GO reduced high glucose induced apoptosis of Ad-MSCs.
We detected the increase in apoptosis of Ad-MSCs in a high glucose environment, which is consistent with our previous results (FigS1-S3)[8].After co-culturing GO with Ad-MSCs, we used flow cytometry to detect the apoptosis of Ad-MSCs, and found that the apoptosis rate was significantly reduced in the same high glucose environment after co-cultivation(Fig.S4 A, B). Then we used the TUNEL kit for apoptosis detection, and obtained Similar results (Fig.S4 C, D). Furthermore, with the increase of GO concentration, the reduction of apoptotic cells was more obvious. The Western blot experiment results showed that the expression of pro-apoptotic molecules BAX, BIM, and STK4 decreased with the increase of GO, and the expression of anti-apoptotic molecule BCL-2 increased with the increase of GO (Fig.S4E,S5), indicating that GO has inhibition of Ad-MSCs apoptosis in high glucose environment.
GO inhibited the expression of Linc00324 in Ad-MSCs.
As we mentioned, Linc00324 has been proved to be related to cell proliferation, invasion and migration in many studies[20, 21], but its role in cell apoptosis is not clear. So, we selected Linc00324 to explore its role in cell apoptosis. After our research, we found that the expression of Linc00324 in Ad-MSCs decreased after increasing the GO concentration during co-cultivation (Fig. 1A, B). So, we chose to establish a cell line that down-regulated and over-expressed Linc00324((Fig. 1C, D) and verify them, and found that under high glucose conditions, the apoptosis of the down-regulated Linc00324 group was significantly reduced, while the over-expression group had the opposite result (Fig. 1E-H, S6). In order to further verify whether GO regulates cell apoptosis through Linc00324, we used a cell line overexpressing Linc00324 to co-culture with GO and found that cell apoptosis did not decrease (Fig. 1J-M). Therefore, Linc00324 was one of the factors that GO regulated apoptosis.
Linc00324 decoyed miR-7977 as a sponge RNA.
In order to determine the mechanism by which Linc00324 regulates Ad-MSCs apoptosis, we performed qPCR to detect the location of Linc00324. Linc00324 is mainly located in the cytoplasm (Fig. 2A, B), which indicates that it may have the function of competing endogenous RNA and can serve as a molecular sponge for miRNA. A search of the miRDB database revealed the 5 miRNAs with the highest binding scores (Fig. 2C). We designed qPCR primers for these five miRNAs to determine whether they are regulated by Linc00324. It is worth noting that the level of miR-7977 was significantly reduced in the Linc00324 overexpression group, but increased in the Linc00324 knockdown group (Fig. 2D). Subsequently, we used RNA-FISH experiments to locate Linc00324 and miR-7977, and found that miR-7977 is also mainly located in the cytoplasm (Fig. 2E). In addition, we predicted the possible sites of Linc00324 in miR-7977, and constructed wild-type (WT) and mutant (MUT) luciferase reporter genes, including firefly and Renilla luciferase sequences (Fig. 2F). According to the result of luciferase assay, miR-7977 mimic reduced the fluorescence of Linc00324 WT, but had no effect on Linc00324 MUT (Fig. 2G).
The endogenous binding between miR-7977 and Linc00324 has been verified by RIP and verified by qPCR analysis. The results showed that compared with the empty vector (MS2) Linc00324 with a mutation (Linc00324-mut) in the miR-7977 targeting site, the Linc00324 RIP in Ad-MSCs was significantly enriched for miR-7977. The vector and another lncRNA-ATB without a predetermined miR-7977 targeting site vector (Fig. 2H). Using biotin-labeled Linc00324 to pull down the affinity of miR-7977 in vitro further confirmed the endogenous binding between miR-7977 and Linc00324 (Fig. 2I).
It is known that miRNA inhibits translation and degrades mRNA in an AGO2 (Argonaute RISC catalytic component 2) dependent manner by binding to the target. In order to prove whether Linc00324 can bind to miR-7977 in this way, we used an RNA pull-down assay to determine whether Linc00324 binds to AGO2. According to the western blot data, Linc00324 instead of AGO2 combined with the antisense control (Fig. 2J). RIP analysis was also performed to verify the interaction of Linc00324 and miR-7977. The enrichment of Linc00324 in the anti-AGO2 group confirmed the binding between Linc00324 and miR-7977 (Fig. 2K, L). In order to further prove the binding mode of Linc00324 and miR-7977, anti-AGO2 RIP was performed in Ad-MSCs with overexpression of miR-7977. The cells transfected with miR-7977 were specifically enriched with endogenous Linc00324 by AGO2 (Fig. 2M), indicating that miR-7977 may be Linc00324-carrying miRNA. Furthermore, we tested the expression level of Linc00324 in the miR-7977 mimics and inhibitors groups, and found that the expression level of Linc00324 decreased in the miR-7977 mimics group, but increased in the inhibitors group (Fig. 2N). Therefore, Linc00324 acted as a sponge for miR-7977.
miR-7977 targeted STK4 to inhibit the apoptosis of Ad-MSCs.
