Chemical-defined medium supporting the expansion of human mesenchymal …

Lots of efforts have been made to uncover the critical components of the platelet lysate for supporting the MSCs expansion [17,18,19,20,21,22,23,24]. Unfortunately, it remains largely unclear what constituents mainly contribute to the MSCs proliferation. To develop the full chemical-defined medium for supporting the proliferation of human MSCs, we investigated the critical regulators, which are responsible for the MSCs proliferation, through analyzing the responses of the MSCs to the platelet lysate via RNA sequencing and bioinformatic analysis, instead of analyzing the platelet lysate itself. Therefore, the transcriptome of the MSCs was analyzed after exposing to different concentrations of human platelet lysate.

The human platelet lysate had the dose effects on the proliferation of human MSCs (Fig.1a). The PCA showed that the human MSCs treated with 1% and 2% platelet lysate had similar gene expression patterns while the 5% platelet lysate conferred human MSCs as a very different pattern (Fig.1b). Therefore, the signal pathways or important genes responsible for human MSC proliferation should also have the dose effects. Gene expression pattern clustering among the samples treated without human platelet (negative control (NC)) or with different concentrations of human platelet lysate (1%, 2%, and 5% PL) was performed (Supplementary Fig.1). Both upregulation and downregulation patterns were further analyzed, because the key components of the platelet lysate might have the positive regulations or negative feedback regulations on the gene expression of MSCs. Thus, the KEGG pathway enrichment analysis was performed for genes from clusters 2, 3, 5, 7, 8, 9, and 11 (Fig.1c, Supplementary Fig.2). Among all these important pathways activated or suppressed by the human PL, the MAPK (mitogen-activated protein kinase) signal pathway was further studied as it has been demonstrated as one key pathway for cell proliferation [25]. Protein-protein interaction network analysis showed that a panel of important growth factors or receptors was involved in the MAPK pathways which might contribute to the supportive role of PL for human MSC proliferation (Fig.1d).

Transcriptome analysis of the human MSCs exposed to different concentrations of human platelet lysate. a Cell number counting after human MSCs exposed to 1%, 2%, and 5% PL in DMEM. Cells were plated onto p12 plates at 2104 cells per well and the cell number was counted 5days later. b PCA of human MSCs exposed to different concentrations of human platelet lysate. c KEGG pathway enrichment of cluster 2 derived from the gene expression pattern clustering analysis. d Protein-protein interaction analysis of the genes derived from the MAPK signaling pathway of cluster 2. NC, negative control, Dulbeccos modified Eagles medium-high glucose; PL, human platelet lysate; An asterisk indicates P<0.05

To analyze these potential important growth factors uncovered from the RNA-sequencing analysis, several basic mediums were firstly tested. Ideally, the basic medium should support the human MSCs alive or proliferation and also have the potential to develop as a full chemical-defined medium. The N2B27 medium is a serum-free medium suitable for neural stem cell and pluripotent stem cell proliferation [26, 27]. Thus, two types of N2B27 were compared, with or without vitamin A. Data showed that the N2B27-VA (N2B27 without vitamin A) could support the human MSC proliferation, much better than the N2B27 with vitamin A or FBS (fetal bovine serum) (Fig.2a). Although the recipe of commercial N2B27 is unavailable to the public, the more original recipe and modified version of N2B27 had been published [28, 29]. Then the BSA (bovine serum albumin) component of the N2B27-VA was replaced by recombinant human albumin, termed as the NB medium (Table1). Data showed that the NB medium was comparable to the N2B27-VA medium (Fig. 2A). Then 12 growth factors uncovered from the RNA-sequencing analysis were added to the NB medium to assess whether they could support the human MSC proliferation. Furthermore, we also tested two factors, BIO (6-bromoindirubin-3-oxime, the GSK3 inhibitor) and OSM (oncostatin M) that were used for cardiac progenitor derivation which also expresses CD105 [30]. Totally, 14 growth factors or chemicals were evaluated. Among them, the Vc (vitamin C), PDGF-AB (platelet-derived growth factor AB), bFGF (basic fibroblast growth factor), and EGF (epidermal growth factor) showed beneficial effects on MSC proliferation while BIO, OSM, and PDGF-BB significantly suppressed their expansion (Fig.2b). The four-factor combination (4F), including Vc, PDGF-AB, bFGF, and EGF, showed similar beneficial effects on MSC proliferation when compared with combining them together (the eleven-factor combination, excluding BIO, OSM, and PDGF-BB, 11F) (Fig.2c). And further eliminating the PDGF-AB did not affect the MSCs proliferation while eliminating the Vc, bFGF or EGF significantly suppressed their proliferation (Fig.2c). Thus, the 3F (3 basic factors), including Vc, bFGF, and EGF, is the minimal combination for supporting the human MSCs expansion, as eliminating anyone of them could significantly inhibit the cell proliferation (Fig.2d). Although the NB medium plus 3F, termed as the NBVbe medium, had longer doubling time than the 5% PL, it was much better than the FBS-based medium (Fig.2e).

