miR-301b~miR-130b—PPARγ axis underlies the adipogenic capacity of mesenchymal stem cells with different tissue origins

Mesenchymal stem cells (MSCs) have been widely used in regenerative medicine and cellular therapy due to their multi-lineage differentiation potential and immunomodulatory function. The applicability of MSCs also depends on their cellular sources and in vivo functions. Here in this study, we systematically compared the morphologic characteristics, immunophenotypes and the adipogenic differentiation of MSCs derived from umbilical cord (UC), adipose tissue (Ad) and bone marrow (BM). We found that the three tissues-derived MSCs displayed decreased adipogenic capacity in the order: Ad-MSC > BM-MSC > UC-MSC, and no morphologic and immunophenotypic differences were observed. Mechanistic investigation revealed a miR-301b~miR-130b—PPARγ axis, whose expression pattern in UC-MSC, Ad-MSC and BM-MSC significantly correlates with their adipogenic capacity. Our results come up with a potential mechanism to elucidate the differential adipogenesis of Ad-MSC, BM-MSC and UC-MSC, which would provide instructional advice for which source of MSCs to choose according to a certain clinical purpose. Furthermore, the miR-301b~miR-130b—PPARγ axis may also be used as a potential therapeutic target for the disorders associated with MSCs-mediated abnormal adipogenesis.


Immunophenotype characterization of the three tissues-derived MSCs.
To determine whether the adherent cells from umbilical cord, adipose tissue and bone marrow met the quantifying criteria of MSCs, we analyzed the expression of surface antigen (CD73, CD90, CD105, CD45, CD34 and CD19). As shown in Fig. 2, stromal cell markers (CD73, CD90 and CD105) were expressed in UC-MSC, Ad-MSC and BM-MSC, with a high positivity rate. However, hematopoietic cell markers (such as CD45, CD34 and CD19) were not expressed. All the surface markers expression detected in the three tissues-derived MSCs conformed to the criteria of MSCs established before 19 and no comparable differences were observed among UC-MSC, Ad-MSC and BM-MSC.

Adipogenic differentiation of UC-MSC, Ad-MSC and BM-MSC. Adipogenic differentiation under
certain condition is a basic feature of MSCs. Multiple comparative studies have been performed about the MSCs' multi-lineage differentiation potential including adipogenic differentiation 15 . However, the adipogenic capacity of UC-MSC, Ad-MSC and BM-MSC is still controversial. Here, passage 3 or passage 4 of UC-MSC, Ad-MSC and BM-MSC were chosen and seeded into 12-well plate followed by adipogenic induction for 14 days when the cells reached 90% confluence. Then the cells were fixed and stained with oil red O. As presented in Fig. 3A-C, UC-MSC, Ad-MSC and BM-MSC all showed increased lipid vesicles at differentiation day 14 compared with their respective controls (differentiation day 0). At differentiation day 0, no obvious differences of oil red O staining among the three tissues-derived MSCs were observed. However, at differentiation day 14, the intensity of oil red O staining can be arranged from strong to weak: Ad-MSC > BM-MSC > UC-MSC (Fig. 3D). Next, the adipogenic differentiation markers (PPARγ, FABP4, PLIN1 and LPL) were detected using semi-quantitative RT-PCR at differentiation day 0 and day 14 (Fig. 4A). And the gray density of the four genes expression were also analyzed and presented in Fig. 4B-D, which, to some degree, further confirmed the adipogenic capacity of the three   MiR-301b~miR-130b-PPARγ expression pattern correlates with adipogenic capacity of UC-MSC, Ad-MSC and BM-MSC. To reveal the potential mechanism underlying the differences of adipogenic capacity of UC-MSC, Ad-MSC and BM-MSC, we unexpectedly found that the original expression of PPARγ (Fig. 4A,B), which is an essential transcription factor for adipogenesis 20 , in the three tissues-derived MSCs, positively correlates with their adipogenic potential (Fig. 4F). Next we want to know what determines the differential expression of PPARγ in UC-MSC, Ad-MSC and BM-MSC. We first turned to microRNAs which are a class of 21-23 nt small non-coding RNAs and can post-transcriptionally regulate gene expression by binding to the 3′-untranslated region (3′UTR) of target mRNAs through their seed sequences 21 . MiRNAs have also been intensively studied in the past few years and can participate in the regulation of many physiological processes including adipogenesis 22 . Thus, using TargetScan and PicTar web-tools, we predicted the miRNAs that may target PPARγ expression. Combined with miRNAs that have been reported to target PPARγ during adipogenesis, we chose some for subsequent experimental validation and found that miR-301b~miR-130b cluster may act as a potential regulator of PPARγ expression and further the adipogenesis of UC-MSC, Ad-MSC and BM-MSC. The binding of miR-301b~miR-130b cluster to the 3′UTR of PPARγ was presented in Fig. 5A. And the expression of PPARγ and miR-301b~miR-130b cluster in UC-MSC, Ad-MSC and BM-MSC were detected by real-time quantitative PCR and presented in Fig. 5B-D. PPARγ expression was in agreement with the semi-quantitative PCR result (Fig. 5B). The expression of miR-301b increased in the order: Ad-MSC < BM-MSC < UC-MSC (Fig. 5C), which is contrary to that of PPARγ. MiR-130b presented high expression in UC-MSC, but displayed similar expression in Ad-MSC and BM-MSC (Fig. 5D). Linear regression analysis showed that miR-301b~miR-130b cluster expression presented significant negative correlation with PPARγ expression and the adipogenic capacity of UC-MSC, Ad-MSC and BM-MSC (Fig. 5E), which, to some degree, revealed a possible mechanism to elucidate the varied adipogenic potentials of the three tissues-derived MSCs.

