Strategies for imaging mitophagy in high-resolution and high-throughput

European Journal of Cell Biology xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

European Journal of Cell Biology journal homepage: www.elsevier.com/locate/ejcb

Research paper

Strategies for imaging mitophagy in high-resolution and high-throughput Deepa Indira1, Shankara Narayanan Varadarajan1, Santhik Subhasingh Lupitha1, Asha Lekshmi, Krupa Ann Mathew, Aneesh Chandrasekharan, Prakash Rajappan Pillai, Ishaque Pulikkal Kadamberi, Indu Ramachandran, Hari Sekar, ⁎ Anurup Kochucherukkan Gopalakrishnan, Santhoshkumar TR Cancer Research Program-1, Rajiv Gandhi Centre for Biotechnology, Thiruvananthapuram, Kerala, India

A R T I C L E I N F O

A B S T R A C T

Keywords: Mitophagy Autophagy Super-resolution imaging High-throughput imaging Apoptosis

The selective autophagic removal of mitochondria called mitophagy is an essential physiological signaling for clearing damaged mitochondria and thus maintains the functional integrity of mitochondria and cells. Defective mitophagy is implicated in several diseases, placing mitophagy as a target for drug development. The identification of key regulators of mitophagy as well as chemical modulators of mitophagy requires sensitive and reliable quantitative approaches. Since mitophagy is a rapidly progressing event and sub-microscopic in nature, live cell image-based detection tools with high spatial and temporal resolution is preferred over end-stage assays. We describe two approaches for measuring mitophagy in mammalian cells using stable cells expressing EGFPLC3 – Mito-DsRed to mark early phase of mitophagy and Mitochondria-EGFP – LAMP1-RFP stable cells for late events of mitophagy. Both the assays showed good spatial and temporal resolution in wide-field, confocal and super-resolution microscopy with high-throughput adaptable capability. A limited compound screening allowed us to identify a few new mitophagy inducers. Compared to the current mitophagy tools, mito-Keima or mito-QC, the assay described here determines the direct delivery of mitochondrial components to the lysosome in real time mode with accurate quantification if monoclonal cells expressing a homogenous level of both probes are established. Since the assay described here employs real-time imaging approach in a high-throughput mode, the platform can be used both for siRNA screening or compound screening to identify key regulators of mitophagy at decisive stages.

1. Introduction Mitophagy, the selective elimination of mitochondria is an important cellular process required for maintaining a pool of ATP generating functional mitochondria. Since mitophagy is selective and a spatially initiated event, live cell imaging methods have been regularly used to track and visualize these dynamic events. The maintenance of functional mitochondria is a highly regulated process that involves constant fission and fusion events; followed by marking with various adaptors that aid in the selective degradation of damaged mitochondria involving lysosomes or proteasome or by both (Ashrafi and Schwarz, 2013; Yang and Yang, 2011; Youle and Narendra, 2011). This process

also plays a critical role in the removal of mitochondria from red blood cells at the time of maturation, and also in T cell differentiation (Gottlieb and Carreira, 2010). Mitophagy received increased attention because of its role in neurodegenerative diseases, hypoxia-mediated cell survival or cell death in cancer and aging-related diseases (Band et al., 2009; Ding and Yin, 2012; Gottlieb and Carreira, 2010). Despite all these, it is difficult to conclude loss or increase in mitophagy contributes to the progression of such diseases. In various types of cancers such as breast, brain, ovarian and prostate, loss of autophagy signaling due to allelic loss of the essential autophagy gene Beclin1 (BECN1) is common (Devenish, 2007; Gong et al., 2013; Zhou et al., 2012b). Studies linking defective autophagy machinery with increased spontaneous

Abbreviations: MAP1LC3, microtubule-associated protein 1 light chain 3; LAMP1, lysosomes associated membrane protein 1; ROS, reactive oxygen species; EGCG, epigallocatechin gallate; CCCP, carbonyl cyanide 3-chlorophenylhydrazone; VDAC1, voltage-dependent anion-selective channel protein 1; PINK1, PTEN-induced putative kinase 1; OPA1, optical atrophy 1; DRP1, dynamin-related protein 1 ⁎ Corresponding author at: Cancer Research Program-1, Rajiv Gandhi Centre for Biotechnology, Poojappura, Thycaud P.O., Thiruvananthapuram, Kerala 695 014, India. E-mail addresses: [email protected] (D. Indira), [email protected] (S.N. Varadarajan), [email protected] (S. Subhasingh Lupitha), [email protected] (A. Lekshmi), [email protected] (K.A. Mathew), [email protected] (A. Chandrasekharan), [email protected] (P. Rajappan Pillai), [email protected] (I. Pulikkal Kadamberi), [email protected] (I. Ramachandran), [email protected] (H. Sekar), [email protected] (A. Kochucherukkan Gopalakrishnan), [email protected] (S. TR). 1 The authors contributed equally. http://dx.doi.org/10.1016/j.ejcb.2017.10.003 Received 13 April 2017; Received in revised form 26 October 2017; Accepted 26 October 2017 0171-9335/ © 2017 Elsevier GmbH. All rights reserved.

Please cite this article as: Indira, D., European Journal of Cell Biology (2017), http://dx.doi.org/10.1016/j.ejcb.2017.10.003

European Journal of Cell Biology xxx (xxxx) xxx–xxx

D. Indira et al.

2.2. Plasmids and generation of stable cells

tumor formation in mouse models suggest that defects in autophagy promote tumorigenesis (Gozuacik and Kimchi, 2004). Deficiency of autophagy and subsequent loss of mitophagic flux to remove damaged mitochondria with higher ROS may lead to increased ROS levels, (Lemasters, 2005) with an increase in tumorigenicity (Novak, 2012). The evidence from cancer and aging place mitophagy as an important target for drug intervention (Taylor and Goldman, 2011). Studies using multiple modes of mitophagy induction has helped to identify the sequence of events and the key regulators such as parkin, PINK1, VDAC1, p62 in mediating the process of mitophagy (Klionsky et al., 2016; Lazarou et al., 2015; Vives-Bauza et al., 2010). However, it is increasingly presumed that additional players do exist that can modulate mitophagy in mammalian cells. Owing to its clinical and biological importance, several assays were developed for the detection of mitophagy (Katayama et al., 2011b; McWilliams et al., 2016; Narendra et al., 2010; Sun et al., 2015; Vives-Bauza et al., 2010; Zhang and Ney, 2010). Since mitophagy is a highly dynamic temporally regulated process, live cell approaches with increased throughput are critical in addressing the complex mechanism of mitochondrial removal as well as to identify small molecule modulators of mitophagy. We describe an image-based approach using the most commonly used probes to visualize autophagosome formation with the inclusion of mitochondria and its fusion with lysosomes as a means of mitophagy detection. Strikingly, we noticed that mitophagy is associated with a decrease in mitochondrial mass with a concomitant increase in lysosomal mass. Hence, a quantitative measure of dual organelle intensity in image-based approach can be utilized as a parameter for quantification purposes; provided, cells with homogenous and stable expression of both fluorescent proteins are ensured. Mitochondrial engulfment by EGFP LC3 has been extensively employed as a tool for mitophagy. However, autophagy-independent aggregation of LC3 and LC3 independent mitophagic response cannot be determined by this approach. The mitochondria-targeted pH-dependent fluorescence of coral-derived protein Keima is the best approach to detect mitophagy in in vivo and in vitro conditions (Katayama et al., 2011b; Sun et al., 2015). The excitation of mito-Keima shifts from 440 nm to 586 nm while delivered to the acidic lysosomes rendering the mito-Keima a ratiometric imaging probe for mitophagy. Since the spectral properties are highly dependent on lysosomal pH, this approach is not fixation compatible. In addition, partial overlap of excitation spectrum at red and green results in orange color at lysosome making it difficult to interpret the images (Williams et al., 2017). A recently described approach, mito-QC employs tandem mCherry −GFP tag fused to the mitochondrial targeting sequence of FIS. The differential acid sensitivity of GFP and mCherry makes this dual color probe as an innovative tool for mitophagy with red and green color in steady state condition changing to red dominating color upon delivery to lysosomes because of increased acid sensitivity GFP over mCherry (McWilliams et al., 2016). mito-QC is fixation compatible and has been used to determine mitophagy in in vivo and in vitro models (McWilliams et al., 2016; Williams et al., 2017). Similar to mito-QC, the methods described here are fixation compatible and utilizes the most widely employed probes for targeting lysosome and mitochondria. In addition, it is possible to track the finer details of rapidly progressing mitophagic process in live-cell super-resolution imaging approach with the help of structured illumination microscopy technique.

