Temporal integration of mitochondrial stress signals by the PINK1:Parkin pathway
The PINK1:Parkin pathway regulates the autophagic removal of damaged and dysfunctional mitochondria. While the response of this pathway to complete loss of ΔΨm, as caused by high concentrations of mitochondrial ionophores, has been well characterized, it remains unclear how the pathway makes coherent decisions about whether to keep or purge mitochondria in situations where ΔΨm is only partially lost or exhibits fluctuations, as has been observed in response to certain types of cellular stress.
To investigate the responses of the PINK1:Parkin pathway to mitochondrial insults of different magnitude and duration, controlled titration of the mitochondrial protonophore, CCCP, was used to manipulate ΔΨm in live cells, and the dynamics of PINK1 and Parkin recruitment was measured by fluorescence microscopy. In contrast to the stable accumulation of PINK1 and Parkin seen at completely depolarized mitochondria, partial depolarization produced a transient pulse of PINK1 stabilization and rapid loss, which was driven by small fluctuations in ΔΨm. As the rate of Parkin dissociation from the mitochondria and phospho-polyubiquitin chain removal was comparatively slow, repetitive pulses of PINK1 were able to drive a slow step-wise accumulation of Parkin and phospho-polyubiquitin leading to deferred mitophagy.
These data suggest that the PINK1:Parkin mitophagy pathway is able to exhibit distinct dynamic responses to complete and partial mitochondrial depolarization. In this way, the pathway is able to differentiate between irretrievably damaged mitochondria and those showing signs of dysfunction, promoting either rapid or delayed autophagy, respectively.
KeywordsCell signalling Mitophagy Parkin PTEN-induced putative kinase 1 (PINK1) Ubiquitin
Carbonyl cyanide m–chlorophenyl hydrazine
Nuclear dot protein 52
Nuclear factor-kappa B
Outer mitochondrial membrane
PTEN-induced putative kinase 1
Translocase of outer membrane:translocase of inner membrane
Ubiquitin specific peptidase
Mitochondrial membrane potential
The mitochondria are vitally important, multifunctional organelles, responsible for ATP production, the regulation of apoptosis, and other cell type-specific processes. A cell’s mitochondrial network is capable of remodelling itself – expanding or diminishing in size to meet the metabolic demands of the cell – or ‘pruning’ and destroying dysfunctional regions of the network by mitophagy to maintain its integrity. While several pathways exist that regulate the autophagic removal of mitochondria from the cell [1, 2, 3, 4], the PTEN-induced putative kinase 1 (PINK1):Parkin pathway is perhaps the most prominent and most closely associated with the destruction of mitochondria in response to stress or loss of mitochondrial membrane potential (ΔΨm) in normal, somatic cells [2, 5]. Indeed, mutation of either the genes encoding the PINK1 and Parkin proteins (PINK1 and PRKN, respectively) can result in the accumulation of dysfunctional mitochondria in animal and cell models  and is associated with juvenile onset autosomal recessive forms of Parkinson’s disease (PD; [7, 8, 9]).
The PINK1 protein itself is a highly labile, nuclear-encoded, mitochondrially-targeted kinase that essentially acts as a sensor of mitochondrial membrane potential (ΔΨm). In normal, polarized mitochondria, PINK1 is translocated into the mitochondria by the translocase of the outer mitochondrial membrane: translocase of the inner membrane (TOM:TIM) complex. There, the amino-terminus of PINK1 undergoes two rounds of proteolytic processing; the mitochondrial targeting sequence is removed by matrix-localized proteases and then a larger portion of the protein is removed by PGAM5-associated rhomboid-like protease (PARL) [10, 11, 12]. The mature 52 kDa PINK1 protein is then retrotranslocated to the cytosol side of the outer mitochondrial membrane (OMM), ubiquitinated, and rapidly degraded by the 26S proteasomes, thereby maintaining a low basal concentration of PINK1. However, as the operation of TIM23 – the component of the TOM:TIM complex that interacts directly with the N-terminus of PINK1  – is ΔΨm dependent, at depolarized regions of a cell’s mitochondrial network, PINK1 accumulates at the OMM. Here, it is able to activate Parkin directly via phosphorylation at Ser65 within the N-terminal Ubl domain and indirectly via production of phospho-ubiquitin moieties [14, 15, 16, 17]. PINK1 and Parkin then operate together to assemble phospho-polyubiquitin (ppUb) chains on OMM substrates with ppUb serving as docking sites for Parkin [18, 19], amplifying the response . These ppUb chains are also bound by autophagy receptors, including optineurin (OPTN) and nuclear dot protein 52 (NDP52) [21, 22]. These bind both ubiquitin and microtubule-associated protein light chain 3 (LC3), promoting the sequestration of ppUb-labelled mitochondria within autophagosomes to be degraded later after fusion with lysosomes.
