Autophagy

ISSN: 1554-8627 (Print) 1554-8635 (Online) Journal homepage: https://www.tandfonline.com/loi/kaup20

Autophagy accounts for approximately one-third

of mitochondrial protein turnover and is protein
selective

Evelyn S. Vincow, Ruth E. Thomas, Gennifer E. Merrihew, Nicholas J.

Shulman, Theo K. Bammler, James W. MacDonald, Michael J. MacCoss & Leo
J. Pallanck

To cite this article: Evelyn S. Vincow, Ruth E. Thomas, Gennifer E. Merrihew, Nicholas J.
Shulman, Theo K. Bammler, James W. MacDonald, Michael J. MacCoss & Leo J. Pallanck (2019)
Autophagy accounts for approximately one-third of mitochondrial protein turnover and is protein
selective, Autophagy, 15:9, 1592-1605, DOI: 10.1080/15548627.2019.1586258

To link to this article: https://doi.org/10.1080/15548627.2019.1586258

View supplementary material Accepted author version posted online: 13

Mar 2019.
Published online: 21 Mar 2019.

Submit your article to this journal Article views: 848

View related articles View Crossmark data

Citing articles: 2 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=kaup20
AUTOPHAGY
2019, VOL. 15, NO. 9, 1592–1605
https://doi.org/10.1080/15548627.2019.1586258

RESEARCH PAPER

Autophagy accounts for approximately one-third of mitochondrial protein turnover

and is protein selective
Evelyn S. Vincow a, Ruth E. Thomas a, Gennifer E. Merrihew a, Nicholas J. Shulmana, Theo K. Bammlerb,
James W. MacDonald b, Michael J. MacCossa, and Leo J. Pallancka
a
Department of Genome Sciences, University of Washington, Seattle, WA, USA; bDepartment of Environmental and Occupational Health Sciences,
University of Washington, Seattle, WA, USA

ABSTRACT ARTICLE HISTORY

The destruction of mitochondria through macroautophagy (autophagy) has been recognised as a major Received 19 April 2018
route of mitochondrial protein degradation since its discovery more than 50 years ago, but fundamental Revised 31 January 2019
questions remain unanswered. First, how much mitochondrial protein turnover occurs through auto- Accepted 11 February 2019
phagy? Mitochondrial proteins are also degraded by nonautophagic mechanisms, and the proportion of KEYWORDS
mitochondrial protein turnover that occurs through autophagy is still unknown. Second, does auto- Autophagy; Drosophila;
phagy degrade mitochondrial proteins uniformly or selectively? Autophagy was originally thought to mitochondria; mitophagy;
degrade all mitochondrial proteins at the same rate, but recent work suggests that mitochondrial protein degradation; protein
autophagy may be protein selective. To investigate these questions, we used a proteomics-based turnover; proteomics; stable
approach in the fruit fly Drosophila melanogaster, comparing mitochondrial protein turnover rates in isotope labelling; turnover
autophagy-deficient Atg7 mutants and controls. We found that ~35% of mitochondrial protein turnover rate
occurred via autophagy. Similar analyses using parkin mutants revealed that parkin-dependent mito-
phagy accounted for ~25% of mitochondrial protein turnover, suggesting that most mitochondrial
autophagy specifically eliminates dysfunctional mitochondria. We also found that our results were
incompatible with uniform autophagic turnover of mitochondrial proteins and consistent with protein-
selective autophagy. In particular, the autophagic turnover rates of individual mitochondrial proteins
varied widely, and only a small amount of the variation could be attributed to tissue differences in
mitochondrial composition and autophagy rate. Furthermore, analyses comparing autophagy-deficient
and control human fibroblasts revealed diverse autophagy-dependent turnover rates even in homo-
geneous cells. In summary, our work indicates that autophagy acts selectively on mitochondrial proteins,
and that most mitochondrial protein turnover occurs through non-autophagic processes.
Abbreviations: Atg5: Autophagy-related 5 (Drosophila); ATG5: autophagy related 5 (human); Atg7:
Autophagy-related 7 (Drosophila); ATG7: autophagy related 7 (human); DNA: deoxyribonucleic acid; ER:
endoplasmic reticulum; GFP: green fluorescent protein; MS: mass spectrometry; park: parkin (Drosophila);
Pink1: PTEN-induced putative kinase 1 (Drosophila); PINK1: PTEN-induced kinase 1 (human); PRKN: parkin
RBR E3 ubiquitin protein ligase (human); RNA: ribonucleic acid; SD: standard deviation; Ub: ubiquitin/
ubiquitinated; WT: wild-type; YME1L: YME1 like ATPase (Drosophila); YME1L1: YME1 like 1 ATPase (human)

Introduction (henceforth ‘autophagy’), a process in which mitochondria are

Mitochondria play critical roles in energy production, calcium sequestered and ultimately undergo lysosomal degradation [14].
homeostasis, metabolite synthesis and apoptosis [1-4]. They are Of the various mitochondrial protein-degradation pro-
also the major cellular source of reactive oxygen species, which cesses, the most extensively studied is autophagy, first identi-
damage mitochondrial DNA, lipids and proteins [3,5]. It is fied in the 1960s. The pioneering work of Fletcher and Sanadi,
therefore unsurprising that progressive accumulation of which found essentially identical turnover rates for three types
damaged mitochondria has been implicated in aging and in of mitochondrial protein as well as mitochondrial lipids [15],
common diseases of the elderly [6–8]. Mitochondria have concluded that mitochondria were ‘turning over as an entity’;
a large repertoire of quality control mechanisms to oppose this soon afterwards, electron micrographs showing mitochondria
deterioration, including multiple pathways for the removal of inside lysosomes provided a plausible mechanism for destruc-
damaged mitochondrial components. Individual mitochondrial tion of whole organelles [16]. Later work revealed that auto-
proteins can be selectively degraded by mitochondrial resident phagy is a highly regulated process involving a complex net-
proteases [9] or the proteasome [10–12], or transported to the work of autophagy-related (ATG) gene factors [17], and
lysosome in mitochondria-derived vesicles [13]. Alternatively, demonstrated multiple pathways leading to autophagic degra-
entire mitochondria can be destroyed via macroautophagy dation of mitochondria [18,19]. In particular, selective

CONTACT Leo J. Pallanck [email protected] University of Washington, W.H. Foege Building S443E, Box 355065, 3720 15th Avenue NE, Seattle, WA
98195-5065, USA
Supplemental data for this article can be accessed here.
© 2019 Informa UK Limited, trading as Taylor & Francis Group
 AUTOPHAGY 1593

destruction of damaged mitochondria was identified and range could be explained by differences in mitochondrial pro-
termed ‘mitophagy’ [20,21]. A more mechanistic understand- tein expression and mitochondrial autophagy rate in the tissues
ing of mitophagy emerged through study of the neurodegen- analysed, and autophagic turnover rates were equally diverse
eration-associated proteins PINK1 (PTEN induced kinase 1) when calculated by comparing protein turnover in homoge-
and PRKN (parkin), which together target dysfunctional neous ATG7−/− or ATG5−/− human fibroblasts with WT cells.
mitochondria for autophagic degradation [22–24], and work Taken together, our findings show that autophagy degrades
in Drosophila demonstrated that this mitophagy pathway is mitochondrial proteins at unequal rates, and that most mito-
required for normal mitochondrial protein turnover chondrial protein turnover in both Drosophila and vertebrates
in vivo [25]. occurs through non-autophagic processes.
While the above findings demonstrate the importance of
autophagy in mitochondrial quality control, there is also evi-
dence that non-autophagic mechanisms contribute substan- Results
tially to turnover of mitochondrial proteins. Although some We measured the contribution of autophagy to mitochondrial
studies that followed Fletcher and Sanadi supported their view protein degradation in vivo by comparing turnover rates of
that mitochondria are degraded as units [26,27], others mitochondrial proteins in WT flies to their turnover rates in
instead found different turnover rates for individual mito- Atg7 null mutants [43]. Atg7 encodes an evolutionarily con-
chondrial components [28–30], and later in vivo proteomic served E1-like enzyme required for autophagic vesicle forma-
studies showed that mitochondrial protein turnover rates are tion [14]. In previous work, we demonstrated markedly
in fact highly diverse [25,31–37]. These wide ranges of turn- impaired mitochondrial protein turnover in Drosophila Atg7
over rates, sometimes spanning more than two orders of null mutants [25], and we now used these well-characterised
magnitude, are consistent with substantial mitochondrial pro- autophagy-deficient flies to quantify the contribution of
tein degradation through non-autophagic mechanisms. autophagy to mitochondrial protein turnover. Briefly, we
Accumulation of mitochondrial proteins after ablation of the used stable isotope labelling followed by mass spectrometry
mitochondrial protease Lon [38] or inhibition of the protea- to measure the rates at which unlabelled proteins were
some [12] further suggests that non-autophagic mechanisms degraded and replaced by labelled proteins. The technique
degrade considerable amounts of mitochondrial protein. and other analyses of the data have been described [25].
Work to date thus indicates that autophagy accounts for
some, but not all, mitochondrial protein turnover; however,
the actual proportion of turnover that occurs via autophagy is Calculating the contribution of autophagy to
unknown. mitochondrial protein turnover
In the last few years, new findings have also begun to blur
the distinction between lysosomal destruction of whole mito- We identified 186 mitochondrial proteins from Drosophila
chondria and targeted degradation of individual mitochon- heads that met quality criteria in both Atg7 mutants and
drial proteins. Work in yeast has demonstrated stress-induced controls (Dataset S1; see Materials and Methods). For each
autophagic degradation of selected mitochondrial proteins protein, we determined the autophagy-dependent turnover
[39,40], and experiments using mammalian cultured cells rate (the fraction of the protein’s total abundance degraded
have suggested that related phenomena may exist in higher via autophagy per unit time), which we then used to calculate
eukaryotes [41,42]. However, whether mitochondrial auto- the contribution of autophagy to the protein’s overall
phagy normally degrades individual mitochondrial proteins degradation.
at different rates in intact metazoans is unknown. The turnover rate of any mitochondrial protein (M) is the
To answer these unresolved questions about the nature of sum of the mitochondrial autophagy rate and all non-
mitochondrial autophagy, we used an approach based on pro- autophagic turnover rates for that protein. Non-autophagic
teomic measurement of protein turnover in the fruit fly turnover includes degradation by mitochondrial proteases, the
Drosophila melanogaster. We calculated the percentage contri- ubiquitin-proteasome system, mitochondria-derived vesicles,
bution of autophagy by comparing mitochondrial protein turn- and any other non-autophagic degradation processes. We
over rates in autophagy-deficient Atg7 (Autophagy-related 7) calculated autophagy-dependent turnover rate by subtracting
null mutant flies to those in normal flies, and the percentage the turnover rate of protein M in Atg7 mutants (in which
contribution of parkin-dependent mitophagy using parkin turnover occurs solely through non-autophagic mechanisms)
(park)null mutants. Approximately 35% of all mitochondrial from the turnover rate of M in WT flies (which includes
protein turnover occurred through autophagy and 25% turnover via both autophagy and non-autophagic mechan-
through parkin-dependent mitophagy, consistent with the isms), as follows:
idea that autophagy primarily degrades mitochondria that autophagy-dependent turnover rate of M ¼ ðWT rate of MÞ
have become dysfunctional. We then investigated whether
 ðAtg7 mutant rate of MÞ
autophagy had differential effects on individual mitochondrial
proteins. We modelled uniform, nonselective mitochondrial We then calculated the ratio of the autophagy-dependent
protein turnover and found that our data were incompatible turnover rate to the protein’s overall degradation rate and
with the model’s predictions. Most importantly, the calculated expressed the result as a percentage. We called this measure
autophagic turnover rates of individual mitochondrial proteins of autophagy’s contribution to degradation percent autophagic
ranged over two orders of magnitude. Only a small part of that turnover.
1594 E. S. VINCOW ET AL.

