MicroRNA Let-7f-5p Promotes Bone Marrow Mesenchymal Stem Cells Survival by Targeting Caspase-3 in Alzheimer Disease Model.

Widespread death of transplanted mesenchymal stem cells (MSCs) hampers the development of stem cell therapy for Alzheimer disease (AD). Cell pre-conditioning might help cope with this challenge. We tested whether let-7f-5p-modified MSCs could prolong the survival of MSCs after transplantation. When exposed to Aβ25-35in vitro, MSCs showed significant early apoptosis with decrease in the let-7f-5p levels and increased caspase-3 expression. Upregulating microRNA let-7f-5p in MSCs alleviated Aβ25-35-induced apoptosis by decreasing the caspase-3 levels. After computerized analysis and the luciferase reporter assay, we identified that caspases-3 was the target gene of let-7f-5p. In vivo, hematoxylin and eosin staining confirmed the success of MSCs transplantation into the lateral ventricles, and the let-7f-5p upregulation group showed the lowest apoptotic rate of MSCs detected by TUNEL immunohistochemistry analysis and immunofluorescence. Similarly, bioluminescent imaging showed that let-7f-5p upregulation moderately prolonged the retention of MSCs in brain. In summary, we identified the anti-apoptotic role of let-7f-5p in Aβ25-35-induced cytotoxicity, as well as the protective effect of let-7f-5p on survival of grafted MSCs by targeting caspase-3 in AD models. These findings show a promising approach of microRNA-modified MSCs transplantation as a therapy for neurodegenerative diseases.

Introduction

Bone marrow-derived mesenchymal stem cells (BM-MSCs) are a class of stem cells with low immunogenicity and are available in large amounts (Lo Furno et al., 2017). They are capable of homing to lesion tissues (De Becker and Riet, 2016), and secreting different neurotrophic growth factors (NTFs) to promote regeneration and neuroprotection (Koniusz et al., 2016; Petrou et al., 2016). Moreover, the use of MSCs in clinical treatments does not raise significant religious or ethical issues (Tanna and Sachan, 2014). Therefore, MSCs hold great potential in stem cell-based therapies for neurodegenerative diseases (Volkman and Offen, 2017).

Alzheimer disease (AD) is common form of dementia, affecting about 36 million people globally (Prince et al., 2016). Extensive investigations have shown that MSCs can cross the blood-brain barrier via intracerebral (transplantation into hippocampus or lateral ventricles), intracarotid, or intravenous injections (Lo Furno et al., 2017). These cells and their extracellular vesicles could stimulate neurogenesis (Oh et al., 2015) and could significantly rescue learning and memory deficits of animals with Aβ deposition reduction and modification of size and number of microglial cells (Duncan and Valenzuela, 2017; Liew et al., 2017). However, MSCs have a low survival rate after transplanted due to the adverse microenvironment [i.e., presence of cytotoxic Aβ oligomers and p-tau, and increased reactive oxygen species (ROS)] (Lee et al., 2015; Oh et al., 2018). Thus, long-term retention of MSCs to achieve meaningful improvement in neurodegenerative diseases is difficult. In our study, we investigated the possibility of modulating MSCs before transplantation to extend their survival under harsh microenvironment, common in AD.

MicroRNAs (miRNAs) are evolutionarily conserved regulators of various physiological and pathological processes. In addition, a discrete group of miRNAs is reported to be functionally linked with cell survival (Chandra et al., 2017). Let-7 family, originally discovered in C. elegans, which are regulators of developmental timing and cell proliferation, have also been reported to suppress tumors by regulating apoptotic genes and cell growth in humans (Chiu et al., 2014; Balzeau et al., 2017). Loss of let-7 results in high self-renewal rates and directly converting normal stem cells to tumor-initiating cells (Chiu et al., 2014). Let-7 inhibits doxorubicin- and paclitaxel-induced apoptosis, as well as ultraviolet B–induced apoptosis (Barh et al., 2010). However, its involvement in Aβ-induced apoptosis remains unknown. In MSCs, let-7 is known to regulate self-renewal depending on high-mobility group AT-hook 2(HMGA2) (Kalomoiris et al., 2016). Let-7f-5p, a member of the let-7 family, exerts a vital role in MSCs differentiation (Davoodian et al., 2017; Han et al., 2018), but its influence on MSCs apoptosis has not yet been studied.

Accordingly, the present study was designed to test the hypothesis that let-7f-5p could regulate the survival of implanted MSCs in AD models. By regulating let-7f-5p levels in MSCs the functional link between let-7p-5p and Aβ-related cell apoptosis was confirmed, for the first time, and the underlying molecular mechanism explored. The effect of let-7f-5p was further investigated in vivo by comparing the apoptosis and survival time of transplanted MSCs. The findings explore the possibility of using miRNA-mediated fortification of MSCs to treat AD.

