Bioreactor-based mass production of human iPSC-derived macrophages enables immunotherapies against bacterial airway infections

Abstract

The increasing number of severe infections with multi-drug-resistant pathogens worldwide highlights the need for alternative treatment options. Given the pivotal role of phagocytes and especially alveolar macrophages in pulmonary immunity, we introduce a new, cell-based treatment strategy to target bacterial airway infections. Here we show that the mass production of therapeutic phagocytes from induced pluripotent stem cells (iPSC) in industry-compatible, stirred-tank bioreactors is feasible. Bioreactor-derived iPSC-macrophages (iPSC-Mac) represent a highly pure population of CD45⁺CD11b⁺CD14⁺CD163⁺ cells, and share important phenotypic, functional and transcriptional hallmarks with professional phagocytes, however with a distinct transcriptome signature similar to primitive macrophages. Most importantly, bioreactor-derived iPSC-Mac rescue mice from Pseudomonas aeruginosa-mediated acute infections of the lower respiratory tract within 4-8 h post intra-pulmonary transplantation and reduce bacterial load. Generation of specific immune-cells from iPSC-sources in scalable stirred-tank bioreactors can extend the field of immunotherapy towards bacterial infections, and may allow for further innovative cell-based treatment strategies.

Fig. 1

Continuous generation of human iPSC-Mac in stirred tank bioreactors. a Scheme of hematopoietic differentiation of human iPSC in stirred tank bioreactors (DASbox system). b Representative pictures of DASbox bioreactor filled with floating MCFCs (left). Images of the 8-blade impeller (right). c Individual cell counts of viable macrophages produced in bioreactors (n = 2 of independent bioreactor runs, mean ± SD.). d Representative macrophage harvest counts (upper graph) and corresponding data from continuous process monitoring (biomass, temperature, pH, and dissolved oxygen (DO) level) for the entire cultivation phase of one 42-day bioreactor run. e Analysis of human cytokines in the medium from the bioreactor (technical duplicates, mean ± SD). f Representative light microscopy of macrophage generation analyzed at harvests 1–5. First row shows brightfield images of non-filtered medium samples from bioreactors (scale bar 500 µm). Second row shows brightfield images of freshly harvested macrophages separated from MCFCs by sedimentation (scale bar 50 µm). Third and fourth row show cytospin (scale bar 50 µm) and flow cytometric analysis of iPSC-Mac (gray line: respective isotype control, red line: CD45, blue line CD14)

Fig. 2

Characterization of iPSC-Mac derived from stirred tank bioreactors. a Representative flow cytometric analysis of iPSC-Mac derived from bioreactor differentiation (pre-gated on viable cells in forward and sidewards scatter (FSC/SSC) analysis, gray line: respective isotype control, green line: cell surface marker). b Representative pictures of either brightfield (upper image, scale bar 200 µm) or cytospin preparations (lower image, scale bar 100 µm) of terminally differentiated iPSC-Mac derived from the bioreactor. c–g Transcriptome analysis of iPSC, iPSC-derived macrophages (iPSC-Mac) and PBMC-derived macrophages (PBMC-Mac) (n = 2, biological replicates). c Unbiased, hierarchical heatmap clustering. d Heatmap of differentially regulated genes (p < 0.001) associated with activation of innate immune response (GO:0002218 left) and pluripotency-associated genes (right). e Volcano plot showing differentially expressed genes between PBMC-Mac and iPSC-Mac (FDR = 0.5). f Gene set enrichment analysis (GSEA) from iPSC-Mac versus PBMC-Mac shows enrichment of genes related to yolk sac-derived macrophages. The gene set was derived from>5-fold upregulated genes of murine yolk sac (YS)-derived macrophages compared to bone marrow-derived macrophages. NES – normalized enrichment score. g Heatmap of >5-fold regulated genes associated with YS macrophages shows predominantly upregulation in iPSC-Mac compared to PBMC-Mac