Since Linc00324 and miR-7977 have a competitive endogenous relationship, Linc00324 has been proven to promote cell apoptosis. It was transferred into miR-7977-mimics and inhibitors in the cells respectively. It was found that the miR-7977-mimics group decreased cell apoptosis under high glucose environment and the miR-7977-inhibitors group increased apoptosis by flow cytometer analysis (Fig. 3A, B) and TUNEL assay (Fig. 3C, D). The WB experiment results showed that the expression of BAX, BIM, and STK4 decreased with the transfection of miR-7977-mimics while the expression of BCL-2 increased (Fig. 3E, F). So, miR-7977 reduced the apoptosis of Ad-MSCs. What is the reason why miR-7977 reduces apoptosis? By predicting the downstream targets of miR-7977 via miRDB MicroRNA Target Prediction (Fig. 4A), we selected STK4, which is also called MST1,closely related to apoptosis[26]. To confirm whether miR-7977 targets STK4, we cloned the 3’UTR sequence of STK4 into the psiCHECK™-2 vector. and constructed a mutant 3'-UTR report that has no binding site to miR-7977(Fig. 4B). The data showed that introduction of miR-7977 diminished luciferase activity of this reporter and the activity of the mutant 3'-UTR reporter gene remained unchanged (Fig. 4C). At the same time, it was verified by qPCR and western blot experiments that miR-7977 had an inhibitory effect on STK4 at mRNA and protein levels (Fig. 4D-F). We added the STK4 inhibitor XMU-MP-1 to the miR-7977 inhibitor group. Flow cytometry analysis showed that XMU-MP-1 inhibited miR-7977 inhibitors apoptosis (Fig. 4G, H), which is consistent with the results of TUNEL analysis (Fig. 4I, J) and Western blot analysis (Fig. 4K, S7). So miR-7977 could directly target the 3’-UTR of STK4 and downregulate STK4 expression and inhibited the apoptosis of Ad-MSCs.
Linc00324 regulated the apoptosis of Ad-MSCs induced by high glucose through miR-7977/STK4.
We further investigated whether the miR-7977/STK4 axis is involved in the regulation of apoptosis by Linc00324 in a high glucose environment. We co-transfected Ad-MSCs with miR-7977 mimics and overexpression Linc00324 and found that miR-7977 mimics can significantly inhibit the apoptosis caused by overexpression Linc00324 via flow cytometry analysis (Fig. 5A, B). The TUNEL assay (Fig. 5C, D) and Western blot experiment (Fig. 5E, Fig.S8) also got consistent results. In addition, we also used miR-7977 inhibitors and down-regulated Linc00324 to co-transfect Ad-MSCs. The results showed that the miR-7977 inhibitors group reversed the decrease in apoptosis caused by downregulation of Linc00324(Fig. 5F-J, Fig S9). Based on the above data, Linc00324 regulated the apoptosis of Ad-MSCs induced by high glucose through miR-7977/STK4.
GO inhibited cell apoptosis and promoted wound healing in diabetic nude mice.
In order to further prove the effect of GO on Ad-MSCs-mediated wound healing of diabetic nude mice, we simulated and established a wound repair model with human skin wounds in diabetic nude mice. We established a wound with a diameter of 1.5 cm, and injected the treated cells with the fluorescent dye CM-Dil into the skin of the wound margin of each group of nude mice by intradermal injection, and tested the cell survival rate after 7 days It was found that the down-regulated Linc00324 and GO mixed culture group had the highest survival rate(Fig. 6A, B). Then we evaluated the wound healing effect of nude mice for 14 days. The results suggest that the GO group is better than the blank control group, and the down-regulated Linc00324 and GO mixed culture group has the best wound healing effect (Fig. 6C, D). Besides, the HE and Masson staining of tissue sections showed that the best group is still the down-regulation group Linc00324 and GO mixed culture group (Fig. 6E, F). We further used tissue immunofluorescence to detect the angiogenesis (CD31) and inflammation (TNF-α) of the wound tissue, and found that the expression of CD31 was high in the down-regulated group and the GO group, while the expression of TNF-α was low, suggesting good wound healing (Fig. 6G-I). We used the tissue of the wound to perform an ELISA test to detect the cytokines related to wound healing and obtained similar results (Fig. 6J). The results of in vivo experiments suggested that GO could inhibit the apoptosis of stem cells in diabetic nude mice, thereby promoting wound healing.