Materials and methods

Ethics approval statement

All experiments were approved by the Laboratory Animal Care Center of Zhengzhou University (see Presentation 1 in Supplementary Material) and were performed according to the guidelines approved by the Institutional Animal Care and Use Committee (IACUC) of Zhengzhou University, China.

Cell culture and Aβ treatment

MSCs were isolated from the bone marrow of 2-month-old male C57BL/6J mice. MSCs were cultured according to the manufacturer's instructions in a complete cell medium consisting of DMEM (HyClone, USA), 10% fetal bovine serum (FBS) (lot NO.1652790 Gibco, USA), and 1% antibiotic-penicillin/streptomycin solution at 37°C in 5% CO2. We used MSCs at passages 7–10 for our experiments. Aβ25−35 (Cat No. A4559; Sigma; IL) oligomer aggregates were prepared at 37°C for 7 days, as previously described (Kwon et al., 2011). For the experiments, the cells were re-seeded in plates and treated with Aβ25−35 (25 μM) for an additional 24 h.

Cell transfection

After 24 h (5,0000 cells/ml) of plating, MSCs were transfected with different tittered lentivirus. Transfection was carried out in the Enhanced Infection Solution with 5 μg/ml polybrene (Shanghai Genechem Co., Ltd. China), and the medium was changed 12 h later. All of the lentivirus contained firefly luciferase gene and puromycin resistance encoding sequence at the same time. Four days after cell transduction, cells were selected on puromycin hydrochloride (5 μg/ml, Sigma-Aldrich, Lyon, France) three times by replacing the culture medium every 2 days. By comparing the screened cell numbers of different tittered lentivirus, we found the most appropriate multiplicity of infection (MOI, 30) with minimum lentivirus and relative large number of screened cells.

The mouse let-7f-5p (mmu-let-7f-5p) sequence (MIMAT0000525) was obtained from the miRBase database. Mmu-let-7f-5p-LV (5′-CAGAATGTCTCAGGTAACCCTCCCTTCCATTCTTGTTAATTAATCAGTGGATTCTTTCCAAATAGCTGGCCGCATGGGCTGAAGACGGACACTGGTGCTCTGTGGGATGAGGTAGTAGATTGTATAGTTTTAGGGTCATACCCCATCTTGGAGATAACTATACAGTCTACTGTCTTTCCCACGGTGGTACACTTCCCATCCA-3′), mmu-let-7f-5p-inhibition-LV (5′-AACTATACAATCTACTACCTCA-3′) and mmu-let-7f-NC-LV were constructed in the same method except for different inserted sequences, as described in our previous report (Jing et al., 2011). The forward primers used for the transduction were as follows: 5′- GAGGATCCCCGGGTACCGGTCAGAATGTCTCAGGTAACCC-3′for mmu-let-7f-5p, and 5′-GGAAAGAATAGTAGACATAATAGC-3′ for mmu-let-7f-5p-inhibition-LV. Mmu-let-7f-NC-LV only contains firefly luciferase gene and puromycin resistance encoding sequence.

Mmu-let-7f-5p-LV, mmu-let-7f-5p-inhibition-LV, and mmu-let-7f-NC-LV were transfected into three groups of MSCs: let-7f-5p upregulation group, let-7f-5p downregulation group, and negative control (NC) group. Another group of MSCs without lentivirus transfection was defined as sham group (SHAM). Successfully transfected cells were selected and cultured in normal medium for subsequent experiments. RT-qPCR was used to quantify the let-7f-5p expression level.

RT-qPCR and western blotting

Cellular RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA). The corresponding cDNA was synthesized using the ReverTra Ace qPCR RT Kit (TOYOBO Co., Ltd.) according to the manufacturer's recommendations. MicroRNA expression was detected using the Hairpin-it™ miRNA qPCR Quantitation Kit (GenePharma Co., Ltd.). RT-qPCR analysis was performed with SYBR Green (TOYOBO Co., Ltd.) and all reactions were run in triplicates. Amplification was performed using the following primers:

let-7f-5p:

forward: 5′-UGAGGUAGUAGAUUGUAUAGUU-3′;

reverse: 5′-AGCUGAUUUCGUCUUGGUA-3′;

U6:

forward: 5′-GCGCGTCGTGAAGCGTTC-3′;

reverse: 5′-GTGCAGGGTCCGAGGT-3′

Caspase-3:

forward: 5′-GACTAGCTTCTTCAGAGGCGA-3′;

reverse: 5′-ATTCCGTTGCCACCTTCCTG-3′

β-actin:

forward: 5′-CCCATCTATGAGGGTTACGC-3′;

reverse: 5′-TTTAATGTCACGCACGATTTC-3′

Let-7f-5p and caspase-3 mRNA levels were normalized to levels of small nuclear RNA U6 and β-actin mRNA, respectively. Relative expression levels were analyzed using 2-ΔΔCT to calculate the CT value.

