The present study was conducted in accordance with the Declaration of Helsinki and the guidelines of the Ethics Committee of Shanghai Jiao Tong University Institutional Animal Care and Use Committee. All experimental protocols were approved by the Ethics Committee of Shanghai Jiao Tong University Institutional Animal Care and Use Committee. The animals were cared for in accordance with the Guide to the Care and Use of Experimental Animals (Vol. 1, 2nd edition, 1993, and Vol. 2, 1984, available from the Canadian Council on Animal Care, Constitution Square, Tower 2, Suite 315, 350 Albert Street, Ottawa, ON K1R 1B1, Canada, or on their Web site at www.ccac.ca), or the Guide for the Care and Use of Laboratory Animals (1996, published by National Academy Press, 2101 Constitution Ave. NW, Washington, DC 20055, USA).

The human normal gastric mucosal cell line GES-1 and human gastric cancer cell line AGS were all purchased from the American Type Culture Collection (ATCC) (Manassas, VA, USA) (24). Next, the cell lines were incubated in the Roswell Park Memorial Institute (RPMI)-1640 medium containing 10% fetal bovine serum (FBS) in an incubator with 5% CO₂ at 37°C.

To detect cellular ROS production, DCFH-DA (Beyotime Institute of Biotechnology, Jiangsu, China) was used. In brief, cells were incubated in FBS-free DMEM containing DCFH-DA (10 µM) for 30 min (25). Then, the cells were washed with PBS. Flow cytometric analyses were performed using a BD FACSCalibur™ flow cytometer (BD Biosciences, San Jose, CA, USA).

To measure the oxidation of cardiolipin, 10-N-nonyl acridine orange (NAO; 2 mmol/l; Molecular Probes, Eugene, OR, USA) was used. In normal cells, NAO interacts with non-oxidized cardiolipin and produces a characteristic green fluorescence (26). However, after cardiolipin is oxidized, NAO cannot bind to it. We evaluate the fluorescence intensity to reflect the cardiolipin oxidation. Image-Pro Plus 6.0; Media Cybernetics, Rockville, MD, USA) was used to obtain the mean densities of the region of interest (27), which was normalized to that of the control group.

Cells in all groups were digested and added into a 24-well Transwell plate where the chambers were then transferred in a manner that may avoid the formation of bubbles. A total of 0.5 ml cell suspension was added into the chambers and they were then incubated in an incubator for 24 h (28). Subsequently, the extra liquids in the upper and lower chambers were discarded and cells in the upper chamber above the Transwell membrane was carefully wiped off using a cotton swab. Cells transferred in the lower chambers were fixed using pre-cooled formalin for 30 min, stained using 1% crystal violet for 10 min and washed in running water until they were totally clean (29). Cells were dried naturally and observed under the microscope. Images were captured and observation results were recorded.

When cell attachment was observed at the plate bottom, a sterile spearhead was used to slightly draw a line in the well, ensuring that the widths of all scratches were the same. Images were captured and the cover of 6-well plate was marked properly to guarantee the same observation field (30). After 24 h, an Olympus inverted microscope was employed for observation of 6 fixed fields of vision.

The cells were washed in PBS and permeabilized for 10 min at 4°C in a solution of 0.1% Triton X-100 and 0.1% sodium citrate in PBS. Cells were washed 3 times in PBS and incubated overnight at 4°C with the primary antibody (30). The primary antibodies used in the present study were as follows: Tomm20 (1:1,000, #ab78547), LAMP1 (1:1,000, #ab24170), F-actin (1:1,000, #ab205) and cyt-c (1:1,000, #ab133504) (all from Abcam, Cambridge, MA, USA), Sirt1 (1:500, #2310), Mfn2 (1:1,000, #11925) (both from Cell Signaling Technology, Inc., Danvers, MA, USA).

Immunoblotting was carried out as described previously (31,32). Briefly, the cells were washed with PBS and lysed in RIPA buffer containing 1.0% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 150 mM NaCl, 50 mM Tris-HCl (pH 8.0), 10 mg/ml pepstatin A, 10 mg/ml leupeptin and 1 mM phenylmethylsulfonyl fluoride (PMSF). After centrifugation for 10 min, supernatant with 20 mg protein for each sample was subjected to immunoblotting analysis (33). Primary antibodies used for immunoblotting were as follows: pro-caspase-3 (1:1,000; #9662), cleaved caspase-3 (1:1,000; #9664; both from Cell Signaling Technology), Bad (1:2,000; #ab90435; Abcam), Bax (1:2,000; #5023; Cell Signaling Technology), caspase-9 (1:1,000; #ab32539), Bnip3 (1:1,000; #ab109362; both from Abcam), LC3II (1:1,000; #3868; Cell Signaling Technology), p62 (1:1,000; #ab56416; Abcam), Beclin1 (1:1,000; #3495), Atg5 (1:1,000; #12994), Sirt1 (1:1,000; #2310), Mfn2 (1:1,000; #11925), FAK (1:1,000; #3285), integrin β1 (1:1,000; #34971), Talin-1 (1:1,000; #4021) (all from Cell Signaling Technology).

