Melatonin protects mesenchymal stem cells from autophagy‐mediated death under ischaemic ER‐stress conditions by increasing prion protein expression

Abstract Object The purpose of this study was to explore whether melatonin could protect mesenchymal stem cells (MSCs) against ischaemic injury, by inhibiting endoplasmic reticulum (ER) stress and autophagy both in vivo and in vitro. Materials and Methods To confirm the protective effect of melatonin against ER stress in MSCs, markers of cell viability, apoptosis and autophagy were analysed. To further investigate the regenerative effect of melatonin‐treated MSCs in ischaemic tissues, a murine hindlimb ischaemic model was established. Results Under oxidative stress conditions, treatment with melatonin suppressed the activation of ER stress–associated proteins and autophagy‐associated proteins acting through upregulation of cellular prion protein (PrPC) expression. Consequently, inhibition of apoptotic cell death occurred. Melatonin also promoted the activation of MnSOD and catalase activities in MSCs. In a murine hindlimb ischaemia model, melatonin‐treated MSCs also enhanced the functional limb recovery as well as neovascularization. These beneficial effects of melatonin were all blocked by knock‐down of PrPC expression. Conclusion Melatonin protects against ER stress/autophagy‐induced apoptotic cell death by augmenting PrPC expression. Thus, melatonin‐treated MSCs could be a potential cell‐based therapeutic agent for ER stress–induced ischaemic diseases, and melatonin‐induced PrPC might be a key molecule in ameliorating ER stress and autophagy.


| INTRODUC TI ON
Mesenchymal stem cells (MSCs) represent one type of stem/progenitor cell-based therapies used to treat ischaemic diseases, such as stroke, cardiovascular disease, peripheral arterial disease, and ischaemic kidney disease. 1 Since these cells have a self-renewal ability, multi-differentiation potential, and an immunomodulatory effect, various studies have investigated the potential benefit of MSC-based therapies in pre-clinical and clinical trials. 1,2 Unfortunately, transplanted MSCs have shown a low survival rate in ischaemic-damaged tissue because a sudden or prolonged blockage of blood flow induces hypoxia, oxidative stress, endoplasmic reticulum (ER) stress and nutrient deprivation. 3,4 To improve the therapeutic efficacy of MSC-based therapies in pathophysiological conditions, a number of different approaches have been explored, including preconditioning, genetic modification, co-transplantation of MSCs with other supporting cells, administration of MSCs along with medication and application of the MSCs with biomaterials. 5 Although these strategies have improved the functionalities of MSCs, it is important to understand the underlying mechanisms behind these strategies in order to safely use MSC-based therapies in ischaemic diseases. The ER is an important organelle for protein folding, calcium homoeostasis, and lipid and carbohydrate metabolism. 6 Endoplasmic reticulum stress occurs in response to a range of cellular stresses, such as ischaemic damage, redox imbalance, oxygen and glucose deprivation and perturbations in calcium homoeostasis. 6,7 To maintain cellular homoeostasis against ER stress, cells have well-orchestrated processes, including ER-assisted degradation, autophagy, the hypoxia response and mitochondrial biogenesis. 6,8 Among these processes, autophagy plays a pivotal role in cell survival, cell growth, energy metabolism and lifespan in several cellular stresses under physiological conditions. 6,8 However, under pathophysiological conditions, severe stress induces excessive autophagy, causing cell death. 9 Therefore, in order to create a successful cell-based therapy for ischaemic diseases, the key is to inhibit the excessive autophagy without eliminating the basal autophagy level.
The normal cellular prion protein (PrP C ) is encoded by the Prnp gene, which plays a physiological role in various cells. 10 Accumulating evidence indicates roles for PrP C in a wide range of cellular processes, such as stress protection, oxidative stress-induced apoptosis, the ER-stress response, cellular differentiation, the immune system, mitochondrial homoeostasis, and metal ion homoeostasis. 11 Knockout of the Prnp gene induces a dysfunction in synaptic transmission and plasticity, memory formation, calcium homoeostasis, neurite outgrowth and metabolism. 10 In stem cell biology, PrP C is essential for cell proliferation, differentiation and self-renewal. [12][13][14] In addition, our previous studies have shown that PrP C is up-regulated by hypoxia preconditioning and melatonin treatment, and that transplanted MSCs facilitate neovascularization in a murine hindlimb ischaemia model. 15,16 Moreover, augmentation of PrP C expression by treatment of MSCs with tauroursodeoxycholic acid inhibited ER stress-mediated cell death. 17 Therefore, an understanding of the role of PrP C may provide insights into the development of stem/progenitor cell-based therapies.
Melatonin, N-acetyl-5-methoxytryptamine, is an indoleamine containing hormone secreted by several tissues including the bone marrow, ovary, testes, gut, placenta and liver, as well as the pineal gland. 18,19 Melatonin has a variety of functions, including the regulation of sleep and circadian rhythms, as well as having anti-inflammatory, antioxidant and anti-cancer effects. [20][21][22][23] Numerous studies have shown that melatonin-treated MSCs can facilitate a therapeutically functional recovery in myocardial infarction, skin wounds, lung ischaemia-reperfusion injury and sepsis-induced kidney injury. [24][25][26][27] In addition, our previous study has shown that melatonin-treated MSCs enhance neovascularization in hindlimb ischaemia through PrP C expression. 16 This study, as an extension of this research, focuses on the effect of melatonin on ER stress in ischaemic disease.
We show here that melatonin protects against autophagy-mediated apoptosis both in vitro and in vivo following ischaemic-induced ER stress and this occurs through the upregulation of PrP C expression.
In addition, we demonstrate the therapeutic effect of melatonintreated MSCs on neovascularization in a murine hindlimb ischaemia model.