Western blotting was used to determine relative protein expression level of caspase-3 and cleaved caspase-3. MSCs cultured in absence or presence of Aβ25−35 for 24 h were lysed in 1 × lysis buffer (Leagene Biotechnology Co., Ltd., Beijing, China) with protease inhibitor and phosphatase inhibitor cocktail at 4°C for 25 min. After testing protein concentration with the BCA Protein Assay Kit (Dingguo, Beijing, China), about 50 μg of protein was separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene difluoride (PVDF) membranes, which was probed with caspase-3 (Cell Signaling Technology, CA, 1:1,000) and cleaved caspase-3 (Cell Signaling Technology, CA, 1:1,000). The membranes were then washed three times in 1 × TBST at room temperature and incubated with a horseradish peroxidase-conjugated rabbit secondary antibody (Dingguo, Beijing, China, 1:5,000) for 90 min. The antibody-antigen complexes were detected using the ECL Western blotting substrate (Dingguo, Beijing, China), and Image J software was used to analyze the protein bands (Original images of westerns see Presentation 1 in Supplementary Material).

Determination of cell viability and apoptosis

Cell viability was evaluated by the MTT assay. Briefly, after treatment with Aβ25−35, MSCs were seeded in 96-well plates and exposed to 10 μl MTT (5.0 g/l, Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) and incubated for 4 h at 37°C. Then, the medium was discarded, and 100 μl DMSO was added into each well. The formazan crystalline product was dissolved, and the optical density was measured at 570 nm with a microplate reader. Cell viability was expressed as a percentage of the value against the non-treated control group.

Apoptosis of MSCs was quantified using the Annexin V-FITC Apoptosis Detection Kit (KeyGEN BioTECH), according to the manufacturer's protocol. After exposure to Aβ25−35, MSCs were washed with PBS, resuspended in 100 μl annexin V-FITC labeling solution, and stained with 5 μl annexin V-FITC and 5 μl propidium iodide (PI) for 30 min in dark at room temperature. Subsequently, samples were analyzed using flow cytometer (Beckman Coulter EPICS~XL, Beckman Coulter, Inc., CA, USA).

Luciferase assay

The potential binding of target genes and miRNAs was predicted with TargetScan (http://www.targetscan.org), miRbase (http://www.mirbase.org), and microRNA.org (http://34.236.212.39/microrna/home.do) miRNA-target prediction databases.

Mutant (MUT) and wild-type (WT) sequences of the Caspase-3 3′UTR were amplified using PCR and cloned into the psiCHECK-2 dual-luciferase vector (purchased from Hanbio Biotechnology Co., Ltd., Shanghai, China). The sequences of the luciferase reporter were as follows:

CASP3-3′UTR-WT: 5′….CAGCCAGCCACAGUGCAGCUACCUCAA.…3′

CASP3-3′UTR-MUT: 5′….GCAGCCAGCCACAGUGCAGaUuCaUuAA.…3′

Chemically synthesized miR-let-7f-5p mimic (5′-UGAGGUAGUAGAUUGUAUAGUU-3′), miR-let-7f-5p inhibitor (5′-AACUAUACAAUCUACUACCUCA-3′) and NC (5′-AGCUGAUUUCGUCUUGGUA-3′) were purchased from Hanbio Biotechnology Co., Ltd., (Shanghai, China). 293T cells were co-transfected with psiCHECK-2-WT + miR-let-7f-5p mimic, psiCHECK-2-WT + miR-let-7f-5p inhibitor, psiCHECK-2-WT + miRNA NC, psiCHECK-2-MUT + miR-let-7f-5p mimic, psiCHECK-2-MUT + miR-let-7f-5p inhibitor, and psiCHECK-2-MUT + miRNA NC, respectively. Luciferase activity was measured using a luminometer and the Dual-Luciferase Reporter Assay Kit (Promega, Madison, WI, USA) 48 h post-transfection, according to the manufacturer's instructions. We used renilla luciferase to normalize the cell number and transfection efficiency.