and the mean percent autophagic turnover for the ribosome

was 63.9% ± 5.4% (p < 3.4 × 10−41 vs. mitochondrial proteins
by Student t test; Figure 1(a-b), Dataset S2). Thus, while the
majority of ribosomal protein degradation is autophagic,
autophagy accounts for approximately one-third of mitochon-
drial protein turnover.

Most mitochondrial autophagy requires parkin

A form of mitochondrial autophagy that has been intensively
studied in recent years is parkin-dependent mitophagy,
a process that detects dysfunctional mitochondria and marks
them for degradation [23]. We previously observed that mito-
chondrial protein turnover was impaired in parkin mutants,
though not as profoundly as in Atg7 mutants [25]. To compare
the influence of parkin and Atg7 more quantitatively, we cal-
culated the contribution of parkin to mitochondrial protein
turnover, comparing turnover rates in parkin mutants and
WT flies (Figure 1(c)). We included only the 168 mitochondrial
proteins that met quality standards in all three relevant geno-
types (parkin, Atg7, and WT; Dataset S1). The mean percent
parkin-dependent turnover in fly heads was 29.5% ± 12.6%
(Figure 1(d)). However, because parkin also takes part in non-
autophagic forms of mitochondrial protein turnover [44–47],
including production of mitochondria-derived vesicles, this
method of calculating parkin-dependent turnover could poten-
tially overestimate its contribution to mitochondrial autophagy.
We therefore recalculated mean percent parkin-dependent
turnover after excluding likely targets of non-autophagic
parkin-dependent degradation (Table S1; see Materials and
Methods). Mean percent parkin-dependent turnover for the
remaining proteins (n = 115) was 25.7% ± 11.8% (Figure 1
(d)). This adjusted mean parkin-dependent turnover value for
Figure 1. Autophagy accounts for approximately one-third of mitochondrial mitochondrial proteins was ~70% of the overall mean autop-
protein turnover. (a) Percent autophagic turnover of individual mitochondrial
proteins and cytosolic ribosomal proteins in Drosophila heads (n = 186 mito-
hagic turnover value derived from Atg7 mutants. Our findings
chondrial and 52 ribosomal proteins). (b) Mean percent autophagic turnover of thus suggest that a large proportion of autophagic mitochon-
mitochondrial and ribosomal proteins (significantly different by Student t test, p drial protein degradation requires parkin, and that quality
< 3.4 × 10−41). Error bars represent SD. (c) Contribution of parkin-dependent
mitophagy to turnover of individual mitochondrial proteins. Unadjusted percent
control surveillance is the major driver of autophagic mito-
parkin-dependent turnover was calculated using 168 mitochondrial proteins chondrial protein degradation.
common to the parkin, Atg7, and WT datasets (Dataset S1). Adjusted percent
parkin-dependent turnover was calculated after excluding 53 putative targets of
nonautophagic parkin-dependent degradation (see Materials and Methods), and
is thus based on 115 proteins. (d) Mean percent parkin-dependent turnover for Mitochondrial autophagy is protein selective
the mitochondrial proteins in panel C. Error bars represent SD.
We next considered the question of whether autophagy
degrades mitochondrial proteins uniformly or selectively.
percent autophagic turnover of M ¼ The idea that autophagy degrades mitochondrial proteins
autophagy-dependent turnover rate of M uniformly dates back to Fletcher and Sanadi’s concept of
 100 mitochondria ‘turning over as an entity’, and we therefore
WT turnover rate of M
called it the classical model of mitochondrial autophagy.
Mitochondrial proteins in fly heads showed a substantial According to the classical model, when a mitochondrion
range of percent autophagic turnover values, but the vast undergoes autophagy, all proteins are degraded simulta-
majority (89%) had less than 50% turnover through auto- neously and equally. As described above, however, recent
phagy (Figure 1(a), Dataset S1). Mean percent autophagic work in yeast and cell culture challenges this idea [39–42].
turnover for all mitochondrial proteins was 36.0% (± 11.7%; To determine whether basal autophagy is protein selective in
Figure 1(b)). For comparison, we calculated percent autopha- an intact metazoan, we compared our Atg7 mutant findings to
gic turnover for proteins of the cytosolic ribosome (n = 52), the predictions of the classical model. The key assumption of
another target of selective autophagy. In striking contrast to the classical model is the idea that mitochondrial proteins all
mitochondrial proteins, 96% of ribosomal proteins undergo autophagic turnover at the same rate (Figure 2(a)).
had percent autophagic turnover values greater than 50%, By contrast, the rate of turnover via other degradation
 AUTOPHAGY 1595

Figure 2. Mitochondrial autophagy is protein selective. (a) Classical model of mitochondrial protein turnover. The total turnover rate of each mitochondrial
protein is the sum of its turnover rate via autophagy (presumed to be the same for all proteins) and its turnover rate via non-autophagic mechanisms
(different from protein to protein). The turnover rate via non-autophagic mechanisms is the sum of turnover by mitochondrial proteases, the ubiquitin-
proteasome system, mitochondria-derived vesicles and any other non-autophagic degradation processes. Note that long-lived proteins are those with low total
turnover rates. (b) Autophagy-dependent turnover rates (h−1) for individual mitochondrial and ribosomal proteins. (c) Mitochondrial proteins have a large
range of autophagy-dependent turnover rates, but cytosolic ribosomal proteins do not. The 5% trimmed ranges (95th/5th percentile values) are displayed; the
full ranges were 253.4-fold for mitochondrial proteins and 2.0-fold for ribosomal proteins. (d) Autophagy-dependent turnover rates of individual mitochondrial
proteins vary substantially even within a specific region of the mitochondrion, including matrix (5% trimmed range 17.2-fold, n = 95) and inner membrane
(IM; 5% trimmed range 24.7-fold, n = 72). (e) Percent autophagic turnover correlates positively rather than negatively with WT turnover rate (h−1).

pathways differs from protein to protein, and thus total turn- to data from another target of autophagy, the cytosolic ribo-
over rates vary. some. The autophagy-dependent turnover rates of ribosomal
The classical model makes two testable predictions. First, proteins fell within a relatively narrow range (full range 2.0-fold
the model predicts that the calculated autophagy-dependent and 5% trimmed range 1.6-fold; Figure 2(b-c)), indicating that
turnover rates of mitochondrial proteins should fall within the large range of rates for mitochondrial proteins reflected
a narrow range because they are all approximations of a single a biological phenomenon rather than experimental noise.
true mitochondrial autophagy rate. Second, the model pre- To test the second prediction, we plotted percent autopha-
dicts that percent autophagic turnover for an individual mito- gic turnover against WT turnover rate for each mitochondrial
chondrial protein should be inversely related to its overall WT protein. The correlation between percent autophagic turnover
turnover rate, because total turnover rate increases as the and WT turnover rate was, surprisingly, significantly positive
contribution of nonautophagic degradation becomes larger. rather than negative (Figure 2(e)). Proteins with low overall
To test the first prediction, we compared autophagy- turnover rates (long-lived proteins), predicted to have
dependent turnover rates for all mitochondrial proteins. The high percent autophagic turnover, actually had very modest
autophagy-dependent turnover rates of mitochondrial proteins percentages of turnover through autophagy. In fact, the mito-
spanned a 253-fold range (Figure 2(b)); when we considered chondrial proteins with the lowest WT turnover rates (n =
only the rates from the 5th to the 95th percentile (the 5% 19, the lowest 10% of the dataset) all had percent autophagic
trimmed range) to limit the impact of extreme values, the turnover values under 50%, and their mean percent autopha-
range was still ~18-fold (Figure 2(c)). Even groups of proteins gic turnover was lower than the mean for the remaining
localising to a particular region of the mitochondrion (e.g. inner proteins in the dataset (28.5% vs. 36.8%; p < 0.005). Our find-
membrane proteins) had large ranges of autophagy-dependent ings thus contradict both of the classical model’s predictions
turnover rates (Figure 2(d)); the range of autophagy-dependent and suggest that, instead of degrading all mitochondrial pro-
turnover rates for complex I proteins alone was 4.1-fold (Figure teins equally, mitochondrial autophagy has differential effects
S1). To put the findings in perspective, we again compared them on individual mitochondrial proteins.
1596 E. S. VINCOW ET AL.

