected in vivo on NSG mice and n ≧ 5 for data collected in vitro (3 different donors). In all the analyses, p less than 0.05 were considered significant (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. “ns” means non-significance).

Figure 5

M-HSCT confers a significant advantage to gene-edited cells when paired with an engraftment enhancer (A and B) The percentage of CXCR4⁺ cells in HSPCs (CD34⁺CD133⁺CD90⁺) (A) and of migrating HSPCs (B), following gene editing (GE; Cas9 RNA, AAVS1 sgRNA, and AAV6-GFP) combined or not with GFP (GE GFP) or CXCR4 (GE CXCR4) mRNA. (C and D) Long-term follow-up of human CD45⁺ (C) and CD45⁺/GFP⁺ (D) cells in the PB after M-HSCT (as in Figure 3D) with CD34⁺ cells gene edited as in (A), in NSGW41 mice. (E–G) The percentage of migrating HSPCs without treatment (E), in the presence of AMD3100 (F) or AMD3465 (G), electroporated with CXCR4 variants. (H) The percentage of human CD45⁺ in the PB of NSG mice at 16 weeks, following the transplantation of CD34⁺ cells electroporated with GFP, ITGA4, KIT, and CD47 mRNA. The results are mean ± SEM, with n ≧ 8 biological replicates for data collected in vivo on NSGW41 mice, with n ≧ 5 biological replicates for data collected in vivo on NSG mice and n ≧ 4 for data collected in vitro (3 biological replicates and 1 technical replicate). In all the analyses, p less than 0.05 were considered significant (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. “ns” means non-significance).