Ghrelin pretreatment enhanced the protective effect of bone marrow-derived mesenchymal stem cell conditioned medium on lipopolysaccharide-induced endothelial cells injury

Shanhui Ge Sun Yat-sen University First A liated Hospital Wanmei He Sun Yat-sen University First A liated Hospital Lishan Zhang Sun Yat-sen University First A liated Hospital Shan Lin Sun Yat-sen University First A liated Hospital Yuling Luo Sun Yat-sen University First A liated Hospital Qingui Chen Sun Yat-sen University First A liated Hospital Mian Zeng (  zengmian@mail.sysu.edu.cn ) Sun Yat-sen University First A liated Hospital https://orcid.org/0000-0001-9179-902X


Background
Acute respiratory distress syndrome (ARDS) is a life-threatening condition which is at least not rare in critically ill patients with high mortality and exerts a substantial disease burden [1].An international investigation revealed that 10.4% of ICU patients developed ARDS and the overall mortality of ARDS was about 40% [2].Therefore, to explore novel therapeutic strategies for ARDS in order to improve prognosis of the condition is of major importance.Lung endothelial barrier injury plays a crucial role in the pathogenesis of ARDS.Under normal condition, a functional alveolar-capillary barrier contributes to the normal gas exchange and removes alveolar edema uid.Under the pathological condition of ARDS, the damage to capillary endothelial cells and the junctions between cells leads to increased pulmonary vascular permeability and alveolar edema [3].Thus, lung endothelial barrier injury is a potential therapeutic target for ARDS.
Bone marrow-derived mesenchymal stem cell (BMSC) is a kind of adult stem cells with multidirectional differentiation potential and self-renewal ability [4].Several studies have reported a bene cial role of BMSCs for treating ARDS [5][6][7].A gradually recognized mechanism of the therapeutic effect is that BMSCs provide protective effects on the injured lung endothelial barrier either directly (via secreting soluble factors) or indirectly (via producing extracellular vesicles) [8].Therefore, instead of the cell itself, conditioned medium (CM) or extracellular vesicles derived from BMSCs are considered as a promosing therapeutic option given the cell-free nature [9][10][11].Recent studies have focused on how to strengthen the protective effects of BMSCs-derived products.Preconditioning BMSCs with hypoxia [12], cytokine [13], chemical drugs [14], and three-dimensional culture [15] were demonstrated to be capable of changing the secretion pro le of BMSCs and play an essential role in the treatments for autoimmune and in ammatory diseases.
Ghrelin was originally discovered as an endogenous hormone involved in the regulation of food intake, bone metabolism, but further researches found its role involved in various physiological functions including cell proliferation and apoptosis [16].We previously found ghrelin alleviated apoptosis and in ammation in acute lung injury [17].It has been reported that ghrelin preconditioning both improved the functional survival of MSCs and enhance its curative effect on ischemic heart disease [18].This suggests ghrelin pretreatment might regulate the the secretion pro le of BMSCs and thus improve its therapeutic effect, but it remains unknown whether this also applies to ARDS.Accordingly, our study aimed to investigate whether ghrelin pretreatment enhanced the protective effect of BMSCs-CM on lipopolysaccharide (LPS)-induced endothelial cells injury..

Mesenchymal stem cell isolation
Primary BMSCs were isolated from Rat using the whole bone marrow adherence method.Three-week-old male Sprague-Dawley rats were provided by the laboratory animal center of Sun Yat-sen University and sacri ced after anesthesia.Femur and tibia were separated and their marrow cavities were ushed using sterile phosphate-buffered saline (PBS) to collect bone marrow cells.Flushing uid was collected in a 15 mL sterile centrifuge tube and centrifuged at 1000 rpm for 5 minutes.The supernatant was discarded and the cell pellet was resuspended with a complete medium.The cell suspension was removed to the T25 culture asks and cultured at 37°C in 5% CO 2 .

Preparation of conditioned medium
The third-passage BMSCs of the logarithmic growth phase were treated with ghrelin for 24 hours after reaching 80% con uence, and then the medium was changed into serum-free DMEM.After 24 hours, the conditioned medium was collected, centrifuged at 1000 rpm for 5 minutes to remove cell debris, and stored at − 80°C after ltration, labeling as BMSCs Ghrelin pretreated -CM (conditioned medium derived from BMSCs pretreated by ghrelin).Conditioned medium of BMSCs in normal condition was collected in the same way except ghrelin treatment and marking as BMSCs-CM (conditioned medium derived from BMSCs).