The mitochondrial autophagy findings are not

explained by genetic compensation
An alternative explanation for both the moderate percen-
tage contribution of autophagy and the wide-ranging
autophagy-dependent turnover rates of mitochondrial pro-
teins is compensatory upregulation of non-autophagic
degradation. Induction of a compensatory degradation
pathway in Atg7 mutants would decrease the effects of
Atg7 inactivation on mitochondrial protein turnover; if
the compensatory pathway were protein selective, its induc-
tion might also increase the range of autophagy-dependent
turnover rates. While compensation probably occurs to
some extent in all mutants, we do not believe that it
accounts for our major findings, based on the following
evidence. First, we looked for upregulation of compensatory
turnover pathways, using both RNA and protein data.
Microarray analysis of the Atg7 mutants showed no change
in expression of genes encoding proteasome subunits, mito-
chondrial proteases, mitochondrial vesicle factors, auto-
phagy factors other than Atg7, or specific mitophagy factors
(Table S2, Dataset S3). We also measured the abundance of
proteins in these same categories, and found significant
changes in only 2 of 58 proteins (~3%; Table S2, Dataset
S4). Second, we investigated whether turnover was less
autophagy dependent in proteins identified as substrates
of non-autophagic turnover processes. We hypothesised
that normal substrates of a given turnover pathway would
be most affected by compensatory induction of that path- Figure 3. The effects of Atg7 ablation are not explained by induction of potential
way. However, orthologues of the mitochondrial proteins compensatory turnover mechanisms. (a) Autophagy-dependent turnover rates
(h−1) of fly orthologues of mitochondrial proteins with (n = 43) and without
identified as proteasome substrates by Wagner et al. [48] (n = 143) evidence of proteasomal turnover (ubiquitinated [Ub] sites) [48]. Red
did not differ from other mitochondrial proteins in percent lines indicate means. (b) Percent autophagic turnover of fly orthologues of
autophagic turnover or autophagy-dependent turnover rate mitochondrial proteins with and without ubiquitinated sites, as in panel A. (c)
Autophagy-dependent turnover rates (h−1) of individual mitochondrial proteins
(Figure 3(a-b)); neither did proteins bearing degron with (n = 24) and without (n = 162) degrons for mitochondrial proteases Lon
sequences for the mitochondrial proteases YME1L (YME1 and YME1L [49,50]. (d) Percent autophagic turnover of mitochondrial proteins
like ATPase) and Lon [49,50] (Figure 3(c-d)). We thus with and without protease degrons (as in panel C). Comparisons of means in
panels A through D are all nonsignificant by Student t test.
found no evidence that our data are explained by genetic
compensation.

(brain, fat body and eye) were available from the public data
Tissue differences in protein expression and repository FlyAtlas [54]. For genes encoding the mitochon-
mitochondrial autophagy rate contribute to the drial proteins in our fly head data, we found that normalised
range of autophagy-dependent turnover rates expression of individual genes in a given tissue varied by as
Another possible explanation for the existence of diverse auto- much as two orders of magnitude (Figure S3, Dataset S1). We
phagy-dependent turnover rates in the fly head samples arises then tested whether this variation in tissue expression corre-
from the fact that they contain multiple tissues. Tissue differ- lated with autophagy-dependent turnover rate. Autophagy-
ences in mitochondrial protein expression and mitochondrial dependent turnover rate correlated negatively with brain
autophagy rate could interact to produce diverse autophagy- expression and positively with fat body expression, and had
dependent turnover rates for individual mitochondrial proteins no relationship with eye expression (Figure 4(a-c)). By con-
(see Figure S2 for a detailed explanation). As there is evidence trast, in proteins from other organelles targeted by autophagy
in vertebrates for both differential tissue expression of mito- (ribosomes, endoplasmic reticulum [ER], and peroxisomes),
chondrial proteins [51] and tissue-specific mitochondrial we found no significant relationship between autophagy-
autophagy rates [52,53], we considered the possibility that the dependent turnover rate and expression in any of the head
range of autophagy-dependent turnover rates could be tissues (Figure 4(d–f)); the findings in mitochondrial proteins
explained in terms of tissue differences within Drosophila therefore appeared to reflect tissue differences in mitochon-
heads. drial autophagy rate rather than in general autophagy. These
We first determined whether fly heads had significant results are consistent with the idea that heterogeneous tissue
differences in mitochondrial protein composition between expression of proteins could explain some observed variation
tissues. Gene expression values for three fly head tissues in autophagy-dependent turnover rates (Figure S2), and they
 AUTOPHAGY 1597

Figure 4. Autophagy-dependent turnover rates of mitochondrial proteins correlate with their tissue expression. Normalised gene expression (tissue enrichment)
values for brain, fat body, and eye were obtained from FlyAtlas [102] and were log10 transformed to normalise skewed distributions. Autophagy-dependent turnover
rates (h−1) were calculated for all mitochondrial proteins with available FlyAtlas data (n = 177), and for proteins from other organelles degraded by autophagy
(ribosomes, ER, and peroxisomes). Autophagy-dependent turnover rates of mitochondrial proteins correlate negatively with brain expression (a) and positively with
fat body expression (b). There is no significant relationship with eye expression (c). Autophagy-dependent turnover rates of other organellar proteins (ribosomes, ER,
and peroxisomes) do not correlate significantly with expression in brain (d), fat body (e), or eye (f); n = 82 other organellar proteins with FlyAtlas data for brain, 79
for fat body, 80 for eye.

also suggest that mitochondrial autophagy rate is substantially mitochondrial protein turnover in a single cell type. We compared
lower in brain than in fat body. mitochondrial protein turnover rates in ATG7−/− and ATG5−/−
While tissue differences in mitochondrial autophagy rate engineered human fibroblasts (originally described by Zhang et al.
contributed to the varying autophagy-dependent turnover [55]) to the corresponding rates in WT cells, and
rates of mitochondrial proteins, they did not provide calculated percent autophagic turnover as described above
a complete explanation for the wide range of rates. (Figure 5(a), Dataset S5). The mean percent autophagic turnover
Correlations between mitochondrial proteins’ autophagy- was 21.8% ± 16.7% when calculated from ATG7−/− cells (n = 211
dependent turnover rates and their expression in a given proteins), and 14.4% ± 10.3% when calculated from ATG5−/− cells
tissue, even when statistically significant, accounted for (196 proteins), consistent with our in vivo finding that most
a minority of the total variance (R2 = 0.01 to 0.26). mitochondrial protein turnover occurs through non-autophagic
Mitochondrial autophagy rate differences in tissues not docu- mechanisms. Also, in both cell lines, the autophagy-dependent
mented in FlyAtlas might account for some of the remaining turnover rates of mitochondrial proteins spanned ranges compar-
variation; however, to explain the observed range of auto- able to those seen in fly heads. In ATG7−/− fibroblasts, the full
phagy-dependent turnover rates, such differences would range of rates was 1659-fold and the 5% trimmed range 11-fold; in
have to be vast (Figure S4). ATG5−/−, the full and trimmed ranges were 1079-fold and 28-fold,
respectively (Figure 5(b-c)). As in previous analyses, we compared
the mitochondrial protein findings to data from another target of
autophagy [56]. Because ribosomal turnover showed little depen-
Autophagy-dependent turnover rates of
dence on autophagy in the fibroblasts, we compared mitochon-
mitochondrial proteins are diverse in a single cell
drial proteins from fibroblasts and fly heads to proteins of the
type
proteasome, which also undergoes autophagic degradation [57].
To test more definitively whether our findings were explained by The range of autophagy-dependent turnover rates was much
tissue differences in mitochondrial autophagy rate and protein larger for mitochondrial than for proteasomal proteins in both
composition, we measured the contribution of autophagy to ATG7−/− and ATG5−/− fibroblasts, and in Atg7 null fly heads as
1598 E. S. VINCOW ET AL.

