Suppression of MicroRNA-203 improves survival of rat bone marrow mesenchymal stem cells through enhancing PI3K-induced cellular activation



As a group of heterogeneous multipotent cells, mesenchymal stem cells (MSCs) have potential in treatment of a variety of clinical diseases. However, the low survival of the transplanted MSCs reduced their therapeutic effects. In this study, we revealed that rno-miR-203 suppressed activity and colony formation and enhanced apoptosis of the rat bone marrow-derived MSCs (BM-MSCs). Using bioinformatics analysis, we found a potential miR-203 binding site within rat phosphatidylinositol 3-kinase (PI3K) 3′UTR, and fluorescent reporter experiments validated the direct and negative regulation of PI3K expression by miR-203 through this site. Ectopic expression of PI3K rescued BM-MSCs from depressed activity induced by miR-203, and suppression of PI3K attenuated the increased BM-MSCs activity by miR-203 inhibitor treatment. Moreover, miR-203 blocking partly protected BM-MSCs from impairment caused by low nutrition. We conclude that inhibition of endogenous miR-203 elevated PI3K expression, which may strengthen PI3K/Akt pathway and promote BM-MSCs activity and survival. © 2014 IUBMB Life, 66(3):220–227, 2014


Mesenchymal stem cells (MSCs) are a group of heterogeneous multipotent cells that can be isolated from many different adult tissues. MSCs have the capacity to self-renew and differentiate into many different cell types particularly osteoblasts, chondrocytes, and adipocytes in culture [1]. After culture expansion and in vivo administration, MSCs home to and engraft to injured tissues and modulate the inflammatory response by suppression of proinflammatory cytokines and upregulation of anti-inflammatory factors [2]. Experimental studies have provided evidence suggesting that MSCs may be useful for the treatment of a variety of clinical disorders [3], such as liver injury [4], cancers [1, 5], severe acute lung injury [6], osteoarticular diseases [7], and cardiac diseases [8].

Despite the benefits, MSC-based therapies is still limited mainly by low survival, engraftment, and homing to damage area, as well as inefficiencies in differentiating into fully functional tissues [9]. Among them, low survival is an evident obstacle. It was suggested that majority of injected MSCs die within several hours of delivery [10, 11]. Poor viability of transplanted MSCs also limited the therapeutic efficiency in injured central nervous system [12]. Therefore, methods to improve their survival or to suppress their apoptosis will be important therapeutically. Genetic engineering of MSCs is being explored as a method to resolve this problem [9]. For example, ectopic expression of HIF-1α protected MSCs against cell death and apoptosis triggered by hypoxic and oxidative stress [13]. Bcl-2-engineered MSCs exhibited reduced apoptosis, which was of value for the regeneration of intervertebral discs using MSC-based transplantation therapy [14]. These studies highlight the effect of gene engineering in elevating the viability and survival of MSCs.

As a large class of small noncoding RNAs, microRNAs (miRNAs) are widely involved in gene regulation. miRNAs are endogenous ∼23-nt RNAs that play important gene-regulatory roles by pairing to the mRNAs of protein-coding genes to direct their post-transcriptional repression [15]. Recently, several studies have reported the roles of miRNA in regulating MSCs viability, aging, and differentiation. For instance, the miRNA miR-378 enhanced osteogenic marker genes expression and promotes BMP2-induced osteogenic differentiation of MSCs [16]. In human MSCs, miR-21 acts as a key molecule determining cell proliferation and differentiation by suppressing SOX2 [17]. Because of the functional mode of miRNAs, we presumed that alteration of special miRNAs could affect the phenotypes of MSCs via regulating related genes, which is largely unknown until now.

