Fig 2: Pathway enrichment analysis of secretomes from primary human coronary artery endothelial cells (HCAEC) and primary human internal thoracic artery endothelial cells (HITAEC) treated with Dulbecco’s phosphate-buffered saline (DPBS), magnesiprotein particles (MPP), primary calciprotein particles (CPP-P), or secondary calciprotein particles (CPP-S) for 24 h. (A) Differentially expressed molecular terms identified among the proteins which are downregulated upon the incubation with CPP-P or CPP-S during the screening of Gene Ontology and Reactome resources through the Database for Annotation, Visualization, and Integrated Discovery (DAVID); (B) Heat map showing the downregulated expression of the ECM components (in particular basement membrane proteins) and notable and stable upregulation of soluble CD59 in HCAECs and HITAECs exposed to CPP-P or CPP-S. Protein abundance is represented as log2-transformed, imputed, and normalised data (see Section 4.4. in Materials and Methods “Proteomic profiling”).

Fig 3: CM from iPS-RPE cells is protective(A and B) Conditioned media (CM) was collected from sCD59 AAV2/5 and/or aflibercept AAV2/5 vector transduced iPS-RPE cells every 48 h. CM was added to confluent 661w cells for 48 h and then incubated with 15% v/v NHS or HI-NHS for 30 min. Again, aflibercept CM (aflibercept concentration = 175.21 ng/mL) did not provide any protection against MAC formation as compared with control (no AAV) CM. CD59 (CD59 concentration = 41.62 ng/mL) and combination (CD59 concentration = 31.23 ng/mL, aflibercept concentration = 62.39 ng/mL) CM did reduce C5-b9 deposition (red) as compared with control CM (∗∗∗p = 0.0004 and ∗p = 0.0449 respectively) (scale bar, 50 μm). Nuclei are marked by Hoechst 33258. n = 3 independent experiments. (C) After 24 h, aflibercept (aflibercept concentration = 822.40 ng/mL) and CD59 (CD59 concentration = 50.3 ng/mL) CM added to bEnd.3 cells in tissue culture inserts increased TEER (Ω.cm2, ∗∗∗p = 0.0005, ∗∗∗∗p < 0.0001 respectively) relative to control CM. Addition of 100 ng/mL VEGF to control CM in the basolateral chamber decreased TEER (∗∗∗p = 0.0003). When added to aflibercept, CD59 or combination (CD59 concentration = 46.644 ng/mL, aflibercept concentration = 446.76 ng/mL) CM, VEGF was prevented from causing increases in endothelial permeability (∗∗∗∗p < 0.0001 for all comparisons).Data analyzed by ordinary one-way ANOVA with Tukey’s multiple comparisons test, n = 9 measurements from three independent experiments, represented as mean ± SEM.

Fig 4: CM from modified ARPE-19 cells reduces neovascular lesions in the JR5558 mouse(A) ARPE-19 cells were transiently transfected with the aflibercept, sCD59, or both plasmids and conditioned media (CM) was harvested after 48 h. This CM was injected intravitreal into pre-screened JR5558 mice. (B) Intravitreal injection of control CM from ARPE-19 cells did not result in any changes to the total area covered by CNV-like lesions after 24 h, when compared with FFA images taken before injection (p = 0.9705; n = 10 eyes post injection). (C) Aflibercept CM (aflibercept concentration = 558.97 ng/mL) significantly reduced lesion area 24 h after intravitreal injection (∗∗p = 0.0095; n = 4 eyes post injection). (D) Similarly, intravitreal injection of CD59 CM (CD59 concentration = 36.81 ng/mL) resulted in a significant reduction in the total area covered by CNV lesions (∗∗∗p = 0.0006; n = 7 eyes post injection). (E) Combination CM (aflibercept concentration = 259.56 ng/mL; CD59 concentration = 7.72 ng/mL) produced similar results (∗∗∗p = 0.0007; n = 9 eyes post injection).Data analyzed by Mann-Whitney test or unpaired two-tailed t test and presented as the percentage change, relative to lesion area pre-injection with error bars indicating SEM.

Fig 5: Conditioned media confers protection from VEGF induced permeability(A) Mouse brain endothelial cells, bEnd.3, were grown to confluence on cell culture inserts with a 0.4-μm pore size. Conditioned media was added to both chambers. VEGF was added to the basolateral chamber of half of the wells. After 24 h, measures of endothelial cell permeability, TEER or FITC-dextran 70kDa (FD70) tracer flux were obtained. (B) After 24 h, addition of 50 ng/mL or 100 ng/mL VEGF to control conditioned media significantly reduced TEER (Ω.cm2) as compared with conditioned media alone. Incubation of VEGF in aflibercept conditioned media (aflibercept concentration = 609.39 ng/mL) results in significantly higher TEER values relative to control conditioned media at VEGF concentrations of 50 ng/mL and 100 ng/mL n = 9 measurements from three independent experiments. (C) In control conditioned media, 100 ng/mL VEGF causes a significant increase in FD70 tracer flux as compared with samples without VEGF (∗p = 0.0286). Again, aflibercept conditioned media prevents increases in endothelial permeability after incubation with 100 ng/mL VEGF (∗∗p = 0.003). RFU, relative fluorescence units. n = 3 independent experiments. (D) Aflibercept conditioned media (aflibercept concentration = 902.5 ng/mL) alone caused increased TEER. Addition of conditioned media containing sCD59 (CD59 concentration = 20.35 ng/mL) or sCD59 and aflibercept (Combo, aflibercept concentration = 533.69 ng/mL, CD59 concentration = 15.47 ng/mL) also prevented VEGF mediated reduction in TEER as compared with control media.∗∗∗∗p < 0.0001 for all comparisons, n = 9 measurements from three independent experiments. Data analyzed by ordinary one-way ANOVA with Tukey’s multiple comparisons test and presented as mean ± SEM.