Human iPSC-derived MSCs (iMSCs) from aged individuals acquire a rejuvenation signature

Abstract

Background: Primary mesenchymal stem cells (MSCs) are fraught with aging-related shortfalls. Human-induced pluripotent stem cell (iPSC)-derived MSCs (iMSCs) have been shown to be a useful clinically relevant source of MSCs that circumvent these aging-associated drawbacks. To date, the extent of the retention of aging-hallmarks in iMSCs differentiated from iPSCs derived from elderly donors remains unclear.

Methods: Fetal femur-derived MSCs (fMSCs) and adult bone marrow MSCs (aMSCs) were isolated, corresponding iPSCs were generated, and iMSCs were differentiated from fMSC-iPSCs, from aMSC-iPSCs, and from human embryonic stem cells (ESCs) H1. In addition, typical MSC characterization such as cell surface marker expression, differentiation capacity, secretome profile, and trancriptome analysis were conducted for the three distinct iMSC preparations-fMSC-iMSCs, aMSC-iMSCs, and ESC-iMSCs. To verify these results, previously published data sets were used, and also, additional aMSCs and iMSCs were analyzed.

Results: fMSCs and aMSCs both express the typical MSC cell surface markers and can be differentiated into osteogenic, adipogenic, and chondrogenic lineages in vitro. However, the transcriptome analysis revealed overlapping and distinct gene expression patterns and showed that fMSCs express more genes in common with ESCs than with aMSCs. fMSC-iMSCs, aMSC-iMSCs, and ESC-iMSCs met the criteria set out for MSCs. Dendrogram analyses confirmed that the transcriptomes of all iMSCs clustered together with the parental MSCs and separated from the MSC-iPSCs and ESCs. iMSCs irrespective of donor age and cell type acquired a rejuvenation-associated gene signature, specifically, the expression of INHBE, DNMT3B, POU5F1P1, CDKN1C, and GCNT2 which are also expressed in pluripotent stem cells (iPSCs and ESC) but not in the parental aMSCs. iMSCs expressed more genes in common with fMSCs than with aMSCs. Independent real-time PCR comparing aMSCs, fMSCs, and iMSCs confirmed the differential expression of the rejuvenation (COX7A, EZA2, and TMEM119) and aging (CXADR and IGSF3) signatures. Importantly, in terms of regenerative medicine, iMSCs acquired a secretome (e.g., angiogenin, DKK-1, IL-8, PDGF-AA, osteopontin, SERPINE1, and VEGF) similar to that of fMSCs and aMSCs, thus highlighting their ability to act via paracrine signaling.

Conclusions: iMSCs irrespective of donor age and cell source acquire a rejuvenation gene signature. The iMSC concept could allow circumventing the drawbacks associated with the use of adult MSCs und thus provide a promising tool for use in various clinical settings in the future.

Keywords: Aged MSC; Aging; Fetal MSCs; Rejuvenation; Secretome; Transcriptome; iMSCs; iPSCs.

Fig. 1

Characterization of fMSCs and aMSCs. a fMSCs and aMSCs express typical MSC surface marker genes CD73, CD90, and CD105 but no hematopoietic marker genes. b fMSCs and aMSCs show MSC-typical surface marker expression detected by FACS. Blue: fluorophore-conjugated antibody against surface antigen. Gray: isotype control. c Confirmation of typical MSC lineage differentiation potential in fMSCs and aMSC. Alizarin Red staining visualized the calcified matrix in red. Adipogenic: Oil Red O staining was used to visualize fat vacuoles of adipocytes in red. Chondrogenic: Cells of pellet culture were stained with Alcian blue to visualize acidic mucosubstances in blue. Pictures were taken using a stereo microscope. d Overlapping and distinct gene expression between fMSCs, aMSCs, and ESCs with e related KEGG pathways and GO terms depicted as bar diagrams with –log(pValue). KEGG pathways are marked in red and GO terms in blue