Animals and cell transplantation

The C57BL/6J-TgN (APP/PS1) ZLFILAS mouse strain (Institute of laboratory animals, Chinese Academy of Medical Sciences, China), harboring two transgenes of PS1 (Hu PS1 deltaE9) and APP (K595N/M596L), was used. Eight-month-old male APPswe/PS1dE9 transgenic Mice (200 ± 20 g) in accordance with the subgroups of MSCs, were divided into the following subgroups: let-7f-5p upregulation group, let-7f-5p downregulation group, NC group, and wild type (WT) group of wild mice transfected with NC group MSCs (n = 9 to 10 animals per group). After removing hair from the head, animals were anesthetized by injecting 4% chloralic hydras (350 mg/kg). Secured in a stereotaxic device (Stoelting Co., Wood Dale, IL, USA), the animal's head was cleaned. Mice were injected with MSCs (3 × 105 cells/10 μl) into the right lateral ventricle (coordinates from bregma: anteroposterior = −0.5 mm, mediolaterally = 1 mm, dorsoventrally = 3 mm). Then, after 5 min, the syringe was slowly removed, and the incision was sutured. Ampicillin (50 mg/kg) was administered for 3 days post-surgery.

Bioluminescence imaging (BLI)

D-luciferin firefly potassium salt substrate (150 μg/ml, XenoLight D-Luciferin K+ Salt, Perkin Elmer) was added to a 96-well plate with different concentrations of MSCs (1.0 × 105, 5 × 104, 2.5 × 104, 1.25 × 104 cell/ml), and signals from the cells were measured immediately. Bioluminescence was detected in anesthetized animals 15 min after intraperitoneal injection of 150 μl D-luciferin (15 mg/ml) using the IVIS Spectrum (Perkin Elmer, Waltham, MA). The signal was captured and analyzed on day 1, 7, 10, 14, and 21 post-op. Regions of interest (ROI) were manually determined, and were recorded as photons per second per square centimeter per steradian (p/s/cm2/sr).

Histological staining and immunofluorescence

MSCs were treated with PKH-26 (Sigma-Aldrich Life Science, CA, USA) for cell membrane labeling before transplantation to evaluate the degree of transplanted cell apoptosis in vivo. APPswe/PS1dE9 transgenic mice were transcardially perfused with 4% paraformaldehyde solution, and their brains were sampled on day 5 post-surgery. Embedded with liquid paraffin, the tissues were cut into 6 μm thick coronal sections. The proportion of apoptotic cells was measured by Image Pro Plus 6.0 (IPP, Media Cybernetics, Carlsbad, CA).

Sections were deparaffinized with xylene and stained with hematoxylin and eosin (H&E) staining according to the manufacturer's instructions. The staining was performed using a Masson trichrome kit (Baso, BA4079), and then examined under a light microscope (Nikon ECLIPSETi-U).

To assess the caspase-3 expression, a rabbit polyclonal antibody against caspase-3 (1:400; Cell Signaling Technology, CA) was used at 4°C overnight. After washing for three times, the cells were incubated with secondary antibody, which was fluorescein isothiocyanate (FITC)-conjugated anti-rabbit IgG (1:100; BOSTER Biological Technology Co., Ltd., Wuhan, China), for 60 min. By calculating the overlap of PHK-26 staining cells (red fluorescence) and FITC-labeling (green fluorescence) cells, we identified the percentage of apoptotic cells among MSCs in AD mice. Slides for normal animal group were prepared identically for immunostaining. Nuclei were stained with diamidino-2-phenylindole (DAPI) and imaged using fluorescence microscopy (Leica, Wetzlar, Germany).

TUNEL immunohistochemistry analysis was performed using the TUNEL Apoptosis Assay kit (Roche, South San Francisco, CA, US). First, tissue slides were deparaffinized and treated with proteinase K (20 μg/ml) for 20 min at 25°C. We equilibrated the tissue sections for 5 min using 100 μl of the kit's equilibration buffer. FITC-12-dUTP Labeling Mix containing Recombinant TdT Enzyme was added to the sections, and the samples were incubated at 37°C in a humidity chamber for 1 h and imaged using fluorescence microscopy (Leica, Wetzlar, Germany).

Statistical analyses

All experiments were performed in triplicate, not technical replicates, and all statistical analyses were performed using SPSS v21. P < 0.05 was considered statistically significant. Data are shown as the mean ± standard error of the mean (SEM). Statistical significance between two groups was analyzed using Student's t-tests. Comparisons between more than two groups were performed by two-way ANOVA with Least Significant Difference (LSD) test to compare between groups.