The relationship between PINK1 and Parkin, and the phospho-polyubiqutin chains that the two proteins synthesize, together constitute two interlinked coherent feed-forward loops (FFLs) – the Parkin activation FFL and phospho-polyubiquitin FFL – followed by a positive feedback loop [19, 20, 23]. Coherent FFLs consist of two activators that coregulate an output, with one of the activators capable of activating the other . Within the context of the PINK1:Parkin pathway, it is likely that these FFLs create an initial delay in the accumulation of phospho-polyubiqutin chains at the OMM, but once a critical concentration of these chains has been reached, the positive feedback produced by the recruitment of Parkin to the chains – their own product – acts as an amplifier, increasing the rate of their production. It is currently hypothesized that this arrangement of feedback loops enables the PINK1:Parkin pathway to essentially filter out transient mitochondrial stress signals, only committing mitochondria to mitophagy if they exhibit a persistent loss of ΔΨm . This has been tested in vitro in various studies using reversible inhibitors of oxidative phosphorylation, such as carbonyl cyanide m–chlorophenyl hydrazone (CCCP), applied at high concentrations to completely depolarize the mitochondria for different periods of time [25, 26]. While these studies have been incredibly revealing, they may not accurately recapitulate the characteristics of mitochondrial stresses experienced in vivo, which are likely to be complex, and the changes in ΔΨm that they produce are likely to be entangled with natural or functionally important variation in mitochondrial activity as part of a cell’s normal operations [27, 28, 29]. It therefore remains an open question as to how the PINK1:Parkin pathway successfully interprets mitochondrial stress signals, especially in situations where ΔΨm is only partially lost or where the cell experiences intermittent exposure to mitochondrial stresses.
In this study, we use a live cell imaging approach to investigate how the PINK1:Parkin pathway processes different stress signals. By monitoring ΔΨm and the expression and localization of PINK1 and Parkin proteins in individual cells, we show that at intermediate levels of mitochondrial stress, where ΔΨm is only partially lost, the temporal dynamics of PINK1 and Parkin differ markedly. While a complete and sustained loss of ΔΨm induced by high concentrations of CCCP leads to stable association of both PINK1 and Parkin with the mitochondria, a partial loss of ΔΨm produces transient pulses of PINK1 accumulation in cells. Loss of mitochondrial PINK1 at the conclusion of these pulses is preceded by a small, partial recovery in ΔΨm even when cells are exposed to constant concentrations of CCCP. Furthermore, when cells are exposed to repeated rounds of mitochondrial stress, occurring on a timescale of minutes or hours, PINK1 is not influenced by prior depolarization events and responds to each event independently, quickly returning to basal levels upon partial or complete recovery of ΔΨm. However, as PINK1 and Parkin operate on different timescales, with Parkin dissociating relatively slowly from phospho-polyubiquitin chains on the OMM after repolarization, repeated rounds of mitochondrial depolarization can cause a step-like accumulation of mitochondrial Parkin if mitochondrial depolarization events are sufficiently frequent or if the mitochondria remain in a persistent state of partial depolarization. In this way, Parkin and phospho-polyubiqitin at the OMM can act as a marker of prior mitochondrial stress, triggering mitophagy if ΔΨm is partially lost for extended periods of time or if fluctuations in ΔΨm occur at high frequency but filter out or “ignore” infrequent, stochastic fluctuations in ΔΨm.
