Imaging mitophagy in the fruit fly

ABSTRACT

Loss-of-function mutations in the genes encoding PRKN/parkin and PINK1 cause autosomal recessive Parkinson disease (PD). Seminal work in Drosophila revealed that loss of park/parkin and Pink1 causes prominent mitochondrial pathology in flight muscle and, to a lesser extent, in dopaminergic neurons. Subsequent studies in cultured mammalian cells discovered a crucial role for PRKN/PARK2 and PINK1 in selective macroautophagic removal of mitochondria (mitophagy). However, direct evidence for the existence of a PINK1-PRKN/PARK2-mediated mitophagy pathway in vivo is still scarce. Recently, we engineered Drosophila that express the mitophagy reporter mt-Keima. We demonstrated that mitophagy occurs in flight muscle cells and dopaminergic neurons in vivo and increases with aging. Moreover, this age-dependent rise depends on park and Pink1. Our data also suggested that some aspects of the mitochondrial phenotype of park- and Pink1-deficient flies are independent of the mitophagy defect, and that park and Pink1 may have multiple functions in the regulation of the integrity of these organelles. Here, we discuss implications of these findings as well as possible future applications of the mt-Keima fly model.

The mt-Keima probe, developed by Atsushi Miyawaki and colleagues, is a mitochondrially targeted form of the fluorescent Keima protein. Two characteristics make mt-Keima a well-suited reporter for mitophagy. First, mt-Keima has a bimodal, pH-sensitive excitation spectrum: it undergoes a shift in optimal excitation wavelength when transported from the neutral to slightly alkaline pH of the mitochondrial matrix to the acidic pH of the lysosome. Second, mt-Keima is resistant to degradation by lysosomal proteases. Thus, the ‘acidic’ mt-Keima signal presumably provides a cumulative measure of mitophagy events that have occurred over the cell’s lifetime. In our recent mt-Keima fly study [1], we performed live dual-excitation ratiometric mt-Keima imaging on flight muscle cells and neurons. Strictly speaking, the imaging was ex vivo and not in vivo, because it was performed on freshly dissected tissue. However, the detected ‘acidic’ mt-Keima signal likely reflected mitophagy events that had occurred in vivo, because the imaging was completed within 30 min after dissection.

Previous mt-Keima studies typically confirmed the lysosomal localization of ‘acidic’ mt-Keima puncta by colabeling with lysosomal dyes, such as LysoTracker. However, as the lysosomotropism of these dyes is pH-dependent, such control experiments are somewhat circular, merely showing that mt-Keima puncta with ‘acidic’ fluorescence properties are indeed in an acidic environment. To provide independent morphological confirmation of the lysosomal localization of ‘acidic’ mt-Keima puncta, we performed correlative light and electron microscopy (CLEM). The CLEM images clearly demonstrate that ‘acidic’ mt-Keima puncta colocalize with lysosomes and unambiguously document the occurrence of mitophagy in Drosophila flight muscle.

When comparing flight muscle and dopaminergic neurons between 1-week-old and 4-week-old mt-Keima flies, we found a clear increase in mitophagy signal in the older flies. In 2015, Finkel and colleagues reported findings in a mt-Keima transgenic mouse model. They compared mitophagy in the dentate gyrus of the brain between young (2–3 months) and old (21–23 months) mice and found a strong decrease in mitophagy signal in the older animals. To resolve these seemingly discrepant effects of aging on mitophagy, a more detailed time course of mitophagy throughout life should be established, with additional assessments at time points between 4 and 8 weeks in the mt-Keima fly and between 3 and 21 months in the mt-Keima mouse. It is conceivable that mitophagy increases with aging up to a certain time point and decreases again in very advanced life stages.

Overall, our data suggest that mitophagy is a relatively rare event under physiological conditions in vivo, at least in Drosophila. In 4-week-old flight muscle, only approximately 2% of the total area of cellular mt-Keima signal was ‘acidic’. This fraction is quite low when taking the cumulative nature of the ‘acidic’ signal into account. This suggests that methods for detection of ongoing mitophagy may have only a low chance of capturing such an event at a snapshot in time.

We crossed the mt-Keima flies with Pink1 loss-of-function mutant flies and 2 different park RNAi fly lines, and found that deficiency of Pink1 and park impair mitophagy in flight muscle and dopaminergic neurons in 3- and 4-week-old flies. In the past years the scarcity of evidence for the existence of Pink1-park-mediated mitophagy in vivo led to concerns that Pink1-park-mediated mitophagy might be an artifact of cell culture systems and the use of mitochondrial toxins. Our data clearly argue against this and demonstrate a crucial role for park and Pink1 in mitophagy in Drosophila in vivo.

Interestingly, we did not detect a significant decrease in mitophagy in 1-week-old Pink1- and park-deficient flies, at an age when mitochondrial abnormalities are already detectable. These early mitochondrial abnormalities may be caused by subtle reductions in mitophagy that are not detectable with mt-Keima imaging, and also point to the mitophagy-independent roles of park and Pink1 in mitochondrial homeostasis and function that have been reported before.

A striking visual message conveyed by our CLEM images is the remarkable difference in size between mt-Keima-positive lysosomes and mitochondria. While the size difference between mitochondria and lysosomes is of course not a novel finding, these images strongly suggest that engulfment and lysosomal degradation of entire mitochondria is unlikely due to physical constraints, and that mitophagy is in all likelihood a piecemeal process, at least in Drosophila flight muscle.

The mt-Keima fly will be a convenient model to determine the impact of physical activity, diet and other environmental factors on mitophagy in vivo and to screen small molecules for effects on this pathway. Given the relative ease of genetic manipulation in Drosophila, this model will also provide an efficient tool to assess how this pathway is affected in vivo by gene mutations linked to PD and various other human neurodegenerative diseases. In recent years, amyotrophic lateral sclerosis and frontotemporal dementia have been linked to several genes encoding proteins involved in selective autophagy, raising the possibility that dysregulation of mitophagy may also play a role in these neurodegenerative diseases.

List of abbreviations

ALS:

amyotrophic lateral sclerosis

CLEM:

correlative light and electron microscopy

FTD:

frontotemporal dementia

PD:

Parkinson disease

Pink1:

PTEN-induced putative kinase 1

Acknowledgments

T.C. is a Postdoctoral Fellow and W.V. a Senior Clinical Investigator of the Research Foundation Flanders (FWO). P.V. is a member of the FENS-Kavli Network of Excellence.

Disclosure statement

No potential conflict of interest was reported by the authors.