| Cell culture
Human adipose-derived MSCs were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). The supplier certified that the MSCs expressed the cell surface markers (CD73 and CD105) and could undergo adipogenic or osteogenic differentiation when cultured with the appropriate differentiation medium. MSCs were maintained in alpha-minimum essential medium (Hyclone, Logan, UT, USA) supplemented with 10% (v/v) foetal bovine serum (Hyclone) and 100 U/mL penicillin/streptomycin (Thermo Fisher Scientific, Waltham, MA, USA). MSCs were incubated in a humidified incubator at 37°C and 5% CO 2 .

| Animal care procedures and experiments
The Institutional Animal Care and Use Committee of Soonchunhyang University approved all surgical interventions and post-operative animal care (IACIC2013-5), and these were performed in accordance with the National Research Council Guidelines for the Care and Use of Laboratory Animals. Animal experiments were performed on 8-week-old male Balb/C nude mice (Biogenomics, Seoul, Korea), which were maintained in a pathogen-free facility under a 12-hours light/dark cycle at 25°C with free access to water and regular laboratory chow in accordance with the regulations of Soonchunhyang University, Seoul Hospital.

| Murine hindlimb ischaemia model
Experiments using a murine hindlimb ischaemia model were performed as previously reported with minor modifications. 15,28 Ischaemia was induced by the ligation and excision of the proximal femoral artery and the boundary vessels of the mice. No later than 6 hours after surgery, PBS, untreated MSCs, or melatonin-treated MSCs transfected with PRNP siRNA or a scrambled siRNA were injected intramuscularly into the ischaemic thigh area (5 × 10 5 cells/100 μL PBS per mouse; five mice per treatment group). Each mouse was injected with cells (or PBS) at five ischaemic sites. Thereafter, melatonin (20 mg/kg/day) was administered intraperitoneally daily for 28 days. The melatonin stock solution was diluted in water, so that the final ethanol concentration was below 1%. Blood perfusion was assessed by measuring the ratio of the blood flow in the ischaemic (left) limb to that in the non-ischaemic (right) limb on post-operative days 0 and 28 using laser Doppler perfusion imaging (LDPI; Moor Instruments, Wilmington, DE, USA).

| Statistical analysis
All data are expressed as the mean ± SE of the mean (SEM) and all of the experiments were analysed using a one-way analysis of variance (ANOVA). Some comparisons of ≥3 groups were made using the Bonferroni-Dunn test. A P value <0.05 was considered statistically significant. These data were statistically analysed using SigmaPlot (Systat Software, San Jose, CA, USA).
Detailed some methods are provided in Appendix .

