Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 33 (5), 1271-80

Systematic Engineering of 3D Pluripotent Stem Cell Niches to Guide Blood Development

Affiliations

Systematic Engineering of 3D Pluripotent Stem Cell Niches to Guide Blood Development

Kelly A Purpura et al. Biomaterials.

Abstract

Pluripotent stem cells (PSC) provide insight into development and may underpin new cell therapies, yet controlling PSC differentiation to generate functional cells remains a significant challenge. In this study we explored the concept that mimicking the local in vivo microenvironment during mesoderm specification could promote the emergence of hematopoietic progenitor cells from embryonic stem cells (ESCs). First, we assessed the expression of early phenotypic markers of mesoderm differentiation (E-cadherin, brachyury (T-GFP), PDGFRα, and Flk1: +/-ETPF) to reveal that E-T+P+F+ cells have the highest capacity for hematopoiesis. Second, we determined how initial aggregate size influences the emergence of mesodermal phenotypes (E-T+P+F+, E-T-P+/-F+, and E-T-P+F-) and discovered that colony forming cell (CFC) output was maximal with ~100 cells per PSC aggregate. Finally, we introduced these 100-cell PSC aggregates into a low oxygen environment (5%; to upregulate endogenous VEGF secretion) and delivered two potent blood-inductive molecules, BMP4 and TPO (bone morphogenetic protein-4 and thrombopoietin), locally from microparticles to obtain a more robust differentiation response than soluble delivery methods alone. Approximately 1.7-fold more CFCs were generated with localized delivery in comparison to exogenous delivery, while combined growth factor use was reduced ~14.2-fold. By systematically engineering the complex and dynamic environmental signals associated with the in vivo blood developmental niche we demonstrate a significant role for inductive endogenous signaling and introduce a tunable platform for enhancing PSC differentiation efficiency to specific lineages.

Conflict of interest statement

Disclosure of potential conflicts of interest

The authors have no conflicts of interest to disclose.

Figures

Fig. 1
Fig. 1
Monitoring mesodermal specification. (A): Four-colour FACS employing a T-GFP cell line sheds light onto pan mesoderm development by tracking the surface expression of E-cadherin, PDGFRα and Flk1. (B): Comparison of undifferentiated cells to cells after 3.75 or 5 days in serum free differentiation media with or without BVT highlights the dynamic progression of these lineage markers and demonstrates that not all phenotypic combinations occur. (C): Populations were sorted singly or in combination as indicated; the average efficiency of colony formation from E−T+P+F+ cells (no. of colonies/no. of seeded cells × 100) was significantly (*) higher than all other sorted fractions (n = 4, except E−T−P+/−F+ n = 2; ANOVA with Tukey’s post hoc analysis α = 0.05).
Fig. 2
Fig. 2
Controlling initial cell aggregate size influences mesodermal specification. (A): The overall effect of endogenously produced factors would depend on the balance of stimulatory or inhibitory regulators that are secreted by the mixture of cell types. A higher local cell density would condition the microenvironment with more endogenous factors than lower density conditions, as demonstrated in Suppl. Figure 2. (B):Two sizes of micropatterned square-pyramidal wells and partial coverage allows similar overall cell densitites with constant volume to be compared. Initial 10-or 100-cell aggregates are shown in 200 and 400 µm inserts respectively, immediately after spinning down the cells and following four days of growth. (C): The mesodermal phenotypes associated most closely with CFC, are shown as a stacked percentage of expression for aggregates that were initially 1–200 cells or non-uniform aggregates, formed from liquid suspension culture (LSC). (D): The phenotypic expression increased with larger aggregate sizes, however, CFC are maximal with 100 cell aggregates. The average number of CFC ± standard error of the mean are shown. Means that do not share a letter are significantly different (n = 6–8; ANOVA with Tukey’s post hoc analysis α = 0.05).
Fig. 3
Fig. 3
Gelatin microparticles do not interfere with cell aggregation and exogenous factors can be delivered locally. (Ai): A schematic of the aggregate formation process with BMP4 loaded gelatin MPs. (Aii): Both type A and type B gelatin MPs with characteristic pI were tested within the cell aggregates as it was expected BMP4 would have different inductive capacities depending on the electrostatic interaction between the gelatin and protein. (B): A range of BMP4 concentrations were tested by calculating the theoretical amount of BMP4 contained/MP. Brachyury (T-GFP) was assessed on day 4.5 (n = 3; one-way ANOVA with Tukey’s post hoc analysis, α = 0.05). Open symbols: daily media exchange. Abbreviations: MP, microparticle; B-MP, BMP4-MP; U-MP, unloaded MP; sB, soluble BMP4; d2ME, daily media exchange from day two. P < 0.05 for comparison of: *daily exchange (open diamond) to sB (filled square) or d2ME of type A/B (filled diamond) gelatin respectively; **24 h sB (filled grey circle) to continuous BMP4 (filled square/diamond); †daily exchange type B (open diamond) to both type A MPs with daily exchange and 24 h sBMP4 (filled grey circle). The MP:cell ratio used to deliver 1–20 ng BMP4 are shown below the axis; representative fluorescent images were taken 24 h after seeding. (C): BMP4 was loaded into type A (square), type B (triangle), and heparinized type A (open square) MPs from 5, 10 or 25 µg/mL stock solutions as indicated and its release into PBS was monitored by ELISA (n = 3). (D): Delivery of 5 ng/well BMP4 with heparinized MPs was as effective as soluble delivery for 100 cell aggregates (n = 4, student’s t-test; 1200 aggregates, 24 well plate, 400 µm). (For better interpretation of the fluoresent images in this figure, the reader is referred to the web version of this article.)
Fig. 4
Fig. 4
Combining local growth factor delivery with low oxygen supports mesodermal development. (Ai): Aggregates formed with blue and red microparticles can be removed and (Aii): encapsulated for further culture. (Aiii): The aggregates respond to media nutrients and soluble factors as this control aggregate seeded with unloaded MPs differentiated with BVT and expressed brachyury (T-GFP). The agarose shell is hightlighted with a dashed white ring. (B): BMP4 and TPO were delivered locally within 100 cell aggregates using heparinized type A gelatin MPs, while VEGF was provided in solution (d0-2) in normoxia. Means that do not share a letter are significantly different (n = 4–8; one-way ANOVA with Tukey’s post hoc analysis α = 0.05). ND- not detected. (C): Assessing the phenotypes of the developing mesoderm at d3.75 as described earlier (ETPF), accurately predicts the general trend in CFC output from day seven cells in both 20% O2 (black bars) and 5% O2 conditions (grey bars). (D): Representative myeloid and erythroid colonies from dual BMP4 and TPO microparticle delivery are shown. Suppl. Figure 5 shows the distribution of colony types for each condition. All scale bars are 200 µm.

Similar articles

See all similar articles

Cited by 21 articles

See all "Cited by" articles

Publication types

MeSH terms

Feedback