Bone marrow-derived MSCs, known as BM-MSCs, remains the principal assembly of MSCs for most preclinical and clinical studies. In this study, we revealed that rno-miR-203 hampered BM-MSCs survival by suppressing proliferation and inducing apoptosis. Expression of a key cell viability-related gene phosphatidylinositol 3-kinase (PI3K, Gene ID: 170911) was directly regulated by miR-203 in BM-MSCs. Further study indicated that the negative role of miR-203 on BM-MSCs survival was through oversuppression of PI3K. This study provides evidence to enhance BM-MSCs viability by gene engineering, which may extend the application of BM-MSCs in cell-based therapy.

Materials and Methods

Isolation, Culture, and Identification of Rat BM-MSCs

Male Wistar rats (4–6 weeks old, approximately 200–220 g body weight) were euthanized by lumbar injection with 5% chloral hydrate at 1 mL/100 g body weight. The animals were maintained under specific pathogen-free conditions and handled in accordance with the NIH Animal Care and Use Committee Regulations. Whole bone marrow was harvested in cell culture flasks by flushing resected femurs and tibias with complete BM-MSCs culture medium [DMEM/F12 (1:1; Gibco, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (PAA from GE Healthcare Life Sciences, Piscataway, NJ, USA), 100 IU/mL penicillin, and 100 μg/mL streptomycin]. The derived bone marrow cells were kept at 37 °C with 5% CO2. Nonadherent hematopoietic cells were discarded 48 h later, and the adherent MSCs were maintained in culture for two passages to generate a BM-MSC population.

The specific cell surface markers for MSCs identification were detected. Briefly, cells were detached by trypsin incubation, rinsed with phosphate buffered saline, and then incubated with fluorescence-labeled mouse anti-rat antibodies (CD90-FITC, CD45-PE, CD34-FITC, or CD29-PE; BioLegend, San Diego, CA, USA). The antibodies-incubated cells were then analyzed using a BD FACSCalibur flow cytometer.

Differentiation of BM-MSCs into adipogenic cells was achieved by using complete BM-MSCs culture medium including adipogenic supplements (1 μM dexamethasone, 0.5 mM 1-methyl-3-isobutyl-xanthine, 0.1 mM indometacin, and 10 μg/mL insulin). The culture medium was replaced every 3 days. After about 8–10 days, cells were fixed in prechilled 4% paraformaldehyde and stained with 0.3% oil red O for accumulation of the lipid droplets. Later, cells were washed with 60% isopropanol to remove excess stain, and then the cells were observed under microscope.

Osteogenic differentiation was induced by using complete BM-MSCs culture medium including osteogenic supplements (0.1 μM dexamethasone, 10 mM sodium β-glycerophosphate, and 50 μg/mL l-ascorbic acid). The culture medium was replaced every 3 days. After about 12–14 days, cells were stained using GenMed Von Kossa Staining Kit (Shanghai, China) for visualization of the calcium.

Vector Constructions

The detailed information of vector construction is available in the Supporting Information.

Retrovirus Packaging and Gene Transduction into BM-MSCs

The retrovirus was packaged using pCDH cDNA (for expression of protein-coding genes and primary miRNAs) or pSIH1 shRNA [for expression of short hairpin RNAs and miRNA “tough decoys” (TuD)] Cloning and Expression Lentivectors (System Biosciences, Mountain View, CA, USA). For target cell infection, the culture medium containing lentiviral particles and 5 μg/μL polybrene infection facilitator (Sigma-Aldrich, St. Louis, MO, USA) was added into BM-MSCs at a multiplicity of infection of 5. Gene expression and cellular phenotypes were detected at 24, 48, or 72 h after transduction.

RNA and Protein Extractions

Large (>200 nt) and small (<200 nt) RNA fractions in BM-MSCs were isolated using the mirVana™ miRNA Isolation Kit (Ambion, Austin, TX, USA) according to the manufacturer's instructions. To extract cellular protein, BM-MSCs were lysed with RIPA lysis buffer (150 mM NaCl, 50 mM Tris-HCl pH 7.2, 1% Triton X-100 and 0.1% SDS) supplemented with Complete Protease Inhibitor Cocktail (Roche, Basel, Switzerland). After centrifugation, the undissolved cell components were removed, and the cellular proteins were obtained.

