of 19
ARTICLE
Clueless/CLUH regulates mitochondrial
fi
ssion by
promoting recruitment of Drp1 to mitochondria
Huan Yang
1
, Caroline Sibilla
2,3,9
, Raymond Liu
4,10
, Jina Yun
1,11
, Bruce A. Hay
4
, Craig Blackstone
2,12
,
David C. Chan
4
, Robert J. Harvey
5,6
& Ming Guo
1,7,8
Mitochondrial
fi
ssion is critically important for controlling mitochondrial morphology, func-
tion, quality and transport. Drp1 is the master regulator driving mitochondrial
fi
ssion, but
exactly how Drp1 is regulated remains unclear. Here, we identi
fi
ed
Drosophila
Clueless and its
mammalian orthologue CLUH as key regulators of Drp1. As with loss of
drp1
, depletion of
clueless
or
CLUH
results in mitochondrial elongation, while as with
drp1
overexpression,
clueless
or
CLUH
overexpression leads to mitochondrial fragmentation. Importantly,
drp1
overexpression rescues adult lethality, tissue disintegration and mitochondrial defects of
clueless
null mutants in
Drosophila
. Mechanistically, Clueless and CLUH promote recruitment
of Drp1 to mitochondria from the cytosol. This involves CLUH binding to mRNAs encoding
Drp1 receptors MiD49 and Mff, and regulation of their translation. Our
fi
ndings identify a
crucial role of Clueless and CLUH in controlling mitochondrial
fi
ssion through regulation
of Drp1.
https://doi.org/10.1038/s41467-022-29071-4
OPEN
1
Department of Neurology, UCLA David Geffen School of Medicine, Los Angeles, CA, USA.
2
Cell Biology Section, Neurogenetics Branch, National Institute of
Neurological Disorders and Stroke, Bethesda, MD, USA.
3
Department of Pharmacology, University College London School of Pharmacy, London, UK.
4
Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, CA, USA.
5
School of Health and Behavioural Sciences,
University of the Sunshine Coast, Sippy Downs, QLD, Australia.
6
Sunshine Coast Health Institute, Birtinya, QLD, Australia.
7
Department of Molecular and
Medical Pharmacology, UCLA David Geffen School of Medicine, Los Angeles, CA, USA.
8
California NanoSystems Institute at UCLA, Los Angeles, CA, USA.
9
Present address: AstraZeneca PLC, Cambridge Biomedical Campus, Cambridge, UK.
10
Present address: Department of Microbiology and Immunology,
UCSF, San Francisco, CA, USA.
11
Present address: Genentech, Inc., South San Francisco, CA, USA.
12
Present address: Department of Neurology,
Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA.
email:
ming
fl
y@ucla.edu
NATURE COMMUNICATIONS
| (2022) 13:1582 | https://doi.org/10.1038/s41467-022-29071-4 | www.nature.com/naturecommunications
1
1234567890():,;
M
itochondria are dynamic organelles, and their mor-
phology and function are governed by a balance
between the opposing actions of fusion and
fi
ssion
1
.
Mitochondrial fusion is controlled by Mitofusin (Mfn) proteins
2
and Opa1
3
, and mitochondrial
fi
ssion is driven by Drp1
4
. Mito-
chondrial
fi
ssion plays pivotal roles in cell division
5
, regulated cell
death
6
, mitochondrial quality control through mitophagy
7
,
8
, and
mitochondrial DNA inheritance
9
,
10
. At the organismal level,
dysregulation of mitochondrial
fi
ssion leads to disrupted meta-
bolic homeostasis
11
, defective brain development
12
, cardiovas-
cular diseases
13
, and cancer
14
. Drp1 dysfunction has also been
implicated in the pathogenesis of several neurodegenerative dis-
eases, including Alzheimer
s disease, Parkinson
s disease (PD),
amyotrophic lateral sclerosis, and Huntington
s disease
15
17
.