Figure 5. Findings in autophagy-deficient cultured cells support the protein selectivity of mitochondrial autophagy. (a) Mean percent autophagic turnover calculated
from ATG7−/− (n = 211 proteins) and ATG5−/− (n = 196) vs. WT human fibroblasts. Error bars represent SD. (b) Autophagy-dependent turnover rates of individual
mitochondrial proteins calculated from ATG7−/− and ATG5−/− fibroblasts (ranges 1659-fold and 1079-fold, respectively). (c) The 5% trimmed ranges (95th/5th
percentile) of autophagy-dependent turnover rates for mitochondrial and proteasomal proteins in ATG7−/− fly heads, ATG7−/− fibroblasts, and ATG5−/− fibroblasts
(n = 16 proteasomal proteins in heads and 33 in both ATG7−/− and ATG5−/− fibroblasts). (d and e) Percent autophagic turnover correlates positively with WT
turnover rate (h−1) for mitochondrial proteins in ATG7−/− fibroblasts (d). ATG5−/− fibroblasts (e) show a trend-level positive correlation between percent autophagic
turnover and WT turnover rate as well.

well (Figure 5(c)). In addition, the percent autophagic turnover presence of mitochondrial protease degron sequences
values of individual mitochondrial proteins did not show an (Figure 6(c-d), Figure S5C–D). The simplest explanation
inverse relationship with WT turnover rate (Figure 5(d-e)), again for the observed diversity of autophagy-dependent turnover
contradicting the classical model of mitochondrial autophagy. In rates in both fly heads and human cells is that some
fact, ATG7−/− fibroblasts, like fly heads, showed a significant posi- mitochondrial proteins undergo more autophagic degrada-
tive correlation between WT turnover rate and percent autophagic tion than others.
turnover (Figure 5(d)), and ATG5−/− fibroblasts showed a trend in
the same direction (Figure 5(e)).
Discussion
As before, we looked for evidence that our findings
could be explained by genetic compensation. Measurement In this work, we addressed two fundamental questions about
of protein abundance in the ATG5−/− fibroblasts revealed mitochondrial autophagy. First, what proportion of mitochon-
modestly increased abundance of most proteasome subu- drial protein degradation occurs through autophagy? Second, is
nits, and proteasome activity was ~25% above WT [55]. the process significantly protein selective? We found that
While the increased proteasome activity might explain the autophagy accounts for approximately one-third of mitochon-
lower mean percent autophagic turnover values in fibro- drial protein turnover in vivo, and that individual mitochon-
blasts compared to fly heads (Figures 1(b), 5(a)), it did not drial proteins are degraded by autophagy at highly divergent
appear to explain the range of autophagy-dependent turn- rates.
over rates. We found no differences in percent autophagic Our finding that autophagy was not the primary mechan-
turnover or auto-phagy-dependent turnover rate in mito- ism of mitochondrial protein turnover is consistent with pre-
chondrial proteins identified as proteasome targets [48] vious work suggesting large-scale mitochondrial protein
(proteins with ubiquitinated sites; Figure 6(a-b), Figure turnover by non-autophagic processes; for instance, protea-
S5A–B). Likewise, the abundance of Lon protease was some inhibition in mammalian cells causes dramatic accumu-
17% higher in ATG5−/− mutants compared to WT, but lation of mitochondrial proteins [12], and yeast mitochondrial
there were no significant associations of percent autophagic proteases can degrade up to ~5% of total mitochondrial pro-
turnover or autophagy-dependent turnover rate with the tein per hour [58]. Studies of mitochondrial DNA turnover
 AUTOPHAGY 1599

Figure 6. The protein-selective effects of autophagy in ATG7−/− and ATG5−/− cells are not explained by genetic compensation. (a) Autophagy-dependent turnover
rates (h−1) of mitochondrial proteins from ATG7−/− fibroblasts with and without evidence of proteasomal turnover (ubiquitinated [Ub] sites) [48]. Red lines indicate
means; n = 44 proteins with ubiquitinated sites, 167 without. (b) Autophagy-dependent turnover rates (h−1) of mitochondrial proteins from ATG5−/− fibroblasts with
(n = 45) and without (n = 151) ubiquitinated sites, as in panel A. (c) Autophagy-dependent turnover rates (h−1) of individual mitochondrial proteins from ATG7−/−
fibroblasts with (n = 31) and without (n = 180) degrons for mitochondrial proteases Lon and YME1L1. (d) Autophagy-dependent turnover rates (h−1) of individual
mitochondrial proteins from ATG5−/− fibroblasts with and without protease degrons (n = 30 proteins with degrons, 166 without). All comparisons of means in panels
A through D are nonsignificant by Student t test.

also suggest that autophagic degradation of mitochondria is [39,40], and cell culture studies have indicated that protein-
not a high-frequency event [59,60]. These findings corrobo- selective mitochondrial autophagy may also occur in verte-
rate the general accuracy of our measurement. It is entirely brates [41,42]; we now demonstrate that basal autophagy
possible, however, that autophagy makes larger contributions degrades mitochondrial proteins at widely differing rates in
to mitochondrial protein turnover during other life stages or an intact metazoan. Multiple factors could contribute to the
under different metabolic conditions. Our technique did have protein selectivity of mitochondrial autophagy. For instance,
the potential to underestimate autophagic turnover if alter- among mitochondria within a single cell, the probability of
native autophagy pathways [61,62] contributed significantly to autophagic degradation may correspond with subcellular var-
mitochondrial protein turnover. However, multiple reports iation in protein composition. Mitochondria in different
indicate that strong inhibition of the Atg7-dependent pathway regions of a cell (e.g. perinuclear vs. peripheral) show differ-
causes profound to total blockade of mitochondrial autophagy ences on multiple measures, including protein content, mem-
[22,42,52,63–65]. brane potential, respiration rate, and reactive oxygen species
Selective autophagic degradation of mitochondrial proteins generation [66–69]. The probability of mitochondrial auto-
has been clearly demonstrated in yeast under stress conditions phagy could likewise differ between subcellular regions,
1600 E. S. VINCOW ET AL.

whether because of environmental factors (local availability of large contribution of parkin-dependent mitophagy under-
autophagic machinery [70]) or intrinsic characteristics (func- scores the idea that autophagy degrades mitochondria primar-
tional specialisation leading to high reactive oxygen species ily when they have become severely dysfunctional. Autophagy
production [68]). Subcellular mitochondrial heterogeneity can thus be seen as a cell’s crucial last line of defence against
could explain why the classical model of mitochondrial mitochondrial damage. Nevertheless, our findings raise new
auto-phagy underestimated percent autophagic turnover in questions about the widely proposed strategy of enhancing
mitochondrial proteins with high WT turnover rates (short- autophagy to treat neurodegenerative disease [78]. Would
lived proteins; Figure 2(e)). These proteins included many drug-stimulated mitochondrial autophagy retain its protein
enzymes involved in fatty acid and amino acid metabolism selectivity? If not, would the change be detrimental? Our
(Dataset S1), which may have been concentrated in work also suggests that enhancing non-autophagic mitochon-
a functionally specialised mitochondrial population with drial protein turnover might be an even better therapeutic
a high probability of autophagy. approach. A deeper understanding of non-autophagic mito-
Differences in mitochondrial protein composition within chondrial protein turnover mechanisms, and their individual
a cell may also arise dynamically as part of quality control. roles in mitochondrial health, is thus an important goal for
Mitochondrial fission has been shown to produce daughter future research.
mitochondria with significant differences in membrane
potential and probability of autophagic degradation, and pos-
sibly even in physical structure [71,72]. These differences Materials and methods
could reflect active or passive damage-based sorting of Drosophila strains and culture
damaged proteins into areas of the mitochondrial network
that will ultimately be isolated and degraded [73–75]. While Fly stocks were maintained on standard cornmeal-molasses
the sorting process itself remains largely speculative, there is food at 25°C. The Atg7d4, Atg7d77, park25, and Pink1rv alleles,
clear evidence that cells can target concentrations of abnormal as well as the UAS-Pink1#2 strain, have been previously
mitochondrial protein; such damage-enriched areas recruit described [43,79,80]. Other strains and alleles were originally
quality control and fission factors to facilitate their own obtained from the Bloomington Stock Center. Atg7 null
removal [76,77]. The net result of these selective sorting and mutants were Atg7d4/Atg7d77 transheterozygotes. The full gen-
degradation processes could be high autophagic turnover otype of parkin mutants was If/CyO; park25/park25. The WT
rates for proteins prone to unfolding or damage. controls were four separate groups of healthy flies with inten-
Differences in autophagic turnover between mitochondrial tionally diverse genetic backgrounds (see protein turnover
proteins could also reflect sorting based on a protein’s identity rate calculations section).
or characteristics. This phenomenon has been demonstrated
in aging yeast, which generate mitochondria-derived com- In vivo stable isotope labelling of flies
partments containing only highly defined subsets of proteins
[39]. If identity-based sorting occurs in other contexts, pro- [5,5,5 – 2H3] leucine (D3-leucine; 99 atom % deuterium) was
teins extremely prone to oxidative damage or unfolding might obtained from Isotec/Sigma-Aldrich (486825). Synthetic com-
be actively directed to autophagy-bound mitochondria at high plete medium without leucine (C-Leu) [81] was supplemented
rates, a preventive approach that could produce more effective with glucose and 60 mg/L D3-leucine. A strain of
quality control than damage sensing alone. Conversely, pro- Saccharomyces cerevisiae auxotrophic for leucine (BB14-3A,
teins that are energetically costly to replace could be sorted Brewer Lab, University of Washington [82]) was grown to
into areas of the mitochondrial network with low probabilities saturation at 30°C, then spun down, flash-frozen in liquid
of degradation. This type of sorting would explain our find- nitrogen, lyophilised, and stored at −80°C until needed.
ings on mitochondrial proteins with low WT turnover rates; Groups of 10 to 50 male flies were selected on the day of
their combination of low total turnover rates and eclosion and housed in perforated plastic flasks, where they
modest percent autophagic turnover is consistent with active received plain yeast paste for 24 h. They were then provided
exclusion from autophagy-bound mitochondria. Many ques- with D3-leucine–labelled yeast paste, which was replaced
tions remain, particularly as to the physical mechanisms of every 2–3 days, and were maintained in humidified containers
protein redistribution, but the advantages of the process are at 25°C. After 120 h or 240 h of labeling (the shortest time
clear. Active redistribution of mitochondrial proteins would points that allowed adequate labeling of mitochondrial pro-
lend additional flexibility and precision to quality control, and teins), flies were flash-frozen in liquid nitrogen. Three biolo-
would facilitate quick adjustments to changing conditions. gical replicates (50–115 heads each) were obtained for each
In summary, our findings suggest that autophagy’s role in genotype and time point.
mitochondrial quality control is different, but no less vital,
than previously envisioned. First, while the contribution of
Sample preparation
autophagy was less than that of all other degradation path-
ways combined, autophagy may still degrade more mitochon- Frozen flies were vortexed to remove heads, and the isolated
drial protein than any other single mechanism. Second, the heads were homogenised in 0.1% RapiGest (Waters
most important function of mitochondrial autophagy may not Corporation, 186001861) solution in 50 mM ammonium
be large-scale protein degradation, but swift removal of unsal- bicarbonate (Fisher Scientific, A643) using a 0.2-mL
vageable parts of the mitochondrial network. The relatively Wheaton micro tissue grinder (Fisher Scientific, 08-414-
 AUTOPHAGY 1601

