Bone marrow mesenchymal stem cells enhance autophagy and help protect cells under hypoxic and retinal detachment conditions

Abstract Our study aimed to evaluate the protective role and mechanisms of bone marrow mesenchymal stem cells (BMSCs) in hypoxic photoreceptors and experimental retinal detachment. The cellular morphology, viability, apoptosis and autophagy of hypoxic 661w cells and cells cocultured with BMSCs were analysed. In retinal detachment model, BMSCs were intraocularly transplanted, and then, the retinal morphology, outer nuclear layer (ONL) thickness and rhodopsin expression were studied as well as apoptosis and autophagy of the retinal cells. The hypoxia‐induced apoptosis of 661w cells obviously increased together with autophagy levels increasing and peaking at 8 hours after hypoxia. Upon coculturing with BMSCs, hypoxic 661w cells had a better morphology and fewer apoptosis. After autophagy was inhibited, the apoptotic 661w cells under the hypoxia increased, and the cell viability was reduced, even in the presence of transplanted BMSCs. In retina‐detached eyes transplanted with BMSCs, the retinal ONL thickness was closer to that of the normal retina. After transplantation, apoptosis decreased significantly and retinal autophagy was activated in the BMSC‐treated retinas. Increased autophagy in the early stage could facilitate the survival of 661w cells under hypoxic stress. Coculturing with BMSCs protects 661w cells from hypoxic damage, possibly due to autophagy activation. In retinal detachment models, BMSC transplantation can significantly reduce photoreceptor cell death and preserve retinal structure. The capacity of BMSCs to reduce retinal cell apoptosis and to initiate autophagy shortly after transplantation may facilitate the survival of retinal cells in the low‐oxygen and nutrition‐restricted milieu after retinal detachment.


| INTRODUC TI ON
Retinal detachment (RD), defined as the separation of neurosensory retina (including photoreceptors) from the underlying retinal pigment epithelium (RPE) which transport nutrients (including glucose) to photoreceptors, is one of the most common sight-threatening eye diseases. [1][2][3] The separation may lower the supply of oxygen and nutrients to photoreceptor out segments, rendering a relative hypoxic milieu which may further impair the energy production required for nutrients transport to photoreceptors, thus diminishing the photoreceptor function. 1,3 Hypoxia can also engender oxidative stress, causing photoreceptor apoptosis. 1,4 In addition, inflammatory cytokines released in RD also contribute to photoreceptor death. 5 Although the anatomic reattachment rate has greatly increased with advances in surgical management, impaired visual acuity may also occur due to photoreceptor apoptosis, necrosis, autophagy, retinal remodelling together with other structural and functional retinal changes. 1,[5][6][7] Attempts to save photoreceptors have been made, including suppressing factors participating in apoptosis, providing neurotrophic factors and modulating inflammation. However, as photoreceptor death is mediated by multiple parallel pathways, strategies aiming at a single factor are not always completely valid. 1,2,8,9 Bone marrow mesenchymal stem cells (BMSCs), acting mainly as the progenitors of all connective tissue cells, can differentiate into several tissue-forming cells and play a vital role in retinal cell regeneration. [10][11][12] Many studies have also demonstrated that BMSCs exert anti-apoptotic and anti-inflammatory effects and represent a promising strategy for the treatment of retinal diseases. Mechanisms underlying the effects of BMSCs include but were not limited to their ability to express a variety of cytokines and neurotrophic factors. 13,14 BMSCs were also reported to actively participate in autophagy to ameliorate ischaemia/reperfusion-induced lung injury and liver fibrosis. 15,16 More important, BMSCs were reported to protect ischaemic retina by increasing autophagy. 17 However, there has been no report on the effects of BMSCs on photoreceptor damage in retinal detachment. Thus, our study was to investigate whether the BMSCs could protect detached and hypoxic retina and whether autophagy participates in the effects of BMSCs on retina.
Our study here suggested that BMSCs play a protective role against photoreceptor death in experimental RD. We also studied the mechanisms by which these stem cells protect photoreceptors in vitro and in vivo.