pEGFP-LC3 vector was described previously (Kabeya et al., 2000). LAMP1-RFP plasmid (Addgene plasmid #1817) was described earlier and procured from Addgene (Sherer et al., 2003). Expression vector for DsRed-Mito (#6975-1) was purchased from Clontech Laboratories. Expression vector for Mito-EGFP (#558718) was purchased from BD Pharmingen™. The cells were transfected with Lipofectamine LTX (Invitrogen, #15338-100) as per the manufacturer’s protocol followed by selection in 500 μg/ml of Geneticin® (Invitrogen, #11811-031) for three to four weeks. Single cell clones stably expressing the first gene was further transfected with the second gene followed by sorting on BD FACSAriaII cell sorter (BD Biosciences) to enrich cells expressing both fluorescent proteins. The single cell clones stably expressing both the transgene were further expanded and validated. 2.3. Chemicals and reagents Drugs used for high-throughput screening are listed in the Supplementary Table 1 with their effective concentration and known mechanism(s) of action. The chemical compounds were obtained from Santa Cruz Biotechnology, Sigma-Aldrich or Calbiochem. Bafilomycin A1 (#B1793) was obtained from Sigma-Aldrich and used at a final concentration of 20 nM. The LysoTracker® Deep Red reagent was procured from Molecular probes™ and used as per standard protocol provided by the manufacturer. 2.4. Fluorescence time-lapse live cell imaging Cells grown on chambered eight well cover glass, Lab-Tek™ (Nunc, #155411), were exposed to drug-containing medium, RPMI 1640, 5% FBS. For live-imaging, cells were incubated in a live cell chamber (Tokai Hit) with optimum CO2, temperature, and humidity. Imaging was carried out using 100X Plan ApoVc oil 1.4 NA objectives on an inverted fluorescence microscope (Nikon Eclipse, Ti) (Nikon Instruments Inc.). Images were captured using RETIGA Exi camera (QImaging) at regular intervals for the indicated time periods. To curtail photo-bleaching, the intensity of light was reduced to less than 25% by intensity iris control. 2.5. High-throughput imaging with automated microscopy An inverted microscope with motorized XY stage and infrared laser based focus compensation system to maintain stable focus for a longer time was employed for high-throughput imaging applications (Nikon Eclipse Ti-PFS). The microscope was supported by 96 well plate live cell incubation chamber to maintain temperature and CO2 throughout imaging periods (Okolab). The four different XY positions were identified for each well and memorized. The stably expressing cells were seeded on 96 well glass bottom plates (Whatman, #7706-2370) and allowed to grow for 24 h. Then the wells were added with different drugs and placed on stage incubator. Images were acquired with Cool SnapHQ camera (Photometrics) using NIS Element AR 3.0 software (Nikon Instruments Inc.) at 15 min interval for 24 h or 48 h. Filter combinations used for EGFP, DsRed, and Hoechst-33342, were described previously (Joseph et al., 2011). All high-throughput imaging was carried out using 40X plan Apo 0.95 objective. Analysis of a sequence of images was performed using NIS element software or open source tools, Squassh3C and SquasshAnalyst as described previously (Rizk et al., 2015).

2. Materials and methods 2.1. Cell lines and maintenance

2.6. High-throughput imaging by BD Pathway 435™ Bioimager

The ovarian cancer cell line OVCAR-8 was procured from National Cancer Institute, USA. The cells were maintained in RPMI 1640 (Invitrogen, #72400-047) containing 10% Fetal Bovine Serum (FBS) (PAN-Biotech GmbH, #3302) in a humidified CO2 (5%) chamber at 37 °C.

For imaging in 96 wells plate format, stably expressing cells were seeded in 96 well glass bottom plates (BD Biosciences, #353219) and after 24 h, medium was removed and drug containing phenol-red free 2

European Journal of Cell Biology xxx (xxxx) xxx–xxx

D. Indira et al.

(caption on next page)

EGFP, the Pathway Bioimager was configured with EGFP Excitation filter of 438 ± 12 nm; dichroic: 458 nm LP, and emission filter 483 ± 15 nm. DsRed/RFP was imaged with excitation filter 545 ± 30 nm; dichroic 570 nm LP, and emission filter 620 ± 60 nm

RPMI 1640 (Invitrogen, #11835-030) supplemented with 5% FBS was added to the cells at appropriate concentration. Plates were imaged by employing BD Pathway 435™ Bioimager (BD Biosciences) using AttoVision™ (version1.6/435; BD Biosciences) software. For imaging of 3

European Journal of Cell Biology xxx (xxxx) xxx–xxx

D. Indira et al.

Fig. 1. Early mitophagic events can be tracked and quantified using Cells expressing MAP1LC3-EGFP and Mito-DsRed. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) (A) Ovarian cancer cell, OVCAR-8 stably expressing EGFP-LC3 and Mito-DsRed were treated with valinomycin and CCCP for 24 h. The LC3 channel, Mito-DsRed channel and merged images are shown for control and treated groups. The homogeneous level of fluorescence expression for both the probe is evident in the image. The scale bar represents 50 μm. (B) The sensor cell line was exposed to valinomycin or CCCP for 36 h. The merged microscopic image (left panel) and representative Squassh3C segmentation images (right panel) are shown to confirm the colocalization in mitophagic cells. (C) The quantified colocalization signal from different experiments is shown in the graph. Increase in colocalization signal is evident in the treated samples. (D) OVCAR-8 EGFP-LC3-Mito-DsRed cells were exposed to valinomycin or CCCP for 36 h. Mean fluorescence intensity of Mito-DsRed from each group was calculated from minimum 100 cells (n = 3). (E) The above-described cell line was exposed to valinomycin or CCCP for 36 h. Percentage of cells showing more than 20 Mito-DsRed LC3 punctae was scored for each group. Values shown are average +/− SD of three independent experiments.

