Mitophagy, the process of clearing damaged mitochondria, is necessary for mitochondrial homeostasis and health maintenance. This article presents some of the latest mitophagy detection methods in human cells, Caenorhabditis elegans, and mice.
Mitochondria are the powerhouses of cells and produce cellular energy in the form of ATP. Mitochondrial dysfunction contributes to biological aging and a wide variety of disorders including metabolic diseases, premature aging syndromes, and neurodegenerative diseases such as Alzheimer’s disease (AD) and Parkinson’s disease (PD). Maintenance of mitochondrial health depends on mitochondrial biogenesis and the efficient clearance of dysfunctional mitochondria through mitophagy. Experimental methods to accurately detect autophagy/mitophagy, especially in animal models, have been challenging to develop. Recent progress towards the understanding of the molecular mechanisms of mitophagy has enabled the development of novel mitophagy detection techniques. Here, we introduce several versatile techniques to monitor mitophagy in human cells, Caenorhabditis elegans (e.g., Rosella and DCT-1/ LGG-1 strains), and mice (mt-Keima). A combination of these mitophagy detection techniques, including cross-species evaluation, will improve the accuracy of mitophagy measurements and lead to a better understanding of the role of mitophagy in health and disease.
Mitophagy is essential for mitochondrial maintenance. Mitochondria intersect multiple cell signaling pathways and are universal sub-cellular organelles responsible for cellular energy production, cell metabolism, and calcium homeostasis1,2,3,4. Mitochondria constantly experience challenges from endogenous and exogenous sources, such as reactive oxygen species (ROS) and mitochondrial toxicants, respectively, which lead to the generation of "aged" and dysfunctional mitochondria. Accumulation of damaged mitochondria decreases the efficiency of ATP production while increasing the amount of harmful ROS, and has been linked to age-related diseases such as metabolic diseases, AD, and PD1,5,6. To prevent mitochondria induced cellular dysfunction, cells need to specifically recognize damaged mitochondria and efficiently remove them through a cellular process termed mitochondrial autophagy (mitophagy). This demonstrated importance of mitophagy in health and disease illustrates the need for accurate and efficient methods to detect mitophagy both in vitro and in vivo.
Mitophagy is a multiple-step process involving many proteins and protein complexes5,7,8. In brief, a damaged mitochondrion is first recognized and engulfed by a double-membraned phagophore, which can originate from the plasma membrane, endoplasmic reticulum, Golgi complex, nucleus, or mitochondrion itself9,10. The spherical phagophore elongates and eventually seals the mitochondria inside, constituting the mitochondrial autophagosome (mitophagosome). The mitophagosome then fuses with the lysosome for degradation, forming an autolysosome in which the damaged mitochondrion is degraded and recycled7,8. Major autophagic proteins also involved in mitophagy include: Autophagy Related 7 (ATG7) and Beclin1 (initiation), Microtubule-Associated Protein 1A/1B-Light Chain 3 (LC3-II) (LGG-1 in C. elegans) and p62 (components of phagophore), and lysosomal-associated membrane glycoprotein 2 (LAMP2)6,7. In addition, there are several essential proteins unique to mitophagy, including PTEN-induced Putative Kinase 1 (PINK-1), Parkin1, Nuclear Dot Protein 52 kDa (NDP52), optineurin, BCL2 Interacting Protein 3 Like (NIX/BNIP3L) (DCT-1 in C. elegans), among others5,6,11.
A common method to detect changes in levels of autophagy is by the ratio of LC3-II/LC3-I or LC3-II/actin. However, this method is nonspecific, as an increase in this ratio may reflect either an increased initiation or an impaired fusion of mitophagosome to lysosome12. Another method is to evaluate the colocalization between an autophagy marker (e.g., LC3) and a mitochondrial protein (e.g., Translocase of Outer Mitochondrial Membrane 20 (TOMM20, which could be degraded by proteasomes)). However, this can only indicate changes in total mitophagy levels and cannot distinguish the step(s) at which blockage occurs. This can be clarified by using lysosomal inhibitors (e.g., E64d+Pepstatin A, termed EP) in parallel to cause the accumulation of mitophagosomes. The difference between the number of mitophagosomes at baseline and the number of mitophagosomes following treatment with EP can indicate mitophagy. These limitations have prompted the development of novel mitophagy detection techniques. In view of the increasing relevance of mitophagy in a wide spectrum of human diseases, we present several robust mitophagy detection techniques which may be useful for researchers. We cover both in vitro and in vivo techniques and recommend combining multiple techniques to verify changes of mitophagy.
