Preclinical quality, safety, and efficacy of a CGMP iPSC-derived myogenic progenitor product for the treatment of muscular dystrophies

Abstract

Pluripotent stem cell (PSC)-derived therapies are in clinical trials of terminally differentiated or transiently required cell types, but to date no PSC-derived trial contributing tissue-specific stem cells or any PSC-based skeletal muscle regeneration trial has been approved. We describe a process in accordance with the Current Good Manufacturing Practice (CGMP) to generate large-scale cryopreserved PAX7-induced myogenic progenitors, which reconstitute both fibers and satellite cells, from PSCs. We subjected the clinical-grade cell product MyoPAXon to biodistribution, toxicity, and tumorigenicity studies in mice under Good Laboratory Practice conditions with no adverse effects and demonstrate long-term engraftment (>1 year) and efficacy in dystrophic mice. Transplantation of 37-60 million MyoPAXon cells into immunosuppressed non-human primates showed human contribution to muscle fibers and satellite cells, with no safety concerns. The US Food and Drug Administration has recently authorized this fully characterized off-the-shelf CGMP product for a first-in-human clinical trial in Duchenne muscular dystrophy, representing the first iPSC-derived tissue-specific stem cell therapy.

Keywords: CGMP product; GLP preclinical studies; MyoPAXon; efficacy; iPS cells; muscle regeneration; muscular dystrophy; myogenic progenitors; non-human primates; safety.

Graphical abstract

Figure 1

MyoPAXon CGMP manufacturing and characterization (A) Overview of MyoPAXon CGMP manufacturing and testing. The master cell bank (MCB) consists of LiPSC-ER2.2 iPSCs, from the source cell bank (SCB), transduced with CGMP lentiviral vectors for conditional expression of PAX7 (108 vials of iPAX7 iPSCs cryopreserved at 2 × 10⁶ cells/vial). One vial of the MCB undergoes myogenic differentiation as described in (B) to produce 68 vials of the working cell bank (WCB; 35 × 10⁶ cells/vial). One vial of the WCB undergoes expansion to produce the MyoPAXon drug product (DP), which is cryopreserved at 50 × 10⁶ per vial (1 vial of WCB produced 150 vials of DP). DP was assessed for stability, GLP mouse safety studies, and non-GLP efficacy and safety studies in mice and NHP recipients. MCB, WCB, and DP underwent extensive quality control (QC). This extensively characterized off-the-shelf DP will be used in the first-in-human phase 1 clinical trial. (B) Detailed scheme outlines the differentiation process. PAX7 induction is initiated by adding doxycycline to the myogenic medium. PAX7⁺ cells emerge in these cultures and begin to proliferate (indicated in orange). PAX7-induced myogenic progenitors are purified based on CD54 expression using CliniMACS, and CD54⁺ myogenic progenitors are frozen as WCB at p1. The last step consists of the expansion of CD54⁺ myogenic progenitors (WCB) from p1 to p4 to produce MyoPAXon DP. (C) Scalability of MyoPAXon, as indicated by growth curves of 3 independent production batches from passage 1 to 4 (P1–P4). VNC, viable nucleated cells. (D) Identity of MyoPAXon DP. Representative flow cytometry plots show the expression of PAX7 (upper left) and the surface markers CD54 and α9β1 (upper right) and the absence of a double-positive population for the pluripotent markers SSEA-4 and TRA-1-81 (lower left) in the MyoPAXon DP. Graph shows respective quantification of flow cytometry data from 3 MyoPAXon productions (lower right). Error bars indicate the mean ± SEM (n = 3). (E) In vitro myogenic differentiation potential of MyoPAXon DP. Representative image shows staining for MHC (red); DAPI (blue). Scale bar: 500 μm. (F) MyoPAXon stability over a period of 36 months. Graph shows viability of cryopreserved MyoPAXon at 3, 6, 9, 12, 18, 21, 24, 27, 30, and 36 months (mo). The x axis indicates hours post-thaw. (G) Confirmation of MyoPAXon DP in vivo regenerative potential. Cells were transplanted into the TA (1 × 10⁶) and quadriceps (3.5 × 10⁶) muscles of NSG mice 1 day after CTX injury. Immunostaining shows LMNA (red), DYS (gray), and DAPI (blue). Scale bars: 500 μm (lower magnification) and 10 μm (close-up). (H) Graph shows in vivo regenerative potential of 3 independent MyoPAXon productions transplanted at 1 × 10⁶ cells in CTX-injured TA muscles of NSG mice. Data are shown as means ± SEM (n = 14 for PV2, n = 24 for CGMP1, and n = 32 for DP). (I) Frequency of integrations near specific genes in myogenic progenitors and parent iPSCs. Each point represents a different gene, with x and y values representing frequencies of cells with integration near that gene in iPSCs (x axis) and myogenic progenitors (labeled MyoPAXon, y axis). Four independent iPSC transductions were performed, and these were differentiated into 7 derivative myogenic progenitors; the plot is of the amalgamated data for all 11 samples. Line of best fit shown in red indicates that frequencies of cells with integrations near particular genes are similar in iPSCs and derived myogenic progenitors. Note its nearness to the diagonal in gray. (J) Frequency of integration sites within 50 kb of a cancer-related gene in MCB and 3 independent transduced iPSC preparations (light green), MyoPAXon and 6 independent myogenic progenitors produced by the same protocol (dark green), and from cell products for the treatment of chronic granulomatous disease (CGD), Wiskott-Aldrich syndrome (WAS), acute lymphoblastic leukemia (CART19 ALL), and β-thalassemia (β-Thal). Data are adapted from Morris et al., Six et al., Boulad et al., and Kohn et al.^,,,