Discussion
MSCs could be isolated from many adult tissues and fetal appendages without too much ethical controversy and have been widely used to repair or regenerate damaged tissues and treat immune diseases in clinical trials also due to their multi-lineage differentiation potential and immunomodulatory function 7,23 . MSCs derived from different tissues may have varied functional properties in spite of the similar morphologic characteristics and immunophenotypes, which make it complicated and confusing about what a source of MSCs to choose for a certain clinical purpose 15,17 . Many comparative studies have been performed to investigate the proliferation, differentiation and immune regulation of MSCs derived from different tissues. For example, Ad-MSCs have been shown to possess higher proliferative and angiogenic capabilities in vitro as compared to the BM-MSCs 15, 24 . Choong et al. reported that BM-MSCs were also shown to undergo senescence at an earlier population doubling than the adipose and umbilical cord tissue-derived stem cells from equine tissue 25 . Placenta-derived MSCs were shown to have a lower potential to undergo adipogenesis but have a higher potential to undergo osteogenesis than BM-MSCs and Ad-MSCs 17 . However, the comparative adipogenic potentials of UC-MSC, Ad-MSC and BM-MSC are still controversial and the underlying molecular mechanism remains to be explored. Here in this study, we systematically compared the morphologic characteristics, surface antigen expression and adipogenic differentiation of UC-MSC, Ad-MSC and BM-MSC. We found that the three tissues-derived MSCs have similar morphology and immunophenotypes (Figs 1 and 2) as reported, but displayed decreased adipogenic capacity in the order Ad-MSC > BM-MSC > UC-MSC (Figs 3 and 4).
Adipogenesis is an important physiological process to produce adipocytes which play a vital role in energy homeostasis and process the largest energy reserve as triglycerol in the body of animals 26 . MSCs mediated adipogenesis is also essential to maintain the local microenvironment of the organisms and has been intensively studied in the past few decades [27][28][29] . The process is tightly controlled by transcription factors and noncoding RNAs [30][31][32] . To reveal the underlying mechanism of differential adipogenic capacity of UC-MSC, Ad-MSC and BM-MSC, we unexpectedly found that PPARγ mRNA expression in the three tissues-derived MSCs positively correlates with their adipogenic capacity. PPARγ has been acknowledged as an essential transcription factor for adipogenic lineage commitment 20,33 .
To further seek what determines the differential expression of PPARγ in UC-MSC, Ad-MSC and BM-MSC, we first turned to the reported miRNAs which play important roles in regulating adipogenesis. Many miRNAs have been identified in different models of adipogenic differentiation, and miR-27 and miR-130b were reported to inhibit PPARγ expression during MSCs-mediated adipogenesis [34][35][36] . MiR-27 and miR-130b were first chosen and their expression was detected in UC-MSC, Ad-MSC and BM-MSC. No significant negative correlation was observed between miR-27 expression and PPARγ expression in the three tissues-derived MSCs ( Supplementary  Fig. 2). Fortunately, miR-130b had inversed expression with that of PPARγ in UC-MSC and Ad-MSC, whereas  (Fig. 5D). MiR-130/miR-301/miR-454 family has been reported to be regulated together to exert a biological effect because of their same seed sequences 37 . Thus, the other family members, miR-130a, miR-301a, miR-301b and miR-454, were also chosen and their expression was detected in the three tissues-derived MSCs. Only miR-301a and miR-301b displayed a similar expression pattern as miR-130b. However, linear regression analysis showed no significant negative correlation between PPARγ expression and the respective expression of all the family members including miR-130b ( Supplementary Fig. 3). Both miR-301b and miR-130b are located in chromosome 22 and belong to the miR-301b~miR-130b cluster 38,39 . When they were analyzed as a cluster, their expression displayed a significant negative correlation with PPARγ expression (Fig. 5E). We first uncovered a miRNA cluster whose expression underlies PPARγ expression and adipogenic potential of the three tissues-derived MSCs.
Our study demonstrated that MSCs derived from umbilical cord, adipose tissue and bone marrow displayed decreased adipogenic capacity in the order Ad-MSC > BM-MSC > UC-MSC. We also established a miR-301b~miR-130b-PPARγ axis whose expression pattern has a significant correlation with the adipogenic potential of the three tissues-derived MSCs, which may provide a potential therapeutic target for the disorders associated with MSCs-mediated abnormal adipogenesis, and provide instructional advice for which source of MSCs to choose according to a certain clinical purpose.