Total RNA was isolated from cells using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). RNA (1 µg) from each sample was reversely transcribed into cDNA using a Reverse Transcription kit (Eurogentec, Fremont, CA, USA). qPCR was performed with primers and matched probes from the Universal Fluorescence-labeled Probe Library (Roche, Basel, Switzerland) (34). The primers used in the present study were as follows: FAK forward primer, 5′-GACTGGCCCAGTGTTCTTCGCTTC and reverse primer, 5′-GCTTCTGACAGAAGGAAAGCCAAC; integrin β1 forward primer, 5′-TCAAACAACTTCATGGTCCCAGTGCTC and reverse primer, 5′-TAAACCTGACTGGTTCGGGGGATTTCT; Talin-1 forward primer, 5′-CGACATGACCAAGTGGGAGCAGAAGTA and reverse primer, 5′-TGGAGCACCTTTAACCTGCTTTCCATC.

The mitochondrial potential was assessed using the probe JC-1, a sensitive fluorescent dye used to detect changes in mitochondrial potential (35). Briefly, after treatment, cells were incubated with 10 mg/ml JC-1 for 10 min at 37°C in the dark and monitored with a fluorescence microscope. Red-orange fluorescence was attributable to a potential-dependent aggregation in the mitochondria. Green fluorescence, reflecting the monomeric form of JC-1, appeared in the cytosol after mitochondrial membrane depolarization.

The level of ATP in cells was determined using the ATP Bioluminescence Assay kit. Briefly, harvested cultured cells were lysed with a lysis buffer, followed by centrifugation for 2 min at 4°C. Finally, the level of ATP was determined by mixing 50 µl of the supernatant with 50 µl of luciferase reagent (36), which catalyzed the light production from ATP and luciferin. The emitted light was linearly related to the ATP concentration and measured using a microplate luminometer.

MTT experiments were performed in 96-well plates. The cells were applied to the scaffold at a density of 10⁴ cells/well. After 2 or 3 days of culture, the samples were washed 3 times with PBS, and then 50 µl of MTT was added to each well (37). The samples were then incubated for 4 h in a humid atmosphere. The MTT solution was removed and 200 µl DMSO was added to each sample and incubated for 10 min. After the addition of Sorensen's buffer, the absorbance was determined at a wavelength of 570 nm.

A TUNEL assay to detect DNA fragmentation in cell nuclei (a marker for apoptosis in testicular tissue), was performed using an In Situ Cell Death Detection kit (Roche, Mannheim, Germany), according to the manufacturer's protocol. DAPI was used to label the nuclei. The results are expressed as apoptotic cells/tubule cross-section (38).

To evaluate the functional role of Yap, siRNA was used to reduce its expression. For Yap, the sense strand siRNA, 5′-GCGACATTCAGGGUGACUAUU-3′ and antisense strand siRNA, 3′-TCGCUGUUCCTCCCACUGAUAAU-5′ were used (39). Briefly, cultured cells were washed with Opti-Minimal Essential Medium without serum or antibiotics and seeded in 6-well plates to 30–40% confluence. The transfection reagent and siRNA were diluted separately in serum-free media, mixed and incubated for 10 min at room temperature to form the siRNA/lipid complex. This complex was then added to each well at a final concentration of 70 nM/well of siRNA. At 48 h after transfection, cells were collected to determine Yap protein expression levels by western blot analysis (40).

Data analysis was conducted using SPSS 19.0 statistical software (SPSS, Inc., Chicago, IL, USA). Measurement data are presented as mean ± standard deviation (SD). The one-way analysis of variance (ANOVA) was conducted to analyze comparisons among multi-groups. P<0.05 was assigned as indicative of statistical significance.

To determine the expression of Yap in gastric cancer in the present study, we carried out qRT-PCR and western blot analysis in normal gastric mucosal cell line GES-1 and gastric cancer cell line AGS. We found abundant Yap expression in AGS but not in the normal gastric mucosal cells (Fig. 1A and B), suggesting that Yap may be involved in the development of gastric cancer. To explain the role of Yap in gastric cancer, we stably knocked down Yap expression using the siRNA technique in AGS cells. The knockdown efficiency was confirmed by western blotting (Fig. 1A and B). Since mitophagy is vital for cancer biological function, we observed the change in mitophagy under Yap deficiency. Loss of Yap reduced mitophagy as evidenced by a lower content of mito-LC3II, p62, Beclin1 and Atg5 (Fig. 1C-G). These data indicated that Yap sustained mitophagy activity. To further provide more evidence for the role of Yap in mitophagy, we conducted the co-staining of mitochondria and lysosomes. As shown in Fig. 1H, loss of Yap reduced the overlap of mitochondria and lysosomes. However, there was no change concerning the mitochondria and lysosomes in the siRNA-control group. Altogether, these data confirmed that Yap is a vital signal for mitophagy activation.