Titration of CCCP to generate states of partial and complete loss of mitochondrial membrane potential
To determine whether cells expressing endogenous Parkin would exhibit a similar response, we first tested whether different levels of ΔΨm could be achieved and reversed in the dopaminergic, neuroblastoma-derived cell line, SH-SY5Y, by titrating CCCP into the cells followed by wash-out (Fig. 1g). As very similar effects to those observed in HeLa cells were achieved we subsequently tested whether a sustained complete or partial loss of ΔΨm could also stimulate a loss of mitochondrial mass in SH-SY5Y cells. This was achieved by staining cells with the ΔΨm-independent mitochondrial dye, MitoTracker Green FM, after incubation in low, medium, and high doses of CCCP for 16 h. Consistent with our observations in Parkin-expressing HeLa cells, we found that SH-SY5Y cells exhibited a dose-dependent response to CCCP with even medium doses producing a statistically significant reduction in staining, indicating that even a partial loss of mitochondrial membrane potential could stimulate mitophagy (Fig. 1h and i).
Partial loss of ΔΨm produces transient pulses of mitochondrial PINK1
Partial recovery of ΔΨm is sufficient to stimulate loss of mitochondrial PINK1
To examine whether the dynamics of exogenous PINK1-EGFP was consistent with endogenous PINK1 proteins, the levels of endogenous PINK1 were measured by western blotting in cells after partial or complete repolarization by incubation with 5, 2.5, or 0 μM CCCP after 1 h in 10 μM CCCP. In agreement with the live cell experiments, unprocessed, full-length PINK1 proteins decreased to near-basal levels after partial repolarization, although this appeared to occur more rapidly at lower CCCP concentrations (Fig. 4g). Consistent with published data, cells exhibited a rapid and complete loss of full-length PINK1 when CCCP was removed entirely [25, 36].
The rate of Parkin loss from the mitochondria is slower than PINK1 after partial or complete recovery of ΔΨm
The autophagy receptor, OPTN, is also retained at Parkin-positive mitochondrial fragments after repolarization of mitochondria
Although fewer Parkin-positive mitochondria were observable by 120 min in pulsed cells compared to continuously treated cells (Additional file 3: Figure S3A), OPTN puncta were still associated with fragmented mitochondria at 120 min, 60 min post-CCCP wash-off (Fig. 7a, c and Additional file 3: Figure S3B) with OPTN-EGFP:mCherry-Parkin colocalization slowly diminishing over time (Fig. 7d). By the same time in the continuously treated cells, OPTN-EGFP puncta had spread over the surface of mCherry-Parkin-coated fragmented mitochondria, forming crescents or rings (Fig. 7e and f), suggesting that although OPTN-EGFP could be retained for a short period of time after mitochondrial repolarization, the process of coating mitochondrial fragments with this autophagy adaptor was arrested and does not go to completion. This may prevent mitochondria that exhibit a transient decrease in ΔΨm from being sequestered within autophagosomes.
The PINK1 response to loss of ΔΨm is stereotyped and is not influenced by prior depolarization events
Pulses of PINK1 activity can promote an incremental accumulation of mitochondrial Parkin and phospho-polyubiquitin
As our previous experiments indicated that mitochondrial EYFP-Parkin levels reach a plateau, becoming near-saturating within 60 min of CCCP treatment (Fig. 5c), we next examined the effects of shorter CCCP pulses on the dynamics of mitochondrial EYFP-Parkin recruitment. The duration of each pulse was reduced to 30 min, which is sufficient to induce PINK1 and Parkin accumulation at the OMM but not long enough for stable recruitment of autophagy receptors or LC3, which takes approximately 35–60 min [37, 39]. For the initial experiment, the pulses were separated by 30 min intervals. Here, we saw a clear increase in the quantity of mitochondrial EYFP-Parkin after successive CCCP pulses (Fig. 9d). However, when the interval between successive pulses was increased to 120 min – the time required for complete removal of phospho-polyubiquitin induced by a prior pulse (Fig. 6d) – the level of mitochondrial Parkin accumulation after each pulse appeared identical in individual cells (Fig. 9e). This was confirmed by statistical analysis, showing that the relative amplitude of the second peak (the maximum level of mitochondrial EYFP-Parkin achieved during each pulse) compared to the first was greater in cells where the pulse interval was only 30 min (Fig. 9f). In order to determine whether this result was meaningful at the level of mitophagy, we exposed cells to the same cycles of CCCP treatment and removal (30 min on:30 min off or 30 min on:120 min off), continuous treatment with CCCP, or a single 30 min pulse of CCCP during a 6 h-long experiment (Fig. 9g). As expected, a greater number of cells exhibited aggregated or cleared mitochondria when cells were pulsed with CCCP at 30 min intervals (80%) than at 120 min intervals (36%) but fewer than those exposed to CCCP continuously (100%; Fig. 9h). Together, these data indicate that the PINK1:Parkin pathway is capable of integrating information about prior disturbances in mitochondrial activity at the level of mitochondrial Parkin recruitment and phospho-polyubiquitin deposition. Furthermore, these data also show that it is not necessary for PINK1 to be continuously associated with the OMM in order to promote Parkin and phospho-polyubiquitin accumulation and that this is a consequence of the difference in the rate of PINK1 and phospho-polyubiquitin degradation upon recovery of ΔΨm.
