Production of platelets in vitro in functionalized 3-dimensional scaffolds mimicking the bone marrow niche

Abstract

The safety, quality and supply of donor-derived platelet units intended for transfusion have improved over the past decades but significant problems still remain. In vitro-derived platelets offer a possible alternative but up-scaling production is hindered by our limited understanding of thrombopoiesis (the release of platelets by their mother cell, the megakaryocyte [MK]). Here, we have developed an integrated strategy aiming to mimic ex vivo the bone marrow physiological niche that promotes thrombopoiesis by mature MK. The screening of a panel of 259 recombinant transmembrane proteins derived from cells known to promote platelet production through direct contact with MK enabled us to show that ACVR1B, CRTAM, MUCEN and BTN1A1 improve platelet production from either cord blood- (ACVR1B) or pluripotent stem cells-derived (CRTAM, MUCEN and BTN1A1) MK. Using two different methodologies, we functionalize either collagen- or silk-based 3-dimensional scaffolds and confirm increased functional platelet production by up to 2-fold. This unbiased approach has allowed us to identify novel proteins whose role in platelet formation was previously unknown and highlights the potential gain of recreating the MK niche to allow in vitro platelets to become a viable alternative for transfusion.

Figure 1.

Recombinant transmembrane proteins can differentially affect platelet production. (A) Schematic representing how recombinant transmembrane proteins (recTMP) are produced from entire ectodomains of membrane-embedded proteins. These are tagged with rat Cd4 domain 3+4, biotin and 6-HIS tag, and immobilized on streptavidin-coated surfaces, megakaryocytes (MK) are cultured within these wells. Analysis of platelet production by flow cytometry of (B) cord blood-derived MK (CBMK) and (C) induced pluripotent stem cell derived-MK (iPSC-MK) incubated with individual recTMP at 10 μg/mL for 72/24 hours (CBMK/iPSC-MK), data represents inverse-variance weighted estimate of protein effect versus log P value; N=3-4.

Figure 2.

ACVR1B, MUCEN, BTN1A1 and CRTAM significantly increase platelet production and proplatelet formation. Analysis of platelet production by flow cytometry of (A) cord blood-derived megakaryocytes (CBMK), N=3-5, and (B) induced pluripotent stem cell-derived megakaryocytes (iPSC-MK), N=3-6, incubated with the indicated individual recombinant transmembrane proteins (recT-MP) at 10 μg/mL for 72/24 hours (h) (CBMK/iPSC-MK). Analysis of platelet production by flow cytometry of (C), CBMK incubated for 72 h with the indicated concentrations of ACVR1B, N=3, and iPSC-MK incubated for 24 h with the indicated concentrations of (D) CRTAM, N=4, (E) BTN1A1, N=3 and (F) MUCEN, N=3. Analysis of proplatelet production on fibrinogen, representative images of (Gi) CBMK and (Gii) iPSC-MK incubated with the indicated individual recTMP at 10 μg/mL for 72/24 h (CBMK/iPSC-MK), DAPI (blue), α-tubulin (green), scale bar =20 μm. Quantification of proplatelet production of (H) CBMK, N=3, and (I) iPSC-MK, N=3, incubated with the indicated individual recTMP at 10 μg/mL for 72/24 h (CBMK/iPSC-MK). All data represents relative values compared to streptavidin-only control (Ctl). All data mean ± standard deviation, repeated measures one-way ANOVA; *P<0.05; **P<0.01; ***P<0.001.

Figure 3.