by automated filter wheel combinations. Images for each well were captured with the respective filter channels using a dry 20X objective with NA 0.70. Images were captured as 2 × 2 or 3 × 3 montages. For evaluation of nuclear condensation, cells were stained with Hoechst33342 at 1 μg/ml (Invitrogen, #H3570) for 5 min before treating with drugs. Post-acquisition image analysis was done using BD software AttoVision™. Cells were segmented using polygon band for demarcating the cell nucleus by optimizing pixel dilation width, level, and erosion factor values. After complete segmentation to identify all the cells as regions of interest (ROI), the output data was analyzed as scatter plots for evaluation of the intensity of red and green fluorescence channels.

been reported as an approach to visualize mitophagy both in live cellbased methods using MAP1LC3 fluorescent fusion probes and in fixed cells by the immuno-localization approach. Even though the live cell approach using genetically encoded MAP1LC3 fluorescent fusion probes is sensitive to mark the early phase of mitophagy; a highthroughput live cell approach for screening mitophagy modulators is yet to be described. A major challenge in converting this method for quantitative mitophagy is a heterogeneous expression of genetically encoded probes. To circumvent this, after transfection of OVCAR 8 cells with MAP1LC3-EGFP and Mito-DsRed plasmids, we have employed repeated flow cytometry sorting, multiple clone expansion, and clonal selection to get cell population with a homogeneous and stable expression of both the probes (Fig. 1A). To validate the cell for mitophagy reporting, valinomycin, an ionophore that depolarize mitochondria and mitochondrial uncoupler CCCP (Carbonyl cyanide m-chlorophenylhydrazone) were employed. As shown in Fig. 1B and C, automated colocalization analysis using Squassh3C/SquasshAnalyst indicated increased colocalization signal in both CCCP and valinomycin treated samples. Mean fluorescent intensity of Mito-DsRed as analyzed using NIS element also showed a significant decrease (Fig. 1D). Also, the percentage of cells with more than 20 MAP1LC3 labeled mitochondrial clusters also revealed the consistent increase in treated cells. The results suggest that stable cells with homogeneous expression of both MAP1LC3 and Mito-DsRed serve as a reliable tool for quantitative mitophagy. We failed to get meaningful and consistent quantitative data in any of these quantitative analyses when pooled colonies with differing expressing levels of MAP1LC3 and Mito-DsRed were employed (Data not shown).

2.7. Laser scanning confocal imaging Cells were grown on chambered cover glass, Lab-Tek™ (Nunc, #155411). The cells were imaged on Nikon A1R spectral confocal microscope using 60X oil 1.4 NA objective. For real-time imaging of mitophagy, cells were placed in a stage incubation chamber (Tokai Hit) to maintain CO2 and temperature. Multiwell time-lapse imaging was carried out using an automated XY stage with focus drift compensation unit, PFS from Nikon using 60× WI objective. 2.8. Super-resolution live cell imaging (SIM) Structured illumination imaging was carried out using commercial TIRF- N-SIM microscope from Nikon (Tokyo, Japan). Cells grown on chambered coverglass were imaged using 100 × 1.49 NA oil immersion objectives using 488 nm laser for GFP and 562 nm laser for Lamp RFP. Images were captured using an EMCCD Camera iXon 897 (Andor, USA). Super-resolution images were reconstructed using NIS element software as per the standard protocol. For time-lapse imaging using N-SIM, cells were maintained at 37 °C and 5% CO2 with an on-stage incubator from Tokai Hit (Japan). To reduce photobleaching during SIM image acquisition, laser power was reduced to < 20% with a minimum exposure time of 100 ms, such that both the channels can be acquired within 200 ms.

3.2. Dynamics of mitophagy by laser scanning confocal imaging High-resolution, real-time imaging was carried out using laser confocal microscope to understand the spatiotemporal regulation of mitophagy. To track the fusion of mitochondria and lysosomes, the lysosomes were also stained with Lysotracker deep red. Since initial studies showed that significant mitophagic events start only after 12 h of CCCP or valinomycin treatment, real-time imaging was initiated from 12 h to reduce unnecessary bleaching. In Fig. 2A and B, distinct MAP1LC3-EGFP cluster formation and engulfment of mitochondrial fragments are visible in the confocal image in valinomycin, and CCCP treated samples. To capture the whole 24 h dynamics of mitophagic events, an imaging interval of 15 min was used (Supplementary Videos 1 and 2). Lysotracker deep red signal even though excited with 633 nm laser line, showed significant photobleaching compared to EGFP and DsRed signal. A few MAP1LC3-EGFP clusters are more active in engulfing mitochondrial fragments and marking its degradation. However, a few of them retained the mitochondria for a longer duration even within the lysosomal clusters. Some of the MAP1LC3-EGFP clusters associated with lysosomes actively fuse with more and more small clusters leading to the creation of larger mitophagic structures (Supplementary Video 1). There was a significant temporal variation in the degradation of mitochondria between autophagolysosomes within a cell. Most of the mitochondrial fragments fuse with lysosomes pre-fused with MAP1LC3; however, a few of them were fused with MAP1LC3

2.9. Western blotting The whole cell extract prepared from the cells were resolved on SDS-PAGE and transferred to a nitrocellulose membrane as per standard protocol. The membrane after blocking with 3% BSA, probed with primary antibody against LAMP1 (#sc-20011), DRP-1 (#sc-101270), OPA1 (#BD-612607), VDAC-1 (#CST-4866s) and p62 (#CST-5114s) followed by respective HRP conjugated secondary antibodies (Santacruz). The bands were developed using ECL reagent. β-actin (#sc47778) served as loading control. 3. Results 3.1. Stable cells expressing MAP1LC3-EGFP and Mito-DsRed to mark early phase of mitophagy Mitochondrial accumulation of the autophagy marker MAP1LC3 has 4

European Journal of Cell Biology xxx (xxxx) xxx–xxx

D. Indira et al.

Fig. 2. Real-time confocal imaging reveals the Dynamics of mitophagy with temporal changes in mean fluorescence intensity of Mito DsRed. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) (A) Ovarian cancer cell, OVCAR-8 stably expressing EGFP-LC3 and Mito-DsRed were stained with Lysotracker deep red. The cells were exposed to valinomycin. Real-time confocal imaging was started after 12 h of valinomycin exposure. Representative snapshots from indicated time points of imaging are shown. The arrow-heads points to areas of colocalization. (B) OVCAR-8 cells stably expressing EGFP-LC3 and Mito-DsRed were stained with Lysotracker deep red. The cells were exposed to CCCP. Real-time confocal imaging was started after 12 h of CCCP treatment. Representative snapshots from indicated time points of imaging are shown. The relative temporal changes of mean fluorescence intensity of three channels of the selected cell are also shown.