PINK1 and Parkin operate together as a molecular switch that is capable of responding to diverse mitochondrial insults and directing an appropriate response at both the level of individual mitochondria and the cell itself. They can stimulate apoptosis by inducing Parkin-dependent ubiquitination of the anti-apoptotic Bcl-2 family protein, Mcl-1, in response to catastrophic and irreparable mitochondrial damage or even salvage mitochondria damaged by ROS by directing the removal of oxidized mitochondrial proteins in mitochondria-derived vesicles, which are trafficked to lysosomes [40, 41]. Of all its functions, the ability of the PINK1:Parkin pathway to direct the autophagic removal of whole mitochondria exhibiting a complete and sustained loss of ΔΨm has received the most attention and study. Despite this, it still remains unclear how the pathway responds and makes coherent decisions about whether to keep or purge mitochondria in situations where ΔΨm is only partially lost or fluctuates over time, as might occur as part of the aging process or through exposure to low levels of mitochondrial stressors.
As a first step towards answering these questions, we investigated how persistent partial loss of ΔΨm affected mitochondrial mass in cells treated with low and intermediate concentrations of the reversible oxidative phosphorylation inhibitor, CCCP. In SH-SY5Y cells, which have an intact PINK1:Parkin pathway, we found that even low doses of CCCP, which caused a slight but measurable decrease in ΔΨm, were capable of stimulating a loss of mitochondrial mass within 16 h, albeit less than that induced by CCCP doses that cause a complete loss of ΔΨm (Fig. 1h and i) – an apparent dose-dependent response to a reduction in ΔΨm. This could possibly be explained by a difference in the response between cells or the generation of heterogeneous populations of partially and fully depolarized mitochondria in individual cells at low and intermediate CCCP doses, as has previously been seen in cells exposed to CCCP for short periods of time , with only the mitochondria exhibiting the lowest ΔΨm being removed. In our experiments, where longer incubation periods allow CCCP to fully equilibrate within cells, it seems more likely that rather than creating mixed populations, the mitochondria will be more uniformly affected by the CCCP, a response supported by our TMRM staining experiments (Fig. 1a). It then becomes probable that the differences we see in mitochondrial mass after incubation with low and high CCCP doses are a product of differences in the kinetics of the PINK1 and Parkin response.
To investigate this possibility, we employed a live cell imaging approach, enabling us to measure the temporal dynamics of PINK1 and Parkin recruitment and loss at mitochondria in cells exposed to different CCCP treatment regimes. Consistent with previous studies showing stable accumulation of full-length endogenous PINK1 in cell populations incubated continuously with high doses of CCCP , we also observed stable accumulation of PINK1-EGFP in individual cells (Fig. 2b). However, in cells exposed to lower doses, PINK1 accumulation was transient, generally lasting less than 90 min (Fig. 2b and c). The reduction in PINK1 was sharp and asynchronous between cells and was found to occur concomitantly with a small increase in ΔΨm (Fig. 3a and b). Like voltage-gated channels, TIM23 undergoes a structural rearrangement, switching from an inactive to an active state if a specific threshold transmembrane electrical potential is exceeded . Fluctuations in ΔΨm in partially depolarized mitochondria may therefore drive periodic bursts of TIM23 activity and inactivity as ΔΨm rises and falls around the threshold, producing corresponding pulses of mitochondrial PINK1 accumulation.
