CRISPR-Cas9 immune-evasive hESCs are rejected following transplantation into immunocompetent mice

Abstract

Although current stem cell therapies exhibit promising potential, the extended process of employing autologous cells and the necessity for donor-host matching to avert the rejection of transplanted cells significantly limit the widespread applicability of these treatments. It would be highly advantageous to generate a pluripotent universal donor stem cell line that is immune-evasive and, therefore, not restricted by the individual's immune system, enabling unlimited application within cell replacement therapies. Before such immune-evasive stem cells can be moved forward to clinical trials, in vivo testing via transplantation experiments in immune-competent animals would be a favorable approach preceding preclinical testing. By using human stem cells in immune competent animals, results will be more translatable to a clinical setting, as no parts of the immune system have been altered, although in a xenogeneic setting. In this way, immune evasiveness, cell survival, and unwanted proliferative effects can be assessed before clinical trials in humans. The current study presents the generation and characterization of three human embryonic stem cell lines (hESCs) for xenogeneic transplantation in immune-competent mice. The major histocompatibility complexes I- and II-encoding genes, B2M and CIITA, have been deleted from the hESCs using CRISPR-Cas9-targeted gene replacement strategies and knockout. B2M was knocked out by the insertion of murine CD47. Human-secreted embryonic alkaline phosphatase (hSEAP) was inserted in a safe harbor site to track cells in vivo. The edited hESCs maintained their pluripotency, karyotypic normality, and stable expression of murine CD47 and hSEAP in vitro. In vivo transplantation of hESCs into immune-competent BALB/c mice was successfully monitored by measuring hSEAP in blood samples. Nevertheless, transplantation of immune-evasive hESCs resulted in complete rejection within 11 days, with clear immune infiltration of T-cells on day 8. Our results reveal that knockout of B2M and CIITA together with species-specific expression of CD47 are insufficient to prevent rejection in an immune-competent and xenogeneic context.

Keywords: CRISPR-Cas9 editing; human embryonic stem cells; immune rejection; transgene insertion; universal cell line; xenogeneic transplantation.

FIGURE 1

Strategy for the generation of immune-evasive human embryonic stem cells. A transgene encoding hSEAP was inserted into the safe harbor CLYBL of wildtype hESCs (1), resulting in the generation of the cell line hESC + hSEAP (2). The hESC + hSEAP had an mCD47 transgene inserted in the B2M loci, rendering B2M non-functional, and the CIITA loci targeted for knockout as well as luciferase randomly integrated by lentiviral transduction (3), which resulted in the generation of the line hESC + hSEAP+2xKO + mCD47 (4). hESC + hSEAP+2xKO + mCD47 had its mCD47 sequence targeted for knockout (5), resulting in the generation of the line hESC + hSEAP+2xKO + mCD47KO (6).

FIGURE 2

Weight data of animals involved in study 1 and Study 2. BALB/c mice underwent weekly weighing throughout both in vivo studies. The graphs depict the mean weight of each experimental group along with the standard deviation (SD) at various time points. Statistical analysis using a two-way repeated measures ANOVA indicated no significant difference in weight among the groups over time.

FIGURE 3

Schedule of in vivo study I. On day 0, animals were subcutaneously injected with either a low number of hESC + hSEAP, a high number of hESC + hSEAP, or wildtype hESCs. To assess hESC survival, blood samples were taken on days 0, 1, 7, 10, and 14 and analyzed for the presence of hSEAP (maximum one blood sample per animal per week). On day 14, the animals were euthanized, and tissue was collected and stored in 4% PFA.

FIGURE 4

Schedule of in vivo study II. On day 0, animals were randomly divided into three groups and injected with either hESCs expressing hSEAP (control), hESC + hSEAP+2xKO + mCD47 V2, or hESC + hSEAP+2xKO + mCD47 V2 KO. Animals were subdivided into groups a and b for blood sampling, which was taken on days 1, 3, 8, 11, 15, 22, and 25 (maximum one blood sample per animal per week). Serum was extracted from blood and analyzed for hSEAP. On days 1, 3, 8, 15, and 25, a subset of animals were euthanized, and tissue was collected for the histological assessment of hESC survival and immune response.

FIGURE 5

Insertion of CD47 and hSEAP and pluripotency profiles of the resulting hESC lines. (A) Long-range PCR amplicon covering the site of insertion in B2M (CD47) and CLYBL (hSEAP). WT, wildtype hSECs with no transgene; V2, hESC + hSEAP+2xKO + mCD47; and KO, hESC + hSEAP+2xKO + mCD47KO. Insertion of hSEAP produces a 4,764 bp product, and CD47 insertion results in a 5,217 bp product. 1Kb plus DNA ladder was used. (B) Flow cytometry data showing the percentage of hESCs positive for OCT4 and SOX2 in the edited hESCs. (C) Immunocytochemical staining for pluripotency markers OCT4, SSEA3, and NANOG on gene-edited hESCs. Scale bars= 100 µm. (D) Immunocytochemical staining of the gene-edited hESCs after 3 weeks of spontaneous differentiation. Markers for each of the three germ layers were used to assess differentiation potential. Mesoderm= smooth muscle actin (SMA), ectoderm = beta tubulin 3 (TUBIII), and endoderm= AFP. Scale bars= 100 µm.

FIGURE 6

Expression of hSEAP, mCD47, and MHC-I on gene-edited hESCs. (A) In vitro detection of hSEAP secreted to cell media by gene-edited hESCs. hSEAP is measured by chemiluminescence and compared to wildtype hESCs by one-way ANOVA **** = p < 0.001. (B) Detection of murine CD47 expression measured by flow cytometry on the generated hESCs and compared to the expression in wildtype hESCs. (C) Binding of murine CD47 to murine SIRPa was assessed by the incubation of fluorescently labeled recombinant murine SIRPa with generated hESCs and wildtype hESCs as the control, followed by flow analysis. (D) Expression of HLA-A, B, and C measured by flow cytometry for generated hESCs and compared to wildtype hESCs.

FIGURE 7

hSEAP and mCD47 expression in gene-edited hESCs after directed neural differentiation. (A) Expression of hSEAP in gene-edited hESCs after 35 days of neural differentiation. Comparison with differentiated wildtype hESCs using Student’s t-test **** = p < 0.0001. (B) mCD47 expression by flow cytometry for gene-edited hESCs and wildtype hESCs after 18 days of neural differentiation.

FIGURE 8

hSEAP in serum samples from in vivo studies I and II. (A) hSEAP levels from serum samples collected from animals during in vivo study I, represented by the chemiluminescent signal, to determine whether differences in hSEAP could be measured in vivo N = 4. (B) hSEAP levels in serum samples from animals collected during in vivo study II, represented by chemiluminescent signal, to test the survival of transplanted gene-edited hESCs. N decreases as the animals are euthanized, according to the study design, but was always above 4. Columns represent the mean value, and error bars represent the standard deviation.

FIGURE 9

Histology of injection site N = 3. (A) H&E staining of the representative injection site from each of the three groups on different days. (B–D) Immunohistochemical staining of the representative injection site from each of the three groups on different days. (B) KU80 (yellow) and DAPI (blue), (C) CD45 (yellow) and DAPI (blue), and (D) CD3 (yellow) and DAPI (blue). (E) Quantification of % CD3 positive cells in the injection site compared to the % of DAPI, calculated from 3–5 images per injection site.