| Ischaemia-mediated ROS induces apoptosis through ER stress in MSCs and in hindlimb ischaemicinjured tissues
ROS is one of the inducers of ER stress which is known to be involved in cell life and death. 29 Our previous study demonstrated that ischaemia-induced ROS-mediated ER stress-triggered apoptosis in ischaemic tissues and led to a deterioration in the quality and efficacy of the transplanted MSCs used as a cell-based therapy for ischaemic diseases. 17 To further investigate the effect of ROS on ER stress in MSCs and ischaemic tissues, the activation and expression of ER stress-as- In addition, the activation and expression of these proteins were also increased in ischaemic-injured tissues ( Figure 1E

| ROS-mediated ER stress activates autophagy in both MSCs and hindlimb ischaemic-injured tissues
Under physiological conditions, ER stress induces autophagy in order to remove misfolded proteins, resulting in the inhibition of cell death; however, under pathophysiological conditions, severe autophagy leads to cell death. 30 To evaluate the effect of ROS-mediated ER stress on autophagy in MSCs, the expression of autophagy marker proteins was examined after treatment of the MSCs with H 2 O 2 . Treatment with H 2 O 2 increased activation of the ER-stress proteins, PERK and IRE1α, but the autophagy inhibitor 3-MA (500 μmol/L) did not significantly decrease this H 2 O 2induced activation of these proteins (Figure 2A,B). ROS-mediated ER stress also increased the expression of the autophagy markers, LC3B and Beclin-1, and 4-PBA (10 μmol/L) an ER-stress inhibitor, suppressed this ROS-induced augmentation of autophagy marker expression (Figure 2A,B). In addition, ROS-mediated ER stress increased the expression of another autophagy marker, ATG7, as well as LC3B and Beclin-1, and decreased the level of p62, which is known to be decreased following autophagy 31 ( Figure 2C,D).
These findings indicate that autophagy is activated following ROS-mediated ER stress. To further examine the effect of ischaemic-induced ER stress on autophagy in ischaemic-injured tissues, the expression of LC3B, Beclin-1, and ATG7 were increased and the level of p62 was decreased ( Figure 2E,F). In addition, immunofluorescence staining for LC3 showed that the level of LC3 was significantly increased in ischaemic tissue, compared with that in normal tissue ( Figure 2G,H). Taken together, these results suggest that ischaemia-induced ER stress activates the autophagy process.

| Melatonin protects cells from oxidative stress through the melatonin-MT 2 -PrP C axis
Our previous study demonstrated that melatonin enhances MSC bioactivities, including proliferation, anti-oxidative effects and immunomodulatory effects, through the upregulation of PrP C expression. 16 Expression of the melatonin receptor 2 (MT 2 ) gene is higher than that of MT 1 in MSCs, and melatonin significantly upregulates MT 2 . 16

| Melatonin attenuated ER stress under ischaemic conditions by regulation of PrP C
To elucidate whether melatonin regulates ER stress through PrP C expression, the activation and expression of ER stress-associated proteins were analysed after treatment of MSCs with H 2 O 2 and melatonin. Melatonin significantly decreased the activation and expression ER stress-associated protein ( Figure 4A-D). To further investigate the protective effect of melatonin in vivo, the activation and/or expression levels of these ER stress-associated proteins were also significantly decreased following transplantation with MSCs treated with melatonin ( Figure 4E-H). However, knock-down of PrP C in the MSCs abolished these inhibitory effects on ER stress-associated proteins. These results suggest that melatonin attenuated ER stress via upregulation of PrP C both in vitro and in vivo.

| Melatonin-induced PrP C protects cells from apoptotic cell death under conditions of ER stress through the regulation of autophagy
To explore whether melatonin protects ER stress-mediated autophagy in MSCs and ischaemic tissues through the regulation of PrP C , we analysed the expression of autophagy markers in the MSCs and ischaemic tissues transplanted with MSCs. Autophagy is regulated through the mTOR and AMPK signalling pathway. 32 In MSCs, treated with H 2 O 2 to induce ROS-mediated ER stress, melatonin activated mTOR and inhibited AMPK phosphorylation ( Figure 5A