The rate of Parkin loss from repolarized mitochondria was markedly slower than that of PINK1 (Fig. 5). Our data suggests (although it does not directly prove) that this is a likely consequence of the slow removal of the ppUb on the OMM of mitochondria after repolarization. These chains have been shown to anchor Parkin at the mitochondria and also serve to recruit autophagy receptors like OPTN, which our data shows is also retained for many minutes after repolarization (Fig. 7). The persistence of ppUb chains is likely a consequence of the phosphorylation, which appears to partially protect polyubiquitin chains from the action of mitochondrially-targeted deubiquitinases including ubiquitin specific peptidase 8 (USP8), USP15, and USP30 , all of which have been shown to regulate Parkin-dependent mitophagy [47, 48, 49, 50]. Our data suggests that it is the disparity in the timing of PINK1 degradation and ppUb removal that enables short but frequent bursts of PINK1 activity to drive incremental accumulation of Parkin and ppUb chains up to a threshold where autophagy receptors can be stably recruited. In this case, the threshold will be reached at a much slower rate than at completely depolarized mitochondria, where PINK1 is stably recruited. In this way the PINK1:Parkin pathway, like other prominent stress-responsive signalling mechanisms, may exhibit different dynamic behaviours in response to insults of differing magnitude , enabling the cell to expedite the removal of irretrievably damaged, non-functional mitochondria but also label, isolate, and temporarily defer the autophagy of partially depolarized mitochondria.
The importance of the phospho-polyubiquitin chain removal rate in shaping the dynamics of the response to mitochondrial insults raises the interesting possibility that the sensitivity of the PINK1:Parkin mitophagy pathway to recurrent changes in ΔΨm could be ‘tuned’ by adjusting the expression level of USPs. Potentially, this could also be achieved by altering the expression of autophagy receptors, including OPTN, which has been shown to be rate-limiting for mitophagy . It could also hypothetically be controlled through the expression of phosphatases that dephosphorylate ppUb. This is likely to be important, given the various demands placed upon mitochondria in different cell and tissue types, and plausible, given that other stress-responsive pathways, such as p53, may show different dynamic behaviours in different cell types . It is also possible that different mitochondrial quality control mechanisms, such as Nix and Bnip3, will be used in preference to PINK1:Parkin in cells that naturally produce large and repeating shifts in their activity to avoid inappropriate mitophagy .
Finally, one potential caveat that was raised during peer review of this manuscript was the effect of FBS on the ability of CCCP to induce mitochondrial depolarization. As shown by Soutar et al, FBS reduces the potency of CCCP such that concentrations of 10 μM or greater are required to completely depolarize mitochondria in cells cultured in the presence of 10% FBS whereas concentrations < 1 μM will achieve comparable effects in the absence of FBS . As all experiments were performed in the presence of 10% FBS, we do not believe that this observation impacts the findings of our study in any way. In many of our experiments, ΔΨm (measured using TMRM) and the localization of fluorescently labelled mitophagy regulators was monitored in the same living cells. We believe that our work and the data presented in Soutar et al emphasize the importance of using consistent cell growth conditions in studies using CCCP that seek to investigate links between changes in ΔΨm and the activity and dynamics of mitophagy regulators.
Our data shows differences in the temporal dynamics of PINK1 and Parkin accumulation and loss at mitochondria exposed to different concentrations of the protonophore, CCCP, with intermediate levels of mitochondrial depolarization giving rise to bursts of PINK1 activity caused by small fluctuations in ΔΨm. By artificially inducing these PINK1 pulses through controlled titration of CCCP, we also show that the pathway can integrate temporally separated periods of mitochondrial depolarization through incremental accumulation of Parkin and ppUb. We propose that this provides a mechanism for cells to discriminate between mitochondria exhibiting different degrees of impairment and control the rate at which they are autophagized. Importantly, it also demonstrates that stable association of PINK1 with the OMM is not essential for mitophagy and likely only occurs at the most heavily damaged mitochondria.