15B). Homogenates were centrifuged at 4°C at 1600 × g for essentially identical turnover rates, occasionally they displayed
10 min, and then at 6500 × g for 10 min, to remove debris and significant differences. Each isoform group was therefore
nuclei. The supernatants were then boiled for 7 min and analysed as a separate protein.
incubated at 60°C for 30 min with dithiothreitol (Sigma- We excluded proteins with excessive inter-replicate varia-
Aldrich, Inc., D0632) at a final concentration of 5 mM. bility of turnover rates. To do this, we calculated the turnover
Iodoacetamide (Sigma-Aldrich, Inc., I1149) was added to rate separately for each biological replicate and determined
a final concentration of 15 mM, and the samples were incu- the coefficient of variation across replicates. Proteins with
bated at room temperature in the dark for 30 min. Trypsin coefficient of variation ≥ 0.25 were excluded from analysis.
(Fisher Scientific, PR-V5111) was added at a ratio of 1 μg Proteins were analysed only if they met inclusion criteria in
trypsin per 50 μg protein, and the samples were incubated for both mutants and WT controls. One ribosomal protein and
1 h at 37°C with shaking. RapiGest was hydrolysed by adding one ER protein had faster turnover in mutants than in con-
hydrochloric acid (Fisher Scientific, SA56-1) to a final con- trols (negative autophagy-dependent turnover rates) and were
centration of 200 mM, followed by incubation at 37°C with outliers by the Tukey method [101]; they were excluded from
shaking for 45 min. The samples were then centrifuged for analysis.
10 min at 4°C at 20,000 × g, and the supernatant was In previous work, we had compared Atg7 and parkin null
collected. mutants to their respective heterozygotes. However, we later
found that both Atg7 and parkin heterozygotes had mild but
significant slowing of mitochondrial protein turnover com-
Mass spectrometry
pared to WT flies. The previously reported effects of Atg7 and
Liquid chromatography and mass spectrometry were per- parkin null mutations on turnover thus represent underesti-
formed as previously described [25]. High-resolution MS mates, and we substituted the current WT dataset as a more
data were processed with BullsEye to optimise precursor appropriate control for both mutants. The WT dataset is
mass information [83]. The MS/MS output was searched a composite derived from four separate groups of healthy
using SEQUEST [84], with a differential modification of flies (w1118, Pink1rv, CyOActGFP/+, and a mixture of CyO/
3.0188325 Da for leucine and a static modification of Hsp70-GAL4 and CyO/UAS-Pink1#2). This approach maxi-
57.021461 Da for cysteine, against a FASTA database contain- mised the number of mitochondrial proteins identified and
ing all protein sequences from FlyBase [85] plus contaminant minimised any influence of genetic background on turnover
proteins. Peptide-spectrum match false discovery rates were rate. Turnover rates are the mean values for all genotypes in
determined using Percolator [86] at a threshold of 0.01, and which the protein was detected, and are highly consistent
peptides were assembled into protein identifications using an across genotypes, as previously reported [25]. Five mitochon-
in-house implementation of IDPicker [87]. drial proteins showed excessive variability of turnover rates
across genotypes (inter-genotype coefficient of variation ≥
0.25) and were excluded from analysis.
Annotation of Drosophila proteins
Drosophila protein localisation was determined from a variety
of resources including gene and protein information databases Calculation of percent autophagic turnover and percent
(FlyBase [88], MitoDrome [89], NCBI [90], UniProt [91]), parkin-dependent turnover
protein localisation prediction algorithms (WoLF PSORT
[92], MitoProt [93], Predotar [94], SignalP [95], NucPred We calculated percent autophagic turnover as described in the
[96], and PTS1 Predictor [97,98]), BLAST [99] and primary text, and we calculated parkin-dependent turnover by com-
literature. paring turnover in parkin mutants with turnover in WT
controls. For the sake of comparability, we restricted analyses
to those proteins that also met quality standards in Atg7
Protein turnover rate calculations mutants. All proteins in parkin-dependent turnover analyses
Turnover rates for fly head proteins were calculated using were thus detected in WT, parkin and Atg7 flies. Five proteins
Topograph software [100] as described in Vincow et al. [25]. had negative parkin-dependent turnover values and were
A protein’s turnover rate was computed based on data from excluded from analysis (final n = 168 proteins). Percent
all peptides detected, and data points from all biological parkin-dependent turnover was calculated for each mitochon-
replicates were pooled for turnover calculations. All proteins drial protein by adapting the equation used for Atg7 mutants:
included in the dataset had at least 15 measurements per percent park-dependent turnover ¼
genotype of percent turnover, derived from at least two pep-
tides. Peptides that could be the product of more than one ðturnover ratein WTÞ  ðturnover rate in parkÞ
 100
gene were excluded from analysis. For a small percentage of turnover rate in WT
genes (2–5%), Topograph grouped peptides corresponding to A more conservative estimate of parkin-dependent autophagic
a single gene into 2–3 non-overlapping ‘isoform groups’. For turnover was also calculated by excluding proteins that
example, isoform group 1 might include peptides mapping appeared to undergo non-autophagic forms of
only to the COX6B-PA isoform, while isoform group 2 pep- parkin-dependent turnover (Table S1). As previously
tides could have come from COX6B-PA, -PB, or -PC. While described [25], proteins were designated potential selective
in most cases the isoform groups for a single protein had parkin targets if their individual percent parkin-dependent
1602 E. S. VINCOW ET AL.