| Cell culture
Rat BMSCs were isolated from bone marrow stroma harvested from the femurs and tibias of female Wistar rats (7 days after birth) as described. 18 The cells were seeded in Dulbecco's modified Eagle's Medium (DMEM)-F12 (HyClone) supplemented with 10% foetal bovine serum (FBS, Gibco) at 37°C in humified air with 5% CO 2 . Rat BMSCs were phenotypically characterized and identified as described, 19 as Figure S1 for BMSC identification. Bone marrow mesenchymal stem cells from the third passage (P3) to P5 were used for transplantation. Mouse BMSCs (P6) were bought from Cyagen BioTech, and P7 and P8 cells were used for coculture assays.
The 661w cell line was a gift from Dr Muayyad R. Al-Ubaidi (University of Oklahoma Health Sciences Center, USA). As a mouse photoreceptor-derived cell line, 661w was commonly used as an in vitro model for photoreceptor damage, because these cells express multiple cell markers of cone photoreceptors. 20,21 The cells were cultured as described previously in high-glucose DMEM (HyClone) supplemented with 10% FBS (Gibco) in a humidified atmosphere of 95% air and 5% CO 2 at 37°C. 22 These cells were generally passaged by trypsinization at a ratio of 1:6 every 3-4 days. For experiments performed in hypoxic conditions, culture dishes and plates were cultured in a sealed, anaerobic workstation (Ruskin Technologies), where the hypoxic condition (1% O 2 , 94% N 2 and 5% CO 2 ), temperature (37°C) and humidity (90%) were kept constant. 23,24 To inhibit autophagy, 3-MA (Sigma) was added to the culture medium to achieve a final concentration of 10 mmol/L 1 hour before the experiments as previously suggested. 25

| In vitro coculture experiments
A Transwell coculture system was used to explore the protective effects of mouse BMSCs under the hypoxic condition. 661w cells were seeded and grown in 24-well (2 × 10 5 cells/well) or 6-well (1 × 10 6 cells/well) Transwell plates (Corning). Bone marrow mesenchymal stem cells at the third passaged were seeded onto the upper or lower well of the Transwell system and cultured for 2 days. The 661w cells were randomly divided into the following groups: (a) normal group: no treatment; (b) hypoxia group: 661w cells cultured under the hypoxic condition; (c) hypoxia +3-MA group: the autophagy inhibitor 3-MA was added 1 hour before the hypoxia treatment; (d) coculture group: BMSCs cocultured with hypoxic 661w cells; and (e) autophagy-inhibited (inhibited) group: the autophagy inhibitor 3-MA was added 1 hour before the coculturing began under hypoxia.

| Cell viability assay
The MTS assay (CellTiter 96 ® AQueous One Solution, Promega) was performed to determine the viability of cells in each group.
Briefly, CellTiter 96 ® AQueous One Solution was added (20 μL/ well), and the cells were incubated for 1 hour at 37°C in a 5% (v/v) CO 2 atmosphere. The optical density at 490 nm (OD 490 ) of each well was measured with a multifunctional microplate reader (VarioSkan, Thermo), using a background control as the blank.
Each group was assayed in triplicate, and this test was performed for three times. The cell viability was expressed as the percentage of the control.

| Mitochondrial transmembrane potential (ΔΨm)
JC-1 (Sigma) staining was used to determine ΔΨm. Briefly, 661w cells in each group were incubated with a JC-1 working solution (0.3 μg/ mL) at 37°C in the dark for 30 minutes and observed under a confocal microscope (Olympus FV1000).

| Caspase activity assays
Caspase-3 activity was measured in 12-well plates by the luminescent method using the Caspase-Glo Assay kit (Keygene Biotech).
Cells of each group were washed twice with PBS, before 1 mL/well of fresh growth medium and 1 μL of detection solution were added.
After incubation for 30 minutes in the dark, luminescence was measured by the microplate reader VarioSkan.

| Monodansylcadaverine staining
Monodansylcadaverine (MDC; Sigma) staining was used to detect autophagosomes in cells of each group. After washing with PBS for three times, cells were incubated with MDC (0.05 mmol/L in PBS) at 37°C for 60 minutes, patterns of green fluorescence were detected by confocal microscopy at an excitation wavelength of 492 nm and an emission wavelength of 520 nm.