tracked in live cells using lysosomal staining. However, as shown in Supplementary Videos 1, 2, and Supplementary Fig. 1, photobleaching of Lysotracker deep red rendered it inferior for long time-lapse application. Moreover, most of the currently available organelle dyes are relatively toxic to live cells than genetically encoded probes. Hence, to detect and track the late or final phase of mitophagy, cells were stably expressed with lysosomal membrane protein LAMP1-Red Fluorescent Protein (LAMP1-RFP) followed by the stable introduction of mitochondrial-EGFP. To validate the cell lines for monitoring mitophagy, the reporter cell line was treated with valinomycin to visualize mitophagy events at fixed time points as well as in time-lapse imaging. At 24 h of treatment with valinomycin, a significant number of cells showed co-localization of mitochondria and lysosomes with a decrease in green fluorescence intensity (Fig. 3). Treatment of bafilomycin along with valinomycin further increased the colocalization. Confocal imaging was carried out after staining the cells with Lysotracker deep red. In Fig. 4, the rapid mitochondrial fragmentation and increase in

initially and subsequently fused with lysosomes (Supplementary Videos 1 and 2). Consistent with previous imaging, confocal imaging also revealed a time-dependent decrease in mitochondrial fluorescence in both CCCP and valinomycin treated samples. Interestingly a biphasic change in MAP1LC3 was observed in CCCP treated cells with the initial gradual increase and later sudden decrease just before cell death (Fig. 2B). 3.3. Cells stably expressing Mito-EGFP and LAMP1- RFP to mark the late phase of mitophagy The above-described method proved suitable for real-time visualization of early events of mitophagy. However, co-localization index or percentage of cells with mitochondrial-MAP1LC3 clusters might also lead to erroneous data because of autophagy independent LC3 aggregation and LC3 independent mitophagy. Damaged mitochondria once marked for degradation is isolated within autophagosomes that finally fuse with lysosomes for its degradation. This event also could be 5

European Journal of Cell Biology xxx (xxxx) xxx–xxx

D. Indira et al.

Fig. 3. Confocal imaging of cells expressing Mito-EGFP and LAMP1-RFP revealed the diverse stages of late mitophagic events. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) OVCAR-8 cells stably expressing Mito–EGFP and LAMP1-RFP were treated with valinomycin for 24 h or 21 h valinomycin followed by 3 h with bafilomycin or bafilomycin alone for 3 h. The representative images of Mitochondria EGFP and Lamp1 RFP are shown. The lysosomes with engulfed mitochondrial EGFP are shown with arrows. The increased lysosomal accumulation of mitochondrial EGFP is evident in Bafilomycin pretreated mitophagic cells. The inset shows a portion of the image zoomed to visualize mitophagy.

6

European Journal of Cell Biology xxx (xxxx) xxx–xxx

D. Indira et al.

Fig. 4. Increase in Lamp1-RFP and decrease in Mito-EGFP marks cells with mitophagy in time-lapse imaging. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) (A) OVCAR-8 cells stably expressing Mito-EGFP and LAMP1-RFP was stained with Lysotracker deep red and exposed to CCCP. Confocal imaging started immediately after CCCP treatment. Representative snapshots from indicated time points of imaging are shown. (B) Increase in intensity of Lamp-1 RFP is evident with time; however, photo-bleaching reduced the Lysotracker Red signal.

7

European Journal of Cell Biology xxx (xxxx) xxx–xxx

D. Indira et al.

Fig. 5. Real-time monitoring of mitophagic events in High-resolution using structured illumination microscopy. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) OVCAR-8 cells stably expressing Mito–EGFP and LAMP1-RFP were treated with CCCP for 12 h. Subsequently, the cells were imaged using Nikon super-resolution microscope (N-SIM) for additional 12 h maintaining the temperature at 37 °C and CO2 at 5%. Representative snapshots from indicated time points of imaging are shown. The decrease in EGFP fluorescence and its engulfment within lysosomes are seen in the image. Arrows indicate the lysosome with engulfed mitochondria.

valinomycin treatment. The later events showing prolonged retention of mitochondrial fluorescence within lysosomes is shown in Supplementary Video 3. An imaging of interval of two minutes could readily identify engulfed mitochondria within lysosomes with sufficient time resolution. As previously noticed, there is a rapid loss of green

lysosomal volume are evident within 6 h of CCCP treatment. However, lysotracker deep red intensity decreased because of photobleaching (Fig. 4B). Further to visualize the later stages of mitophagy and to analyze the retention time of mitochondria within lysosomes, one-hour imaging with an interval of two minutes was carried out after 24 h of

8

European Journal of Cell Biology xxx (xxxx) xxx–xxx

D. Indira et al.

fluorescence once mitochondrial fragments are inside the lysosomes (Zhou et al., 2012a). However, live cell imaging with valinomycin also supported that occasionally a few lysosomes retains mitochondrial green fluorescence for a prolonged time allowing for their detection even in steady state level analysis.

exposed to mitophagy inducers. The results described above suggest that at least for drug-induced mitophagy, the parameter of mitochondrial or lysosomal mass may form a quantifiable output of mitophagy if the stable homogeneous expression of both the transgene is established. We have further imaged the cells using a high-throughput BD Pathway 435™ Bioimager (BD Biosciences) followed by automated segmentation as described. The output information of red and green fluorescence yielded robust data for the quantitative analysis of mitophagy. A scatter plot of mitochondrial EGFP intensity against the lysosomal red intensity of segmented cells treated with a different drug is shown in Fig. 8A and B. High-throughput imaging also enabled time-dependent quantitative analysis of mitophagy in the same well with repeated imaging based on mitochondrial to lysosomal intensity (Supplementary Fig. 2).

3.4. High-resolution imaging of mitophagy using structured illumination microscopy reveals real-time capabilities and temporal resolution Confocal and widefield microscopy, revealed the generation of smaller unresolvable mitochondrial fragments and its fusion with MAP1LC3 vacuoles and fusion with lysosomes during mitophagy. To track such events with better resolution in real-time, we employed structured illumination microscopy (N-SIM). With the imaging condition described, the structured illumination imaging with an interval of 5 min revealed the temporal changes of mitophagy in valinomycin treated samples (Fig. 5 and Supplementary Video 3).

3.6. Simultaneous visualization of cell death and mitophagy reveals increased mitophagy in surviving cells In general, mitophagy is considered as a survival signaling despite their co-existence with other forms of cell death like apoptosis. Live cell-based methods for simultaneous detection of mitophagy and apoptosis are essential to discriminate these possibilities. Earlier we have developed a successful high-throughput protocol for caspase activation employing cells stably expressing ECFP-EYFP linked with DEVD amino acid sequence in between (Achuthan et al., 2011; Joseph et al., 2011). The results from the study indicated a high correlation with nuclear chromatin condensation and caspase activity. Consequently, we have analyzed whether the system described here for autophagy and mitophagy can also be adapted for simultaneous live cell visualization of cell death and mitophagy. The cells expressing MitoEGFP and LAMP1-RFP were stained with Hoechst 33342 as described before treating with the drug. Both fixed time single imaging, as well as time-lapse imaging, substantiated the potential of this approach in visualizing mitophagy and the hallmark of apoptosis, condensation of chromatin. Analysis of a large sequence of events from the high-throughput imaging data displays an increased mitophagy in cells surviving from apoptosis in later hours of drug treatment (Supplementary Fig. 3). This indicates that mitophagy even if associated with apoptosis, its role in conferring survival advantages to surviving fraction cannot be ruled out. Another possibility is increase in mitophagy may contribute for apoptosis resistance. Further studies using these tools will unravel such unresolved questions in cell biology.