turnover was greater than their percent autophagic turnover log-ratio (M value) and the average log spot intensity (A
(n = 53). value) are not. We rescale the M and A values based on an
estimate of the intra-spot correlation, and then fit
a conventional linear model based on the rescaled M and
Analysis of tissue-specific gene expression
A values. Since the M and A values can be expressed as linear
Tissue expression of genes encoding mitochondrial and other combinations of the log expression values for each spot, it is
organellar proteins was determined using FlyAtlas microarray straightforward to fit the model and then compute any con-
data [54]. If more than one probeset matched a given gene, we trast of interest.
selected the best match based on signal level, exclusivity and/ For this experiment, we were only interested in the comparison
or location within the gene. Expression was analysed in all between fed Atg7 nulls and fed controls. However, we fitted
FlyAtlas tissues represented in the head: brain, fat body and a linear model using all the array data, and then computed an
eye. We were primarily interested in the proportion of a given empirical Bayes adjusted contrast between the two groups of
mitochondrial protein that was expressed in each tissue (in interest. Using all arrays and the empirical Bayes adjustment
other words, how much of the protein’s total abundance was allowed us to borrow information both across arrays and across
found in that tissue). We therefore analysed the relative tissue genes, increasing power to detect differences. The p values were
expression (enrichment) of genes encoding the proteins of corrected for multiple testing using Benjamini and Hochberg’s
interest. To do this, we normalised each tissue RNA value to original procedure (false discovery rate = 0.1) [107]. After analysis
the RNA value for whole fly, as described by Robinson was complete, we mapped the probe sequences to Drosophila
et al. [102]. transcripts using BLAST+ [99], FlyBase [88], and Entrez Gene
Of 186 mitochondrial proteins, 4 were mitochondrially [90]. Of the 91 probes that showed significant difference in expres-
encoded and 5 had no usable RNA data, leaving 177 proteins sion between genotypes, 69 matched known gene transcripts, and
with RNA measurements. For other organellar proteins (ER, we obtained Gene Ontology information for these using
ribosome and peroxisome), the number of genes with usable FlyBase [88].
RNA values varied slightly by tissue because a few genes were
not reliably detected in fat body or eye. The final number of
Measurement of protein abundance
genes was 82 for brain, 79 for fat body and 80 for eye.
We measured protein abundance from the same raw mass spectro-
metry data used in the turnover study. Abundance comparisons
Microarray analysis
were made at the second time point (240 h), when differences
Gene expression in Atg7 mutants was evaluated by reanalysis between genotypes were most marked. For this analysis we used
of microarray data originally published by Erdi et al. [103]. the original heterozygote control flies rather than WT flies (see
Raw data were downloaded from the ArrayExpress website ‘Protein turnover rate calculations’ above). While the composite
(study E-MEXP-3352) using the Bioconductor ArrayExpress control group approach was appropriate for measurement of
package [104]. Microarray analysis was performed using turnover (more consistent and less noisy than abundance [25]),
4 × 44 k format Drosophila Gene Expression Microarrays measurement of relative protein abundance required mutant and
(Agilent, G2519F-021791). Atg7 mutants (third instar larvae) control samples that had been run at the same time. We calculated
were compound heterozygous for null mutations (Atg7d14/ fold change in protein abundance between mutant and control
Atg7d77), and controls were heterozygotes (Atg7d14/ using Skyline [108] and MSstats [109]. Prior to MSstats analysis,
CG5335d30). CG5335 is an unrelated gene located within an we computed total abundance (labelled plus unlabelled) for each
exon of Atg7. Only data from fed larvae were used. peptide using an R script. The statistical significance of intergroup
Experimental design and procedures were otherwise as pre- differences was calculated using a linear mixed model, then
viously described [103]. adjusted for multiple comparisons by the Benjamini–Hochberg
The data had been collected using two-colour arrays, but procedure with a false discovery rate of 0.05.
for this analysis we wished to compare samples that were
not hybridised on the same array, so we analysed the data
Analysis of human ATG7 and ATG5 null fibroblast data
using a separate-channel analysis method implemented in
the Bioconductor limma package [105,106]. Briefly, we We analysed previously published sets of protein turnover
background corrected the individual channel data using rates from engineered human fibroblasts lacking either
the ‘normexp’ method, with an offset of 50. This estimates ATG7 or ATG5 [55]. For each mutant vs. WT comparison,
an array-wide background value by modelling the data as we included proteins that met either of two criteria: the
a convolution between a lognormal and an exponential protein was marked ‘selected’ in both genotypes, or its corre-
distribution. After subtracting the background estimate, lation coefficient for kdeg was ≥ 0.7 in both genotypes.
we normalised within each array using a locally weighted ‘Selected’ means that the protein was chosen for analysis in
(loess) fit, and then normalised between arrays using a so- the original publication on these datasets [55]. Approximately
called ‘Aquantile’ normalisation, which performs a quantile 12% of mitochondrial proteins in the ATG7−/− dataset and
normalisation based on the average spot intensity, while 15% in the ATG5−/− dataset had negative autophagy-
keeping the log-ratios of each spot unchanged. dependent turnover rates (faster turnover in the mutant),
The separate-channel analysis is based on the observation and were excluded from analysis. We classified a protein as
that while the log-ratios for a given spot are correlated, the mitochondrial if it met any of the following criteria:
 AUTOPHAGY 1603

1) The protein was identified as mitochondrial in the human References

version of MitoCarta 2.0 [110] based on evidence types known,
[1] Wang C, Youle RJ. The role of mitochondria in apoptosis. Annu
GFP [green fluorescent protein], proteomics, or strong computa- Rev Genet. 2009;43:95–118.
tional. Proteins of the cytosolic ribosome were excluded. [2] McBride HM, Neuspiel M, Wasiak S. Mitochondria: more than
2) The protein had ‘mitochondrion’ or ‘mitochondria[l]’ in just a powerhouse. Curr Biol. 2006 Jul 25;16(14):R551–60.
its UniProt protein name or NCBI gene name. [3] Wallace DC. Mitochondria and cancer. Nat Rev Cancer. 2012
3) The protein had an amply documented mitochondrial Oct;12(10):685–698.
[4] Lill R, Hoffmann B, Molik S, et al. The role of mitochondria in
function (e.g. complex IV assembly). cellular iron-sulfur protein biogenesis and iron metabolism.
Biochim Biophys Acta. 2012 Sep;1823(9):1491–1508.
[5] Mari M, Morales A, Colell A, et al. Mitochondrial glutathione:
Identification of proteasome and mitochondrial protease features, regulation and role in disease. Biochim Biophys Acta.
substrates 2012 Oct 30;1830(5):3317–3328.
[6] Bratic A, Larsson NG. The role of mitochondria in aging. J Clin
We identified probable substrates of mitochondrial proteases in Invest. 2013 Mar;123(3):951–957.
both fly head and human fibroblast data based on the presence [7] Kong Y, Trabucco SE, Zhang H. Oxidative stress, mitochondrial
of degron sequences for mitochondrial proteases Lon and dysfunction and the mitochondria theory of aging. Interdiscip
YME1L (YME1L in Drosophila, YME1L1 in Homo sapiens) Top Gerontol. 2014;39:86–107.
[49,50]. We obtained sequences for the fly proteome from [8] Dupuis L. Mitochondrial quality control in neurodegenerative
diseases. Biochimie. 2014 May;100:177–183.
FlyBase and for the human proteome from NCBI, and searched [9] Bulteau AL, Bayot A. Mitochondrial proteases and cancer.
both proteomes for degrons using an algorithm written in Biochim Biophys Acta. 2011 Jun;1807(6):595–601.
Python 2.7. Criteria for the YME1L degron were based on the [10] Neutzner A, Youle RJ, Karbowski M. Outer mitochondrial mem-
hidden Markov model described by Rampello et al. [50]. Lon brane protein degradation by the proteasome. Novartis Found
degron motifs were FhhF and FhhFP, where ‘h’ is a hydrophobic Symp. 2007;287:4–14; discussion 14–20.
amino acid on the Janin scale [111]. A protein was designated as [11] Clarke KJ, Adams AE, Manzke LH, et al. A role for ubiquitinyla-
tion and the cytosolic proteasome in turnover of mitochondrial
degron-bearing if any of its isoforms had at least one degron. uncoupling protein 1 (UCP1). Biochim Biophys Acta. 2012
We identified proteins as proteasome substrates based on the Oct;1817(10):1759–1767.
presence of ubiquitinated sites as reported by Wagner et al. [48]. [12] Margineantu DH, Emerson CB, Diaz D, et al. Hsp90 inhibition
These were sites that showed altered abundance of ubiquitinated decreases mitochondrial protein turnover. PLoS One. 2007;2(10):
peptides after proteasome inhibitor treatment. For the human e1066.
ATG5−/− and ATG7−/− fibroblasts, we compared mitochondrial [13] Soubannier V, McLelland GL, Zunino R, et al. A vesicular trans-
port pathway shuttles cargo from mitochondria to lysosomes.
proteins with one or more ubiquitinated sites to the remaining Curr Biol. 2012 Jan 24;22(2):135–141.
mitochondrial proteins. For fly heads, we did the same using [14] Mizushima N. Autophagy: process and function. Genes Dev. 2007
Drosophila orthologues of the Wagner et al. data. We obtained Nov 15;21(22):2861–2873.
these orthologues with the DRSC Integrative Ortholog [15] Fletcher MJ, Sanadi DR. Turnover of rat-liver mitochondria.
Prediction Tool [112] (DIOPT) v6.0.3, minimum score 5. Biochim Biophys Acta. 1961 Aug;5(51):356–360.
[16] Ashford TP, Porter KR. Cytoplasmic components in hepatic cell
lysosomes. J Cell Biol. 1962 Jan;12:198–202.
Acknowledgments [17] Xie Z, Klionsky DJ. Autophagosome formation: core machinery
and adaptations. Nat Cell Biol. 2007 Oct;9(10):1102–1109.
We thank J.K. Chung, A. Duttaroy, and T. Neufeld for providing fly [18] Tolkovsky AM. Mitophagy. Biochim Biophys Acta. 2009 Sep;1793
stocks. We thank the Brewer Lab at the University of Washington for (9):1508–1515.
BB14-3A yeast and for their gracious assistance with yeast culture. We [19] Ney PA. Mitochondrial autophagy: origins, significance, and role
thank the Rabinovitch Lab for sharing unpublished data and providing of BNIP3 and NIX. Biochim Biophys Acta. 2015 Oct;1853(10Pt
feedback. We are also grateful to the University of Washington’s B):2775–2783.
Proteomics Resource for mass spectrometry facilities and members of [20] Lemasters JJ Selective mitochondrial autophagy, or mitophagy, as
the Pallanck Lab for critical reading of the manuscript. a targeted defense against oxidative stress, mitochondrial dysfunc-
tion, and aging. Rejuvenation Res. Spring 2005;8(1):3–5.
[21] Kim I, Rodriguez-Enriquez S, Lemasters JJ. Selective degradation
Disclosure statement of mitochondria by mitophagy. Arch Biochem Biophys. 2007 Jun
15;462(2):245–253.
No potential conflict of interest was reported by the authors.
[22] Narendra D, Tanaka A, Suen DF, et al. Parkin is recruited selec-
tively to impaired mitochondria and promotes their autophagy.
J Cell Biol. 2008 Dec 1;183(5):795–803.
Funding [23] Narendra DP, Jin SM, Tanaka A, et al. PINK1 is selectively
stabilized on impaired mitochondria to activate Parkin. PLoS
This work was supported by National Institutes of Health (National Institute
Biol. 2010 Jan;8(1):e1000298.
of General Medicine) grants [1RO1GM086394] and [1R01GM104990] to LJP.
[24] Rub C, Wilkening A, Voos W. Mitochondrial quality control by the
Pink1/Parkin system. Cell Tissue Res. 2017 Jan;367(1):111–123.
[25] Vincow ES, Merrihew G, Thomas RE, et al. The PINK1-Parkin
ORCID pathway promotes both mitophagy and selective respiratory chain
Evelyn S. Vincow http://orcid.org/0000-0002-7449-3173 turnover in vivo. Proc Natl Acad Sci USA. 2013 Apr 16;110
Ruth E. Thomas http://orcid.org/0000-0001-9648-2339 (16):6400–6405.
Gennifer E. Merrihew http://orcid.org/0000-0003-4903-0318 [26] Gross NJ. Control of mitochondrial turnover under the influence
James W. MacDonald http://orcid.org/0000-0002-7328-7626 of thyroid hormone. J Cell Biol. 1971 Jan;48(1):29–40.
1604 E. S. VINCOW ET AL.