Full chemical-defined medium development. a Basal medium comparison. Human MSCs were plated onto p12 plates at 2104 cells per well and the cell number was counted 5days later. bd Growth factor and chemical screening. Human MSCs were plated onto p12 plates at 2104 cells per well and the cell number was counted 5days later. e Cell doubling time analysis for the human MSCs cultured with the FBS, 5% PL, and NBVbe medium. The cells were continuously cultured for 5 passages. FBS. fetal bovine serum; NB, the N2B27-VA medium with the BSA (bovine serum albumin) replaced by the recombinant human albumin; PL, platelet lysate; BIO, 6-bromoindirubin-3-oxime; OSM, oncostatin M; Vc, vitamin C; PDGF, platelet-derived growth factor; bFGF, basic fibroblast growth factor; HGF, hepatocyte growth factor; EGF, epidermal growth factor; IGF, insulin-like growth factor 1; BDNF, brain-derived neurotrophic factor; NGF, nerve growth factor; VEGFA, vascular endothelial growth factor A; TGFb2, transforming growth factor beta 2; 11F, Vc + PDGF-AA + PDGF-AB + bFGF + HGF + EGF + IGF + BDNF + NGF + VEGFA + TGFb2; 4F, Vc + PDGF-AB + bFGF + EGF; 3F, Vc + bFGF + EGF. An asterisk indicates P<0.05; N.S. indicates non-significant

To further confirm that the NBVbe medium could really support the expansion of human MSCs, human MSCs were characterized after cultured with the NBVbe medium for 5 passages. They showed more fibroblast-like morphology (Fig.3a) and were positive for MSC cell surface marker CD73, CD90, and CD105 while negative for HLA-DAR, CD45, CD34, CD19, and CD11b (Fig.3b). The human MSCs cultured with the NBVbe medium could form more colonies than the platelet lysate medium (Fig.3c, d) and differentiate into adipocytes, osteocytes, and chondrocytes (Fig.3e).

Characterization of human MSCs cultured with the NBVbe medium. a The human MSCs cultured with NBVbe medium had the different cell morphology from them cultured with the 5% PL medium. Scale bar, 100m. b Cell surface marker analysis by flow cytometry. c Colony formation assay for human MSCs cultured with the NBVbe medium or 5% PL medium. Scale bar for the upper panel, 1cm. Scale bar for the lower panel, 200m. d Colony number counting for human MSCs cultured with the NBVbe medium or 5% PL medium for 5 passages. e The human MSCs were directed to adipocytes, osteocytes, and chondrocytes after cultured with the NBVbe medium for 5 passages. Left: Oil Red O staining and anti-FABP4 immuno-staining for the differentiated adipocytes; middle: Alizarin Red S staining and anti-Osteocalcin immunostaining for the differentiated osteocytes; right: Alcian blue staining and anti-Aggrecan immunostaining for the differentiated chondrocytes. Scale bar, 200m. An asterisk indicates P<0.05

Because the MSCs cultured in the NBVbe medium were less proliferative than the PL medium, we were wondering whether further improvements could be made. RNA sequencing showed that the MSCs cultured with the NBVbe medium had a very different gene expression pattern when comparing them in culture medium containing 1%, 2%, or 5% PL (Fig.4a). And gene clustering analysis indicated that the NBVbe medium might be more closed to the medium containing 1% platelet lysate (Fig.4b). Indeed, there were still a large amount of differentially expressed genes between the MSCs cultured in the NBVbe medium and 5% PL (Fig.4c). KEGG pathway enrichment analysis for these differentially expressed genes indicated that they were mostly involved into the cell cycle and DNA replication pathways (Fig.4d). Protein-protein interaction network analysis uncovered 15 cytokines might support the proliferation of human MSCs (Supplementary Fig.3A, 3B). Unfortunately, these cytokines could not further promote the MSCs proliferation in both the NB medium or the NB medium plus 3F (Supplementary Fig.3C, 3D). Interestingly, the DNA repair pathway was also activated by the PL medium (Fig.4d), which indicated that the 5% PL might induce genome instability, the adverse effects alongside the fast cell proliferation. RNA sequencing also revealed that the cells cultured with 5% PL had more alternative splicing events than those cells cultured with the NBVbe medium (Fig.4e, f). Furthermore, the DNA exon sequencing revealed that the MSCs cultured with 5% PL might have more SNP (single-nucleotide polymorphism) events (Fig.4g). Thus, the NBVbe medium might introduce less DNA mutations than PL.