3.5. High-throughput, real-time automated microscopic imaging and compound screening The results shown above indicate that both the cell lines developed are excellent tools to analyze two distinct phases of mitophagy because of relatively stable expression of the transgene. Since autophagy and mitophagy are a highly dynamic process involving multiple steps from initiation to execution, end-stage assays even if carried out in live cells cannot always detect mitophagic events at fixed time imaging. So, we decided to adopt an automated microscopy aided live cell imaging in a multi-well format with multi-point time-lapse imaging in the confocal and non-confocal mode as described. The small molecules used in the current study are listed in Supplementary Table 1. After image acquisition, it is hard to maintain the same segmentation features and Region of Interest (ROI) for a sequence of events because of spatial displacement of cells over time. Therefore, ROIs were drawn on individual images and analyzed for co-localization or relative fluorescence intensity. For EGFP-MAP1LC3-Mito-DsRed cells, ROI was drawn on MAP1LC3 channel to mark the LC3 punctae and analyzed the co-localization index with Mito-DsRed using NIS element software. The untreated cells showed a correlation of coefficient (Pearson coefficient) of less than 0.42 in most of the cells. The cells with an average of 0.6 or more Pearson coefficient values were scored as mitophagic. An initial analysis of a large number of time-lapse sequences helped us to identify seven positive hits from 50 diverse chemical compounds employed for screening. These include benzisoxazole, procaspase activating compound 1, pro-caspase activating compound 2, MIRA-1, EGCG, Gemcitabine and chrysin apart from the known mitophagy inducers like starvation, valinomycin, and CCCP. Representative confocal images from the multiwell time-lapse sequences for the selected drugs are shown in Fig. 6 with the colocalization signal (Fig. 6B). A representative time-lapse imaging shows the dynamics of mitophagy from 12 h of PAC1 treatment (Supplementary Video 4). Even though we were unable to perform automated analysis using the sequence of images, the tool is highly suitable to visualize initiation and execution phase of mitophagy in real-time mode with sufficient temporal resolution. The second method employed, a cell expressing Mito-EGFP and LAMP1-RFP was also adaptable for real-time imaging. Consistent with the above-described method, most of the positive hits identified from the first screen also emerged as the positive hits in this approach. A representative image of PAC1 treated cells with and without bafilomycin is shown in Fig. 7A. MIRA1 treated images are shown in 7B. Since we have observed a dramatic change in mitochondria to lysosomal mass in cells treated with the positive hits, critical mitochondrial and lysosomal structural proteins were analyzed by western blot. As shown in Fig. 7C, most of the leads increased endogenous LAMP1 substantiating the imaging results. Interestingly a consistent decrease in mitochondrial structural protein VDAC1, DRP1, and OPA1 were noticed in cells

4. Discussion Several methods have been described to detect the events of mitophagy in cells and a few for quantitative mitophagy applications. Most of the quantitative methods utilize mitochondrial and lysosomal stains such as MitoTracker and LysoTracker dyes with differing spectral characteristics to visualize lysosomal delivery of mitochondrial components (Dolman et al., 2013; Mauro-Lizcano et al., 2015). Dye-based approaches, though easy to adapt in cultured cells, their non-specific loading to other organelles, physicochemical state dependent loading, cell toxicity are serious concerns. We have observed that MitoTracker and lysosomal dyes fail to accumulate on cells with oxidized mitochondria (Unpublished). Consistent with this a recent report by Padman et al. showed redistribution of mitochondrial dyes (DiOC6, TMRM, MTR, and MTG) from mitochondria to lysosomes after CCCP treatment (Padman et al., 2013). Genetically encoded fluorescent protein based approaches are non-toxic methods and also allows to detect mitophagy in live cells with sufficient temporal resolution to track the dynamics of mitophagy (Chen et al., 2017). MAP1LC3 based probes and its localization with mitochondria have been extensively employed for detecting autophagy events. An innovative approach was described for mitophagy, exploiting the pH-dependent differential emission of coralderived fluorescent protein Keima targeted at mitochondria. 9

European Journal of Cell Biology xxx (xxxx) xxx–xxx

D. Indira et al.

Fig. 6. Automated microscopic imaging from representative treated wells showing the dynamics of mitophagy. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) (A) OVCAR-8 cells stably expressing EGFP-LC3 and MitoDsRed were treated with indicated drugs and imaged using an automated microscope in live-cell incubation chamber using confocal scanning from 12 h onwards. Representative snapshots from indicated time points of imaging are shown. Compared to control progressive loss of Mito-DsRed and increased colocalization with EGFP- LC3 is evident in the images. (B) Squassh3C/SquasshAnalyst analysis results of colocalization signal for the treated cells is shown in the graph. A significant increase in colocalization signal is evident in cells exposed to the indicated lead compounds compared to control.

Mitochondrial keima is a dual excitation fluorophore that shifts its excitation from 440 nm to 586 nm from neutral to the acidic environment so that delivery to lysosomes can be tracked by dual excitation imaging (Katayama et al., 2011a). A quantitative high-throughput

mitophagy detection using this approach is yet to be described. Other popular methods for autophagy include stable cell lines expressing green fluorescent protein-microtubule-associated protein 1 light chain 3B (GFP-LC3) vesicles or luciferase-based assays (Bampton et al., 2005; 10

European Journal of Cell Biology xxx (xxxx) xxx–xxx

D. Indira et al.

Fig. 7. Increase in LAMP-1 and decrease in mitochondrial proteins are evident in cells treated with lead mitophagy inducing compounds. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) (A) OVCAR-8 cells stably expressing Mito–EGFP and LAMP1-RFP were treated with PAC-1 for 24 h or 21 h PAC-1 followed by 3 h with bafilomycin. The representative images of Mitochondria EGFP and Lamp RFP are shown. Arrows marks lysosomes with engulfed mitochondria. (B) The above-described cells were treated with MIRA-1 for 24 h. Representative image of mitochondria and lysosomes are shown with pixel intensity map. The arrow defines the area from which the pixel intensity map is acquired. (C) OVCAR-8 stably expressing Mito–EGFP and LAMP1-RFP were treated with indicated drugs for 48 h. The whole cell extract prepared was used for SDS PAGE electrophoresis and then transferred to nitrocellulose paper by western blotting. Probed with indicated antibodies. β-actin served as loading control. The decrease in DRP1, OPA1, and increase in LAMP1 is evident in cells treated with mitophagy inducers.

