RNA-Based Strategies for Cardiac Reprogramming of Human Mesenchymal Stromal Cells

Abstract

Multipotent adult mesenchymal stromal cells (MSCs) could represent an elegant source for the generation of patient-specific cardiomyocytes needed for regenerative medicine, cardiovascular research, and pharmacological studies. However, the differentiation of adult MSC into a cardiac lineage is challenging compared to embryonic stem cells or induced pluripotent stem cells. Here we used non-integrative methods, including microRNA and mRNA, for cardiac reprogramming of adult MSC derived from bone marrow, dental follicle, and adipose tissue. We found that MSC derived from adipose tissue can partly be reprogrammed into the cardiac lineage by transient overexpression of GATA4, TBX5, MEF2C, and MESP1, while cells isolated from bone marrow, and dental follicle exhibit only weak reprogramming efficiency. qRT-PCR and transcriptomic analysis revealed activation of a cardiac-specific gene program and up-regulation of genes known to promote cardiac development. Although we did not observe the formation of fully mature cardiomyocytes, our data suggests that adult MSC have the capability to acquire a cardiac-like phenotype when treated with mRNA coding for transcription factors that regulate heart development. Yet, further optimization of the reprogramming process is mandatory to increase the reprogramming efficiency.

Keywords: cardiac differentiation; cardiac reprogramming; mRNA; mesenchymal stromal cells (MSC); miRNA.

Figure 1

Phenotype-related and functional characterization of mesenchymal stromal cells (MSC): (A) Flow cytometric measurements revealed a high expression of common MSC surface markers (CD29, CD44, CD73, CD90, CD105), while very low levels were found for hematopoietic surface markers (CD45 and CD117). Representative flow cytometry charts of adipose tissue-derived MSC (adMSC) demonstrate the expression level of surface markers. Blue histograms represent measurement of CD surface marker with corresponding isotype control, shown in red. (B) Tri-lineage differentiation assay indicated adipogenic, osteogenic, and chondrogenic differentiation of MSC. Detection of adipocytes was performed by labelling of FABP4, while osteocytes and chondrocytes were identified by fluorescence staining of osteocalcein and aggrecan, respectively. Scale bar: 50 µm. Results in (A) are shown as mean ± SEM, obtained by analysis of three different donors for each MSC cell type.

Figure 2

Comparative microarray analysis of undifferentiated dental follicle stem cells (DFSCs), bone marrow (BM) MSC, and adMSC. (A) Comparison of signal intensity for .cel files (blue) and .chp files (red) after normalization demonstrates sufficient data quality. (B) MSC from different sources are clearly distinct in regard to their transcription profile. A high patient-dependent variety was found for BM MSC, while adMSC and DFSCs demonstrate a more homogenous distribution. (C) Venn diagram visualizes expressed genes overlapping between different MSC cell types. (D) The numbers of up- and down-regulated transcripts is significantly differentially expressed in all three cell types.

Figure 3

miRNA transfection and programming efficiency in MSC. (A) Uptake of miRNA was determined using Cy3-labelled miRNA and flow cytometry. (B) Detection of dead cells revealed low cytotoxicity induced by miRNA transfection. (C) Relative expression of cardiac marker genes among all tested cell types, four weeks after transfection and cultivation under different culture conditions. Reprogramming efficiency with cardiac induction medium I, II and myo-miRNAs (miR-1, miR-499, miR-208, miR-133) resulted in an up-regulation of cardiac specific markers in all types of MSC, while most profound up-regulation was found for cardiac induction medium II. Among tested MSC, the strongest increase of cardiac gene expression was observed for adMSC. Note, no beneficial effects on cardiac programming were observed following myo-miRNA transfection. Data are shown as mean ± SEM, obtained from three donors for each MSC type. Statistical analysis was performed using ANOVA test, followed by Bonferroni post-hoc analysis. * p ≤ 0.5, ** p ≤ 0.05, *** p ≤ 0.001.

Figure 4

mRNA-based cardiac programming of adMSC. (A) Concentration-dependent expression of transfected mRNAs was evaluated with mRNA coding GFP. The quantative flow cytometry analysis demonstrated maximum transfection efficiency of ~80% when ≤ 1000 ng mRNA were applied. (B) Representative scatterplots of control cells (left) and cells transfected with GFP mRNA (right). (C) Corresponding microscopy images of cells expressing GFP following mRNA treatment. (D) Cytotoxic effects were only induced when mRNA amounts higher than 1000 ng were used for transfection. (E) Compared to untreated control cells, higher gene expression levels of selected cardiac markers were detected for all reprogramming conditions, in particular for α-actinin. (F) Immunolabeling of cells using anti α-actinin antibody results in a faint fluorescence signal in cells transfected with MESP1 and GATA4, MEF2C, and TBX5 (GMT) mRNAs, Scale Bar: 25 µm. (G) Moreover, GMT treated cells also demonstrated protein expression of MEF2C, an early cardiac transcription factor. Flow cytometry and qRT-PCR data are shown as mean ± SEM, obtained from three different donors. Statistical analysis was performed using one-way ANOVA. * p ≤ 0.5, ** p ≤ 0.05, *** p ≤ 0.001.

Figure 5

Transcriptome based comparison of reprogrammed adMSC. (A) Quality control of microarray data. Box plot of signal intensity of performed microarrays on .cel (blue) and .chp files normalization (red) confirm good data quality. (B) Principal component analysis (PCA) demonstrates clustering of treated groups, clearly showing the impact of respective reprogramming conditions on the transcriptomic profile compared to control cells (blue). Yet, cells subjected to MESP1 (green), GMT (purple) or cardiac induction medium II solely (red) remain distinguishable. (C) Up-and down-regulated transcripts and corresponding Venn diagram (D) showing the impact of reprogrammed cells compared to control. Most differentially expressed transcripts were regulated by all three reprogramming treatments (2828 genes), while 1816 transcripts are shared by GMT vs. control and MESP1 vs. control. (E) Detailed comparison of common and distinct up-regulated (red) and down-regulated (green) transcripts among the three reprogrammed groups. The differences found for optimized medium vs. MESP1 and GMT transfections are much more prominent than the differences between MESP1 and GMT.

Figure 6

The impact of reprogramming on cardiac-differentiation pathways. Up-regulated and down-regulated transcripts of respective programming conditions are labelled in red or green color. (A,B) Strongest up-regulation of transcripts involved in cardiac development ((A) heart development, (B) cardiac progenitor differentiation) was mainly found in GMT reprogrammed cells, followed by MESP1 treatment and cardiac induction medium II. Key cardiac transcription factors and signaling molecules were significantly up-regulated, including TBX5, GATA4, MEF2C, HAND2, BMP4, and IGF.