Cell culture, plasmids, and transfections
HeLa (CCL-2) human cervical adenocarcinoma cells and SH-SY5Y (CRL-2266) human neuroblastoma cells were obtained from ATCC (Manassas, VA). HeLa cells were cultured in DMEM supplemented with 10% FBS, 200 mM L-glutamine, and 100 units/mL penicillin and streptomycin. SH-SY5Y cells were cultured in a 1:1 mix of Eagle’s minimum essential medium and F12 Medium supplemented with 10% FBS and 100 units/mL penicillin and streptomycin. Both cell lines were maintained at 37 °C in a humidified 5% CO2 atmosphere. Cells were seeded in either 60 mm tissue culture treated dishes at a density of 5 × 106 cells/dish for biochemical analysis or 35 mm glass-bottomed dishes (Celvis) at a density of 1 × 105 cells/dish for live cell microscopy. All experiments were performed in the appropriate cell growth medium supplemented with 10% FBS.
The pMito-mCherry plasmid construct was produced by cloning mCherry cDNA into the pShooter-MITO vector (Life Technologies, USA), in-frame with the mitochondrial targeting presequence of subunit VIII of human cytochrome c oxidase. pEYFP-Parkin was produced by cloning the wild type, full-length human Parkin cDNA sequence into pEYFP-C1 (Takara, USA), in frame with EYFP. pmCherry-Parkin was subsequently produced by replacing the EYFP coding sequence with the mCherry sequence. pPINK1-EGFP was produced by cloning the wild type, full-length human PINK1 cDNA sequence into pEGFP-N1 (Takara, USA), in frame with EGFP. The pOPTN-EGFP (#27052) expression vector was purchased from Addgene. Cells were transfected with the indicated plasmid constructs using a linear polyethylenimine-based transfection reagent (Polyscience, USA) 24 h after cells were plated.
The primary antibodies used for immunofluorescent staining and immunoblotting-based experiments were as follows: actin (A2066, Sigma), DNA (61,014, Progen Biotechnik GmbH), K48 polyubiquitin (05–1307, EMD Millipore), K63 polyubiquitin (Apu3, 05–1308, EMD Millipore), mono- and polyubiquitinated conjugates monoclonal antibody (FK2, Enzo Lifesciences), Parkin (ab15954, Abcam), phospho-ubiquitin Ser65 (#37642; Cell Signaling), PINK1 (D8G3, Cell Signaling), and TOM20 (D8T4N, Cell Signaling). Detection of primary antibody-binding was performed using anti-mouse Alexa Fluor 555 conjugated antibodies (Life Technologies) or anti-rabbit Alexa Fluor 647 conjugated antibodies (Life Technologies) for immunofluorescent staining experiments and HRP-conjugated anti-rabbit or anti-mouse secondary antibodies (Santa Cruz Biotechnology), as appropriate, for immunoblotting.
Cells for immunoblotting were lysed in RIPA buffer containing both protease and phosphatase inhibitor cocktail. The protein concentration of the samples was determined by BCA assay. After normalizing protein concentrations, the samples were boiled in Laemmli buffer and the proteins resolved by SDS-PAGE prior to western blotting for the indicated proteins.
Time lapse microscopy and immunofluorescence
Live cells were imaged using a Nikon Ti-Eclipse wide field microscope equipped with a CFI Plan 40x oil immersion NA 1.30 objective, Intensilight epifluorescence illuminator, computer-controlled stage (Nikon), CoolSNAP MYO camera (Photometrics), and a full environmental enclosure with CO2, humidity, and temperature control (InVivo Scientific). The microscope was controlled using Elements software (Nikon). Images were acquired at the indicated intervals (typically 3 min). EGFP and EYFP was excited through a 465–495 nm excitation filter and emitted light was detected through a 515–55 nm barrier filter reflected from a 505 nm dichroic mirror. TMRM and mCherry fluorescence was excited through a 535–550 nm excitation filter and emitted light was detected through a 610–675 nm barrier filter reflected from a 565 nm dichroic mirror. For experiments involving the measurement of ΔΨm, cells were incubated with 10 nM TMRM for 15 min at 37 °C prior to imaging. TMRM concentration was kept constant even if CCCP concentration was varied during the experiment.