Results

Aβ25−35 induced cell apoptosis of MSCs

When treated with Aβ25−35 (25 μM) for 24 h, MSCs showed increased cell death after Annexin V/PI staining (Figure 1A), as compared to the untreated cells mainly for the data of early apoptosis were different statistically (P < 0.001). But the difference between the data of late apoptosis/necrosis was not statistically significant (Figure 1B). We also detected the expressions of caspase-3, a key moderator in apoptosis, and cleaved caspase-3, the marker of apoptotic pathway activation. Western blotting showed much higher levels of caspase-3 protein in Aβ25−35-treated cells, as compared to those in untreated cells (P < 0.01). Interestingly, the expression of cleaved caspase-3 protein increased dramatically from almost zero in untreated cells to such a high degree in Aβ25−35-treated cells (P < 0.001) (Figures 1D,E). However, let-7f-5p expression in the Aβ25−35-treated cells decreased (P < 0.01) (Figure 1C). Thus, we first identified that Aβ25−35 could induce cell apoptosis of MSCs, mainly early apoptosis, with caspase-3 activation and let-7f-5p suppression.

Figure 1

Aβ25−35 (25 μM, 24 h) induce MSCs apoptosis. (A) FACS analysis of apoptotic MSCs stained with Annexin V/PI in the presence or absence of Aβ25−35 treatment. Viable cells (lower left quadrant, annexin V−PI−), early apoptotic cells (lower right quadrant, annexin V+PI−), late apoptotic cells or necrotic cells (upper right quadrant, annexin V+PI+). (B) The statistical data from (A). Aβ25−35 exposure induced early apoptosis, but not late apoptosis/necrosis. (C) The let-7f-5p levels in MSCs with and without Aβ25−35 treatment. (D) The caspase-3 and cleaved caspase-3 levels in MSCs with and without Aβ25−35 treatment. (E) The statistical data from (D) All experiments were repeated at least three times independently. **P < 0.01 and ***P < 0.001. FACS, Flow cytometry analysis.

Anti-apoptosis effect of let-7f-5p on Aβ25−35-treated MSCs

To explore the role of let-7f-5p in Aβ25−35-treated cells, we transfected MSCs with mmu-let-7f-5p-inhibition-LV, mmu-let-7f-5p-LV, and mmu-let-7f-NC-LV, respectively, and then induced them with Aβ25−35. The levels of let-7f-5p in MSCs were subsequently detected by RT-qPCR (Figure 2A). Compared to the untreated group, Aβ25−35 (25 μM) exerted a significant inhibitory effect on the growth of MSCs, as determined from the MTT assay (P < 0.01). However, the cytotoxic effects could be partly attenuated by upregulating let-7f-5p expression (P < 0.05), but were amplified with let-7f-5p downregulation (P < 0.01) (Figure 2B). Similarly, FACS using Annexin V/ PI staining showed that Aβ25−35 increased early apoptosis of MSCs (P < 0.001), which was ameliorated by let-7f-5p upregulation prior to Aβ25−35 treatment (P < 0.001) (Figures 2C,D). The data indicated that transfected let-7f-5p significantly increased cell survival after Aβ25−35 exposure (Figure 2D). These data suggest that let-7f-5p exerts an anti-apoptosis effect.

Figure 2

Let-7f-5p regulates MSCs survival under Aβ25−35 exposure (25 μM, 24 h). (A) Let-7f-5p expression of MSCs after screened by puromycin hydrochloride. Data showed that MSCs were successfully transfected with mmu-let-7f-5p-inhibition-LV, mmu-let-7f-5p-LV, and mmu-let-7f-NC-LV. (B) Cell viability detected using MTT, when MSCs were exposed to Aβ25−35. (C) FACS analysis of apoptotic MSCs stained with Annexin V/PI following Aβ25−35 exposure for 24 h. Viable cells (lower left quadrant, annexin V−PI−), early apoptotic cells (lower right quadrant, annexin V+PI−), late apoptotic cells or necrotic cells (upper right quadrant, annexin V+PI+). (D) The statistical data from (C). *P < 0.05, **P < 0.01, and ***P < 0.001. FACS, Flow cytometry analysis.

Caspase-3 is the direct target of let-7f-5p

To further investigate the molecular mechanism underlying let-7f-5p-mediated anti-apoptosis of MSCs, we used different algorithms to predict potential targets of let-7f-5p. Among the candidates, we noted that the seed sequence of let-7f-5p is complementary to that of 3′ UTR of caspase-3 (Figure 3A). Furthermore, this putative binding was verified using the dual luciferase assay. WT-CASPASE-3−3′ UTR and MUT-CASPASE-3−3′ UTR luciferase reporters were generated and co-transfected with miR-let-7f-5p inhibitor, miR-let-7f-5p mimic, or miRNA NC into 293T cells, respectively. Let-7f-5p co-transfection significantly suppressed the activity of luciferase reporter containing WT-CASPASE-3−3′ UTR but had no effect on the activity of the MUT-CASPASE-3−3′ UTR vector (P < 0.05) (Figure 3B). In summary, these results proved that caspase-3 was a target of let-7f-5p.