Experimental design
To explore the therapeutic effects of BMSCs Ghrelin pretreated -CM, EA.hy926 cells were divided into four groups, including the control groups (no treatment or treated with DMSO), LPS group (LPS treatment for 24 hours), LPS + BMSCs-CM group (BMSCs-CM was added into the culture system after the endothelial cells were stimulated by LPS for 0.5 hour and co-cultured until 24 hours), and LPS + BMSCs Ghrelin pretreated -CM group (BMSCs Ghrelin pretreated -CM was added in the same way with LPS + BMSCs-CM group).To further con rm the role of AKT/GSK3β pathway in the protective effects of BMSCs Ghrelin pretreated -CM on endothelial cells, LY294002 at a concentration of 10 µM was added before LPS treatment.Finally, we compared the gene expression pro les between BMSCs and BMSCs Ghrelin pretreated to select the potential therapeutic molecules.

Cell viability assay
Cell viability assay was conducted using cell counting kit-8 reagent (CK04, DOJINDO, Japan).Endothelial cells were seeded in 96-well plates (2000 cells per well) and cocultured with conditioned medium derived from BMSCs pretreated with gradient concentrations of ghrelin for 24 hours.BMSCs were seeded in 96well plates (5000 cells per well) and cultured with serum-free DMEM containing gradient concentrations of ghrelin for 24, 48, and 72 hours.For the CCK8 assay, the cells in 96-well plates were added with serum-free DMEM containing 10% CCK8 solution.After incubation at 37°C for 2 hours, the absorbance at 450 nm was measured with a microplate reader (Sunrise, TECAN) Scratch wound healing assay Endothelial cells were seeded in 6-well plates (5×10 4 cells per well) and grew until 90% con uence.Scratch was made by a 200 µL sterile pipette tip and PBS was used to wash off the exfoliated cells.Then serum-free DMEM, LPS, and conditioned medium were added as described before.The photos of scratches were taken at 0 hour and 24 hours, respectively.The areas of scratches were measured and analyzed using Image J software.Wound healing rates (%) = (scratch area at 0 h-scratch area at 24 h)/ scratch area at 0 h × 100%.

Mitochondrial membrane potential assay
Endothelial cells were seeded in 6-well plates (5×10 4 cells per well) and treated as described before.
Mitochondrial membrane potential was detected using a Mitochondrial membrane potential assay kit with JC-1 (Beyotime, China).Endothelial cells were washed once using PBS and then incubated with a JC-1 probe (10 µM) at 37°C for 20 minutes.After incubation, cells were washed twice with JC-1 staining buffer to remove excess JC-1 probe.The green and red uorescence intensity of endothelial cells was observed under a uorescent microscope (Leica DMI8) and analyzed by Image J software.

Apoptosis assay by ow cytometry
Apoptosis rates of endothelial cells were detected by a FITC Annexin V Apoptosis Detection Kit I (BD Pharmingen™, USA).Endothelial cells were collected and centrifuged at 1000 rpm for 5 minutes.Then cells were washed twice with precooled PBS and resuspended in Annexin V binding buffer at a concentration of 1×10 6 cells/ml.FITC Annexin V and Propidium Iodide (PI) staining solution were added, blended gently, and incubated with endothelial cells at RT (25°C) for 15 minutes without illumination.The percentages of Annexin V and PI-positive cells were analyzed by ow cytometry within one hour.

Quantitative real-time PCR
Total RNA was extracted from endothelial cells using RNA-Quick Puri cation Kit (RN001, ESscience, China) and the purity was detected by the A260/A280 ratios.Then 1000 ng RNA was reverse transcribed to cDNA using cDNA Synthesis SuperMix (NovoProtein, China).Quantitative real-time PCR was performed by SYBR qPCR Mix (NovoProtein, China) to analyze the expression of GAPDH, interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α), and interleukin-1β (IL-1β) (primers of each gene were shown in Table 1).

RNA sequencing analysis
Total RNA was extracted from the BMSCs and BMSCs pretreated with ghrelin and residual DNA was removed using DNase I.The RNA was successively puri ed by RNA Clean XP magnetic beads and enriched by Oligo (dT) magnetic beads.The enriched product (mRNA) was divided into small fragments (150-300 nt) and cDNA was synthesized using Reverse Transcriptase SuperScript II, RNaseH, and DNA polymerase, sequentially.The cDNA was puri ed by AmPure XP Bead and repaired the terminal.Then A was added to the 3' and the sequencing joints were connected.The cDNA library was established after puri cation and qPCR ampli cation.The quali ed library was pooling according to the effective concentration and the amount of target data, and the sequencing is conducted on the HiSeq platform according to PE 150 strategy.The TPM (Transcripts Per Million) was utilized to calculate the expression of genes and those with |log2 (fold change) |≥1 and p-value < 0.05 were considered statistically signi cant genes.Gene Ontology (GO) (http://www.geneontology.org/) and KEGG (http://www.genome.jp/kegg/)analysis were performed to discuss the functions and enriched pathways of statistically signi cant genes.