Quantitative RT-PCR

The detailed information for quantitation of miRNA and protein-coding genes is available in the Supporting Information.

Cell Counting Kit-8 Cell Activity Assay

BM-MSCs were seeded and transduced into a 96-well plate. At 24, 48, and 72 h after transduction, 10 μL Cell Counting Kit-8 (CCK-8; Dojindo, Kumamoto, Japan) was added to the cells. The absorbance at 450 nm was measured after 2 h using a Synergy 2 Multi Mode Microplate Readers (BioTek, Winooski, VT, USA).

Colony Formation Assay

To measure the colony formation, 100 BM-MSCs were seeded into each well of a 12-well plate. The culture medium was replaced every 3 days. On the 12th day after seeding, colonies were counted only if the contained more than 30 cells. The colonies were finally stained using 2% crystal violet.

Flow Cytometry Apoptosis Analysis

For apoptosis detection, BM-MSCs were stained with Annexin V and propidium iodide (PI) using an Alexa Fluor 488 Annexin V/Dead Cell Apoptosis Kit (Invitrogen) according to the manufacturer's instructions. The stained cells were analyzed using a BD FACSCalibur flow cytometer.

Prediction of miRNA Targets

The TargetScanHuman (Release 6.2, was used to predict the potential targets of miR-203.

Fluorescent Reporter Assay

BM-MSCs were coinfected by the lentivirus with a RFP reporter vector (pCDH1/RFP-PI3K-wtUTR or -mutUTR) and with miR-203 overexpression vector or miR-203 TuD. In each group, identical amounts of green fluorescent protein (GFP) containing lentiviral vector were applied to ensure a parallel GFP expression. After 48 h, the cells were harvested and lysed with RIPA lysis buffer, and the fluorescent activities were measured using a Synergy 2 Multi Mode Microplate Readers (BioTek).

Western Blotting

The proteins were resolved on an SDS denaturing polyacrylamide gel and then transferred onto a nitrocellulose membrane. Antibodies to PI3K or an endogenous control GAPDH (Abcam, Cambridge, MA, USA) were incubated with the membranes overnight at 4°C. The membranes were then washed and incubated with horseradish peroxidase-conjugated secondary antibodies. Protein expression was assessed by enhanced chemiluminescence, and the bands were captured by a FluorChem FC2 Imaging System (Alpha Innotech, Kasendorf, Germany).

Statistical Analyses

All of the experiments were performed in triplicate. The hypothesis test for significance between two groups used Student's t-test; for three or more groups, a one-way analysis of variance was used, followed by Student-Newman-Keuls q-test for comparing two groups. All the data processing was performed using GraphPad Prism v5.0 software, and the statistical significance was set at P ≤ 0.05.


Confirmation of the Isolated Cells as BM-MSCs

At the beginning of the study, three criteria were introduced to confirm that the bone marrow-derived cells were BM-MSCs [18]. First, the cells are plastic-adherent and exhibit a typical morphological character of MSCs under standard culture conditions (Fig. 1A). Second, most of the cells express CD29 and CD90 and are absent for the expression of hematopoietic cell surface markers such as CD34 and CD45 (Fig. 1B). Third, when cultured in specific stimulus, the cells can differentiate into adipocytes and osteocytes (Fig. 1C). The validation of the BM-MSCs ensured the reliability of the data from the following experiments.

Figure 1.