Animals (male and female mice) were born and bred in an accredited animal facility, in accordance and approval of the NIH Animal Care and Use Committee. Euthanasia methods must be consistent with all national and institutional regulations.
1. Detection of Mitophagy in Human Cells
2. Detection of Mitophagy in C. elegans
NOTE: The nematode C. elegans provides a platform to assay mitophagy at the organismal level. Two strains can be used to monitor mitophagy: (1) mitochondria-targeted Rosella (mtRosella) or (2) mitophagy receptor DCT-1 fused with GFP along with autophagosomal marker LGG-1 fused with Discosoma sp. red fluorescent protein (DsRed)5,17.
3. Detection of Mitophagy in Mice
NOTE: Previous methods to detect mitophagy in mice were cumbersome, insensitive, and difficult to quantify. A transgenic mouse model expressing the mitochondrial-targeted form of the fluorescent reporter Keima (mt-Keima) can now be utilized to assess levels of mitophagy in a wide range of physiological and pathophysiological conditions.
Detection of Mitophagy in Human Cells:
Using the procedure presented here, human HeLa cells were transfected with mt-Keima plasmid. Healthy cells demonstrated a well-organized mitochondrial network (GFP, 488 nm) with few incidences of mitophagy (RFP, 561 nm). However, cells pretreated with a mitochondrial uncoupler FCCP (30 µM for 3 h) exhibited a profound increase in mitophagy incidence (Figure 1A). Mitophagy was also measured using the co-localization between LAMP2 and COXII (Figure 1B).
Detection of Mitophagy in C. Elegans
We examined the body-wall muscle cells of transgenic nematodes expressing DsRed fused with LGG-1 and either mtRosella or DCT-1 fused with GFP under normal and mitophagy-inducing conditions such as oxidative stress. Transgenic animals were exposed to paraquat (8 mM for 2 days). Mitophagy induction is indicated by the decreased GFP/DsRed ratio of mtRosella fluorescence (Figure 2A). Additionally, the elevated number of co-localization events between DCT-1::GFP and DsRed::LGG-1 signals signifies mitophagy stimulation in response to oxidative stress (Figure 2B).
Detection of Mitophagy Using the mt-Keima Mice:
Imaging analysis of in vivo mitophagy provides insight into mitophagy in normal and pathophysiological conditions. Using the protocol described above, mitophagy occurrence in different organ tissues, such as liver and cerebellum (Figure 3), can be visualized with confocal mitophagy.
Figure 1. Different methods to evaluate mitophagy in human cells. (A) Human HeLa cells were transfected with mt-Keima plasmid for three days, followed by imaging at both 488 nm and 561 nm. As a positive control, ceflls were treated with FCCP (30 µM) 3 h prior to imaging. (B) Images of human primary fibroblasts stained with LAMP2 (RFP) and COXII (GFP). (C) Images of human fibroblasts stained with anti-TOM20 (red) and anti-LC3 (green) antibodies. The bars on the right show the readout of mitochondria co-localizing with autophagosomes. Co-localization is increased by energetic stress (glucose-free galactose media) and the addition of lysosomal inhibitors E64d and Pepstatin A (EP) for 24 h. Representative scale bars for (A), (B), and (C) are labeled on the last figure of each panel, independently. Quantitative data are shown in mean ± S.E.M. ***p < 0.001; t-test. Please click here to view a larger version of this figure.
Figure 2. In vivo detection of mitophagy in C. elegans. (A) Transgenic nematodes expressing mtRosella in body-wall muscle cells were treated with 8 mM paraquat. Mitophagy stimulation is signified by the decreased ratio between pH-sensitive GFP to pH-insensitive DsRed (n = 50; ***p < 0.001; t-test). Arrowheads point out intestinal autofluorescence. Scale bars denote 20 µm. Acquisition details: Exposure time, 200 ms; Contrast, medium. Images were acquired using a 10X objective lens. Quantitative data are shown in mean ± S.E.M. values. (B) Transgenic nematodes co-expressing mitophagy receptor DCT-1 fused with GFP together with the autophagosomal protein LGG-1 fused with DsRed in body-wall muscle cells were exposed to 8 mM paraquat. Mitophagy induction is signified by co-localization of GFP and DsRed signals (for each group of images DCT-1 is shown in green on top, autophagosomes in red below, and a merged image at the bottom; n = 30; ***p < 0.001; t-test). Sale bars denote 20 µm. Acquisition details: Resolution, 1,024 X 1,024; Master gain, Track1: 562 and Track2: 730; Emission filters, Track1 Channel1: 639 nm and Track2 Channel2: 519 nm; Laser intensity, Track1 (543 nm): 12.9% and Track2 (488 nm): 39.4%. Images were acquired using a 40X objective lens. Quantitative data are shown in mean ± S.E.M. Please click here to view a larger version of this figure.