With respect to PD, we and others have shown that
PINK1
and
parkin
, mutations that lead to recessive forms of PD
18
,
19
, function
in the same pathway to regulate mitochondrial integrity
20
24
and
mitophagy
25
27
. Overexpression of
drp1
rescues mitochondrial
and tissue defects of
PINK1
or
parkin
mutants in
Drosophila
22
,
23
.
Drp1 also facilitates segregation of the damaged parts of mito-
chondria, which are targeted for mitophagy mediated by PINK1
and Parkin
7
,
8
,
25
.
Drp1 is a member of the dynamin superfamily of GTPases. It is
localized predominantly in the cytosol and is recruited onto
mitochondria via interactions with Drp1 receptors anchored to
the outer mitochondrial membrane (OMM). In mammalian cells,
these receptors include mitochondrial
fi
ssion factor (Mff) and
mitochondrial dynamics proteins of 49 and 51 kDa (MiD49 and
MiD51)
28
32
. Once recruited onto the OMM, Drp1 co-assembles
with these receptors, forming an oligomeric ring to constrict
mitochondria and drive
fi
ssion
32
.
clu
gene orthologues can be found in evolutionarily distant
eukaryotes, including yeast
33
, ameba
34
,
35
,
Arabidopsis
36
,
Droso-
phila
(
clueless
, referred to as
clu
hereafter)
37
,
38
and mammals
(
CLUH
)
39
,
40
. In all of the above species, cells with loss of
clu
orthologues show a phenotype of clustered mitochondria in the
perinuclear region, in contrast to wild-type cells in which mito-
chondria are dispersed throughout the cytoplasm. However, the
mechanism underlying this phenotype remains unclear. Clu
orthologues have been identi
fi
ed as RNA-binding proteins in
yeast
41
,
Drosophila
42
, and mammalian cells
39
,
43
. Clu orthologues
also share an evolutionarily conserved domain structure
37
, with a
highly conserved N-terminal Clu domain, and a C-terminal tet-
ratricopeptide repeat (TPR) domain that is responsible for
binding RNAs
42
. While primarily cytoplasmic,
Drosophila
Clu
protein is found in granules juxtaposed with mitochondria in
female germline cells. These are thought to be ribonucleoprotein
(RNP) complexes formed as an adaptation to metabolic
changes
37
,
44
. In mammalian cells, CLUH speci
fi
cally binds mul-
tiple mRNAs of nuclear-encoded, mitochondria-destined proteins
that belong to critical metabolic pathways, such as oxidative
phosphorylation (OXPHOS), tricarboxylic acid (TCA) cycle, and
fatty acid oxidation
39
,
45
. CLUH is distributed throughout the
cytosol under unstressed conditions. In contrast, during starva-
tion, CLUH and its bound mRNAs form large RNP particles that
function as compartments within which CLUH regulates trans-
lation and stability of these mRNAs, controls mTORC1 signaling,
and modulates metabolic rewiring
39
,
45
,
46
. In addition, Clu
associates with mitochondrial proteins, including TOM20, Porin,
and PINK1, as well as mitochondrially localized Parkin following
mitochondrial damage
47
. Importantly,
clu
interacts genetically
with
PINK1
and
parkin
in
Drosophila
, with previous reports
showing that overexpression (OE) of
clu
rescues
PINK1
but not
parkin
mutant phenotypes
47
,
48
.
In this study, we identify
Drosophila
Clu and human CLUH as
key upstream regulators of Drp1 to control mitochondrial
fi
ssion.
We show that Clu and CLUH promote the recruitment of Drp1
onto mitochondria from the cytosol. This involves CLUH binding
to mRNAs encoding Drp1 receptors Mff and MiD49 and reg-
ulation of their translation. In addition, we found that similarly to
drp1
OE,
clu
OE rescues mitochondrial and tissue defects of
parkin
null mutants, in addition to
PINK1
null mutants, in
Drosophila
. Our study provides new mechanistic insights into
how Drp1 activity and mitochondrial
fi
ssion are regulated.