Statistical analysis
Continuous data were expressed as mean ± standard deviation ().One-way ANOVA and post hoc tests were utilized to compare differences among the groups.All the experimental data were statistically analyzed by SPSS 25.0.A p-value less than 0.05 was considered as statistically signi cant.

Surface markers identi cation of BMSCs
At passage 3, BMSCs adhered to the plate with uniform morphology and spindle shape (Fig. 1a).

Optimum concentration selection of ghrelin
To determine the optimal concentration of ghrelin for pretreatment, we rst detected the effect of different concentrations of ghrelin on the viability of BMSCs under a serum deprivation condition.When cultured in a serum-free medium, the viability of BMSCs decreased in a time-dependent manner, and the best viability of BMSCs was observed when the cells were treated by 100 nM and 1000 nM ghrelin (Fig. 2a).We further collected the conditioned medium from BMSCs stimulated by different concentrations of ghrelin for 24 hours and co-cultured with endothelial cells.The results showed that the best viability of endothelial cells was observed when co-cultured with conditioned medium derived from 100 nM ghrelin pretreated BMSCs (Fig. 2b).According to these results, we chose the 100 nM ghrelin as the appropriate concentration for the following experiments.
BMSCs Ghrelin pretreated -CM promotes the migration ability of endothelial cells Scratch wound healing assay was utilized to analyze the migration ability of endothelial cells.As exhibited in Fig. 3a ~ b, LPS signi cantly inhibited endothelial cells migration, and BMSCs-CM treatment could enhance endothelial cells migration.Compared with the BMSCs-CM, BMSCs Ghrelin pretreated -CM better-promoted wound closure after scratching.

BMSCs Ghrelin pretreated -CM alleviates LPS-induced apoptosis of endothelial cells
To evaluate the effect of BMSCs Ghrelin pretreated -CM on LPS-induced apoptosis, three methods were used to detect the apoptosis of endothelial cells.As shown in Fig. 3c ~ d, LPS treatment signi cantly decreased the mitochondrial membrane potential of endothelial cells (p 0.05)..Compared with the LPS group, the mitochondrial membrane potential in BMSCs-CM and BMSCs Ghrelin pretreated -CM groups was lower (p < 0.05).However, no signi cant difference was found between BMSCs-CM and BMSCs Ghrelin pretreated -CM groups.
Next, AnnexinV and PI dual staining was used to label the apoptotic cells (Fig. 3e ~ f).In comparison with the control group, LPS induced obvious apoptosis in endothelial cells (7.33% versus 50.93%, respectively, p < 0.01).The percentages of apoptotic cells in BMSCs-CM and BMSCs Ghrelin pretreated -CM groups were higher than that of the LPS group.Moreover, the apoptotic rate of endothelial cells in the BMSCs Ghrelin pretreated -CM group was signi cantly lower compared with the BMSCs-CM group(38.94%versus 46.59%, respectively, p < 0.01).Furthermore, the results of western blotting showed that BMSCs Ghrelin pretreated -CM reduced the expression of Bax protein and increased the expression of Bcl-2 protein compared with the LPS group and BMSCs-CM group (Fig. 3g ~ i).
The above results indicated that BMSCs Ghrelin pretreated -CM effectively reduced LPS-induced apoptosis of endothelial cells.

