Tissue procurement and processing
Skin samples (≥ 10 cm2) were obtained in accordance with the German Medicines Act (“Arzneimittelgesetz”) and the German Act on Organ and Tissue Donation, Removal and Transplantation (“Transplantationsgesetz”) as discard tissues from plastic surgeries (abdominoplasties and mastopexies) from donors aged ≤ 50 years who had given written informed donor consent. Tissues from donors who tested serologically positive for HIV1/2, HBV, HCV, HTLV1/2 or syphilis were discarded.
Skin processing and stem cell production took place in an EU-GMP grade A cabinet in a grade B clean room under laminar air flow (“A in B”). Skin tissue was freed from excess subcutaneous tissue, disinfected, washed, dissected into equal pieces (about 2.5 cm2) and subjected to two-step enzymatic digestion using collagenase (Collagenase NB 6 GMP Grade, Nordmark, Uetersen, Germany) followed by animal component-free trypsin (Recombinant Trypsin Solution, Biological Industries, Beit Haemek, Israel) . After filtration and washing/centrifugation of the filtrates, pellets were resuspended in an in-house MSC-favoring growth medium (Ham’s F‑10 supplemented with fetal calf serum, L‑glutamine, fibroblast growth factor‑2, HEPES, hydrocortisone, insulin, glucose and phorbol myristate acetate), pooled, and incubated in C6 cell culture plates in a cell culture incubator at 3.1% CO2 (as instructed by the Ham’s F‑10 manufacturer, Biochrom, Berlin, Germany), 90% humidity, 37°C.
Assessment of cell confluency and morphology
During the whole cell expansion and subcultivation process, cell confluency and morphology were assessed visually under a phase-contrast microscope by comprehensively trained lab assistants strictly applying the four eyes principle (i.e. cross-checked by the Head of Production). Sets of standardized photographs of cultures at different stages of confluency as well as typical and untypical or artificial cell morphology are used as reference for comparison. A photograph displaying a cell monolayer at 70% confluency is shown in (Fig. S1 [see Additional file 2]).
Cell expansion
Cells were expanded as unsegregated (“mixed”) cell cultures in up to 30 parallel batches from C6 well via T25 and T75 to 8×T175 culture flasks by serial passaging (six passages). During cultivation, regular, protocol-defined medium changes facilitated depletion of non-adherent cells from the culture. Having grown to 70% confluency, cells were harvested using recombinant trypsin and transferred to the next larger culture vessel at a seeding density of 3×104 cells/cm2. When 8×T175 flasks had been reached, four flasks were harvested to be cryopreserved as primary mixed culture for internal back-up purposes. The remaining four flasks were further subcultivated to 16×T175 flasks for magnetic separation of the ABCB5+ cells, respectively (Fig. 1).
Isolation of ABCB5+ cells and batch pooling
After reaching 75% confluency, ABCB5+ MSCs were isolated using magnetic beads (micromer® TC1 Epoxy, Micromod, Rostock, Germany) coated with a mouse anti-human monoclonal antibody directed against the sequence 493‑508 (RFGAYLIQAGRMTPEG) of the ABCB5 extracellular loop 3 [4] (Maine Biotechnology Services, Portland, Maine; GMP purification: Bibitec, Bielefeld, Germany; ICH Q5 [14] virus depletion and safety study: Charles River, Erkrath, Germany). The manufacturing protocol allows for multiple successive isolation cycles, by using only 12 of 16×T175 flasks for ABCB5+ cell isolation, while the remaining four flasks are further subcultivated to generate 16×T175 flasks (see Fig. 1). This cycle may be repeated until passage number 16 [7], provided that no changes in cell morphology or growth behavior occur. If required, the cryopreserved back-up mixed culture can be thawed and subcultivated for further cell production and isolation.
All ABCB5+ cell isolates from one and the same donor tissue that had been isolated in parallel (i.e. on the same day at the same passage number) were pooled to gain one drug substance batch. Each batch was then divided into multiple aliquots of 2×106–18×106 cells, which were cryopreserved (CryoStor® CS10 freeze medium, BioLife Solutions, Bothell, Washington) and stored in the vapor phase of liquid nitrogen.