Results
clu
OE suppresses, and
clu
loss-of-function exacerbates,
PINK1
and
parkin
null mutant phenotypes in
Drosophila
. We and
others
fi
rst demonstrated that
PINK1
and
parkin
function in a
common pathway to regulate mitochondrial integrity and quality
in
Drosophila
20
,
21
.
Drosophila
muscle contains organized, high-
density mitochondria that
fi
ll the spaces between myo
fi
brils. Loss
of
PINK1
or
parkin
leads to severe mitochondrial and tissue
defects, including thoracic indentation that re
fl
ects underlying
muscle degeneration (Supplementary Fig. 1a
c), cell death
(compare Fig.
1
a to Fig.
1
b, d), and disrupted tissue integrity
(compare Fig.
1
a
to Fig.
1
b
,d
). Previous work has suggested that
OE of
clu
rescues
PINK1
null but not
parkin
null mutant
phenotypes
47
,
48
. In contrast, by using the strong muscle driver
Mef2-GAL4, we found that OE of
clu
suppressed all the above
defects in
parkin
null mutants, in addition to
PINK1
null mutants
(Fig.
1
c, c
,e,e
, quanti
fi
ed in g, h; Supplementary Fig. 1a
c).
Using the same GAL4 driver, we previously showed that
drp1
OE
similarly suppresses phenotypes due to loss of either
PINK1
or
parkin
22
,
23
.
PINK1
null and
parkin
null mutants are both viable, as are
double null mutants
20
,
24
.
clu
null mutants are also viable, though
they have a greatly reduced lifespan, dying 3
6 days after
eclosure
37
. In contrast, we found that double null mutants of
PINK1 clu
and
parkin clu
were both lethal, with death during late
pupal stages. The lethality of
parkin clu
double null mutants has
also been noted previously
48
. These synthetic lethal interactions
are reminiscent of previous work by us and others, which showed
that loss of
drp1
function in
PINK1
or
parkin
null mutant
background also results in lethality in
Drosophila
22
,
23
.
To bypass the lethality issue in order to study phenotypes of
PINK1 clu
and
parkin clu
double mutants, we utilized a
clu
partial
loss-of-function (
clu
d00713
hypomorphic) mutant in the
PINK1
or
parkin
null mutant background. With reduced Clu protein
levels
37
,
clu
d00713
mutants did not show a greatly reduced
lifespan
37
or muscle disintegration (Supplementary Fig. 1d, e),
in contrast to
clu
null mutants.
clu
d00713
mutants showed more
elongated mitochondria (Fig.
1
i, j, quanti
fi
ed in o; Supplementary
Fig. 1d
d
,e
e
, quanti
fi
ed in f), as with
PINK1
or
parkin
null
mutants (Fig.
1
k, l). Strikingly,
PINK1 clu
d00713
(Fig.
1
m) and
parkin clu
d00713
double mutants (Fig.
1
n-n
) showed highly
elongated, interconnected, and enlarged mitochondria, which
were much more severe than any single mutant alone (Fig.
1
j
l,
quanti
fi
ed in o). In addition, while
clu
d00713
mutants showed ATP
levels comparable to those in wild-type
fl
ies,
PINK1 clu
d00713
and
parkin clu
d00713
double mutants showed a greater decrease in
ATP levels than was observed in either
PINK1
or
parkin
null
mutant alone (Fig.
1
p). Therefore, like
drp1
,
clu
OE suppresses,
and
clu
loss-of-function exacerbates,
PINK1
and
parkin
null
mutant phenotypes in
Drosophila
(Fig.