Figure 3. In vivo detection of mitophagy in mt-Keima transgenic mice. Representative images of mt-Keima signals in the liver (upper) and cerebellum area (including Purkinje cells, lower panel) of a mt-Keima mouse (458 nm, green; 561 nm, red). Scale bar denotes 10 µm. Please click here to view a larger version of this figure.
Accurate measurement of mitophagy is technically demanding. Here, we have presented several robust techniques which allow for both qualitative detection of mitophagy and quantification of mitophagy levels in the most common laboratory experimental models.
To acquire replicable data, an experimental design with at least three biological repeats is necessary. All researchers involved in experimentation and analysis must be blinded to experimental group identities. Furthermore, imaging fields must be randomly chosen during image acquisition. For cell culture and C. elegans studies, at least three biological repeats should be performed. For mt-Keima mouse studies, it is recommended to use a sufficient number of mice to achieve statistical significance. For mammalian cells, analyze 30-100 cells for each experiment, and run at least 3 different experiments. Quality control at these steps enables replicable results. Detection of mitophagy in yeast has recently been summarized elsewhere25.
There are several critical steps in the in vitro mitophagy detection techniques. Transfection efficiency, which depends on the quality of reagents and the DNA used, is vital during transfection of the mt-Keima protein. Thus, optimization of the transfection efficiency in different conditions (e.g., using different cell lines) is necessary. For the LAMP2/COXII method, antibody specificity and optimal primary antibody dilution fold affect the quality of the images and should be tested before beginning experiments. The same applies to the high-content imaging method, where antibody quality and dilution are crucial. The automated microscopy system and the customized analysis protocol allows for the imaging and analysis of at least 1,500 cells/condition, which makes the results extremely robust compared to other in vivo imaging based methods. Additionally, the analysis protocol provides data on further mitochondrial parameters such as length and total area on a cell by cell basis, which is highly useful in mitophagy research.
In some cases, it is not recommended to assay the colocalization of LAMP2 with mitochondrial outer membrane proteins such as TOMM20. The very early stages of mitophagy involve Parkin-mediated ubiquitination of mitochondrial outer membrane proteins, including TOMM20 and Mitofusin 1 (MFN1). Parkin conjugates multiple different ubiquitin chain linkages on these proteins including K48-linked ubiquitin chains, which promote degradation of the protein by the 26S proteasome. Therefore, these ubiquitinated mitochondrial outer membrane proteins can still be degraded even though mitophagy is blocked6. This can skew the data, resulting in false positives in overall data interpretation. In addition, elimination of mitochondrial DNA (mtDNA nucleoids) is a second indicator of mitophagy, and it can be quantified by immunofluorescence using an anti-DNA antibody6.
Some critical steps for monitoring in vivo mitophagy in C. elegans are listed below:
1. Transfer of transgenic animals to new plates every 24–48 h is recommended to avoid the outgrowth of progeny, which leads to a mixed population or starvation, which itself may trigger mitophagy. Therefore, non-starved nematodes should be used.
2. Levamisole is a mild anesthetic which is used to immobilize nematodes. Avoid anesthetics that could affect mitochondrial activity, such as sodium azide. Sodium azide blocks components of the mitochondrial respiratory chain and perturbs energy generation, leading to mitochondrial and oxidative stress. Thus, sodium azide is likely to trigger mitophagy.
3. Specimens should not be allowed to dry out during microscopic examination. Therefore, the use of M9 buffer instead of water is recommended to ensure favorable osmotic conditions.
4. Intestinal autofluorescence increases with age in C. elegans. Thus, body-wall muscle cells close to the intestine should be avoided during microscopic examination.