Fig. 3

Antimicrobial activity of iPSC-Mac generated in stirred tank bioreactors. a Scanning raster electron microscopy depicting phagocytosis of latex beads (green) by iPSC-Mac (red) at different time points. b Rate of phagocytosis by terminally differentiated iPSC-Mac (blue) and PBMC-Mac (green) after 2 h of incubation with GFP-labeled P. aeruginosa (PAO1) at 4 or 37 °C (n = 3 of biological replicates, mean ± s.e.m, two-way ANOVA with Sidak’s multiple comparisons test, ns denotes not significant, representative data and gating strategy provided in Supplementary Figure 3a). c, d Heatmap of commonly regulated genes in iPSC-Mac and PBMC-Mac associated with (c) inflammatory response (GO:0006954; σ/σ_max = 0.13, P < 0.05) and (d) innate immune response (GO:0045087, σ/σ_max = 0.13, P < 0.05) after 1 h exposure to PAO1. e Venn-diagramm of >5-fold upregulated genes in iPSC-Mac (blue) and PBMC-Mac (green) after pathogen contact. f, g Top-ranked GOs associated with f biological processes and g molecular function (according to human gene atlas) of top 100 upregulated genes after 1 h pathogen contact (without pre-selection, EnrichR)

Fig. 4

Pulmonary infection and simultaneous macrophage transplantation in hIL-3/GM-CSF KI mice. a Scheme of pulmonary transfer of P. aeruginosa (PAO1) and simultaneous transplantation of iPSC-Mac (PiMT) into hIL-3/GM-CSF KI mice. b Rectal temperature and disease score 24 h post infection of hIL-3/GM-CSF KI mice infected with PAO1 (infected) or infected and transplanted (infected + PiMT). Disease score and temperatures measured before the experiment served as control values (n = 6 animals/group, mean ± s.e.m). c Change in body weight after 24 h. Values are normalized to the respective weights before infection (n = 6 animals/group, mean ± s.e.m). d Lung function measured by head-out body plethysmography. Values measured pre infection served as control values (n = 6 animals/infected and infected + PiMT groups, n = 12 animals for control values, mean ± s.e.m). e Colony forming units (CFU) of PAO1 per left lung after 24 h (n = 6 animals/group, mean ± s.e.m). f Images of bronchio-alveolar lavage fluid (BALF) samples and BALF absorbance at 650 nm (n = 3 animals/group, mean ± s.e.m). g Flow cytometric analysis of BALF and lung. Percentage of mouse granulocytes (determined as GR1⁺ cells) in BALF and Lung (n = 3 animals/group, mean ± s.e.m, exemplarily data and gating strategy provided in Supplementary Figure 4c). h Right lung histology. Left: Two representative images of infected (left) and infected + PiMT (right) mice. Scale bars: 500 µm upper row and 100 µm lower row. Right: histological scoring (n = 3 animals/group, mean ± s.e.m). i Representative diagram of BALF and lung of one animal per group stained with hCD45. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns denotes not significant; statistical significances were assessed using one-way ANOVA with Dunnett’s multiple comparisons test (b, d, g) or Student’s t-test (c, e, f, h))

Fig. 5

Pulmonary infection and therapeutic transplantation of iPSC-Mac in hIL-3/GM-CSF KI mice. a Scheme of pulmonary infection with P. aeruginosa (PAO1) and therapeutic transplantation of iPSC-Mac (PiMT) derived from bioreactors into hIL-3/GM-CSF KI mice. b Disease score of hIL-3/GM-CSF KI mice infected with PAO1 (infected, red), infected and transplanted (infected + PiMT, blue) and control mice receiving PBS twice (control, black). Left: Disease score over time. Right: Disease score 24 h post infection (n = 3 animals/group, mean ± s.e.m). c Infected mice (red dot) and infected + PiMT mice (blue dot) 24 h after infection. d Representative trajectory of mouse activity 24 h post infection for hIL-3/GM-CSF KI mice infected with PAO1 (infected, red), infected and transplanted (infected + PiMT, blue) and control mice receiving PBS twice (control, black) analyzed by video documentation and manual tracking. e Rectal temperature analyzed 24 h post infection (n = 3 animals/group, mean ± s.e.m). f Change in body weight after 24 h. Values are normalized to the respective weights before infection (n = 3 animals/group, mean ± s.e.m). g Colony forming units (CFU) of PAO1 per lung after 24 h (n = 3 animals/group, mean ± s.e.m). h Images of BALF samples. i Levels of human cytokines in BALF of infected + PiMT mice (n = 4: 2 animals and 2 technical duplicates, mean ± s.e.m). j Lung histology images of infected (left), infected + PiMT (middle) and control (right) mice. Scale bars: 500 µm upper row and 100 µm lower row. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns denotes not significant; statistical significances were assed using one-way ANOVA with Dunnett’s multiple comparisons test (b, e, f) or Student’s t-test (g))