BMSCs Ghrelin pretreated -CM decreases the expression of the in ammatory cytokines of endothelial cells
As the results of qPCR showed (Fig. 4), in ammatory cytokines TNF-α, IL-1β, and IL-6 exhibited increased expression after LPS treatment, and the expression of above cytokines was downregulated in BMSCs-CM and BMSCs Ghrelin pretreated -CM groups when compared with the LPS group with no signi cant difference between the BMSCs-CM and BMSCs Ghrelin pretreated -CM groups.
AKT /GSK3β signaling pathway mediated the protective effect of BMSCs Ghrelin pretreated -CM To investigate the protective mechanism of BMSCs Ghrelin pretreated -CM on endothelial cells, we utilized western blotting to assess the phosphorylation of AKT and GSK3β protein.LPS decreased the phosphorylation of AKT and GSK3β protein compared with the control group, whereas BMSCs Ghrelin pretreated -CM signi cantly increased the phosphorylation of AKT and GSK3β protein (Fig. 5a,b,c).Then, PI3K/AKT inhibitor LY294002 was added to block activation of AKT signaling pathway, we found that BMSCs Ghrelin pretreated -CM didn't increase the expression of p-AKT and p-GSK3β protein in the existence of LY294002 (Fig. 5d,e,f).These results indicated that BMSCs Ghrelin pretreated -CM mediated the activation of the AKT /GSK3β signaling pathway.
We further detected whether the protective effects of BMSCs Ghrelin pretreated -CM was disturbed by LY294002.We treated endothelial cells with LY294002 1h before LPS stimulation, and the anti-apoptotic effect of BMSCs Ghrelin pretreated -CM was attenuated (Fig. 6a,c).Meanwhile, LY294002 partially reversed the effects of BMSCs Ghrelin pretreated -CM on endothelial cells' migration promotion and in ammation cytokines reduction (Fig. 6b,d ~ g).These results suggested that BMSCs Ghrelin pretreated -CM protected endothelial cells from LPS induced injury in part through the activation of the AKT /GSK3β signaling pathway.

In uence of ghrelin on the genes expression pro le of BMSCs
To explore the potential mechanism by which BMSCs Ghrelin pretreated -CM activates the AKT /GSK3β signaling pathway in endothelial cells, we performed transcriptome sequencing on BMSCs with and without ghrelin treatment and analyzed the relationship between differentially expressed genes (DEGs) and the function of endothelial cells.In total, we found 88 upregulated genes and 405 downregulated genes in the ghrelin group (Fig. 7a).The GO and KEGG analysis exhibited that DEGs were mainly related to the biological processes including cytokines and chemokines interaction, osteoclast differentiation, immune responses, and defenses to microbial infections(Fig.7b,c).We further discovered 5 upregulated DEGs both associated with activation of AKT signaling pathways and endothelial function: Wnt5a, S100b, Bmp2, Id4, and Pthlh (Fig. 7d), most of which encoded secreted proteins.