[27] Khan AA, Wilson JE. Studies of turnover in mammalian subcel- [48] Wagner SA, Beli P, Weinert BT, et al. A proteome-wide, quanti-
lular particles: brain nuclei, mitochondria and microsomes. tative survey of in vivo ubiquitylation sites reveals widespread
J Neurochem. 1965 Feb;12:81–86. regulatory roles. Mol Cell Proteomics. 2011 Oct;10(10):M111
[28] Swick RW, Rexroth AK, Stange JL. The metabolism of mitochon- 013284.
drial proteins. 3. The dynamic state of rat liver mitochondria. [49] Shi H, Rampello AJ, Glynn SE. Engineered AAA+ proteases reveal
J Biol Chem. 1968 Jul 10;243(13):3581–3587. principles of proteolysis at the mitochondrial inner membrane.
[29] Lusena CV, Depocas F. Heterogeneity and differential fragility of Nat Commun. 2016 Oct;27(7):13301.
rat liver mitochondria. Can J Biochem. 1966 May;44(5):497–508. [50] Rampello AJ, Glynn SE. Identification of a degradation signal
[30] Von Hungen K, Mahler HR, Moore WJ. Turnover of protein and sequence within substrates of the mitochondrial i-AAA Protease.
ribonucleic acid in synaptic subcellular fractions from rat brain. J Mol Biol. 2017 Mar 24;429(6):873–885.
J Biol Chem. 1968 Apr 10;243(7):1415–1423. [51] Pagliarini DJ, Calvo SE, Chang B, et al. A mitochondrial protein
[31] Dai DF, Karunadharma PP, Chiao YA, et al. Altered proteome compendium elucidates complex I disease biology. Cell. 2008 Jul
turnover and remodeling by short-term caloric restriction or 11;134(1):112–123.
rapamycin rejuvenate the aging heart. Aging Cell. 2014 Jun;13 [52] Sun N, Yun J, Liu J, et al. Measuring in vivo mitophagy. Mol Cell.
(3):529–539. 2015 Nov 19;60(4):685–696.
[32] Karunadharma PP, Basisty N, Dai DF, et al. Subacute calorie [53] McWilliams TG, Prescott AR, Allen GF, et al. mito-QC illumi-
restriction and rapamycin discordantly alter mouse liver proteome nates mitophagy and mitochondrial architecture in vivo. J Cell
homeostasis and reverse aging effects. Aging Cell. 2015 Aug;14 Biol. 2016 Aug 1;214(3):333–345.
(4):547–557. [54] Chintapalli VR, Wang J, Dow JA. Using FlyAtlas to identify better
[33] Karunadharma PP, Basisty N, Chiao YA, et al. Respiratory chain Drosophila melanogaster models of human disease. Nat Genet.
protein turnover rates in mice are highly heterogeneous but 2007 Jun;39(6):715–720.
strikingly conserved across tissues, ages, and treatments. Faseb J. [55] Zhang T, Shen S, Qu J, et al. Global analysis of cellular protein
2015 Aug;29(8):3582–3592. flux quantifies the selectivity of basal autophagy. Cell Rep. 2016
[34] Kruse SE, Karunadharma PP, Basisty N, et al. Age modifies Mar 15;14(10):2426–2439.
respiratory complex I and protein homeostasis in a muscle [56] Mathis AD, Naylor BC, Carson RH, et al. Mechanisms of in vivo
type-specific manner. Aging Cell. 2016 Feb;15(1):89–99. ribosome maintenance change in response to nutrient signals.
[35] Kim TY, Wang D, Kim AK, et al. Metabolic labeling reveals Mol Cell Proteomics. 2017 Feb;16(2):243–254.
proteome dynamics of mouse mitochondria. Mol Cell [57] Hoeller D, Dikic I. How the proteasome is degraded. Proc Natl
Proteomics. 2012 Dec;11(12):1586–1594. Acad Sci USA. 2016 Nov 22;113(47):13266–13268.
[36] Price JC, Guan S, Burlingame A, et al. Analysis of proteome [58] Augustin S, Nolden M, Muller S, et al. Characterization of pep-
dynamics in the mouse brain. Proc Natl Acad Sci USA. 2010 tides released from mitochondria: evidence for constant proteo-
Aug 10;107(32):14508–14513. lysis and peptide efflux. J Biol Chem. 2005 Jan 28;280
[37] Nelson CJ, Li L, Jacoby RP, et al. Degradation rate of mitochon- (4):2691–2699.
drial proteins in Arabidopsis thaliana cells. J Proteome Res. 2013 [59] Collins ML, Eng S, Hoh R, et al. Measurement of mitochondrial
Jul 5;12(7):3449–3459. DNA synthesis in vivo using a stable isotope-mass spectrometric
[38] Suzuki CK, Suda K, Wang N, et al. Requirement for the yeast gene technique. J Appl Physiol (1985). 2003 Jun;94(6):2203–2211.
LON in intramitochondrial proteolysis and maintenance of [60] Poovathingal SK, Gruber J, Lakshmanan L, et al. Is mitochondrial
respiration. Science. 1994 May 13;264(5161):891. DNA turnover slower than commonly assumed? Biogerontology.
[39] Hughes AL, Hughes CE, Henderson KA, et al. Selective sorting 2012 Oct;13(5):557–564.
and destruction of mitochondrial membrane proteins in aged [61] Nishida Y, Arakawa S, Fujitani K, et al. Discovery of Atg5/
yeast. Elife. 2016;5. Atg7-independent alternative macroautophagy. Nature. 2009 Oct
[40] Abeliovich H, Zarei M, Rigbolt KTG, et al. Involvement of mito- 1;461(7264):654–658.
chondrial dynamics in the segregation of mitochondrial matrix [62] Juenemann K, Reits EA. Alternative macroautophagic pathways.
proteins during stationary phase mitophagy. Nat Commun. 2013 Int J Cell Biol. 2012;2012:189794.
Nov;4. [63] Matsuda N, Sato S, Shiba K, et al. PINK1 stabilized by mitochon-
[41] Hamalainen RH, Manninen T, Koivumaki H, et al. Tissue- and drial depolarization recruits Parkin to damaged mitochondria and
cell-type-specific manifestations of heteroplasmic mtDNA activates latent Parkin for mitophagy. J Cell Biol. 2010 Apr 19;189
3243A>G mutation in human induced pluripotent stem cell- (2):211–221.
derived disease model. Proc Natl Acad Sci USA. 2013 Sep [64] Sumpter R Jr., Sirasanagandla S, Fernandez AF, et al. Fanconi
17;110(38):E3622–30. anemia proteins function in mitophagy and immunity. Cell. 2016
[42] Le Guerroue F, Eck F, Jung J, et al. Autophagosomal content May 5;165(4):867–881.
profiling reveals an LC3C-dependent piecemeal mitophagy [65] Kawajiri S, Saiki S, Sato S, et al. PINK1 is recruited to mitochon-
pathway. Mol Cell. 2017 Nov 16;68(4):786–796 e6. dria with parkin and associates with LC3 in mitophagy. FEBS Lett.
[43] Juhasz G, Erdi B, Sass M, et al. Atg7-dependent autophagy pro- 2010 Mar 19;584(6):1073–1079.
motes neuronal health, stress tolerance, and longevity but is dis- [66] Daniele JR, Esping DJ, Garcia G, et al. High-throughput charac-
pensable for metamorphosis in Drosophila. Genes Dev. 2007 Dec terization of region-specific mitochondrial function and
1;21(23):3061–3066. morphology. Sci Rep. 2017 Jul 27;7(1):6749.
[44] Tanaka A, Cleland MM, Xu S, et al. Proteasome and p97 mediate [67] Wikstrom JD, Twig G, Shirihai OS. What can mitochondrial
mitophagy and degradation of mitofusins induced by Parkin. heterogeneity tell us about mitochondrial dynamics and
J Cell Biol. 2010 Dec 27;191(7):1367–1380. autophagy? Int J Biochem Cell Biol. 2009 Oct;41(10):1914–1927.
[45] Yoshii SR, Kishi C, Ishihara N, et al. Parkin mediates [68] Kuznetsov AV, Margreiter R. Heterogeneity of mitochondria and
proteasome-dependent protein degradation and rupture of the mitochondrial function within cells as another level of mitochon-
outer mitochondrial membrane. J Biol Chem. 2011 Jun 3;286 drial complexity. Int J Mol Sci. 2009 Apr 24;10(4):1911–1929.
(22):19630–19640. [69] Ferreira R, Vitorino R, Alves RM, et al. Subsarcolemmal and
[46] McLelland GL, Soubannier V, Chen CX, et al. Parkin and PINK1 intermyofibrillar mitochondria proteome differences disclose
function in a vesicular trafficking pathway regulating mitochon- functional specializations in skeletal muscle. Proteomics. 2010
drial quality control. Embo J. 2014 Feb 18;33(4):282–295. Sep;10(17):3142–3154.
[47] Hammerling BC, Najor RH, Cortez MQ, et al. A Rab5 endosomal [70] Choe SC, Hamacher-Brady A, Brady NR. Autophagy capacity and
pathway mediates Parkin-dependent mitochondrial clearance. Nat sub-mitochondrial heterogeneity shape Bnip3-induced mitophagy
Commun. 2017 Jan 30;8:14050. regulation of apoptosis. Cell Commun Signal. 2015;13(1):37.
 AUTOPHAGY 1605