The live cell experiments depicted in Fig. 7 and the imaging of all immunofluorescently stained samples was conducted using a Zeiss LSM700 confocal laser scanning microscope equipped with a Definite Focus II autofocus system, an Incubator XL S1 full environmental enclosure with CO2, humidity, and temperature control, and a Plan-Apochromat 63x NA 1.4 oil immersion DIC M27 objective (Carl Zeiss). Alexa Fluor 555 and mCherry fluorescence was excited using a 555 nm laser and emitted light was detected through a LP560 filter. Alexa Fluor 647 fluorescence was excited using a 639 nm laser and emitted light was detected through a LP640 filter. EGFP and EYFP fluorescence was excited using a 488 nm laser and emitted light was detected through a SP555 filter. DAPI (NucBlue™ Fixed Cell ReadyProbes™ Reagent; ThermoFisher Scientific) fluorescence was excited using a 405 nm laser.
For the live cell microscopy experiments conducted on the Nikon Ti-Eclipse, 14-bit Nikon nd2 timecourse image files were acquired using Elements software (Nikon) and analysed using Fiji . For each recorded fluorescence channel, the images were thresholded to remove background signal. For experiments involving PINK1-EGFP or TMRM, the fold change in fluorescence throughout the entire mitochondrial network was measured at each time point in individual cells. For experiments involving EYFP-Parkin, the fold change in EYFP fluorescence was measured in a representative region of the mitochondrial network at each time point. Line scan measurements were made using the intensity profile tool in Elements. For the live cell microscopy experiments conducted on the Zeiss LSM700, 16-bit image files were acquired using Zen software (Zeiss). Pearson’s correlation coefficient and line scan intensity profile measurements were also made using the same software. For the experiments described in Fig. 1e-f, which involved the immunofluorescent staining of mtDNA nucleoids, we utilized an anti-DNA antibody that has been previously shown to preferentially stain mtDNA [37, 56]. The number of mtDNA puncta was quantified for 20 cells per condition. Vehicle treated control cells contained an average of 76 (±24) puncta per cell. Overall, > 3000 puncta were counted across 60 cells.
Statistical analyses were performed in GraphPad Prism 7 (GraphPad, USA) using the tests indicated in the figure legends.
We would like to thank Lynn Boyd, Jason Jessen, and members of the Nelson laboratory for helpful discussions and input on the project. We would also like to thank Tiara Griffen for assistance with molecular cloning.
This study was conceived, designed, and coordinated by DEN. JLB performed and analysed the experiments shown in Figs. 1a-d + g-i, 3, 4a-f, 5, 7, and 9g + h. MCS performed and analysed the experiments shown in 1e + f, 4g, 6, 8e + f, and S1. WAR performed and analysed the experiments shown in Figs. 8a-d and 9a-f. DEN and APN performed and analysed the experiments shown in Fig. 2. LCW performed and analysed the experiments shown in 6A. DEN wrote the manuscript. All authors read and approved the final manuscript.
This work was supported by start-up funds from Middle Tennessee State University (MTSU) to DEN. It was also supported by internal grants from MTSU Faculty Research and Creative Activity Committee and the MTSU Foundation (Special Projects) to DEN. JLB and DEN were supported by funds from the MTSU Molecular Biosciences PhD Program. LCW and APN were supported by scholarships from MTSU URECA.
Ethics approval and consent to participate
Consent for publication
The authors declare that they have no conflicts of interest with the contents of this article.
- 23.Ordureau A, Sarraf SA, Duda DM, Heo JM, Jedrychowski MP, Sviderskiy VO, Olszewski JL, Koerber JT, Xie T, Beausoleil SA, et al. Quantitative proteomics reveal a feedforward mechanism for mitochondrial PARKIN translocation and ubiquitin chain synthesis. Mol Cell. 2014;56(3):360–75.CrossRefGoogle Scholar
- 38.Adamson A, Boddington C, Downton P, Rowe W, Bagnall J, Lam C, Maya-Mendoza A, Schmidt L, Harper CV, Spiller DG, et al. Signal transduction controls heterogeneous NF-kappaB dynamics and target gene expression through cytokine-specific refractory states. Nat Commun. 2016;7:12057.CrossRefGoogle Scholar
- 54.Soutar MPM, Kempthorne L, Annuario E, Luft C, Wray S, Ketteler R, Ludtmann MHR, Plun-Favreau H. FBS/BSA media concentration determines CCCP's ability to depolarize mitochondria and activate PINK1-PRKN mitophagy. Autophagy. 2019:1–10.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.