Figure 2

MyoPAXon efficacy in NSG-mdx^4Cv and NSG-FKRP^P448L mice (A) Overview of transplantation studies in NSG-mdx^4Cv and NSG-FKRP^P448L mice. One day after 12-Gy hindlimb irradiation, 1 × 10⁶ cells were injected into the TA muscles and assessed 2 months later for in situ force measurements and engraftment analysis. (B) MyoPAXon in vivo regenerative potential in NSG-mdx^4Cv mice. Representative image shows engrafted area (left), as indicated by staining for LMNA (red), DYS (gray), and DAPI (blue). Scale bar: 500 μm. At right, engraftment quantification for both CGMP batches of MyoPAXon (CGMP1, red dots; DP, blue dots). Graph reports mean ± SEM (n = 16). (C) MyoPAXon engraftment in NSG-FKRP^P448L mice. Representative image of the engrafted area (left) and respective quantification (right). Graph reports mean ± SEM (n = 13, CGMP1, red dots; DP, blue dots). Scale bar: 500 μm. (D) MyoPAXon rescues α-DG glycosylation in NSG-FKRP^P448L mice. Representative immunostaining of cell-injected TA muscles. LMNA (red), IIH6 (green), and DAPI (blue). Scale bar: 500 μm. (E and F) Functional effect changes following cell transplantation in TA muscles of NSG-mdx^4Cv (E) and NSG-FKRP^P448L (F) mice. Specific force (sF₀: F₀ normalized to cross-sectional area) of muscles transplanted with CGMP1 (red dots) or DP MyoPAXon (blue dots) or injected with PBS (control). Graph reports mean ± SEM (n = 16 and n = 9 for E and F, respectively). ∗p < 0.05; ∗∗p < 0.01. (G) Specific force in TA muscles from non-dystrophic NSG control mice. Measurements were performed in both non-irradiated and irradiated PBS-injected TA muscles assessed 2 months later. Graph shows mean ± SEM (n = 15 for no X-ray and n = 7 for X-ray + PBS) (H) Representative images show satellite cell engraftment in transplanted NSG-mdx^4Cv (left) and NSG-FKRP^P448L (right) TA muscles. Donor-derived satellite cells (arrowheads) were identified as cells located below the basal lamina (Lam, gray) co-expressing LMNA (red) and PAX7 (green). DAPI in blue stains nuclei. Scale bar: 20 μm. (I) Graph bars show respective quantification of human satellite cell engraftment indicated as percentage (from H). Data are shown as mean ± SEM (n = 15 for NSG-mdx^4Cv and n = 9 for NSG-FKRP^P448L) Absolute values for donor (PAX7⁺LMNA⁺ and recipient (PAX7⁺LMNA⁻) satellite cells were 35.3 ± 3.7 and 7 ± 0.8, respectively, in NSG-mdx^4Cv, and 11 ± 2.1 and 2 ± 0.5, respectively, in NSG-FKRP^P448L.

Figure 3

GLP safety study to assess toxicity, tumorigenicity, and biodistribution (A) Overview of GLP safety studies. Mice were divided into 4 cohorts, receiving intramuscular administration of either vehicle or the MFD of 3.5 × 10⁶ MyoPAXon DP in CTX pre-injured muscles (1 day prior to transplant). Mice were assessed at 11 weeks (toxicity; group 1) or at 25–26 weeks (tumorigenicity; group 2) post-administration. (A–C) Histopathology of selected tissues (Table S4) from the toxicity (B) and tumorigenicity (C) studies. Representative images of hematoxylin and eosin staining in the targeted muscle quadriceps, lung, liver, and kidney of NSG mice following treatment with vehicle or MyoPAXon. Scale bar: 200 μm. (D and E) MyoPAXon biodistribution. PCR-based quantification of human Alu DNA in tissues from animals treated with cells (D) or vehicle (E). Graph shows mean ± SEM (n = 44 for D, except for the ovaries and testis, n = 22 and n = 8 for E, 4 males and 4 females. Points represent the average of each sample run in technical triplicate.. ILN, inguinal lymph node; MLN, mesenteric lymph node.