[71] Twig G, Elorza A, Molina AJ, et al. Fission and selective fusion [93] Claros MG, Vincens P. Computational method to predict mito-
govern mitochondrial segregation and elimination by autophagy. chondrially imported proteins and their targeting sequences. Eur
Embo J. 2008 Jan 23;27(2):433–446. J Biochem. 1996 Nov 1;241(3):779–786.
[72] Barsoum MJ, Yuan H, Gerencser AA, et al. Nitric oxide-induced [94] Small I, Peeters N, Legeai F, et al. Predotar: A tool for rapidly
mitochondrial fission is regulated by dynamin-related GTPases in screening proteomes for N-terminal targeting sequences.
neurons. Embo J. 2006 Aug 23;25(16):3900–3911. Proteomics. 2004 Jun;4(6):1581–1590.
[73] Youle RJ, van der Bliek AM. Mitochondrial fission, fusion, and [95] Petersen TN, Brunak S, von Heijne G, et al. SignalP 4.0: discri-
stress. Science. 2012 Aug 31;337(6098):1062–1065. minating signal peptides from transmembrane regions. Nat
[74] Shirihai OS, Song M, Dorn GW 2nd. How mitochondrial dynamism Methods. 2011;8(10):785–786.
orchestrates mitophagy. Circ Res. 2015 May 22;116(11):1835–1849. [96] Brameier M, Krings A, MacCallum RM. NucPred–predicting
[75] Abeliovich H, Dengjel J. Mitophagy as a stress response in mam- nuclear localization of proteins. Bioinformatics. 2007 May 1;23
malian cells and in respiring S. cerevisiae. Biochem Soc Trans. (9):1159–1160.
2016 Apr 15;44(2):541–545.
[76] Burman JL, Pickles S, Wang C, et al. Mitochondrial fission facil- [97] Neuberger G, Maurer-Stroh S, Eisenhaber B, et al. Prediction of
itates the selective mitophagy of protein aggregates. J Cell Biol. peroxisomal targeting signal 1 containing proteins from amino
2017 Oct 2;216(10):3231–3247. acid sequence. J Mol Biol. 2003 May 2;328(3):581–592.
[77] Yang JY, Yang WY. Bit-by-bit autophagic removal of parkin-labelled [98] Neuberger G, Maurer-Stroh S, Eisenhaber B, et al. Motif refine-
mitochondria. Nat Commun. 2013;4:2428. ment of the peroxisomal targeting signal 1 and evaluation of
[78] Khalil B, El Fissi N, Aouane A, et al. PINK1-induced mitophagy taxon-specific differences. J Mol Biol. 2003 May 2;328(3):567–579.
promotes neuroprotection in Huntington’s disease. Cell Death [99] Camacho C, Coulouris G, Avagyan V, et al. BLAST+: architecture
Dis. 2015;6:e1617. and applications. BMC Bioinformatics. 2009;10:421.
[79] Greene JC, Whitworth AJ, Kuo I, et al. Mitochondrial pathology [100] Hsieh EJ, Shulman NJ, Dai DF, et al. Topograph, a software plat-
and apoptotic muscle degeneration in Drosophila parkin mutants. form for precursor enrichment corrected global protein turnover
Proc Natl Acad Sci USA. 2003 Apr 1;100(7):4078–4083. measurements. Mol Cell Proteomics. 2012 Nov;11(11):1468–1474.
[80] Park J, Lee SB, Lee S, et al. Mitochondrial dysfunction in [101] Tukey JW. Exploratory data analysis. Reading (MA): Addison-
Drosophila PINK1 mutants is complemented by parkin. Nature. Wesley; 1977.
2006 Jun 29;441(7097):1157–1161. [102] Robinson SW, Herzyk P, Dow JA, et al. FlyAtlas: database of gene
[81] Sherman F Getting started with yeast [Internet]. [cited 2003 22 Aug]. expression in the tissues of Drosophila melanogaster. Nucleic
http://dbb.urmc.rochester.edu/labs/sherman_f/StartedYeast.html. Acids Res. 2013 Jan;41(Database issue):D744–50.
[82] McCune HJ, Danielson LS, Alvino GM, et al. The temporal program [103] Erdi B, Nagy P, Zvara A, et al. Loss of the starvation-induced
of chromosome replication: genomewide replication in clb5{Delta} gene Rack1 leads to glycogen deficiency and impaired autopha-
Saccharomyces cerevisiae. Genetics. 2008 Dec;180(4):1833–1847. gic responses in Drosophila. Autophagy. 2012 Jul 1;8
[83] Hsieh EJ, Hoopmann MR, MacLean B, et al. Comparison of (7):1124–1135.
database search strategies for high precursor mass accuracy MS/ [104] Kauffmann A, Rayner TF, Parkinson H, et al. Importing
MS data. J Proteome Res. 2010 Feb 5;9(2):1138–1143. ArrayExpress datasets into R/Bioconductor. Bioinformatics. 2009
[84] Ducret A, Van Oostveen I, Eng JK, et al. High throughput protein Aug 15;25(16):2092–2094.
characterization by automated reverse-phase chromatography/electro- [105] Smyth GK, Altman NS. Separate-channel analysis of two-channel
spray tandem mass spectrometry. Protein Sci. 1998 Mar;7(3):706–719. microarrays: recovering inter-spot information. BMC
[85] Marygold SJ, Crosby MA, Goodman JL. Using flybase, a database of Bioinformatics. 2013;14:165.
drosophila genes and genomes. Methods Mol Biol. 2016;1478:1–31. [106] Ritchie ME, Phipson B, Wu D, et al. limma powers differential
[86] Kall L, Canterbury JD, Weston J, et al. Semi-supervised learning expression analyses for RNA-sequencing and microarray studies.
for peptide identification from shotgun proteomics datasets. Nat Nucleic Acids Res. 2015 Apr 20;43(7):e47.
Methods. 2007 Nov;4(11):923–925.
[107] Benjamini Y, Hochberg Y. Controlling the false discovery rate:
[87] Zhang B, Chambers MC, Tabb DL. Proteomic parsimony through
a practical and powerful approach to multiple testing. J Royal Stat
bipartite graph analysis improves accuracy and transparency.
Soc Ser B (Methodological). 1995;57(1):289–300.
J Proteome Res. 2007 Sep;6(9):3549–3557.
[88] Attrill H, Falls K, Goodman JL, et al. FlyBase: establishing a gene [108] MacLean B, Tomazela DM, Shulman N, et al. Skyline: an open
group resource for drosophila melanogaster. Nucleic Acids Res. source document editor for creating and analyzing targeted pro-
2016 Jan 4;44(D1):D786–92. teomics experiments. Bioinformatics. 2010 Apr 1;26(7):966–968.
[89] Sardiello M, Licciulli F, Catalano D, et al. MitoDrome: a database of [109] Choi M, Chang CY, Clough T, et al. MSstats: an R package for
Drosophila melanogaster nuclear genes encoding proteins targeted to statistical analysis of quantitative mass spectrometry-based proteo-
the mitochondrion. Nucleic Acids Res. 2003 Jan 1;31(1):322–324. mic experiments. Bioinformatics. 2014 Sep 1;30(17):2524–2526.
[90] Brown GR, Hem V, Katz KS, et al. Gene: a gene-centered infor- [110] Calvo SE, Clauser KR, Mootha VK. MitoCarta2.0: an updated
mation resource at NCBI. Nucleic Acids Res. 2015 Jan;43 inventory of mammalian mitochondrial proteins. Nucleic Acids
(Database issue):D36–42. Res. 2016 Jan 4;44(D1):D1251–7.
[91] UniProt Consortium. UniProt: a hub for protein information. [111] Janin J. Surface and inside volumes in globular proteins. Nature.
Nucleic Acids Res. 2015 Jan;43(Database issue):D204–12. 1979 Feb 8;277(5696):491–492.
[92] Horton P, Park KJ, Obayashi T, et al. WoLF PSORT: protein [112] Hu Y, Flockhart I, Vinayagam A, et al. An integrative approach to
localization predictor. Nucleic Acids Res. 2007 Jul;35(Web ortholog prediction for disease-focused and other functional
Server issue):W585–7. studies. BMC Bioinformatics. 2011 Aug 31;12:357.