(Shen et al., 2011). A few of these assays carried out in a highthroughput mode enabled identification of novel autophagy inducers. Similarly, the mito-Timer, a fluorescent probe that changes its color depending on the age of mitochondria proved as an excellent tool for

Eng et al., 2010; Farkas et al., 2009; Kimura et al., 2007; Mizushima and Kuma, 2008; Zhou et al., 2012a). A successful adaptation of autophagy assay by image-based high-throughput method enabled hierarchal clustering of currently used anticancer drugs into distinct groups 11

European Journal of Cell Biology xxx (xxxx) xxx–xxx

D. Indira et al.

Fig. 8. OVCAR-8 cells stably expressing Mito–EGFP and LAMP1-RFP is High-throughput adaptable for compound screening based on red and green signal intensity. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) OVCAR-8 expressing Mito–EGFP and LAMP1-RFP were treated with valinomycin for 48 h and imaged in high-throughput mode as described. The merged images of Mito-EGFP and LAMP1-RFP channels for control well and valinomycin treated wells are shown (A). The corresponding scatter plot for red and green fluorescence intensity of the entire well after automated segmentation and analysis is shown below (B).

and software-based co-localization analysis after proper segmentation allow quantification of the mitophagy process. One of the major advantages of this system is its ability to monitor mitophagy with high temporal resolution. The mitophagic flux can be determined in live mode even with a large number of compounds in a single experiment either in wide-field or confocal imaging as described here. Also, live cell staining of chromatin using Hoechst 33342 also enables to distinguish apoptotic and mitophagic cells in a population of treated cells. In the second method, we have used stable cells expressing the lysosomal membrane protein LAMP1-RFP and Mitochondria with EGFP. Earlier, lysosomes and EGFP-LC3 combinations were employed to detect autophagy in live cells that indicated that in the highly acidic environment the engulfed EGFP quickly loses its fluorescence (Kimura et al., 2007). Since the approach is real-time compatible, even steadystate imaging also can visualize snapshots of mitophagy. Quite interestingly we have noticed that occasionally the mitochondria engulfed within lysosomes retain fluorescence for a longer time. This is consistent with the previous study that indicated that MAP1LC3–EGFP punctae retain fluorescence signal for sufficient time within lysosomes (Bampton et al., 2005). Even with a limited number of compound screening, we have observed highly heterogenic degradation pattern of Mito-EGFP signal within lysosomes. A vital distinction of the assay is the ability to image multiple XY positions within a single well or multiple wells that will allow visualizing the rare and spatially distinct events of mitophagy. Interestingly we have observed that increased mitophagy alters the ratio of cellular lysosomal to mitochondrial mass. This gradual change in lysosomal and mitochondrial mass helped us to render the assay quantitative after segmentation in any of the channels

studying mitochondrial biogenesis both in vitro and in vivo (Ferree et al., 2013; Hernandez et al., 2013). However, it is yet to be explored as a tool to study dynamics of mitophagy. The selective elimination of damaged mitochondria by autophagosome called mitophagy primarily act as a regulated mechanism of maintaining the functional integrity of mitochondria. Even though their associations with cytotoxic cell death were reported, its primary role as a cell death inducer in the absence of other forms of cell death such as apoptosis or necrosis is yet to be determined. The recent findings indicate that upon mitochondrial damage, mitochondria-localized PINK1 recruits E3-ligase parkin to mitochondria marking it for degradation (Matsuda and Tanaka, 2010; Shiba-Fukushima et al., 2012; Vives-Bauza et al., 2010). This event is followed by ubiquitination of mitofusin by parkin leading to its proteasome-dependent degradation (Gegg et al., 2010). Mitofusin degradation blocks fusion events and mitophagy is activated. Based on this signaling cascade involved in mitophagy, multiple methods were described for mitophagy detection as a steady state approach either by immunofluorescence or by live cells expressing mitophagy related proteins (Larsen et al., 2010; Zhu et al., 2011). However, it has been increasingly shown that the process of mitophagy is a complex and highly dynamic process that progresses rapidly once initiated. Thus, it is challenging to detect the process of mitophagy in the end-stage methods. In the present study, we have employed two separate live cell approaches to specifically detect mitophagy in cancer cells. The first approach using stable cells expressing both EGFP-LC3 and Mito-DsRed is sensitive for image-based detection of early changes of mitophagy, marking of mitochondria with LC3 aggregates. Conventional imaging 12

European Journal of Cell Biology xxx (xxxx) xxx–xxx

D. Indira et al.

Authors’ contributions

in automated mode. Consistent with this, loss of mitochondrial mass was reported previously in reticulocyte mitophagy (Zhang and Ney, 2010). We have successfully identified several mitophagy inducers using this image-based approach. A few of the compounds emerged as the positive hits such as valinomycin, CCCP were previously reported to induce mitophagy in cancer cells. We have identified PAC-1, PAC-2, Benzisoxazole, EGCG and MIRA-1 as the new mitophagy inducers using both the screening approaches described. PAC-1 was initially identified as the first direct caspase activator subsequently shown to have potent antitumor activity in in vitro and in vivo animal models (Peterson et al., 2009, 2010). We have recently shown that PAC-1 induced apoptosis involves Bax and Bak-independent mitochondrial permeabilization (Seervi et al., 2011). MIRA-1 is a known p53 activator that reactivates the mutant p53. Benzisoxazole was identified as a Hsp90 Inhibitor that binds the N-terminal domain of Hsp90, efficiently displacing geldanamycin from the ATP binding site (Gopalsamy et al., 2008). In general, the positive hits identified show various structural and functional properties. Staining the cells with Hoechst also enables trouble-free segmentation as well as to discriminate apoptotic and mitophagic cells. Interestingly we have observed increased mitophagy in cells surviving from apoptosis after treatment with Benzisoxazole and EGCG indicating that mitophagy plays a role in allowing cancer cell survival from cytotoxic cell death. Continuous monitoring of surviving fractions revealed the slow progression of mitophagy process with significant loss of mitochondrial mass concomitant with an increase in lysosomal mass. Further studies using the model described here may unravel the key players that maintain the mitophagy process in surviving cells and the fate of the cells. Overall, the methods described here for mitophagy detection in live cell mode easily discriminate apoptotic, autophagic and mitophagic cells with sufficient temporal resolution. We have imaged the cells for up to 48 h at an interval of 15 min after drug treatment without any inhibition of cell proliferation. For the first time, we show that the emerging structured illumination microscopy can provide both improved spatial resolution to resolve finer features of mitophagy and sufficient temporal resolution to track the whole process of mitophagy within a single cell. The results described here suggest that since the process of autophagy and mitophagy are a highly dynamic process, real-time imagebased approaches and ability to track fine details with ultra-high resolution in live cells provide more robust data than the steady state level detection methods. This assay system also presents an opportunity to monitor both mitophagy and cell death by apoptosis with good temporal resolution. Further studies using these tools may unravel whether mitophagy signaling is strictly a survival mechanism or it also contributes to cell death without initiation of apoptosis signaling. A disadvantage of the approach is the requirement of the stable introduction of two separate probes and difficulty in generating monoclonal cells expressing both the probes. However, once stable monoclonal cells are generated, accurate quantification and analysis of longtime dynamics of mitophagy are possible using diverse imaging platforms including high-throughput screening approaches.