Discussion
In this study, we for the rst time explored the effects of BMSCs Ghrelin pretreated -CM on LPS-induced endothelial injury.Our results showed that ghrelin pretreatment signi antly enhance the protective effects of BMSCs-CM on LPS-induced endothelial barrier dysfunction, which was partly mediated by activation of the AKT/GSK3β signaling pathways.Our study provided the theoretical basis for a new therapeutic strategy of ARDS.
Currently, the mortality of ARDS remains high, about 40%, and the main therapeutic strategies were lungprotective mechanical ventilation and anti-infection treatment, and there is lack of intervention targeting at the pathogenesis [19].The damage of alveolar epithelial cells and alveolar endothelial cells caused by various pathogenic factors is the key process of the ARDS.LPS, also known as endotoxin, is a constituent of the outer wall of Gram-negative bacteria cells and is often used to induce ARDS injury model [20].In this study, EA.hy926 cells were treated with LPS to induce ARDS-related endothelial injury, and we found that LPS signi cantly increased apoptosis, elevated the expression level of TNF-α, IL-1β, and IL-6, and impaired the migration ability of the endothelial cells.These results indicated that LPS treatment effectively induced endothelial dysfunction.
Known as pluripotent stem cells with immunomodulatory capacity, MSCs secrete cytokines, growth factors, and bioactive factors to reduce the in ammatory response and repair damaged tissues.The conditioned medium derived from MSCs was demonstrated to attenuated the endothelial barrier dysfunction in different conditions, including hypoxia [21], ischemia [22], diabetes [23], and wound healing [24].In this study, we found that BMSCs-CM reduced the apoptosis and in ammatory factors expression, and improved the migration ability of endothelial cells, which was consistent with previous studies.
Ye et al [25] demonstrated that BMSCs highly expressed growth hormone secretagogue receptor (GHSR), which was the receptor of ghrelin.Ghrelin promoted the viability, osteogenesis, and chondroblast differentiation of MSC by regulating the ERK1/2 signaling pathway [25,26].As reported, ghrelin preconditioning enhanced the therapeutic effects of MSCs[18], then we assumed that ghrelin preconditioning might also impact the effects of MSCs on the endothelial injury.First, we detected the viability of endothelial cells co-cultured with conditioned medium derived from different concentrations of ghrelin pretreated BMSCs with LPS treatment, and 100 nM ghrelin-pretreated BMSCs-CM better improved the viability of endothelial cells compared with BMSCs-CM.Then, therapeutic effects of BMSCs Ghrelin pretreated -CM on endothelial cell injury under LPS attack were observed, and BMSCs Ghrelin pretreated -CM remarkably reduced the apoptosis and improved the migration of endothelial cells in comparison with BMSCs-CM.Taken together, our results suggested that ghrelin preconditioning enhanced the ability of BMSCs-CM to repair the endothelial injury.
AKT/GSK3β signaling pathway plays an essential role in regulating the process of apoptosis, in ammation, proliferation, and migration [27,28].Si et al [29]found that adrenomedullin alleviated the apoptosis induced by hypoxia and serum deprivation via activating the AKT/GSK3β signaling pathway.In LPS-stimulated macrophages, the activation of the PI3K/AKT/GSK3β pathway reduced the nuclear translocalization of NF-κB and then downregulated the expression of IL-6 and IL-12 [30].Furthermore, AKT/GSK3β pathway was con rmed to mediate the repair of endothelial barrier dysfunction in ARDS [31].Therefore, we assessed the activation of the AKT/GSK3β pathway using western blotting.The results showed that BMSCs Ghrelin pretreated -CM activated the AKT/GSK3β pathway, and the protective effects of BMSCs Ghrelin pretreated -CM were partly reversed by the PI3K/AKT inhibitor LY294002, which indicated that BMSCs Ghrelin pretreated -CM alleviated the endothelial injury induced by LPS partly via activating the AKT/GSK3β pathway.
Finally, we tried to investigate the change of gene expression pro le in BMSCs pretreated by ghrelin using RNA-seq sequencing.According to our previous results, we hypothesized that ghrelin preconditioning promotes the secretion of certain proteins which acts on endothelial cells and activates the AKT/GSK3β pathway to exert protective effects.We discovered ve upregulated genes in ghrelin preconditioning BMSCs which were both related to the activation of the AKT pathway [32][33][34][35][36][37] and the endothelial function: Wnt5a, S100b, Bmp2, Id4, and Pthlh.Wnt5a, known as secretory protein, is an important regulator of the non-classical Wnt pathway and plays a double-sided role on endothelial cells.Zhang et al [38]found that exosomes derived from MSCs reduced hypoxia-induced apoptosis of rat pulmonary artery endothelial cells by increasing the expression of Wnt5a protein while Breton-Romero et al [39] demonstrated that the activation of Wnt5a contributed to the endothelial dysfunction.S100b is a secretory calcium-binding protein that was initially found to be secreted by glial cells, and subsequently detected in non-neural tissues such as chondrocytes and adipose tissue [40].Exogenous S100b protein promoted angiogenesis by facilitating the release of VEGF and tube formation of human umbilical vein endothelial cells [41].Bmp2 (bone morphogenetic protein 2), a member of the TGF-β superfamily, regulated bone formation through autocrine and paracrine mechanisms.Recent studies reported that Bmp2 also had an impact on endothelial cells and angiogenesis.Zuo et al [42]revealed that Bmp2 promoted the proliferation, migration, and angiogenesis of human umbilical vein endothelial cells through p38, ERK, and AKT/mTOR pathways.In addition, ID4 (DNA binding inhibitor 4) and Pthlh (parathyroid hormone-associated protein) were mainly related to the pro-angiogenesis ability of cancer [37,43].Taken together, our results indicated that ghrelin preconditioning upregulated the expression of Wnt5a, S100b, Bmp2, Id4, and Pthlh, which probably contribute to the improvement of endothelial dysfunction.

Limitations
Our study has several limitations.First, a endothelial cell line but not primary endothelial lung cells was utilized to simulate the pulmonary endothelial system in vitro.Moreover, the pathway inhibitor-LY294002 used in this study might has off target effects, and siRNA methodology would be a better method to silence AKT.Besides, we only con rmed the protective effects of BMSCs Ghrelin pretreated -CM in vitro, lack of further con rmatory animal experiments.

Conclusion
In summary, we demonstrated that ghrelin pretreatment strengthen the protective effects of BMSCs-CM by inhibiting LPS-induced apoptosis and in ammation, and improving the migration of endothelial cells in vitro via activation of AKT/GSK3β pathway, which might be associated with ghrelin-induced upregulation of Wnt5a, S100b, Bmp2, Id4, and Pthlh genes in BMSCs.Nevertheless, the detailed mechanism involving in the therapeutic effects of ghrelin-pretreated BMSCs-CM needs further studies to elucidate.

Figures
Figures

Figure 2 Selection
Figure 2