Next, we used MTT, caspase-3 activity and wound healing assay to examine the influence of mitophagy on gastric cancer viability and migration. Meanwhile, FCCP, an activator of mitophagy, was used as the gain-of-function assay in Yap-deleted AGS cells. Compared to the control group, loss of Yap reduced cellular viability as revealed by a lower OD value and increased caspase-3 activity (Fig. 1I and J). However, re-activation of mitophagy by FCCP under Yap deficiency reversed cellular viability (Fig. 1I and J). Similarly, results from the wound healing assay also uncovered that Yap promoted cancer mobilization (Fig. 1K and L). Loss of Yap reduced the distance of cancer mobilization and this tendency was reversed by FCCP (Fig. 1K and L). Collectively, these data indicated that Yap maintained gastric cancer viability and mobilization via mitophagy.

To further investigate the mechanism by which mitophagy sustains cellular viability, we firstly detected the cellular apoptotic rate under Yap deficiency. Loss of Yap increased the number of TUNEL-positive cells (Fig. 2A and B). However, this harmful effect of Yap deficiency was reversed by re-activation of mitophagy. These data indicated that Yap preserved gastric cancer cell viability via blocking cellular apoptosis.

Since mitophagy is a self-clearing system for mitochondria; therefore, we proposed that Yap-regulated mitophagy has the ability to reduce the caspase-9-related mitochondrial apoptosis pathway. Firstly, we measured the change in proteins related to the caspase-9 apoptosis pathway. Loss of Yap increased caspase-3 and −9 and Bax expression but reduced c-IAP1 content (Fig. 2C-G). These data indicated that Yap deletion initiated caspase-9-related cellular apoptosis. However, upon reactivation of mitophagy by FCCP, the apoptotic markers were abated (Fig. 2C-G).

Apart from the downstream protein alteration, we also observed that Yap loss aggravated cytochrome c (cyt-c) leakage from mitochondria via immunofluorescence (Fig. 2H). However, this trend was reversed by mitophagy activation by FCCP (Fig. 2H). The cyt-c release was derived from mitochondrial oxidative stress. In healthy cells, mitochondrial cardiolipin retains cyt-c in the inner membrane of mitochondria. However, upon oxidation, cardiolipin may liberate cyt-c contributing to cyt-c leakage into the cytoplasm via mPTP. Therefore, we used NAO to observe cardiolipin oxidation. As shown in Fig. 2I and J, loss of Yap increased the expression of oxidized cardiolipin (CL). However, activation of mitophagy may reduce cardiolipin oxidation. Since cardiolipin oxidation is derived from excessive ROS, we aimed to ascertain whether Yap-regulated mitophagy had the ability to alleviate cellular oxidative stress. As shown in Fig. 2K and L, Yap deletion caused ROS overproduction via flow cytometric analysis. Upon re-activation of mitophagy in Yap-deleted gastric cells by FCCP, the ROS content was reduced. Notably, we used the exogenous H₂O₂ to mimic the ROS outburst in the control gastric cancer cells, which was used as the positive control group. Meanwhile, NAC, a ROS scavenger, was used to remove the oxidative stress in Yap-deleted cells, which was used as the negative control group. Exogenous H₂O₂ not only increased the cellular ROS content (Fig. 2L) but also induced cardiolipin (Fig. 2J). However, NAC had the ability to reduce oxidative stress (Fig. 2L) and (Fig. 2J) in Yap-deleted cells. Altogether, these data indicated that Yap was the housekeeper for gastric cancer cell survival via activation of mitophagy which blocked the caspase-9-related apoptotic pathways.

Apart from cellular apoptosis, cellular migration is another factor for the development of gastric cancer. Since cancer migration is related to the F-actin-mediated cytoskeletal rearrangement, F-actin may contribute to the formation of lamellipodia which directs the cancer cells to move in some direction. In the present study, we used the cell membrane dye DiO to observe the change in lamellipodium. We found that loss of Yap impaired the formation of lamellipodia (Fig. 3A and B). The shorter and less lamellipodium appeared in response to the Yap loss (Fig. 3A and B), which was accompanied by a lower number of migrated cells via Transwell assay (Fig. 3C and D). However, activation of mitophagy reversed the formation of lamellipodium and therefore facilitated cellular migration. These data indicated that Yap regulated lamellipodium-based migration via mitophagy.