| Melatonin-treated MSCs enhance neovascularization in a murine hindlimb ischaemia model
To examine whether melatonin-treated MSCs could promote the formation of neovessels in a murine hindlimb ischaemia model, blood perfusion and tissue recovery were assessed following transplantation of MSCs into the ischaemic-injured tissue. To stimulate the transplanted MSCs over the long-term, an intraperitoneal injection of melatonin was performed every day for 28 days following surgery. The blood perfusion ratio was analysed using LDPI at postoperative day 0, 3, 7, 14, 21 and 28. The blood perfusion ratio was significantly higher in limbs transplanted and maintained with the F I G U R E 1 Endoplasmic reticulum stress-mediated apoptotic cell death in MSCs and ischaemic tissue. A, Western blot of p-PERK, PERK, p-eIF2α, eIF2α and ATF4 after treatment of MSCs with hydrogen peroxide (200 μmol/L) for 0, 2, 4, 6 and 8 hours. B, The levels of p-PERK, p-eIF2α and ATF4 were normalized to those of PERK, eIF2α and β-actin, respectively. Values represent the mean ± SEM. *P < 0.05; **P < 0.01 vs untreated MSCs. C, Western blot of p-IRE1α, IRE1α, p-JNK, JNK and CHOP after treatment of MSCs with hydrogen peroxide (200 μmol/L) for 0, 2, 4, 6 and 8 hours. D, The levels of p-IRE1α, p-JNK and CHOP were normalized to those of IRE1α, JNK and β-actin, respectively. Values represent the mean ± SEM. *P < 0.05; **P < 0.01 vs untreated MSCs. E, Western blot of p-PERK, PERK, p-eIF2α, eIF2α and ATF4 in ischaemic-injured tissues from the murine hindlimb ischaemia model, 4, 12 and 24 hours after surgery. F, The expression of p-PERK, p-eIF2α and ATF4 were normalized to those of PERK, eIF2α and α-tubulin, respectively. Values represent the mean ± SEM. *P < 0.05; **P < 0.01 vs normal tissue. G, Western blot of p-IRE1α, IRE1α, p-JNK, JNK and CHOP in ischaemic-injured tissues from the murine hindlimb ischaemia model, 4, 12 and 24 hours after surgery. H, The expression of p-IRE1α, p-JNK and CHOP were normalized to those of IRE1α, JNK and α-tubulin, respectively. Values represent the mean ± SEM. *P < 0.05; **P < 0.01 vs normal tissue. I, Western blot of BCL2, BAX, cleaved caspase-3, and cleaved PARP-1 after treatment of MSCs with hydrogen peroxide (200 μmol/L) for 0, 2, 4, 6 and 8 hours. J, All protein expression levels were normalized to that of β-actin. Values represent the mean ± SEM. *P < 0.05; **P < 0.01 vs untreated MSCs. K, Western blot of BCL2, BAX, cleaved caspase-3 and cleaved PARP-1 in ischaemic-injured tissues from the murine hindlimb ischaemia model, 4, 12 and 24 hours after surgery. L, All protein expression levels were normalized to that of α-tubulin. Values represent the mean ± SEM. *P < 0.05; **P < 0.01 vs normal tissue melatonin-treated MSCs than with any other group ( Figure 7A,B).