5. If paraquat fails to induce mitophagy: a. increase paraquat exposure time on worms, b. increase paraquat concentration, or c. prepare a fresh working solution of paraquat since its efficiency declines over time.
6. If increased matricidal hatching is observed in worms exposed to paraquat, either: a. reduce the period of paraquat treatment, b. reduce paraquat concentration, c. supplement plates with fluorodeoxyuridine (FUdR) (final concentration of 100-400 μM), an inhibitor of DNA synthesis that prevents egg hatching, or d. conduct the experiment in older worms (e.g., 4-day-old worms).
This procedure describes the steps for imaging mitophagy in the liver using the mt-Keima transgenic mice, as shown in Figure 3. Tissues other than the brain and liver have been analyzed using the mt-Keima system14. Dual-excitation ratio imaging in liver tissues is obtained via two sequential excitations. All images must be obtained from freshly excised organs. Tissues examined should be kept cold and processed as quickly as possible. Liver slices can be obtained at any age. When imaging mitophagy, optimize laser powers and exposure times for each organ during microscopic analysis. Laser power should be set at the lowest output that allows clear visualization of the mt-Keima signal14. However, there are some limitations for the mt-Keima transgenic mice. Due to the instability of Keima as well as the loss of a pH gradient across the lysosomal membrane under fixation, this mouse model is not suitable for cryo-sectioning and immunohistochemistry. Another limitation is that mt-Keima, or matrix aggregates of this protein, may affect unknown mitochondrial or cellular functions, even though it does not change mitochondrial oxygen consumption rate14. Furthermore, the time and cost associated with mt-Keima mice render it less desirable for high-throughput analysis than the aforementioned cell culture and C. elegans methods. Besides the methods mentioned here, other ways to quantitatively study mitophagy in mt-Keima mice include Western blot or FACS. To note, in addition to the mt-Keima transgenic mice, there is another mitophagy mouse model, the "mito-QC", a transgenic mouse model with a pH-sensitive fluorescent mitochondrial signal26. Due to the complexity and dynamics of mitophagy, we recommend combining the aforementioned fluorescence methods with other common mitophagy detection methods, such as electron microscopy and Western blotting, to strengthen the findings.
The development of new mitophagy detection techniques such as those mentioned here will have broad applications, from mechanistic studies of mitophagy to high throughput drug screens. It should be noted that mitophagy is easily affected by both endogenous and exogenous fluctuations (e.g., cell culture conditions such as cell density, starvation, hypoxia, etc.)1,5,8. Therefore, it is paramount that the same experimental conditions are maintained for all experiments. It is important to use a combination of different mitophagy detection methods in both in vitro and in vivo studies to verify the correct interpretation of mitophagy status. Further understanding of the mechanisms of mitophagy, achieved through techniques mentioned herein, will allow for the development of new, more precise methods of mitophagy detection.
The authors have nothing to disclose.
We thank Dr. Atsushi Miyawaki and Dr. Richard J. Youle for sharing the mt-Keima plasmid and mt-Keima integrated Hela cells. We thank Raghavendra A. Shamanna and Dr. Deborah L. Croteau for the critical reading of the manuscript. This research was supported by the Intramural Research Program of the NIH (VAB), National Institute on Ageing, as well as a 2014-2015 and a 2016-2017 NIA intra-laboratory grant (EFF, VAB). EFF was supported by HELSE SOR-OST RHF (Project No: 2017056) and the Research Council of Norway (Project No: 262175).
AUTHOR CONTRIBUTIONS:
EFF designed the manuscript and prepared the draft; KP, NS, EMF, RDS, JSK, SAC, YH, and ED wrote different sections of the paper; NT, JP, HN, and VAB revised the manuscript and provided expertise.