Identification of the bone marrow-derived cells as BM-MSCs. A: The second passage of rat bone marrow-derived adherent cells (×40). B: The expression of cell surface markers for MSCs identification were measured by incubating the cells with fluorescence-labeled antibodies and detecting by flow cytometry. C: For differentiation detection, cells were maintained in special induction media. The differentiation capability of BM-MSCs into adipogenic or osteogenic cells was evaluated by oil red O [2] or Von Kossa [3] staining, respectively (×100). The uninduced cells were the control [1]. The arrows indicate the red lipid droplets [2] or the black calcium deposition [3]. [Color figure can be viewed in the online issue, which is available at]

miR-203 Inhibits BM-MSCs Survival

Given that miR-203 plays a tumor suppressor role in many types of cancer cells, we asked here whether miR-203 affects the survival of BM-MSCs. The retroviral vectors expressing primary miR-203 or the TuD miR-203 inhibitor were introduced respectively to enhance (Fig. 2A) or suppress (Fig. 2B) miR-203 level in BM-MSCs. In the cell activity detection, when miR-203 level was elevated, the BM-MSCs activity decreased at 24, 48, and 72 h after transduction (Fig. 2C). Reasonably, an enhanced cellular activity was observed when the endogenous miR-203 was inhibited (Fig. 2D). BM-MSCs with overexpressed miR-203 showed a depressed colony formation activity, and a strengthened colony formation by TuD-203-treated BM-MSCs was obtained (Fig. 2E). Next, we determined whether BM-MSCs apoptosis was also influenced by miR-203, because forced expression of miR-203 led to a high apoptosis ratio of BM-MSCs and inhibition of miR-203 alleviates its apoptosis (Fig. 2F). The above data indicated that miR-203 was an adverse factor of BM-MSCs survival.

Figure 2.

miR-203 obstructs proliferation and enhances apoptosis of BM-MSCs. A and B: Primary miR-203 (pri-203; A) or miR-203 TuD (TuD-203; B) expression vectors were transduced into BM-MSCs by recombinant retrovirus, and miR-203 levels were measured using quantitative RT-PCR. U6 snRNA was the endogenous normalizer. C and D: miR-203 was overexpressed (C) or suppressed (D) in BM-MSCs, and cellular activity was measured using CCK-8 assay at 24, 48, and 72 h after transduction. E: miR-203 was overexpressed or suppressed, and colony formation by BM-MSCs was detected. Crystal violet-stained cell colonies are shown. F: miR-203 was overexpressed or suppressed, and BM-MSCs apoptosis was detected using Annexin V and PI staining and flow cytometry analyses (*P < 0.05). [Color figure can be viewed in the online issue, which is available at]

PI3K, a Key Regulator of Cell Proliferation, Is a miR-203 Direct Target and Is Negatively Regulated by miR-203 in BM-MSCs

To reveal the mechanism of miR-203-induced BM-MSCs phenotypes alteration, we next seek and confirm the miR-203 functional target gene. With the help of the TargetScanHuman database, we predicted that PI3K, an important member of the PI3K/Akt pathway, was a potential target of miR-203 because its 3′UTR bears a miR-203 binding site (Fig. 3A). When the miR-203 binding site within PI3K 3′UTR was merged downstream of a RFP coding region, miR-203 reversely regulate RFP intensity (Fig. 3B). However, the expression of RFP merged with mutated miR-203 binding site (Fig. 3A) did not respond to the alterations of miR-203 level (Fig. 3B). In BM-MSCs, the endogenous PI3K expression is also negatively controlled by miR-203 on both mRNA (Fig. 3C) and protein (Fig. 3D) levels.

Figure 3.

miR-203 directly targets and negatively regulates PI3K. A: As was predicted by the TargetScanHuman database, wild-type PI3K 3′UTR (PI3K wt UTR) bears a potential miR-203 binding site. The sequence of the mutated PI3K 3′UTR (PI3K mut UTR) used in the fluorescent reporter assay is also shown. The underlined nucleotides are the mutated ones. B: The wild-type or mutated miR-203 binding sequence was cloned downstream of RFP coding region to form the reporter vectors. BM-MSCs were cotransducted with one of the reporter vectors and miR-203 expression (pri-203) or suppression (TuD-203) vectors. At 48 h after transfection, cells were harvested, and RFP intensity was measured. GFP was the normalizer. C and D: miR-203 was overexpressed or suppressed in BM-MSCs, and PI3K mRNA or protein levels were detected using quantitative RT-PCR (C) or Western blot (D) assays, respectively. β-Actin and GAPDH were the endogenous controls (*P < 0.05).