| Retinal detachment induction and treatment administration
RD in rats was induced as previously described. 2 Briefly, an anterior chamber puncture was performed via the corneal limbus to lower the intraocular pressure, and approximately one half of the retina was detached mechanically by the subretinal injection

| TUNEL assay and ONL thickness ratio evaluation
Prior to staining with fluorescent dye DAPI (Sigma), which was often used to show the nuclei of cells, a terminal dUTP nick-end labelling

| Immunohistochemistry
Three sections were incubated with 1% BSA for 1 hour to block nonspecific binding. Subsequently, the sections were incubated with primary antibody (the concentrations are listed in Table 1) overnight at 4°C. Alexa Fluor 633-conjugated goat anti-mouse IgG or Alexa Fluor 488-conjugated goat anti-mouse IgG (1:800, Invitrogen) was used as the secondary antibody, and the mixture was incubated at room temperature for 1 hour. The nucleus was counterstained with DAPI, and images of the retina were taken with an epifluorescence microscope (Olympus IX71).

| Western blot analysis
Western blotting was performed to evaluate the key players in the apoptotic cascade and photoreceptor and glial cell markers to delineate the protective mechanisms of the cells. Retinas from eyes in those three groups were dissected from the RPE choroid at 3 days, 1 week, 2 weeks and 4 weeks after BMSC transplantation.
Proteins were collected according to a method described previously. 8 The concentration of each sample was determined using the BCA method. Protein samples were then separated with 10% SDS-PAGE gels and transferred to PVDF membranes (0.45-μm pores; Millipore). After blocking with 5% non-fat milk in TBST, the membranes were incubated with a primary antibody (Table 1) at 4°C overnight. The membranes were then incubated for 60 minutes at room temperature with an HRP-labelled anti-rabbit or anti-mouse secondary antibody (1:5000, Millipore). Protein bands were detected by enhanced chemiluminescence (ECL, Millipore) and exposure to film (Aermei-film). Densitometry analysis was performed with Image-Pro Plus, and protein levels were normalized to those of actin.

| Statistics
All statistical analyses were performed using SPSS 19.0 software (IBM, IL). Numerical data are presented as the means and standard deviation (±SD). Differences between two groups were analysed using t tests or Mann-Whitney U tests, while multiple groups were analysed by one-way ANOVA or Kruskal-Wallis tests. P < .05 was considered a significant difference.

| Autophagy plays a protective role in hypoxiatreated 661w cells
When cultured under hypoxic conditions, 661w cells showed significant morphological changes, especially after 24 hours, and some cells were even rounded and floating ( Figure 1A). The cell viability decreased as the hypoxic time extended, falling below 50% of that of normal cells after 48 hours ( Figure 1B). The rate of cell apoptosis mildly increased after 2 hours in hypoxia and gradually increased as the low-oxygen exposure extended; at 48 hours, the proportion of necrotic cells surpassed that of apoptotic cells, and necrosis became the main reason underlying the observed decrease in viability ( Figure 1C).
The hypoxia condition was previously shown to induce autophagy in 661w cells. 24 We confirmed this in our study ( Figure 1D) and further inhibited autophagy with 3-MA to study its protective role in hypoxic 661w cells. Cells were incubated with 3-MA, an autophagosome-lysosome fusion inhibitor, 1 hour before the hypoxic conditions were introduced. When 3-MA was added to the normoxic culture, no significance difference was observed between the two groups ( Figure 2). However, after 8 hours in hypoxia, both autophagy-related protein expression and MDC staining (green puncta revealed MDC-labelled autophagosomes) showed that autophagy was up-regulated in the hypoxia group and suppressed in hypoxic cells treated with the 3-MA inhibitor ( Figure 2). Upon analysing the cellular morphology, viability, apoptosis rate and ΔΨm, hypoxia was shown to exert a detrimental effect on the cells. When autophagy was inhibited, the cells showed no significant changes under the normoxic condition. Compared with those in the hypoxia group, cells in the hypoxia +3-MA group were more morphologically altered and had a lower viability and a higher apoptotic rate (P < .05, Figure 3). This indicated that inhibiting autophagy in a hypoxic environment increases cell apoptosis and decreases cell viability, and autophagy may play a protective role in the early stage of hypoxia stress.