The lamellipodium consists of F-actin, a kind of stress fiber (41). We found that Yap loss caused F-actin downregulation via immunofluorescence assay (Fig. 3E and F). However, activation of mitophagy reversed the F-actin expression despite deletion of Yap. To ascertain the underlying mechanism underlying F-actin collapse under Yap inhibition, we focused on the ROS-induced oxidative stress. Previous research suggests that F-actin is regulated by oxidative stress and excessive ROS may impair F-actin synthesis (42). In the present study, we found that exogenous H₂O₂ treatment reduced the F-actin immunofluorescence intensity which was similar to the results noted in the Yap-deleted cells (Fig. 3E and F). However, removal of ROS via NAC in Yap-deleted cells, restored the F-actin content, comparable to the effect of mitophagy reactivation in Yap-deleted cells.

Apart from cytoskeletal rearrangement, adhesive proteins are also implicated in cellular migration (43). The expression of the adhesive markers FAK, integrin β1 and Talin-1 were analyzed using qPCR (Fig. 3G-I) and western blot analyses (Fig. 3J-M). The results revealed that Yap inhibition reduced FAK, Integrin β1 and Talin-1 expression at the mRNA and protein levels. In contrast, mitophagy activation by FCCP exerted opposite effects in relation to Yap deletion. Altogether, these results indicatedan important role of Yap in modulating gastric cancer migration via mitophagy.

To ascertain the mechanism by which Yap sustains mitophagy activity, we focused on Mfn2 expression which is a novel receptor of mitophagy (44). We demonstrated that loss of Yap expression reduced Mfn2 expression when compared to the control group (Fig. 4A and C). Furthermore, several studies have suggested that Sirtuin 1 (SIRT1) is the upstream inducer of Mfn2 expression (45), and therefore, we aimed to ascertain whether SIRT1 is involved in Mfn2 regulation by Yap. Firstly, we found that loss of Yap was associated with the downregulation of SIRT1 (Fig. 4A and B). Furthermore, SIRT1 activator SRT1720 was used and it reversed SIRT1 expression albeit Yap deletion (Fig. 4A and B). Moreover, application of SRT1720 further restored siRNA-Yap-reduced Mfn2 expression (Fig. 4A-C). These data established the regulatory role of SIRT1 in Mfn2 expression.

To provide further evidence for the role of SIRT1 in Mfn2 expression, we co-stained for SIRT1 and Mfn2. As shown in Fig. 4D, compared to the control group, loss of Yap reduced the SIRT1 and Mfn2 expression. However, SIRT1 activator SRT1720 reversed the SIRT1 expression and subsequent upregulated the Mfn2 content. Next, to demonstrate whether the SIRT/Mfn2 pathway was responsible for mitophagy activation, we examined the proteins involved in mitophagy. As shown in Fig. 4A and E-G, compared to the Yap-silenced group, reactivation of SIRT1 reversed the mitophagy activity as evidenced by increased mito-LC3II, Atg5 and p62. Altogether, these findings demonstrated that Yap signaled mitophagy via the SIRT1/Mfn2 pathway.

Finally, to ascertain whether SIRT1 and Mfn2 are involved in gastric cancer viability, we observed mitochondrial function under SIRT1 activation. Firstly, mitochondria are the center for cellular energy metabolism. However, loss of Yap reduced ATP production (Fig. 5A). This harmful effect was reversed by SIRT1 reactivation. Since the mitochondrial membrane potential is the source of ATP production (33), we consequently assessed the change in mitochondrial potential. Compared to the control group, loss of Yap reduced the membrane potential, which was restored via SIRT1 activator (Fig. 5B and D). Furthermore, loss of Yap is associated with increased mPTP opening rate. In contrast, reactivation of SIRT1 suppressed the extensive mPTP opening (Fig. 5C). The mPTP opening provides a channel for proton leakage into the cytoplasm, finally leading to membrane potential collapse.

Apart from energy production, mitochondria also produce excessive ROS to induce cellular apoptosis (46). In the present study, we found that cellular anti-oxidant factors such as GSH and SOD were downregulated in response to the loss of Yap (Fig. 5E and G). In contrast, MDA, an end product of lipid peroxidation, was upregulated (Fig. 5F), indicative of cellular oxidative stress. However, recovery of SIRT1 exerted an opposite effect and reduced the levels of these markers of cellular oxidative stress (Fig. 5E-G). Furthermore, we also assessed caspase-3 activity to ascertain whether SIRT1/Mfn2 is involved in cellular damage. Restoration of SIRT1 via its activator reduced the caspase-3 activity when compared to the Yap deficiency group (Fig. 5H). Collectively, these data indicate that the SIRT1/Mfn2 pathway regulated by Yap is vital for mitochondrial function and cellular viability.