Collagen scaffolds functionalized with recombinant transmembrane proteins increases platelet production. Addition of recombinant transmembrane proteins (recTMP) to modified collagen cell supports via biotin and the multiple binding sites of streptavidin. (A) Scheme for attaching a biotin-tagged recTMP via addition of a biotin to collagen with 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide, hydrochloride/ N-hydroxysuccinimide (EDAC/NHS) crosslinker in the presence of amine-polyethylene glycol (PEG)-biotin (EZ-link), amine provides the moiety to crosslink to collagen with a PEG extender and biotin end, recTMP synthesized with a biotin incorporating sequence can be bound in the presence of streptavidin. (B) Analysis of methods of adsorption, crosslinking with PEG biotin in excess and limiting the concentration of crosslinker (x linker limited), or by adding excess crosslinker and limiting the amount of amine-PEG-biotin (side chain limited), quantified by Cd4 enzyme-linked immunosorbant assay (ELISA), data represents optical density (OD), N=3. (C) 5-bromo-4-chloro-3'-indolyphosphate p-toluidine salt/nitro-blue tetrazolium chloride (BCIP/NBT) dye deposition on collagen support material in triplicate of; adsorbed, crosslinker-limited, or side chain-limited attachment of recTMP. (D) Quantification of candidate recTMP using Cd4 ELISA, data represents optical density, N=3. Quantification of relative platelet production on functionalized collagen scaffolds of (E) cord blood-derived megakaryocytes (CBMK) with 0.01 mg/ mL ACVR1B and (F) induced pluripotent stem cell-derived MK (iPSC-MK) with either 0.178 mg/mL MUCEN, 0.355 mg/mL BTN1A1, 0.71 mg/mL CRTAM or all 3 recTMP together, N=4, data compared to streptavidin-only control (Ctl.). All data mean ± standard deviation, Repeated measures one-way ANOVA with Bonferroni post hoc; *P<0.05. scaff: scaffold; Adsorb: adsorption.

Figure 4.

Effect of functionalized silk films on ex vivo platelet production. (A) Silk films were prepared by dispensing a silk fibroin solution onto a polydimethylsiloxane (PDMS) mold. When the solution dries, a silk film is formed that contains a dispersion of recombinant transmembrane proteins (recTMP). The film was finally soaked in the culture medium before cell seeding. Representative immunofluorescent images of (Bi) cord blood-derived megakaryocytes (CBMK) and (Bii) induced pluripotent stem cell-derived MK (iPSC-MK) cultures on silk films functionalized with the indicated recTMP (scale bar =30 μm). (C, D) The positive recTMP supported increased proplatelet formation by CBMK and released platelet particles by iPSC-MK compared to the controls, N=3. All data mean ± standard deviation, paired student t test; *P<0.05. Ctl.: control.

Figure 5.

Effect of 3-dimensional silk scaffold functionalization on ex vivo platelet production. (A) The silk sponge was prepared inside a flow chamber by dispensing an aqueous silk solution mixed with salt particles. After leaching out the salt, the resulting porous silk sponge was cultured with cord blood-derived megakaryocytes (CBMK) or induced pluripotent stem cell-derived MK (iPSC-MK) and perfused to allow platelet collection. (B, C) Representative confocal microscopy analysis of CFSE⁺ CBMK and iP-SC-MK adhering onto functionalized silk scaffolds upon seeding (green = MK; grey = silk scale bar =100 μm). After 24-hour incubation into the system the presence of MK elongating proplatelet shafts could be appreciated in the presence of the positive recombinant transmembrane proteins (recTMP), but not of the control (green = MK, grey = silk, scale bars =100 μm). (D, E) The silk bone marrow was perfused with culture medium for 8 hours and released platelets collected into gas-permeable bags and analyzed by immunofluorescence staining of β1-tubulin (scale bar =10 μm). (F, G) Platelet count was assessed by flow cytometry by mixing samples with counting beads and expressed as fold increase relative to the control (N=3). (H, I) β1-tubulin staining of ex vivo-collected platelets in adhesion on type I collagen over 60 minutes. Platelets were functional as they spread and showed the reorganization of microtubules (scale bars =10 µm). No difference was observed among control or positive recTMP. All data mean ± standard deviation, paired student t test; *P<0.05. 3D: 3-dimensional; Ctl.:control.