| In vitro coculturing with BMSCs showed a positive effect on hypoxic cells partially by increasing autophagy
After culturing under the hypoxic condition, obvious morphological alterations of 661w cells were found, as they displayed less adhesion and more cell death than cells under the normal condition. However, when the cells were cocultured with BMSCs (coculture group), significantly more cells with normal morphology and increased cell viability (although slightly lower than that of normal cells) were observed ( Figure 4A). As shown in Figure 4B where

| BMSC migrated under retina and the transplantation attenuated photoreceptor damage in retinal detachment
RD damaged the retina, as thinning ONL thickness, shortened photoreceptor outer segments, and disordered arrangement was observed after 1 day of RD. Even 28 days after the retina was reattached, the damage was still observable. The entire retinal layer was thinned and undulated, especially the ONL, which had scattered nuclei. When immunofluorescence analysis was applied to detect the expression of rhodopsin, which reflects the photoreceptor state, detached and reattached retinas were disordered and showed weaker fluorescence compared with the normal retina ( Figure 6A).
However, these histological changes were ameliorated in the BMSC transplantation group, with a thicker ONL being observed, especially at 2 and 4 weeks after transplantation. Furthermore, photoreceptors were more regularly arranged and thicker, as revealed by immunohistochemistry, with rhodopsin expression being increased to nearly a normal level after the retina was reattached ( Figure 6B,C).
The BMSCs were marked to track their destiny after transplantation in the RD model. After transplantation, these cells could be found under the retina where they can migrate. These results were shown in Figure S2. (cells/mm 2 , P < .05) and day 7 (cells/mm 2 , P < .05) after treatment compared to those in the blank group. However, after 14 days, the number of apoptotic cells reduced rapidly, and no difference was observable between the groups ( Figure 7A).

| BMSCs inhibit retinal detachment-induced photoreceptor apoptosis
Caspase family members, which are central regulators of apoptosis, were shown to be activated in the retina in the early stage of RD. 26 Three days after treatment, the BMSC group showed significantly lower cleaved caspase-3 levels than the other two groups (P < .05), and 1 week after transplantation, this level decreased to nearly that in the normal retina ( Figure 7B). Significant differences in caspase-8 and cleaved caspase-9 levels were also observed between retinas treated with BMSCs and the other two groups (P < .05; Figure 7C). As the activity and protein levels of caspases decreased to nearly the normal level in the later stage, the three groups showed no significant differences.

| Autophagy in photoreceptors could be activated by BMSCs
To study the effect of BMSCs on autophagy in RD, the expression levels of autophagic signalling markers were examined by Western blot analysis. The expression of the microtubule-associated protein

| D ISCUSS I ON
In this study, we showed the beneficial effects of BMSCs on retinal be cytokines produced by the paracrine effect of the BMSCs. 12,30 Differentiation was also found to be another important aspect, as the transplanted BMSCs expressed several functional cell markers; however, the functions of the differentiated cells remain controversial. [35][36][37] Recently, studies have shown that BMSCs can regulate autophagy. 15,16 Autophagy, one of the important mechanisms underlying sur-  40 When prolonged and augmented by calpain or TNF-α inhibitor, autophagy could significantly reduce photoreceptor apoptosis and protect retina. 38,41 In our in vitro study, autophagy in 661w cells was induced under hypoxic conditions, which was consistent with previous researches of others, 24 and inhibiting autophagy with 3-MA resulted in more apoptotic cells.