In addition, the percentage of limb salvaged was significantly higher in limbs transplanted and maintained with the melatonin-treated MSCs than with any other group ( Figure 7C). In ischaemic-injured tissues, the secretion of the angiogenic cytokines, such as human VEGF, HGF and FGF, was significantly higher in mice transplanted and maintained with the melatonin-treated MSCs than with any other group ( Figure 7D). Immunofluorescent staining for CD31 (capillary) and α-SMA (arterioles) revealed that the densities of capillaries and arterioles were significantly increased in mice transplanted and F I G U R E 2 ER stress-mediated autophagy in MSCs and ischaemic tissue. A, Western blot of p-PERK, PERK, p-IRE1α, IRE1α, LC3B and Beclin-1 after pre-treatment of MSCs with 3-MA (500 μmol/L; an autophagy inhibitor) or 4-PBA (10 μmol/L; an ER-stress inhibitor) followed by treatment with hydrogen peroxide (200 μmol/L). B, The levels of p-PERK, p-eIF2α, LC3BII and Beclin-1 were normalized to those of PERK, eIF2α, LC3BI and β-actin, respectively. Values represent the mean ± SEM. **P < 0.01 vs untreated MSCs, ## P < 0.01 vs MSCs treated with H 2 O 2 , and $ P < 0.05; $$ P < 0.01 vs 3-MA-pretreated MSCs treated with H 2 O 2 . C, Western blot of LC3B, Beclin-1, p62 and ATG7 in cells after treatment of MSCs with hydrogen peroxide (200 μmol/L) for 0, 2, 4, 6 and 8 hours. D, The levels of LC3BII were normalized to LC3B1 whereas the levels of Beclin-1, p62, and ATG7 were normalized to β-actin. Values represent the mean ± SEM. *P < 0.05; **P < 0.01 vs untreated MSCs. E, Western blot of LC3B, Beclin-1, p62, and ATG7 in ischaemic-injured tissues from the murine hindlimb ischaemia model, 4, 12 and 24 hours after surgery. F, The levels of LC3BII were normalized to LC3B1 whereas the levels of Beclin-1, p62 and ATG7 were normalized to α-tubulin. Values represent the mean ± SEM. *P < 0.05; **P < 0.01 vs normal tissue. G, Immunofluorescent staining for DAPI (blue) and LC3B (green) in normal and ischaemic tissues. Scale bar = 30 μm. H, Autophagic cells were quantified as the percentage of LC3Bpositive cells per total DAPI-stained cells. Values represent the mean ± SEM. **P < 0.01 vs normal tissue maintained with the melatonin-treated MSCs than with any other group ( Figure 7E-H). Inhibition of PrP C expression significantly prevented these beneficial effects of melatonin on neovascularization.
However, the densities of capillaries and arterioles at the sites injected with melatonin-treated MSCs transfected with PRNP siRNA were significantly lower than those at the sites injected with MSCs  Figure 6F-H); this effect also led to the increased apoptosis in ischaemic-injured tissues (see, Figure 6I,J).
In addition, oxidative stress increased the expression PrP C (see, Figure 3A). Therefore, PRNP siRNA might suppress the physiological increase in PrP C , thereby reducing the functional recovery at sites injected with melatonin-treated MSCs transfected with PRNP siRNA compared with that at the sites injected with MSCs alone. These results indicate that transplantation with MSCs stimulated with melatonin improves neovascularization and functional recovery through the expression of PrP C .