Figure 4

Long-term studies show persistent muscle engraftment Transplantation studies were performed in CTX pre-injured TA muscles of NSG mice. (A) Representative images show persistent engraftment of engineering run PV2 at 13 months post-injection, as indicated by staining of muscle sections to visualize LMNA (red), DYS (gray), and DAPI (blue). Scale bars: 500 μm (left) and 100 μm (right). Graph shows engraftment quantification at 1 and 13 (from A) months after injection. Data are shown as mean ± SEM (n = 8 for 1 month and n = 7 for 13 months). (B) Persistent engraftment of the DP at 6 months post-injection. Representative images show donor-derived myofibers, as indicated by staining for LMNA (red), DYS (gray), and DAPI (blue). Scale bars: 500 μm (left) and 100 μm (right). Graph shows respective engraftment quantification. Data are shown as mean ± SEM (n = 8). (C) Representative images of muscle sections stained to visualize Ki67 (green), LMNA (red), and DAPI (blue). The white star indicates an LMNA⁺Ki67⁺ cell. Scale bar: 50 μm. (D) Quantification of donor-derived proliferative cells. Graph reports the percentage of LMNA⁺Ki67⁺ nuclei among the LMNA⁺ cell population (from C). Data are shown as mean ± SEM; n = 8 for PV2 1 month (637 nuclei analyzed), n = 7 for PV2 13 months (446 nuclei analyzed); n = 16 for DP 1 month (1,566 nuclei analyzed) and n = 8 for DP 6 months (1,393 nuclei analyzed).

Figure 5

Transplantation of large doses of MyoPAXon into muscles of NHP recipients (A) Overview of experimental design. Four cynomolgus macaques were enrolled in this study (pilot 19JP7, followed by 20HP6, 20AP25, and 20AP21). Between 37 and 60 × 10⁶ MyoPAXon cells were injected into the EDB muscles of these 4 recipients (Table S6). CTX injections were performed during the same surgery prior to cell injection. Immunosuppression consisted of anti-CD20, anti-CD-154, and tacrolimus. Assessment was performed 28 days post-transplantation. (B and C) Representative images display muscle injected with saline (Ctrl) and EDB muscle injected with 40 × 10⁶ RFP-labeled MyoPAXon cells. Staining shows RFP (red) and TRA-1-85 (gray), both MyoPAXon-specific markers. DAPI (blue) shows nuclei. Scale bars: 100 μm (B) and 50 μm (C). (D) Abundant and widespread engraftment in the transplanted EDB muscle of NHP recipient 20AP25. Representative stitched images show immunostaining for RFP (red), TRA-1-85 (gray), and DAPI (blue). Scale bar: 500 μm. (E) Representative images show myofiber engraftment, as indicated by staining for RFP (red) and DYS (gray), in transplanted EDB muscles from NHP recipients 19JP7, 20HP6, and 20AP25. DAPI (blue). Scale bar: 100 μm. (F) Engraftment quantification in EDB muscles injected with 37–40 × 10⁶ cells (3 distinct NHP recipients). Data show the total number of DYS⁺RFP⁺ donor-derived myofibers. (G) Engraftment quantification in EDB muscles injected with 60 × 10⁶ cells (2 distinct NHP recipients). Data show the total number of DYS⁺RFP⁺ donor-derived myofibers. (H) Transplanted cells contribute to the muscle satellite cell compartment of NHP EDB muscles. Left: donor-derived satellite cell (arrowheads) identified based on the localization below the basal lamina (Lam, gray) and the co-expression of RFP (red) and PAX7 (green). DAPI (blue). Scale bar: 10 μm. Right: graph shows quantification of human satellite cells indicated as percentage. Data are shown as mean ± SEM (n = 7). Absolute values for donor (PAX7⁺RFP⁺) and recipient (PAX7⁺RFP⁻) satellite cells were 4 ± 0.7 and 144 ± 19.5, respectively. (I) Biodistribution of H2B-RFP-labeled MyoPAXon in NHP recipients. PCR-based quantification of RFP transcripts in negative control (RFP⁻; MyoPAXon PV2 cells), positive control (RFP⁺; NHP H2B-RFP-labeled CyMN.2 myogenic progenitor cells), heart, kidney, lung, liver, APB, and diaphragm muscles from the 2 distinct engrafted NHP recipients (20HP6 and 20AP25). Graph shows mean ± SEM (n = 2 independent biospecimens per tissue site and n = 3 independent replicates for controls). Each point represents the average of the sample run in technical triplicate.