F I G U R E 5
Coculturing with BMSCs increased autophagy in 661w cells, and the effect of BMSCs was weakened when autophagy was inhibited. (A) MDC staining showed higher number of autophagosome in hypoxic cells, as compared to normal cells, and the autophagosome was further increased in the cocultured group. Magnification: 20×. (B) In the cocultured group, the expression of LC3II and the ratio of LC3II/ LC3I were obviously increased, while the expression of p62 was decreased, compared with those in the hypoxia and normal groups. (C) 3-MA, an autophagy inhibitor strongly diminished the protective effects of BMSCs. In the autophagy-inhibited coculture (inhibited) group, the cell viability was decreased to nearly that in the hypoxia group. These assays were repeated for three times. *: P < .05, compared with the other two groups. #: P < .05, compared with cocultured cells F I G U R E 6 BMSC transplantation ameliorated cell death during retinal detachment. (A) In the detached retina, decreased ONL thickness, shortened photoreceptor outer segments and disordered arrangement were observed after 1 d of RD. Even after the retina was reattached, damage was still observable, as the whole retinal layer was thinned and undulated, especially in the ONL, and scattered nuclei were observed. Compared to normal retina, detached and reattached retinas were disordered and showed decreased expression of rhodopsin, as revealed by weak green fluorescence. Magnification: the photograph of detached retina from HE staining was 4×, while others (including inset) were 20×. (B) Histological changes were ameliorated in the BMSC-transplanted group, as a thicker ONL was observed, especially at the 2 and 4 wk after transplantation. Magnification: 20×. (C) Photoreceptors were more regularly arranged and thicker in BMSC-transplanted groups, as detected by immunohistochemistry (magnification: 40×), with rhodopsin expression being increased to nearly normal levels after the retina was reattached. These assays were repeated for three times. *: P < .05, compared with the blank and NC groups F I G U R E 7 BMSC transplantation attenuated apoptosis and activated autophagy in photoreceptors in retinal detachment. (A) TUNEL staining showed that the subretinal transplantation of BMSCs significantly reduced the number of TUNEL-positive cells in the ONL at 3 (cells/mm 2 , P < .05) and 7 d (cells/mm 2 , P < .05) after treatment compared to those in the blank and NC groups. Magnification: 40×. (B and C) At 3 d after transplantation, the treated group showed significantly lower cleaved caspase-3 levels than the other two groups (P < .05), and at 1 wk after transplantation, this level decreased to nearly that in the normal retina. Significant differences in caspase-8 and cleaved caspase-9 levels were also found between retinas treated with BMSCs and the other two groups. (D) The expression of the microtubuleassociated protein LC3-II was decreased, and p62 expression was increased at 6 and 10 d after RD compared with those at 3 d after RD, suggesting that autophagy levels peaks on day 3 after RD. Three days after treatment, significantly increased LC3-II expression and decreased p62 levels were found in the BMSC-treated group compared to those in the other two groups. However, after 7 d post-treatment, the levels did not differ significantly among the groups. These assays were repeated for three times Autophagy and apoptosis, both the activation of which are dependent on Fas, are interconnected, and autophagy in photoreceptors in RD could negatively regulate the caspase-dependent apoptosis, while inhibiting autophagy in retina could increase the activation of caspase-8. 38,40,42,43 In our study, the decrease in caspase-8 in Fas-mediated extrinsic apoptotic signalling was more obvious than caspase-9, suggesting that the apoptosis-attenuating effect of BMSCs may be exerted through inhibiting extrinsic apoptosis.
Autophagy was also found to be the mechanism by which stem cells play a protective role and improve their own reproductive potential. In damaged liver cells, MSC ameliorated liver fibrosis via activating autophagy, 16 and human placenta-derived mesenchymal stem cells also promoted hepatic regeneration by up-regulating HIF-1α and activating autophagy. 44 BMSCs also attenuated lung injury by increasing the autophagy level via the PI3K/Akt signalling pathway. 15 In this study, coculturing with BMSCs induced more intense autophagy than that in the hypoxia group. To further investigate the role of autophagy, 3-MA treatment, which inhibited autophagy, strongly diminished the protective effects of BMSCs, as determined by assessing cellular morphology, viability and apoptosis. The paracrine mechanism or exosome may involve in the autophagy induction, and this should be further studied. In vivo, autophagy was activated in the BMSC-treated group 3 days after treatment, but the autophagy level was not significantly changed after 7 days, indicating that BMSCs may promote autophagy in the early stages after transplantation in detached retinas.

CO N FLI C T O F I NTE R E S T
The authors have no competing interests to declare.

AUTH O R CO NTR I B UTI O N S
XL and JX performed the research. XL and LY analysed the data and wrote the paper. YL, YH and ZL contributed to the acquisition and analysis of data. YZ and GS designed the research study and revised the paper.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used to support the findings of this study are available from the corresponding author upon request.