Fig. 2

Characterization of MSC-derived iPSCs a Clustering analysis of primary MSCs and corresponding MSC-iPSCs and ESCs. As similarity measure Pearson correlation was used. b Expression of pluripotency-associated genes detected by microarray in iPSCs of fetal and aged background compared to primary MSCs and the ESC line H1

Fig. 3

Characterization of fMSC-iMSCs, aMSC- iMSCs, and ESC-iMSCs. a Morphology of iMSCs compared to native MSCs. b Flow cytometry-based analysis of MSC surface marker in iMSCs. Blue/Purple: iMSCs labeled with antibody specific to marker. Gray: isotype control. c Differentiation potential of iMSCs in vitro. Osteogenic: Alizarin Red S staining; adipogenic: Oil red O staining; chondrogenic: Alcian Blue staining. d Gene expression of MSC, hematopoietic, and pluripotency markers in iMSCs compared to primary MSCs and ESC H1. Representative images of n = 3 experiments

Fig. 4

Distinct and overlapping gene expression patterns between iMSCs and primary MSCs isolated from donors of distinct ages. a Venn diagram-based on expressed genes detected by microarrays of one sample each of fMSC-iMSCs, aMSC-iMSCs, and ESC-iMSCs. MSC-related genes were expressed in all three iMSC preparations. b Clustering dendrogram of Illumina gene expression experiments based on Pearson correlation. One cluster consists of iPSCs and ESCs (red box), and the other separated cluster contains primary MSCs as well as all three iMSC preparations (blue box). c Venn diagram of iMSCs and MSCs from fetal and aged donors based on expressed genes of one sample each d Heatmap showing clustering of primary MSCs and iMSCs for genes related to DNA damage repair. e Heatmap showing clustering of primary MSCs and iMSCs for aging-related genes

Fig. 5

Identification of rejuvenation and aging signature. a Heatmap of genes expressed in MSCs but not in MSC-iMSCs and iPSCs (the aging signature) and of genes expressed in MSC-iMSCs and iPSCs but not in MSCs (the rejuvenation signature). b List of genes of the rejuvenation and aging signature. c Heatmap of genes in the rejuvenation and aging signature including published datasets [24, 25]. d Real-time PCR confirmation of a subset from rejuvenation and aging signature genes employing external MSC and additional iMSCs samples. Relative mRNA expression is shown normalized to expression in fMSCs

Fig. 6

Protein association network related to aging and rejuvenation signatures of gene expression in iMSCs and parental MSCs. Based on the Biogrid database protein interaction, networks have been constructed from the rejuvenation signature (a) and aging signature (b) listed in Fig. 5b. Most of the proteins coded by genes from both signatures can be connected to a network with interactions reported in the Biogrid database. Community clustering of the network via edge-betweenness reveals several communities with densely connected nodes and fewer connections across communities. Communities are indicated by dedicated color shading; edges between communities are colored red. a The rejuvenation signature PAN includes communities characterized by INHBE (blue), TP53, CDKN1C, IL32 (light blue), CDK10 (petrol), ELAVL1 (purple), DNMT3B (yellow), and EEF1A2 (green). b The aging signature PAN includes communities characterized by HSPA5 (blue), CCDC8 (purple), CAT (petrol), EYA2 (yellow), APP (green), and TGFB1 (red)

Fig. 7

Comparative analyses of the secretome of fMSCs, iMSCs, and aMSCs. a Cytokine expression of fMSC1, fMSC2, fMSC3, and iMSCs (fMSC-iMSCs, aMSC-iMSCs, and ESC-iMSCs) detected using membrane-based cytokine arrays. Expression plot of the most abundant cytokines shared between fMSCs and iMSCs; threshold of expression in comparison to reference spots was set to 20%. b KEGG pathway analysis of common cytokines between fMSCs and iMSCs–log(pValue). c GO terms associated with common secreted cytokines between fMSCs and iMSCs as –log(pValue). d Cytokine expression of fMSCs, iMSCs, and aMSCs (aMSC5, aMSC6, aMSC7)