Drug product formulation
The final drug product was manufactured as an investigational medicinal product to be tested in various clinical trials (Table 1). It was formulated as a ready-to-use suspension, with total cell count and packaging size depending on the intended application. The required quantity of cryopreserved drug substance aliquots, originating from one and the donor skin tissue, were thawed, pooled, washed and suspended in HRG solution (Ringer’s lactate solution containing 2.5% human serum albumin and 0.4% glucose) at a target concentration of 1×107 cells/ml. The cell suspension was filled in one or multiple 1‑ml polycarbonate or 10‑ml polypropylene syringes (BD PlastipakTM, Becton Dickinson, Heidelberg, Germany), as required, which were sealed with polyethylene closing cones (Combi-Stopper, B. Braun, Melsungen, Germany).
In-process controls and release tests
In-process and release tests followed validated GMP-compliant procedures according to the requirements of the European Pharmacopoeia, if applicable. A schematic overview is given in Fig. 1, the specifications are given in Table S1 [see Additional file 1].
Cell cycle analysis
Immediately before the magnetic separation of ABCB5+ cells, a sample containing about 2×106 cells of the mixed culture were fixed with ice-cold 70% ethanol (added dropwise while vortexing) for at least 30 min at 4°C. After washing in 0.02% EDTA, cells were resuspended in RNase A (Thermo Fisher)/propidium iodide solution and analyzed by flow cytometry (BD AccuriTM C6; Becton Dickinson, Heidelberg, Germany) using standard gating strategies.
Determination of ABCB5+ cell content
Following the ABCB5+ cell isolation, but before the enzymatic detachment of the microbeads (which leads to a transient loss of ABCB5 from the cell surface), ABCB5+ cell content was determined after incubation with an Alexa Fluor® 647-coupled donkey anti-mouse secondary antibody (Thermo Fisher, Langenselbold, Germany) targeting the anti-ABCB5 antibody used for the cell isolation. To discriminate between ABCB5+ cells and free bead-antibody complexes, calcein acetoxymethylester (staining metabolically active cells) was added to the cell-secondary antibody suspension before incubation. Fluorescence was measured by flow cytometry (BD AccuriTM C6, Becton Dickinson). Free, unbound bead-antibody complexes were excluded from the ABCB5+ cell content calculation by gating only events with high calcein fluorescence.
Mycoplasma testing
Supernatant and cell suspension samples were spiked with internal control DNA, and genomic DNA was isolated using the Microsart AMP Extraction Kit (Minerva Biolabs, Berlin, Germany). Isolated DNA was subjected to qPCR including positive and negative controls and 10CFUTM Sensitivity Standards for Mycoplasma (M.) orale, M. fermentans and M. pneumoniae (Microsart ATMP Mycoplasma Kit; Minerva Biolabs).
Endotoxin testing
After the ABCB5+ cell isolation, microbead detachment and washing/centrifugation of the cell suspension, a supernatant sample was diluted 1:10 with Endosafe® Limulus Amebocyte Lysate Reagent Water and transferred into an Endosafe® PTSTM cartridge (limit of detection 0.05 EU/ml), which was loaded into an Endosafe® PTSTM reader (all from Charles River, Charleston, South Carolina). The endotoxin level of the sample was calculated based on the change in optical density analyzed against an internal standard curve.
Determination of live cell count and vitality
Propidium iodide solution (1 mg/ml) was added to a cell suspension sample from each drug substance batch and from each produced drug product to stain dead cells. Fluorescence was measured using flow cytometry (BD AccuriTM C6) and cell count and vitality, defined as percentage of propidium iodide-excluding cells, calculated.
Determination of viability, CD90 expression and bead residues
Viability, CD90 expression and microbead residues, which might result from insufficient bead detachment or cell washing, were analyzed in parallel. From each drug substance batch, a sample was incubated with calcein acetoxymethylester and an Alexa Fluor® 647-conjugated anti-human CD90 antibody (BioLegend, London, UK). Fluorescence was analyzed by flow cytometry (BD AccuriTM C6), and the viability of the cell population (defined as percentage of metabolically active cells, i.e. cells converting calcein acetoxymethylester to calcein) and the content of CD90+ cells were calculated. To detect residual beads, a sample of ABCB5 antibody-conjugated beads was used to define a gate in the FSC/SSC dot plot. To exclude (false-positive) signals, only calcein-negative events in that gate were used to calculate bead residues.