The RNA profile comparison between the human MSCs cultured with NBVbe or 5% PL. a Venn diagram showing the numbers of genes differentially regulated in human MSCs cultured with NBVbe, 1% PL, 2%PL, or 5% PL medium. b Hierarchical clustering analysis of mRNA profiles in human MSCs cultured with NBVbe, 1% PL, 2%PL, or 5% PL medium. c Differentially expressed genes between the human MSCs cultured with NBVbe and 5%PL. d KEGG pathway enrichment for differentially expressed genes between the human MSCs cultured with NBVbe and 5% PL. e Alternative splicing events detected by RNA sequencing. f Alternative splicing events compared with the human MSCs cultured with DMEM via RNA sequencing analysis. g SNV (single-nucleotide variant) and InDel (insertion and deletion mutations) detected by DNA exon sequencing. NC, negative control, human MSCs cultured with DMEM; DEGs, differentially expressed genes; A5SS, alternative 5 splicing site; A3SS, alternative 3 splicing site; MXE, mutually exclusive exons; RI, retained intron; SE, skipped exon

Thus, the full chemical-defined medium for expanding the human MSCs was developed. Then we were wondering whether it also supports the isolation of MSCs from human umbilical cord tissues. Several dissociation enzymes, reagents, and different combinations were assessed to compare their efficiency on the isolation of human MSCs. Among them, the collagenase B was the most efficient enzyme for human MSC isolation from the umbilical cords in the NBVbe medium (Fig.5a). Concentration optimization showed that the 2mg/mL collagenase B was the most efficient concentration (Fig.5B). The isolated human MSCs were fibroblast-like cells and could be expanded with the NBVbe medium (Fig.5c). They were positive for CD73, CD90, and CD105 expression and negative for HLA-DR, CD45, CD34, CD19, and CD11b expression (Fig.5d). Directed differentiation assay showed that these human MSCs could be differentiated into adipocytes, osteocytes, and chondrocytes (Fig.5e). Therefore, the NBVbe medium could be used for the isolation of human MSCs from the umbilical cords.

The NBVbe medium supports the human MSC isolation from the umbilical cords. a Cell isolation reagents comparison. Cell number was counted 30days after the cell isolation. b Human MSCs were isolated with the different concentration of collagenase B. Cell number was counted 30days after the cell isolation. c Human MSCs isolated from the umbilical cord, passage 0 and 6 cultured with the NBVbe medium. d Flow cytometry analysis of the human MSCs after isolated and cultured with the NBVbe medium. e Human MSCs were isolated and expanded with the NBVbe medium and then differentiated into adipocytes, osteocytes, and chondrocytes. Left: Oil Red O staining and anti- FABP4 immunostaining for the differentiated adipocytes; middle: Alizarin Red S staining and anti-Osteocalcin immunostaining for the differentiated osteocytes; right: Alcian blue staining and anti-Aggrecan immunostaining for the differentiated chondrocytes. Scale bar, 200m. An asterisk indicates P<0.05. GCD, gentle cell dissociation reagent; SR, ReLeSR; CDB, cell dissociation buffer, enzyme-free; ACF, animal component-free cell dissociation kit

In conclusion, we have developed the full chemical-defined medium for isolation and expansion of the human MSCs. The human MSCs cultured with the NBVbe medium had the self-renewal and differentiation abilities with expressing MSC markers, although the bFGF and vitamin C had been demonstrated as effective factors to support the human MSCs proliferation in the conventional medium [31,32,33]. The NBVbe medium we developed here is the first chemical-defined medium supporting the proliferation of human MSCs with the formulation available to the public. Moreover, this would be very helpful for molecular mechanism studies of the human MSCs, including the cell identity characterization. However, the NBVbe medium was not as good as the human platelet in supporting the proliferation of human MSCs. Thus, more effects should be made to further optimize the chemical-defined medium.

Go here to see the original:

Chemical-defined medium supporting the expansion of human mesenchymal ...