DI, SSL, SNV, and TRS contributed to the designing and execution of experiments and preparation of the manuscript. KAM, AC, and PRP developed the stable cell lines, AL and IPK carried out the western blot experiments, IR, HS, and AKG performed confocal imaging and superresolution imaging and analysis of the data sets. Conflict of interest The authors have no conflict of interest to declare. Acknowledgments We thank Prof. Noboru Mizushima for the expression vector pEGFPLC3. This work was supported by the Department of Biotechnology, Govt. of India (BT/PR17786/MED/32/541/2016). DI and SSL are funded by Junior Research Fellowship from Council of Scientific and Industrial Research, Govt. of India. KAM and AL are funded by Junior Research Fellowship from Indian Council of Medical Research, Govt. of India. SNV is funded by Junior Research Fellowship from University Grants Commission, Govt. of India. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ejcb.2017.10.003. References Achuthan, S., Santhoshkumar, T.R., Prabhakar, J., Nair, S.A., Pillai, M.R., 2011. Druginduced senescence generates chemoresistant stemlike cells with low reactive oxygen species. J. Biol. Chem. 286, 37813–37829. Ashrafi, G., Schwarz, T., 2013. The pathways of mitophagy for quality control and clearance of mitochondria. Cell Death Differ. 20, 31–42. Bampton, E.T., Goemans, C.G., Niranjan, D., Mizushima, N., Tolkovsky, A.M., 2005. The dynamics of autophagy visualised in live cells: from autophagosome formation to fusion with endo/lysosomes. Autophagy 1, 23–36. Band, M., Joel, A., Hernandez, A., Avivi, A., 2009. Hypoxia-induced BNIP3 expression and mitophagy: in vivo comparison of the rat and the hypoxia-tolerant mole rat, Spalax ehrenbergi. FASEB J. 23, 2327–2335. Chen, L., Ma, K., Han, J., Chen, Q., Zhu, Y., 2017. Monitoring mitophagy in mammalian cells. Methods Enzymol. 588, 187–208. Devenish, R.J., 2007. Mitophagy: growing in intricacy. Autophagy 3, 293–294. Ding, W.-X., Yin, X.-M., 2012. Mitophagy: Mechanisms, Pathophysiological Roles, and Analysis. Dolman, N.J., Chambers, K.M., Mandavilli, B., Batchelor, R.H., Janes, M.S., 2013. Tools and techniques to measure mitophagy using fluorescence microscopy. Autophagy 9, 1653–1662. Eng, K.E., Panas, M.D., Hedestam, G.B.K., McInerney, G.M., 2010. A novel quantitative flow cytometry-based assay for autophagy. Autophagy 6, 634–641. Farkas, T., Høyer-Hansen, M., Jäättelä, M., 2009. Identification of novel autophagy regulators by a luciferase-based assay for the kinetics of autophagic flux. Autophagy 5, 1018–1025. Ferree, A.W., Trudeau, K., Zik, E., Benador, I.Y., Twig, G., Gottlieb, R.A., Shirihai, O.S., 2013. MitoTimer probe reveals the impact of autophagy, fusion, and motility on subcellular distribution of young and old mitochondrial protein and on relative mitochondrial protein age. Autophagy 9, 1887–1896. Gegg, M.E., Cooper, J.M., Chau, K.-Y., Rojo, M., Schapira, A.H., Taanman, J.-W., 2010. Mitofusin 1 and mitofusin 2 are ubiquitinated in a PINK1/parkin-dependent manner upon induction of mitophagy. Hum. Mol. Genet. 19, 4861–4870. Gong, C., Bauvy, C., Tonelli, G., Yue, W., Delomenie, C., Nicolas, V., Zhu, Y., Domergue, V., Marin-Esteban, V., Tharinger, H., 2013. Beclin 1 and autophagy are required for the tumorigenicity of breast cancer stem-like/progenitor cells. Oncogene 32, 2261–2272. Gopalsamy, A., Shi, M., Golas, J., Vogan, E., Jacob, J., Johnson, M., Lee, F., Nilakantan, R., Petersen, R., Svenson, K., 2008. Discovery of benzisoxazoles as potent inhibitors of chaperone heat shock protein 90. J. Med. Chem. 51, 373–375. Gottlieb, R.A., Carreira, R.S., 2010. Autophagy in health and disease. 5. Mitophagy as a way of life. Am. J. Physiol.-Cell Physiol. 299, C203–C210. Gozuacik, D., Kimchi, A., 2004. Autophagy as a cell death and tumor suppressor mechanism. Oncogene 23, 2891–2906. Hernandez, G., Thornton, C., Stotland, A., Lui, D., Sin, J., Ramil, J., Magee, N., Andres, A., Quarato, G., Carreira, R.S., Sayen, M.R., Wolkowicz, R., Gottlieb, R.A., 2013. MitoTimer: a novel tool for monitoring mitochondrial turnover. Autophagy 9, 1852–1861.

5. Conclusions Cancer cells stably expressing a homogenous level of genetically encoded fluorescent probes of mitochondria, lysosomes and autophagosome marker MAP1LC3 were developed as a sensor for quantitative mitophagy. A successful image-based High-throughput approach for mitophagy using these sensors helped to identify several new mitophagy inducers. The approach revealed the spatiotemporal dynamics of mitophagy from beginning to end at single cell level in real-time mode employing diverse microscopic techniques including confocal imaging and structured illumination based super-resolution imaging. 13