Figure 3

Caspase-3 was verified as a functional target of let-7f-5p. (A) Diagram of putative target site for let-7f-5p in the caspase-3 mRNA 3′ UTR. (B) MUT and WT sequences of the Caspase-3 3′UTR and dual luciferase reporter gene assay. The psi-CHECK2 containing WT or MUT 3′ UTR of caspase-3 were co-transfected with miR-let-7f-5p inhibitor, miR-let-7f-5p mimic or miRNA NC. (C) Western blotting assay was used to compare the expression levels of caspase-3 and cleaved caspase-3 in MSCs, following transfection with mmu-let-7f-5p-LV, mmu-let-7f-5p-inhibition-LV and mmu-let-7f-NC-LV. (D) Bar graph from (C). (E) The relative mRNA expression levels of caspase-3 were determined by RT-qPCR. *P < 0.05, **P < 0.01, and ***P < 0.001. NC, negative control; MUT, mutant; WT, wild-type; UTR, untranslated region.

This functional targeting binding was confirmed by transfecting mmu-let-7f-5p-LV, mmu-let-7f-5p-inhibition-LV, and mmu-let-7f-NC-LV in MSCs. Three groups of MSCs with different let-7f-5p levels were thus constructed: let-7f-5p upregulation group, let-7f-5p downregulation group and NC group. After treating the different groups with Aβ25−35 for 24 h, we examined the expression of caspase-3 at both mRNA and protein levels. Western blotting illustrated that let-7f-5p could suppress the expression of caspase-3 as well as cleaved caspase-3 (P < 0.001) and inhibiting let-7f-5p resulted in increased expression of caspase-3 and cleaved caspase-3 at the same time (P < 0.01) (Figures 3C,D). Data of RT-qPCR showed the same pattern of caspase-3 expression (all P < 0.01) (Figure 3E), thus proving a negative correlation between caspase-3 and let-7f-5p.

Anti-apoptotic effect of let-7f-5p on transplanted MSCs in APP/PS1 mice

Based on the results in vitro, we examined the anti-apoptotic effect of let-7f-5p on transplanted MSCs in vivo. The APPswe/PS1dE9 transgenic mice were divided into three groups by transplanting three different microRNA-modified MSCs (as described in results section Materials and Methods), and the same number of wild counterparts were transplanted with NC MSCs as WT group. MSCs were labeled in red using PKH-26 before transplantation, and histology analysis was performed to evaluate the apoptotic rates by calculating the proportion of positive cells. H&E staining of four groups showed that there were foreign cells in right ventricles (injected position) with different percentages of dead cells while there were no foreign cells in left ventricles (Figure 4A). As shown in Figure 4B, the TUNEL-positive cells in the let-7f-5p downregulation group were the highest among the four groups in total (P < 0.05), followed by the NC group (P < 0.001, compared to WT group). Increased let-7f-5p levels significantly decreased the percentage of TUNEL-positive cells (P < 0.01), but these were still higher than WT group (Figure 4B). Consistently, immunofluorescence analysis of caspase-3 revealed a similar trend as that observed with TUNEL staining (Figure 4C). Thus, the data indicated the let-7f-5p could inhibit apoptosis of MSCs in vivo.

Figure 4

Cell apoptosis of grafted MSCs detected by histological analysis. (A) Hematoxylin and eosin (H&E) staining of brain tissues after MSCs transplantation for 5 days (original magnification 100X). Black arrowheads: transplanted MSCs; Asterisks: lateral ventricles. (B) TUNEL staining apoptotic cells in green. Composite fluorescence images combining the signals from TUNEL (green), DAPI (blue), and PKH-26 (red) (original magnification 200X) are shown. Yellow arrowheads: transplanted MSCs without apoptosis; white arrowheads: apoptotic MSCs; (C) Immunofluorescence analysis shows the levels of caspase-3. DAPI (blue DNA in nuclei), transplanted MSCs (red), and cells expressed caspase-3 (green). (original magnification 200X). Yellow arrowheads: transplanted MSCs without apoptosis; white arrowheads: apoptotic MSCs; *P < 0.05, **P < 0.01, and ***P < 0.001. LV, left ventricle; RV, right ventricle; TUNEL, terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling; DAPI, 4′,6-diamidino-2-phenylindole.