miR-203 Affects the Growth and Apoptosis of BM-MSCs by Regulating PI3K Expression

To evaluate whether the adverse effects of miR-203 on BM-MSCs survival was due to its regulation of PI3K, miR-203 was first overexpressed, and PI3K was subsequently strengthened to save the oversuppressed endogenous PI3K in BM-MSCs. We found that ectopic expression of PI3K could rescue miR-203-induced depressed BM-MSCs activity. The enhanced cell activity by miR-203 inhibitor could also be restored by the following treatment of PI3K siRNA (Fig. 4A). Similarly, the alteration of colony formation (Fig. 4B) and apoptosis (Fig. 4C) by miR-203 overexpression or inhibition was also attenuated by PI3K enhancement or suppression, respectively. The above data indicate that miR-203 affects BM-MSCs survival by negatively regulating PI3K.

Figure 4.

miR-203 inhibits proliferation and promotes apoptosis of BM-MSCs via oversuppression of PI3K. A: BM-MSCs were first treated with miR-203 expression vectors (pri-203) and then with PI3K overexpression vectors (PI3K) 24 h later. In another group, BM-MSCs were treated similarly with miR-203 inhibitor (TuD-203) and subsequently with short-hairpin RNA to PI3K (shR-PI3K). PI3K expression was detected by Western blotting, and cell activity was detected by CCK-8 assay 48 h later. B and C: BM-MSCs were treated as described in (A), and cell colony formation (B) and apoptosis (C) were detected (*P < 0.05). [Color figure can be viewed in the online issue, which is available at]

Inhibition of miR-203 Rescues the Depressed BM-MSCs Activity Caused by Low-Serum Cultural Condition

Low-nutrition microenvironment is one of the adverse conditions that BM-MSCs may endure in the organisms in clinical therapy. When the serum concentration in the cultural medium was reduced to 2%, the BM-MSCs activity significantly decreased, and the ratio of apoptosis and dead cells increased to ∼60% after 72 h when compared with those in the cultural medium with 10% serum. Moreover, when they were treated with miR-203 inhibitors, the BM-MSCs in low-serum medium exhibited a higher activity and lower apoptosis ratio, although these phenotypes were not restored to the normal level (Fig. 5). These results highlight that inhibition of miR-203 strengthens the survival ability of BM-MSCs.

Figure 5.

Inhibition of miR-203 alleviates the cell activity depression induced by low-serum condition. A: BM-MSCs were treated with miR-203 inhibitor (TuD-203) or the control vector pSIH1 and maintained in BM-MSCs culture medium with 2% fetal bovine serum. At 72 h later, cell activity was measured by CCK-8 assay. B: BM-MSCs were treated as described in (A), and cell apoptosis was detected using Annexin V and PI staining and flow cytometry analyses (*P < 0.05). [Color figure can be viewed in the online issue, which is available at]


More evidences have indicated that MSCs exhibited growth arrest and low survival activity when they are endogenously recruited and home to sites of inflammation and injury [9]. Thus, it is necessary to prepare MSCs for clinical therapies via genetic modification. In this study, we revealed a miR-203-PI3K cell survival and apoptosis pathway, in which miR-203 hampered BM-MSCs survival by oversuppression of PI3K. First, miR-203 was indicated to be a negative regulator of BM-MSCs activity and survival. Peripheral and direction injection of single MSC suspension into the injured organ are still preferred routes in regeneration therapies [19]. Thus, to exist and grow independently is a significant capacity for BM-MSCs in the target tissues. We suggested that suppression of miR-203 not only enhanced activity and alleviate apoptosis of BM-MSCs in vitro but also, by colony formation detection, promoted the proliferation of the individual cells that lack of the cell–cell support and interaction. Second, PI3K emerged to be a potential functional target of miR-203 in BM-MSCs. The direct targeting of miR-203 on PI3K mRNA was confirmed by fluorescent reporter assay. Third, miR-203 ectopic expression- or inhibition-induced BM-MSCs phenotype alteration could be alleviated by subsequent expression or suppression of PI3K. This provided evidence to prove that miR-203-induced depressed survival activity of BM-MSCs was through its targeting and regulation of PI3K.