Methods
Isolation and culture of human MSCs. Human umbilical cord, adipose tissue and bone marrow from healthy donors were collected from the affiliated hospital of Jining Medical University. All human MSC studies were approved by the Ethics Committees of the hospitals and the Institutional Review Board of Jining Medical University and carried out in accordance with their approved guidelines and all participants provided written informed consents. MSCs were isolated from the three tissues as previously described [40][41][42] . All the cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) (Gibco). Adipogenic differentiation. Adipogenic differentiation was performed according to the method described elsewhere with minor modifications 40 . The MSCs were seeded into 12/6-well plates. Upon reaching confluence, the MSCs were changed into adipogenic medium composed of DMEM with 10% FBS, 1 µM dexamethasone, 0.01 mg/mL insulin, 100 µg/mL indomethacin and 0.5 mM 3-isobutyl-1-methyl-xanthine for 14 days (all purchased from Sigma, St. Louis, Mo, USA).
Oil red O staining. The cells were rinsed with PBS twice in the plates after discarding the supernatant and fixed with 4% paraformaldehyde at room temperature for 20 min. The cells were then washed with PBS and stained with oil red O (Solarbio) for 20 min followed by washing with PBS. Lipid vesicles were observed and photographed using an IX71 Olympus microscope (Olympus, Tokyo, Japan). Quantification of the staining was performed using Image-Pro Plus software.

RNA extraction and RT-PCR analysis.
Total RNA was extracted from cell samples using TRIzol Reagent (Invitrogen) and quantified using the NanoDrop 2000 spectrophotometer (Thermo Scientific, Bremen, Germany). The first strand of cDNA was synthesized using M-MLV reverse transcriptase (Invitrogen) according to the manufacturer's instruction. Oligo (dT) was used as the primers for reverse transcription of mRNA. Stem-loop RT primers were used for the reverse transcription of miRNAs. Actin and U6 were used as their respective controls. Semi-quantitative PCR and real-time quantitative PCR were performed in ABI PCR machines using pre-stained mix (Tiangen) and SYBR Premix (TransGen Biotech) respectively. The primers used for reverse transcription and RT-PCR were listed in Table 1.
Statistical Analysis. Student's t-test (two-tailed) was performed to analyze the data. Statistical significance was set at P < 0.05, as indicated by an asterisk (*P < 0.05; **P < 0.01).