Microbiological examination
Microbiological examination was carried out by a certified academic contract laboratory. Samples from every single cryovial of each drug substance batch and from each produced drug product were diluted in NaCl-peptone buffer, inoculated in BACT/ALERT® BPN (anaerobic) and BACT/ALERT® BPA (aerobic) culture bottles (bioMérieux, Nürtingen, Germany), and incubated in the BACT/ALERT® 3D60 (bioMérieux) microbial detection system. Positive samples were seeded onto solid culture medium immediately after detection. After 7 days of incubation, all negative samples were seeded onto solid culture medium.
IL‑1RA secretion assay
Human monocytic cells (THP‑1; LGC Standards, Wesel, Germany) were differentiated into macrophages by incubation in differentiation medium containing 150 nmol/ml phorbol 12‑myristate 13‑acetate for 48 h at 37°C, 5% CO2. In two wells of a 24‑well plate, 1×105 macrophages were co-cultivated with 2×104 ABCB5+ MSCs. In one of the two wells, M1 polarization was stimulated by adding 50 IU/ml interferon‑γ (Imukin®, Clinigen Healthcare, Burton-upon-Trent, UK) at the start of cocultivation, and 50 IU/ml interferon‑γ and 20 ng/ml lipopolysaccharides from Escherichia coli O111:B4 (Sigma-Aldrich) after 24 h. After 48 h, supernatants were collected and analyzed for IL‑1RA using a colorimetric sandwich ELISA kit (Quantikine®, R&D Systems, Abingdon, UK) according to the manufacturer’s instructions.
VEGF secretion assay
ABCB5+ MSCs (3×105) were seeded into a culture dish in stem cell medium and incubated under hypoxic conditions (1% O2) at 37°C. After 48 h, supernatants were collected and analyzed using the Invitrogen VEGF Human ELISA Kit (Thermo Fisher) according to the manufacturer’s instructions.
Tube formation assay
ABCB5+ MSCs (1×105 and 1.5×105) were seeded in two wells of a 24‑well plate coated with Geltrex® basement membrane matrix (Thermo Fisher) and incubated in stem cell medium for 19‑22 h at 37°C, 3.1% CO2. Tube structures were photographed using an inverted microscope (40× final magnification; DM IL LED, Leica, Wetzlar, Germany) equipped with a digital camera (DFC320, Leica) and semi-quantitatively classified into six (A–F) categories, ranging from A = tubular branches of several cells forming a defined network-like structure to F = no tubular branches visible (Fig. S2 [see Additional file 2]), with A–C being considered as successful angiogenic differentiation. For assay validation, human umbilical vein endothelial cells (Thermo Fisher) and human skin melanoma cells (SK‑MEL‑28, ATCC® HTB‑72TM, LGC Standards, Wesel, Germany) were used as positive and negative controls, respectively.
Determination of MSC markers
In addition to routine determination of CD90+ cell content (see above In-process controls and release tests) the batches intended for use at study sites in the Czech Republic in the multinational clinical trial on peripheral artery disease (NCT03339973, Table 1) were additionally tested for expression of CD73, CD105 und CD44 [5, 15], as required by the Czech national drug authority (State Institute for Drug Control, SÚKL). To this end, cell suspension samples were incubated with a fluorescein isothiocyanate (FITC)-conjugated anti-human CD73 antibody, an Alexa Fluor® 647-conjugated anti-human CD105 antibody and a FITC-conjugated anti-human CD44 antibody (all from BioLegend), respectively. Fluorescence was measured by flow cytometry (BD AccuriTM C6) and percentages of CD73+ CD105+ and CD44+ cells were calculated. Pre-specified acceptance criteria were CD73: ≥95% [5]; CD105: determined and declared; CD44: determined and declared.