In most cases of tissue injury, there can be hypoxia and ischemia, leading to a lack of nutrients. Therefore, it is important for BM-MSCs to endure and survive in low-nutrition microenvironment [20]. Our study indicated that in low-serum cultural medium, BM-MSCs exhibited obviously arrested activity and high apoptosis rate. Importantly, pretransduction of miR-203 into BM-MSCs could partly alleviate the low-nutrition-induced cell damage. This result provides a potential way to strengthen the BM-MSCs survival in cell-based clinical therapy.

The rat miR-203 (MIMAT0000876) was located at chromosome 6q32. The sequence of rat miR-203 is the same as human miR-203 (MIMAT0000264). Generally, human miR-203 acts as a tumor suppressor in various human cancers. For example, miR-203 inhibits the proliferation and invasion of U251 glioma cells via suppressing the expression of PLD2 [21], and low miR-203 level is associated with unfavorable prognosis in human glioma [22]. miR-203 also functions as a tumor suppressor in human breast cancer [23], melanoma [24], and basal cell carcinoma [25]. Adverse to the cancer cells, low survival and proliferation of BM-MSCs may be an unfavorable phenotype in cell-based clinical therapy. According to our experiments, miR-203-induced oversuppression of PI3K expression is possibly a mechanism of low survival of BM-MSCs, and inhibition of this miRNA could improve BM-MSCs survival.

The PI3K-Akt pathway is closely related with high proliferation and low apoptosis of various types of cells. This pathway plays key regulatory roles in MSC survival, proliferation, and some other phenotypes [26]. Previous studies have revealed the benefit of MSCs from PI3K-Akt pathway activation in the tolerance of oxidative stress [27], hypoxia, and serum deprivation [28]. Our study also revealed a method of elevating the PI3K/Akt pathway and enhancing BM-MSCs activity.

Importantly, we should pay more attention here to the potential malignant transformation and carcinogenesis mediated by BM-MSCs, especially the gene-engineered and PI3K/Akt-activated ones. The MSCs functions in tumors are controversial. Some investigators found that MSCs inhibit tumor growth, whereas others reported that MSCs promote tumor growth [3, 5]. Basically, there are two main roles of MSCs in cancer: one is indirect involvement via the tumor-modulatory effect of MSCs and the other is direct involvement via malignant transformation of the MSCs themselves [1]. In our research, miR-203 inhibitor-modified BM-MSCs have additional carcinogenic risk for two reasons. One is that genetic manipulations either by viral or by nonviral transgene delivery methods can cause malignant transformation of MSCs [1]. The other reason is that PI3K/Akt pathway, which is associated with tumor initiation and progression, was strengthened in BM-MSCs. This cell signal pathway is possibly associated with MSC-induced cancer progression [29]. Therefore, care must be taken not to introduce cancer to patients who received miR-203 inhibitor-treated BM-MSC therapy. As another point of view, overexpression of miR-203 may reduce the MSC-mediated cancer progression by suppression of PI3K/Akt pathway, which needs to be elucidated in future.

Collectively, our study indicated that miR-203 negatively regulates BM-MSCs activity and survival. Inhibition of endogenous miR-203 releases expression of its direct target PI3K from oversuppression and then promotes BM-MSCs survival. This knowledge may shed light on the mechanisms by which miRNAs affect BM-MSCs activity and may also help to improve the application of BM-MSCs in clinical therapy.


This work was supported by the China Postdoctoral Science Foundation (No. 2013M541185), the National Natural Science Foundation of China (Nos. 81270528 and 81170444), and the Natural Science Foundation of Tianjin (Nos. 08JCYBJC08400, 12JCZDJC25200, and 11JCZDJC27800).