1
q). Together, these
results show that
clu
acts in parallel to the
PINK1-parkin
pathway
to regulate mitochondrial morphology and function.
clu
regulates mitochondrial morphology in
Drosophila
.To
investigate if
clu
suppresses
PINK1
and
parkin
null mutant
phenotypes by regulating mitochondrial
fi
ssion or fusion, we
ARTICLE
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-022-29071-4
2
NATURE COMMUNICATIONS
| (2022) 13:1582 | https://doi.org/10.1038/s41467-022-29071-4 | www.nature.com/naturecommunications
examined mitochondrial morphology in response to muscle-
speci
fi
c
clu
RNAi
48
or
clu
OE.
clu
RNAi resulted in signi
fi
cantly
elongated mitochondria, as visualized by mitoGFP (Fig.
2
a, b)
and with transmission electron microscopy (TEM) (Fig.
2
d
d
,
e
e
). Mitochondrial elongation was also seen in
clu
hypomorphic
mutants as shown above (Fig.
1
j, o; Supplementary Fig. 1e
e
, f).
clu
OE led to mitochondrial fragmentation as visualized by
mitoGFP (Figs.
1
f and
2
c). Under TEM, muscle with
clu
OE had
mitochondria that were smaller in size and with uneven cristae
density (Fig.
2
f
f
). Changes in mitochondrial morphology are
statistically signi
fi
cant (Fig.
2
g, h). These results show that
clu
regulates mitochondrial morphology in a pro-
fi
ssion manner as
drp1
does.
Overexpression of
drp1
suppresses adult lethality, tissue
damage, and mitochondrial defects of
clu
null mutants in
Drosophila
. To further explore whether
clu
regulates
Good mito quality
Healthy tissue
Compromised mito quality
Damaged tissue
PINK1
parkin
clu
Lethality
Further compromised mito quality
Severely damaged tissue
PINK1
parkin
clu
PINK1
parkin
clu
PINK1
parkin
clu
PINK1
parkin
clu
Good mito quality
Healthy tissue
WT
clu hypo
PINK1
n
ull
PINK1
n
ull + clu hypo
parkin null
par
k
in null + clu hypo
0%
25%
50%
75%
100%
125%
% ATP levels
P
= 0.3603
P
= 0.0047
P
= 0.0008
P
= 0.0034
P
= 0.0002
WT
clu hypo
PINK1
n
ull
PINK1
n
ull + clu hypo
pa
rkin nul
l
par
k
in null
+cluhypo
0%
100%
200%
300%
400%
500%
Relative mito size
(ATP5A)
P
= 0.0037
P
= 0.0123
P
= 0.0033
P
= 0.0286
P
= 0.0087
WT
parkin null
par
ki
n
n
ull + clu OE
PINK
1
n
ull
PINK1
n
ull + clu
O
E
clu OE
0%
25%
50%
75%
100%
125%
% Disintegrated muscle
P
= 3.5 x 10
-5
P
= 0.0002
P
= 4.0 x 10
-6
P
= 0.3739
P
= 0.0006
WT
parkin null
par
ki
n
n
ull + clu OE
PINK1
n
ull
PINK1
n
ull + clu
O
E
clu OE
0%
25%
50%
75%
100%
125%
% TUNEL-positive muscle
P
= 6.3 x 10
-7
P
= 1.1 x 10
-5
P
= 7.4 x 10
-18
P
= 0.1172
P
= 0.0004
WT
parkin
null
PINK1
null
clu
OE
PINK1
null
+
clu
OE
parkin
null
+
clu
OE
a
MitoGFP
TUNEL
Toluidine
Blue
c
bdf
e
a’
c’
b’
d’
f’
e’
WT
PINK1
null
parkin
null
clu
hypomorph
PINK1
null +
clu
hypomorph
parkin
null +
clu
hypomorph
ATP5A
ik
jl
mnn’
p
q
h
go
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-022-29071-4
ARTICLE
NATURE COMMUNICATIONS
| (2022) 13:1582 | https://doi.org/10.1038/s41467-022-29071-4 | www.nature.com/naturecommunications
3
mitochondrial morphology by inhibiting fusion or promoting
fi
ssion, we investigated whether
clu
regulates
mfn
(also known as
Marf
)or
drp1. clu
null mutant
fl
ies are short-lived, dying
3
6 days after eclosure, with about 50% survival on Day 3, less
than 25% survival on Day 4, about 5% survival on Day 5, and 0%
survival after Day 6 (Fig.