Let-7f-5p might enhance survival of MSCs in Vivo

As the lentiviral vector transduced into MSCs containing the luciferase gene, it was possible to monitor resettlement and survival of MSCs in vivo. BLI analysis showed the correlation between the bioluminescent signal and tested MSCs number in three groups indicating high transfected efficacy of lentiviral. In contrast, bioluminescent signal in PBS group without cells was absent (Figures 5A,C). The mice were also tested by BLI to determine survival of the injected MSCs on day 1, 7, 10, 14, and 21 post-implantation. After 1 day, BLI showed that the labeled cells in the brain of all groups had similar bioluminescent signal. As shown in Figure 5B, the bioluminescent signals from MSCs in NC group could not be detected until day 14 after transplantation. However, signals from the let-7f-5p downregulation group could only be detected up to 7 days post-transplantation, while those from the let-7f-5p upregulation group could be detected up to day 14 but not in day 21 post-transplantation. The bioluminescent signals in mice transplanted with let-7f-5p overexpressed MSCs were generally higher than those transplanted with other groups of MSCs. Interestingly, the signal of let-7f-5p upregulation group increased upon injection, reached the highest on day 7, and dropped gradually after the following day. However, the other two groups showed a clear decline in the signal over a period of time (Figure 5D). This finding suggests that let-7f-5p could moderately prolong the retention of transplanted xenograft MSCs.

Figure 5

Let-7f-5p prolonged the retention of transplanted MSCs. (A) BLI of varying numbers of MSCs in vitro and linear relationship analysis between bioluminescence intensity and the cell number. (B) BLI of grafted MSCs in APPswe/PS1dE9 transgenic mice on different days after transplantation in vivo. (C) Time-course bioluminescence intensity curves from (A). (D) Time-course bioluminescence intensity curves from (B). The data in (C,D) are presented as mean ± standard deviation (SD). BLI, bioluminescence imaging.

Discussion

Excessive accumulation of Aβ may induce significant cytotoxicity in neurons, and is a key pathogenic factor of AD (Gupta et al., 2016; Kong et al., 2016). In the present study, Aβ25−35 was used to investigate the mechanism of Aβ neurotoxicity. Our data showed that Aβ25−35-treated cells exhibited a notable increase in cell early apoptosis, along with the upregulation of let-7f-5p expression. Using lentivirus transfection, we identified an inverse relationship between Aβ25−35-induced cell apoptosis and let-7f-5p levels from MTT assay and FACS analysis. Furthermore, caspase-3, a hub in the apoptotic pathway, was confirmed as a target gene of let-7f-5p, and the potential binding site was confirmed using the luciferase assay. These results were further confirmed in vivo. After transplanting into the AD mice, MSCs overexpressing let-7f-5p showed the lowest apoptotic rate, as determined by histological analysis, and the longest survival time, determined using bioluminescence, while the let-7f-5p knockdown resulted in the highest apoptotic rate and the shortest lifespan.

AD is characterized by Aβ plaque accumulation, tau hyperphosphorylation, and progressive loss of neurons. Despite various research and clinical trials on AD, no effective treatment has been developed till date (Park S. E. et al., 2017a). MSCs are a potential AD therapeutic option, due to their multi-target effect. Although MSCs was originally anticipated to replace dysfunctional cells via differentiation, the current paradigm is that MSCs could secrete various cytotropic factors that may reduce amyloid burden, modulate inflammation, and increase endogenous neurogenesis in AD (Kim et al., 2012; Lee et al., 2012). Some MSC-based treatment studies have shown promising results in AD animal models (Kim et al., 2015; Park et al., 2016). The intracerebroventricular route has been shown to be the optimal route to enhance MSC-based treatment effects, as compared to other routes for AD treatments, based on its advantages, such as the activation of neurogenesis (Park S. E. et al., 2017a). Therefore, we chose intracerebroventricular route for transplanting MSCs in mice in this study.

However, the harsh microenvironment in the AD condition in brain inevitably results in cell apoptosis, thus hampering the therapeutic efficacy of MSCs (Wang et al., 2017). For translating stem cell therapy into clinics, therefore, it is a prerequisite to optimize cell survival and function after engraftment (Lee et al., 2015). Till date, modifying MSCs (gene or microRNA modifications) before transplantation, chemically activating signaling pathways, and addition of growth factors, have been elucidated to augment survival of MSCs (Lee et al., 2015; Park W. S. et al., 2017b; Oh et al., 2018).

