Cost-Effective Mechanical Aggregation of Cardiac Progenitors and Encapsulation in Matrigel Support Self-Organization in a Dynamic Culture Environment

Abstract

Human iPSC-derived self-organized cardiac tissues can be valuable for the development of platforms for disease modeling and drug screening, enhancing test accuracy and reducing pharmaceutical industry financial burden. However, current differentiation systems still rely on static culture conditions and specialized commercial microwells for aggregation, which hinders the full potential of hiPSC-derived cardiac tissues. Herein, we integrate cost-effective and reproducible manual aggregation of hiPSC-derived cardiac progenitors with Matrigel encapsulation and a dynamic culture to support hiPSC cardiac differentiation and self-organization. Manual aggregation at day 7 of cardiac differentiation resulted in 97% of beating aggregates with 78% of cTnT-positive cells. Matrigel encapsulation conjugated with a dynamic culture promoted cell migration and the creation of organized structures, with observed cell polarization and the creation of lumens. In addition, encapsulation increased buoyancy and decreased coalescence of the hiPSC-derived cardiac aggregates. Moreover, VEGF supplementation increased over two-fold the percentage of CD31-positive cells resulting in the emergence of microvessel-like structures. Thus, this study shows that the explored culture parameters support the self-organization of hiPSC-derived cardiac microtissues containing multiple cardiac cell types. Additional stimuli (e.g., BMP) in long-term scalable and fully automatized cultures can further potentiate highly structured and mature hiPSC-derived cardiac models, contributing to the development of reliable platforms for high-throughput drug screening and disease modeling.

Keywords: Matrigel encapsulation; WNT signaling modulation; cardiac differentiation; dynamic culture; manual aggregation; self-organization.

Figure 1

Manual induction of aggregation is reliable and cost-effective when performed at day 7. (A) Aggregate size distribution for mechanical aggregation performed at day 7, day 12, or day 15 of cardiac differentiation. Aggregation at day 7 and day 12 promoted a higher frequency of aggregates with a size of 50 and 100 µm. At day 15, aggregation yielded a higher variance population. The lines correspond to centered sixth order polynomials that model size distribution. Error bars, SEM, n = 3. (B) Percentage of beating aggregates and percentage of cells expressing cTnT (determined by flow cytometry) at day 18 of cardiac differentiation. Aggregation at day 7 resulted in an average of 97% aggregates with a beating phenotype and 53% cTnT-positive cells. Error bars, SEM, n = 3. * p-value < 0.05, ** p-value < 0.01, (ANOVA with Tukey’s test). (C,D) Aggregates at day 10 (C) and day 15 (D) of differentiation after aggregate induction at day 7. Some aggregates appeared to develop structures with different depths (red arrows). Scale bar: 100 µm.

Figure 2

Matrigel encapsulation and a dynamic culture environment promotes cell migration and organization. (A) Scheme describing the experimental procedure. The WNT signaling modulation protocol was initiated in monolayer (I) with aggregate formation performed manually at day 7 (II). Encapsulation of aggregates with Matrigel was performed at day 9 (III) with encapsulated aggregates being cultured in dynamic conditions up to 30 days (IV). (B) H&E staining and immunohistochemistry panel of anti-cTnT, anti-CD31, anti-CD34, and anti-SMA markers of control and encapsulated aggregates after culture in dynamic conditions (n = 3). H&E staining highlighted organized structures in encapsulated aggregates, in some cases showing cell polarization with the creation of lumens (black arrows) and cell migration into the Matrigel matrix. Scale bar: 100 µm, top and bottom, 20 µm mid.

Figure 3

Aggregate characterization by analysis of lineage markers using real-time PCR, flow cytometry, and calcium transient microscopy. (A) Analysis of lineage markers gene expression by real-time PCR. Most genes expression was not significantly different between the monolayer (MONO), aggregates (AGG), and aggregates supplemented with VEGF (VEGF). Addition of VEGF significantly increased the expression of CD31. Error bars, SEM, n = 3. * p-value < 0.05, ** p-value < 0.01 (2-way ANOVA with Tukey’s test). (B) Flow cytometry showed an increase in cTnT-positive cells for aggregates without VEGF compared with the monolayer and aggregates supplemented with VEGF. Error bars, SEM, n = 6 for MONO, n = 4 for AGG and n = 3 for VEGF. ** p-value < 0.01 (Welch’s t-test). (C) CD31 protein expression increased over two-fold when a double dosage of VEGF was supplemented compared with the monolayer and aggregates without VEGF. Error bars, SEM, n = 7 for MONO, n = 6 for AGG, and n = 3 for VEGF. * p-value < 0.05 (Welch’s t-test). (D) Calcium transient analysis of aggregates supplemented with VEGF. The average time between calcium transients demonstrates a strong statistically significant decrease from 9.1 s to 4.0 s when isoproterenol (+Iso) was used. Carbachol (+Carb) showed no statistically significant effects (7.3 s). Error bars, SEM, n = 10 for VEGF, n = 8 for +Iso, n = 7 for +Carb. ** p-value < 0.01, *** p-value < 0.001 (Welch’s t-test). (E) Representative calcium transient profiles. Left, aggregates supplemented with VEGF (8.5 s, SD ± 0.1), middle, when 1 µM isoproterenol was supplemented (4.0 s, SD ± 0.1), and right, when 1 µM carbachol was supplemented (8.0 s, SD ± 0.1).

Figure 4

Confocal fluorescence microscopy of aggregates supplemented with VEGF showed the presence of microvessel-like structures. Representative 3D projection of aggregates supplemented with VEGF stained with cTnT and DAPI or CD31 and DAPI. Supplementation of VEGF maintained tissue complexity and cTnT expression, while resulting in the development of CD31-positive microvessel-like structures. Scale bar: 25 µm.