mt-Keima mouse | Jackson Laboratory | ||
Lipofectamine 2000 DNA transfection reagent | Thermofisher | #11668027 | |
Opti-MEM medium (Gibco) | Thermofisher | #31985062 | serum-free medium |
mtKemia plasmid: pCHAC-mt-mKeima | addgene | #72342 | |
COXII antibody (mouse) | abcam | #ab110258 | |
LAMP2 antibody (rabbit) | NOVUS | #CD107b | |
goat-anti-rabbit with wavelength 568 nm of red fluorescent protein (RFP) | Thermofisher | #Z25306 | Alexa Fluor 568 dye |
goat-anti-mouse with wavelength 488 nm of green fluorescent protein (GFP) | Thermofisher | #Z25002 | Alexa Fluor 488 dye |
prolong gold antifade with DAPI | Invitrogen | #P36931 | |
6-well plate | SIGMA Corning Costar |
#CLS3516 | |
4-well chamber slide | THermofisher, Nunc Lab-Tek | #171080 | |
Nunc F 96-well plate | Thermofisher | #152038 | |
LC3B antibody rabbit | NOVUS | #NB100-2220 | |
DNA antibody | Progen Biotechnik | #anti-DNA mouse monoclonal, AC-30-10 | |
DAPI | Thermofisher | #D1306 | antifade mounting medium with DAPI |
IN Cell analyzer (fluorescent reader ) | GE Healthcare Life Sciences | #IN Cell analyzer 2200 | |
Eclipse TE-2000e confocal microscope | Nikon | #TE-2000e | |
Colocalization software | Volocity | #Volocity 6.3 | alternative Zeiss ZEN 2012 software |
IN Cell Investigator Software | GE Healthcare Life Sciences | #28408974 | |
cell culture medium | Thermofisher | #DMEM–Dulbecco's Modified Eagle Medium | |
Penicillin-Streptomycin (10,000 U/mL) | Thermofisher | #15140122 | |
Fetal Bovine Serum | Sigma-Aldrich | #12003C-1000ML | |
Cell culture Incubator | Thermofisher | #Thermo Forma 3110 CO2 Water Jacketed Incubator | |
epifluorescence microscope | Zeiss | Zeiss Axio Imager Z2 | |
camera | Olympus | Olympus DP71 | |
confocal microscope | Zeiss | Zeiss Axio Observer Z1 | |
confocal software | Zeiss | ZEN 2012 | |
image analysis software | Image J | colocalization analysis, etc | https://imagej.nih.gov/ij/ |
statistical analysis software | GraphPad Software Inc., San Diego, USA | GraphPad Prism software package | |
material to make a worm pick | Surepure Chemetals | #4655 | The pick is made of 30 gauge 90% platinum 10% iridium wire |
IR: N2;Ex[pmyo-3 TOMM-20::Rosella] | Material inquiry to Tavernarakis Nektarios | Maintain transgenic animals at 20 °C | |
IR: N2; Ex[pdct-1 DCT-1::GFP; pmyo-3 DsRed::LGG-1] | Material inquiry to Tavernarakis Nektarios | ||
pmyo-3 TOMM-20::Rosella | Material inquiry to Tavernarakis Nektarios | ||
pdct-1 DCT-1::GFP | Material inquiry to Tavernarakis Nektarios | ||
pmyo-3 DsRed::LGG-1 | Material inquiry to Tavernarakis Nektarios | ||
Paraquat solution | see supplementary data for preparation | ||
M9 buffer | see supplementary data for preparation | ||
M9-levamisole buffer | see supplementary data for preparation | ||
Glass Microscope Slides and Coverslips | Fisher Scientific | #B9992000 | |
Surgical forceps | STERIS Animal Health | 19 Piece Canine Spay Pack Economy | |
Surgical scissors | STERIS Animal Health | 19 Piece Canine Spay Pack Economy | |
1x PBS | Thermofisher | #AM9625 | 10x PBS needs to be diluted to 1x PBS by using ddH2O |
shaker | Fisher Scientific | #11-676-178 | Thermo Scientific MaxQ HP Tabletop Orbital Small Open Air Platform Shaker Package A |
2% agarose pad | see supplementary data for preparation | ||
Vibroslice blades | World precision instruments | #BLADES-2 | single-edge blade |
metal plate | MSC | #78803988 | 0.012 in thick x 6 in wide x 12 in long, 430 Stainless Steel Sheet |
Triton X-100 | detergent | ||
Methyl viologen dichloride hydrate | Sigma-Aldrich | #856177 | paraquat |
Incubator for nematodes | AQUALYTIC | Incubator to maintain 20 °C | |
Dissecting stereomicroscope | Olympus | SMZ645 | |
Confocal microscope | Zeiss | AxioObserver Z1 | For nematodes (step 2) |
epifluorescence microscope | Zeiss | AxioImager Z2 | For nematodes (step 2) |
UV crosslinker | Vilber Lourmat | BIO-LINK – BLX-E365 | UV light source; 356 nm |