| D ISCUSS I ON
Ischaemic-injured sites have harsh tissue conditions, including low oxygen levels, restricted nutrients, and high levels of ROS, with the result that effectiveness of the transplanted stem/progenitor cells used for regenerative medicine to treat ischaemic diseases is decreased. Among the pathophysiological conditions, ischaemia/ROSinduced ER stress is known to induce apoptotic cell death in host tissues as well as in the transplanted stem/progenitor cells. 6,7,17,29 To solve this issue, several studies have suggested that melatonin could be used to protect against ER stress in neurotoxic states, liver diseases, immunotoxic states, pulmonary fibrosis and diabetes. 33 Our previous study revealed that melatonin could enhance MSC functionality for the treatment of ischaemic disease by upregulating the express of PrP C . 16 To further examine whether melatonin protects against ER stress through the upregulation of PrP C , we have confirmed the protective effect of melatonin on autophagy-mediated cell death in MSCs with ischaemia-induced ER stress. Our results showed that melatonin-induced PrP C expression decreased ER stress/autophagy-mediated apoptosis both in vitro and in a murine hindlimb ischaemia model.
Various reports have indicated that PrP C , which is expressed in many cells beyond the central nerve system, 34 contributes to the regulation of several cellular process, including differentiation, proliferation, stress protection, myelin maintenance, mitochondrial homoeostasis and interaction with signal transduction pathways. 11,35 In particular, PrP C has been shown to have a protective effect against ROS-mediated oxidative stress. 11 PrP C suppresses ROS production by regulating the activities of superoxide dismutase and glutathione peroxidase. 36,37 In fact, our previous study revealed that TUDCA-induced PrP C increased MnSOD expression via phosphorylation of Akt and activation of its signalling pathway. 17 In keeping with the concept of PrP C regulating the activity of enzymes that control ROS levels it has been shown in a PrP-null mouse, that total SOD activity was significantly decreased through modulation of mitochondrial complex I. 38 Since ischaemic-injured tissues represent a ROS-rich oxidative stress environment, reductions in ROS levels are important for the successful transplantation of stem/progenitor cells in cell-based therapies. One study has shown that PrP C enhances MSC engraftment to bone marrow. 39 Hypoxia-induced PrP C expression has also been shown to facilitate the survival of transplanted MSCs at ischaemic-damaged sites in hindlimb model of ischaemia. 15 Our data indicated that melatonin increased the activa- Melatonin is known to be a powerful ROS scavenger through this regulation of antioxidant enzymes. 40,41 Several studies have investigated the protective effect of melatonin in MSCs with ischaemic injury in order to increase their efficacy in transplantation.
This study also showed that ROS/ischaemia triggered the activation of ER stress-associated proteins in both MSCs in vitro, and in ischaemic-injured tissues in vivo. Prolonged activation of PERK and IRE1α induce apoptotic cell death under pathophysiological conditions. 44 In addition to activation of the unfolded protein response ER stress stimulates autophagy. Autophagy is cellular F I G U R E 5 Melatonin inhibits oxidative ER stress-mediated autophagy via PrP C expression. A, Western blot of p-mTOR, mTOR, p-AMPK and AMPK after treatment of melatonin-pretreated MSCs with hydrogen peroxide (200 μmol/L). B, The levels of p-mTOR and p-AMPK were normalized to those of mTOR and AMPK, respectively. Values represent the mean ± SEM. **P < 0.05; **P < 0.01 vs untreated MSCs, ## P < 0.01 vs MSCs treated with H 2 O 2 , $$ P < 0.01 vs MSCs treated with melatonin, and && P < 0.01 vs MSCs treated with melatonin and si-PRNP. C, Western blot of LC3B, Beclin-1, p62 and ATG7 after treatment of melatonin-pretreated MSCs with hydrogen peroxide (200 μmol/L). D, The levels of LC3BII were normalized to LC3B1 whereas the levels of Beclin-1, p62 and ATG7 were normalized to β-actin. Values represent the mean ± SEM. *P < 0.05; **P < 0.01 vs untreated MSCs, # P < 0.05; ## P < 0.01 vs MSCs treated with H 2 O 2 , $ P < 0.05; $$ P < 0.01 vs MSCs treated with melatonin, and & P < 0.05; && P < 0.01 vs MSCs treated with melatonin and si-PRNP. E, Western blot of LC3B, Beclin-1, p62 and ATG7 in ischaemic-injured tissues 1 days after injection with PBS, MSCs (MSC), MSCs treated with melatonin (MSC + Melatonin), MSCs treated with melatonin and PRNP siRNA (MSC + si-PRNP + Melatonin), or MSCs treated with melatonin and scrambled siRNA (MSC + si-Scr + Melatonin) in a murine hindlimb ischaemia model. F, The levels of LC3BII were normalized to LC3B1, whereas the levels of Beclin-1, p62, and ATG7 were normalized to α-tubulin. Values represent the mean ± SEM. *P < 0.05; **P < 0.01 vs normal, ## P < 0.01 vs PBS, $ P < 0.05; $$ P < 0.01 vs MSC, and && P < 0.01 vs MSC + si-PRNP + Melatonin Values represent the mean ± SEM. *P < 0.05; **P < 0.01 vs normal, # P < 0.05; ## P < 0.01 vs PBS, $ P < 0.05; $$ P < 0.01 vs MSC, and && P < 0.01 vs MSC + si-PRNP + Melatonin. I, TUNEL assay in ischaemic-injured tissues from the groups in E-H. Scale bar = 50 μm. J, Apoptotic cells were quantified as the percentage of TUNEL (green)-positive cells per total DAPI-stained cells. Values represent the mean ± SEM. *P < 0.05; **P < 0.01 vs PBS, # P < 0.05 vs MSC, $ P < 0.05; $$ P < 0.01 vs MSC + Melatonin process that is essential for cell survival, development, differentiation and cellular homoeostasis. 