Aβ could damage neurons to stimulate pathogenesis of AD in vitro (He et al., 2017; Shanmuganathan et al., 2018), but its influence in MSCs has not yet been reported. For the first time, we found that overexpression of let-7f-5p increased cell viability and reduced the level of early apoptosis after Aβ25−35 treatment, indicating that overexpression of let-7f-5p partly rescued Aβ25−35-induced neurotoxicity in vitro. Apoptosis, a form of programmed cell death, is a highly controlled cell suicide process, associated with the activation of caspases, a dedicated set of proteolytic enzymes (Evan and Vousden, 2001). Caspases lead to DNA fragmentation by inducing cleavage of specific proteins in the early stages of apoptosis, and caspase-3 acts as the final effector in this process. First reported in Drosophila that dme-miR-14 can regulate cell proliferation and cell death, the involvement of miRNAs in regulating apoptosis has accumulated evidence during the past decade (Chandra et al., 2017). Consistent with previous reports, we identified caspase-3 as a target gene of let-7f-5p. The let-7 family was also found to be crucial in promoting cell survival by inhibiting protein kinase B or Akt, key protein involved in the activation of PI3K-Akt pathway (Zhuang et al., 2011) (Huat et al., 2015).

Survival of transplanted MSCs in various CNS models was successively reported to be limited up to several months in immune-deficient or immune-suppressed rodent models (Dezawa et al., 2004; Lu et al., 2006; Constantin et al., 2009; Zhang et al., 2011; Segal-Gavish et al., 2016). One report indicated survival of MSCs after 45 weeks (Zhang et al., 2006). However, in the non-immune suppressed mice, some studies reported survival of MSCs for only few days following transplantation (Munoz et al., 2005; Wakabayashi et al., 2010). In the APPswe/PS1dE9 transgenic mice without immune suppression, we found that the exogenous MSCs with the let-7f-5p augment were retained in brain more than 2 weeks.

Growing evidence has proven that let-7f played an important role in cell growth, migration, invasion, and angiogenesis in tumors (Yan et al., 2015). In our study, we observed decreased number of apoptotic MSCs in the brain after upregulation of let-7f-5p, together with the decline in caspase-3 expression. This anti-apoptosis effect has been reported in other let-7 family microRNA. Ham and his colleague revealed the let-7b-mediated pro-survival effect in transplanted mesenchymal stem cells for cardiac regeneration (Cheng et al., 2015). Let-7b expression upregulated the cell cycle-related proteins, CCND1 and CDK4, resulting in excessive proliferation of epidermal stem cells (Qin et al., 2012). Apoptosis has also been linked to the ability of let-7 to target Bcl-2, cyclin-dependent kinase 5, and Fas proteins (Barh et al., 2010). However, in the cardiovascular system, for example, let-7c overexpression enhances endothelial apoptosis mediated by oxidized-LDL through suppression of Bcl-xl (Qin et al., 2012). These results contradict our present study, but the different roles of let-7 family in cell death have been previously reported in cancer (Wang et al., 2015; Di Fazio et al., 2017). The difference between Aβ25−35-induced apoptosis and other kind of cell death might also account for these controversial results.

In general, more studies are required to confirm these data. Firstly, let-7f-5p alone may not be sufficient to completely regulate MSCs survival. Therefore, whether other miRNAs are involved in the Aβ-induced apoptosis remains to be further elucidated. In addition, as the transplanted cells did not survive for more than 1 month, repeated administrations may be essential. Moreover, we only detected the bioluminescent signals of injected MSCs on day 1, 7, 10, 14, and 21 post-implantation. Such long intervals make it impossible to identify the exact time of survival. More frequent observations should be made in the following experiments. More importantly, further studies on the efficacy and safety of MSC-based therapy in AD are needed. Next, we aim to explore whether let-7f-5p-modified MSCs could ameliorate dementia and pathology in the AD animals. We hope that more efficacious cell-replacement therapies will be developed in the near future to substantially restore disease-disrupted brain circuitry.

Overall, our study has provided evidence that by targeting caspase-3, let-7f-5p could improve MSCs survival in AD models, both in vitro and in vivo. We believe that MSC-based therapy will be a promising approach for neurodegenerative diseases in the future.

Author contributions

YJ and DZ conceived and supervised the study. LH and YZ designed experiments. LH, RZ, KW, and YLu performed experiments. YLi organized the database. RD and YY performed the statistical analysis. LH wrote the first draft of the manuscript. YZ, RZ, and KW wrote sections of the manuscript. All authors contributed to manuscript revision, read, and approved the submitted version.

Conflict of interest statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.