European Journal of Cell Biology xxx (xxxx) xxx–xxx

D. Indira et al.

relationship of procaspase-activating compound 1 (PAC-1) and its cellular co-localization with caspase-3. J. Med. Chem. 52, 5721–5731. Peterson, Q.P., Hsu, D.C., Novotny, C.J., West, D.C., Kim, D., Schmit, J.M., Dirikolu, L., Hergenrother, P.J., Fan, T.M., 2010. Discovery and canine preclinical assessment of a nontoxic procaspase-3–activating compound. Cancer Res. 70, 7232–7241. Rizk, A., Mansouri, M., Ballmer-Hofer, K., Berger, P., 2015. Subcellular object quantification with Squassh3C and SquasshAnalyst. Biotechniques 59, 309–312. Seervi, M., Joseph, J., Sobhan, P., Bhavya, B., Santhoshkumar, T., 2011. Essential requirement of cytochrome c release for caspase activation by procaspase-activating compound defined by cellular models. Cell. Death. Dis. 2, e207. Shen, S., Kepp, O., Michaud, M., Martins, I., Minoux, H., Metivier, D., Maiuri, M., Kroemer, R., Kroemer, G., 2011. Association and dissociation of autophagy, apoptosis and necrosis by systematic chemical study. Oncogene 30, 4544–4556. Sherer, N.M., Lehmann, M.J., Jimenez-Soto, L.F., Ingmundson, A., Horner, S.M., Cicchetti, G., Allen, P.G., Pypaert, M., Cunningham, J.M., Mothes, W., 2003. Visualization of retroviral replication in living cells reveals budding into multivesicular bodies. Traffic 4, 785–801. Shiba-Fukushima, K., Imai, Y., Yoshida, S., Ishihama, Y., Kanao, T., Sato, S., Hattori, N., 2012. PINK1-mediated phosphorylation of the Parkin ubiquitin-like domain primes mitochondrial translocation of Parkin and regulates mitophagy. Sci. Rep. 2. Sun, N., Yun, J., Liu, J., Malide, D., Liu, C., Rovira, I.I., Holmstrom, K.M., Fergusson, M.M., Yoo, Y.H., Combs, C.A., Finkel, T., 2015. Measuring in vivo mitophagy. Mol. Cell 60, 685–696. Taylor, R., Goldman, S.J., 2011. Mitophagy and disease: new avenues for pharmacological intervention. Curr. Pharm. Des. 17, 2056–2073. Vives-Bauza, C., Zhou, C., Huang, Y., Cui, M., de Vries, R.L., Kim, J., May, J., Tocilescu, M.A., Liu, W., Ko, H.S., 2010. PINK1-dependent recruitment of Parkin to mitochondria in mitophagy. Proc. Natl. Acad. Sci. 107, 378–383. Williams, J.A., Zhao, K., Jin, S., Ding, W.X., 2017. New methods for monitoring mitochondrial biogenesis and mitophagy in vitro and in vivo. Exp. Biol. Med. 242, 781–787. Yang, J.-Y., Yang, W.Y., 2011. Spatiotemporally controlled initiation of Parkin-mediated mitophagy within single cells. Autophagy 7, 1230–1238. Youle, R.J., Narendra, D.P., 2011. Mechanisms of mitophagy. Nat. Rev. Mol. Cell Biol. 12, 9–14. Zhang, J., Ney, P., 2010. Reticulocyte mitophagy: monitoring mitochondrial clearance in a mammalian model. Autophagy 6, 405–408. Zhou, C., Zhong, W., Zhou, J., Sheng, F., Fang, Z., Wei, Y., Chen, Y., Deng, X., Xia, B., Lin, J., 2012a. Monitoring autophagic flux by an improved tandem fluorescent-tagged LC3 (mTagRFP-mWasabi-LC3) reveals that high-dose rapamycin impairs autophagic flux in cancer cells. Autophagy 8, 1215–1226. Zhou, W.-H., Tang, F., Xu, J., Wu, X., Yang, S.-B., Feng, Z.-Y., Ding, Y.-G., Wan, X.-B., Guan, Z., Li, H.-G., 2012b. Low expression of Beclin 1, associated with high Bcl-xL, predicts a malignant phenotype and poor prognosis of gastric cancer. Autophagy 8, 389–400. Zhu, J., Dagda, R.K., Chu, C.T., 2011. Monitoring mitophagy in neuronal cell cultures. Neurodegener.: Methods Protoc. 325–339.

Joseph, J., Seervi, M., Sobhan, P.K., Retnabai, S.T., 2011. High throughput ratio imaging to profile caspase activity: potential application in multiparameter high content apoptosis analysis and drug screening. PLoS One 6, e20114. Kabeya, Y., Mizushima, N., Ueno, T., Yamamoto, A., Kirisako, T., Noda, T., Kominami, E., Ohsumi, Y., Yoshimori, T., 2000. LC3, a mammalian homologue of yeast Apg8p, is localized in autophagosome membranes after processing. EMBO J. 19, 5720–5728. Katayama, H., Kogure, T., Mizushima, N., Yoshimori, T., Miyawaki, A., 2011a. A sensitive and quantitative technique for detecting autophagic events based on lysosomal delivery. Chem. Biol. 18, 1042–1052. Katayama, H., Kogure, T., Mizushima, N., Yoshimori, T., Miyawaki, A., 2011b. A sensitive and quantitative technique for detecting autophagic events based on lysosomal delivery. Chem. Biol. 18, 1042–1052. Kimura, S., Noda, T., Yoshimori, T., 2007. Dissection of the autophagosome maturation process by a novel reporter protein, tandem fluorescent-tagged LC3. Autophagy 3, 452–460. Klionsky, D.J., Abdelmohsen, K., Abe, A., Abedin, M.J., Abeliovich, H., Acevedo Arozena, A., Adachi, H., Adams, C.M., Adams, P.D., Adeli, K., 2016. Guidelines for the use and interpretation of assays for monitoring autophagy. Autophagy 12, 1–222. Larsen, K.B., Lamark, T., Øvervatn, A., Harneshaug, I., Johansen, T., Bjørkøy, G., 2010. A reporter cell system to monitor autophagy based on p62/SQSTM1. Autophagy 6, 784–793. Lazarou, M., Sliter, D.A., Kane, L.A., Sarraf, S.A., Wang, C., Burman, J.L., Sideris, D.P., Fogel, A.I., Youle, R.J., 2015. The ubiquitin kinase PINK1 recruits autophagy receptors to induce mitophagy. Nature 524, 309–314. Lemasters, J.J., 2005. Selective mitochondrial autophagy, or mitophagy, as a targeted defense against oxidative stress, mitochondrial dysfunction, and aging. Rejuvenation Res. 8, 3–5. Matsuda, N., Tanaka, K., 2010. Uncovering the roles of PINK1 and parkin in mitophagy. Autophagy 6 (7), 952–954. Mauro-Lizcano, M., Esteban-Martínez, L., Seco, E., Serrano-Puebla, A., Garcia-Ledo, L., Figueiredo-Pereira, C., Vieira, H.L., Boya, P., 2015. New method to assess mitophagy flux by flow cytometry. Autophagy 11, 833–843. McWilliams, T.G., Prescott, A.R., Allen, G.F., Tamjar, J., Munson, M.J., Thomson, C., Muqit, M.M., Ganley, I.G., 2016. mito-QC illuminates mitophagy and mitochondrial architecture in vivo. J. Cell Biol. 214, 333–345. Mizushima, N., Kuma, A., 2008. Autophagosomes in GFP-LC3 transgenic mice. Autophagosome Phagosome 119–124. Narendra, D., Kane, L.A., Hauser, D.N., Fearnley, I.M., Youle, R.J., 2010. p62/SQSTM1 is required for Parkin-induced mitochondrial clustering but not mitophagy; VDAC1 is dispensable for both. Autophagy 6, 1090–1106. Novak, I., 2012. Mitophagy: a complex mechanism of mitochondrial removal. Antioxid. Redox Signal. 17, 794–802. Padman, B.S., Bach, M., Lucarelli, G., Prescott, M., Ramm, G., 2013. The protonophore CCCP interferes with lysosomal degradation of autophagic cargo in yeast and mammalian cells. Autophagy 9, 1862–1875. Peterson, Q.P., Hsu, D.C., Goode, D.R., Novotny, C.J., Totten, R.K., Hergenrother, P.J., 2009. Procaspase-3 activation as an anti-cancer strategy: structure- activity

14