45 Under physiological conditions, the basal level of autophagy plays an important role in cellular adaptation to nutrient starvation, endo-lysosomal degradation, the removal of misfolded and long-lived proteins, and other cellular stresses caused by detrimental cellular substances and invading pathogens. 8 However, excessive autophagy, initiated by long-term stress, triggers cellular damage, resulting in the induction of apoptotic cell death. 8,46 However, some studies have suggested that inhibition of autophagy protects against ischaemic-induced injury, whereas other studies indicate that activation of autophagy inhibits apoptosis. 33 pathways. 46 This study showed melatonin protected against ER stress/autophagy-mediated apoptotic cell death by appropriately regulating the expression of pro-and anti-apoptotic proteins. Of particular importance, these beneficial effects were blocked by silencing of PrP C . These results are in good agreement with our previous findings showing that melatonin regulated apoptosis-associated proteins under oxidative stress conditions by upregulating PrP C . 16 In the liver of leptin-deficient mice, melatonin has also F I G U R E 7 Assessment of functional recovery and neovascularization in a murine hindlimb ischaemia model. A, Blood perfusion was assessed using laser Doppler perfusion imaging analysis in the ischaemic limbs of mice injected with PBS, MSCs (MSC), melatonin-treated MSCs (MSC + Melatonin), MSCs treated with melatonin and PRNP siRNA (MSC + si-PRNP + Melatonin), or MSCs treated with melatonin and scrambled siRNA (MSC + Scr-siRNA + Melatonin). B, The blood perfusion ratio was measured by LPDI analysis (blood flow in the left ischaemic limb/blood flow in the right non-ischaemic limb). Values represent the mean ± SEM. *P < 0.05; **P < 0.01 vs PBS, and ## P < 0.01 vs MSC. C, Distribution of the different outcomes (limb salvage, toe loss and foot necrosis) in each of the treatment groups at post-operative day 28. D, Secretion of human VEGF, HGF and FGF from the ischaemic-injured tissues in each group was assessed by ELISA. Values represent the mean ± SEM. **P < 0.01 vs normal, ## P < 0.01 vs PBS, $$ P < 0.01 vs MSC, and && P < 0.01 vs MSC + si-PRNP + Melatonin. E-G At post-operative day 28, the formation of capillaries and arterioles were analysed using immunofluorescent staining for CD31 (green; scale bar = 50 μm; E) and α-SMA (red; scale bar = 100 μm; G). The capillary (F) and arteriole (H) densities were quantified as the number of CD31positive cells and α-SMA-positive cells, respectively. Values represent the mean ± SEM. *P < 0.05; **P < 0.01 vs PBS, # P < 0.05; ## P < 0.01 vs MSC, $$ P < 0.01 vs MSC+si-PRNP + Melatonin F I G U R E 8 Schematic representation of the proposed mechanisms by which melatonin protects against ER stress/autophagy-induced cell death through upregulation of the cellular prion protein (PrP C ) levels. Treatment of MSCs with melatonin inhibits ER stress, autophagy and apoptosis via PrP C upregulation. Moreover, melatonin increases the activation of MnSOD and catalase activity. Transplantation of melatonin-treated MSCs improves the functional recovery and vessel formation in ischaemic diseases through this melatonin-mediated PrP C expression been shown to inhibit oxidative damage through a reduction in ER stress and autophagy. 50 This study therefore suggests that melatonin-stimulated PrP C protects against ischaemic injury and apoptotic cell death by inhibiting the ER stress-autophagy axis both in vitro and in vivo. Of particular note, we demonstrated for the first time that PrP C is involved in melatonin-mediated ER stress and autophagy inhibition both in MSCs and in a hindlimb ischaemia model (Figure 8). Our findings therefore indicate that melatonin is a key regulator of ER stress and autophagy. In addition, our results suggest that melatonin-induced PrP C is a crucial protective molecule for cell survival against apoptosis induced by ER stress-mediated autophagy. These observations provide important insights into other ischaemia-induced cell death mechanisms that are inhibited by melatonin-induced PrP C expression. The funders had no role in the study design, data collection or analysis, decision to publish, or preparation of the manuscript.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interests.

AUTH O R CO NTR I B UTI O N S
JHL: study concept and design, acquisition of data, analysis and interpretation of data, drafting of the manuscript. YMY, YSH and SKJ: acquisition of data, analysis and interpretation of data, statistical analysis. SHL: study concept and design, acquisition of data, analysis and interpretation of data, drafting of the manuscript, procurement of funding, study supervision.