3
a, red line; Supplementary Tables 1 and
2). In contrast, wild-type
fl
ies can live over 100 days (Fig.
3
a,
black line).
clu
OE fully rescued the adult lethality of
clu
null
mutants (Fig.
3
a, blue line), con
fi
rming that the lethality of
clu
null mutants was indeed due to lack of
clu
.
mfn
RNAi did not
rescue the lethality of
clu
null mutants. Strikingly, however,
drp1
OE signi
fi
cantly rescued the lethality due to lack of
clu
(Fig.
3
a,
Fig. 1
clu
overexpression suppresses, and
clu
loss-of-function exacerbates,
PINK1
and
parkin
null mutant phenotypes in
Drosophila
.a
f
Confocal
microscopy images of the thoracic muscle labeled with mitoGFP (green) and TUNEL (red).
a
’–
f
Toluidine Blue staining of plastic sections of embedded
thoraces.
clu
overexpression (OE) suppresses muscle death (
a
f
) and tissue disintegration (
a
’–
f
)in
parkin
null mutants (
park
25
/
dpk
21
), in addition to
PINK1
null mutants (
PINK1
5
). Expression of UAS-mitoGFP (
a
f
) and UAS-
clu
(
c
,
c
,
e
,
e
,
f
,
f
) were driven by Mef2-GAL4.
g
,
h
Quanti
fi
cation of muscle death (
g
)
and tissue disintegration (
h
). For each genotype, 3 male
fl
ies were analyzed (mean ± SEM,
n
=
3). For each
fl
y, 30
50 muscle pieces were analyzed, and the
percentage of TUNEL-positive muscle (
g
) or that of disintegrated muscle (
h
) was calculated.
i
n
Mitochondria in the muscle are labeled using a mouse
anti-ATP5A antibody. Mitochondria are more elongated in
clu
d00713
homozygotes (
j
), as compared to those in wild-type (WT)
fl
ies (
i
).
PINK1
5
(
k
) and
park
25
/dpk
21
(
l
) single mutants show elongated mitochondria with vacuolation.
clu
d00713
homozygotes in the
PINK1
5
or
park
25
/dpk
21
background show
exacerbated mitochondrial defects, including enlargement, severe vacuolation, as well as irregular shape and distribution (
m
n
).
o
Quanti
fi
cation of the
relative mitochondrial sizes in (
i
n
). For each genotype, 3 male
fl
ies were analyzed (mean ± SEM,
n
=
3). For each
fl
y, 30
50 mitochondria were analyzed
using Fiji/ImageJ and the average mitochondrial size was calculated.
p
Results of ATP measurements using whole
fl
y lysates.
clu
hypo (hypomorph):
clu
d00713
homozygotes. Experiments were performed in triplicate (mean ± SEM,
n
=
3).
q
A schematic illustration of the genetic interactions between
clu
and the
PINK1
parkin
pathway in
Drosophila. PINK1
and
parkin
function in the same pathway to regulate mitochondrial integrity. Loss of either
PINK1
or
parkin
results in severe mitochondrial dysfunction and tissue damage.
clu
overexpression suppresses phenotypes due to loss of
PINK1
or
parkin
. Partial loss of
clu
function (
clu
d00713
mutants) exacerbates either
PINK1
null or
parkin
null mutant phenotypes. Complete loss of
clu
function (
clu
f04554
mutants) in
PINK1
or
parkin
null mutant background results in lethality.
a
f
,
i
n
Scale bars: 5
μ
M.
g
o
,
p
One-way ANOVA with post hoc Tukey
s HSD test.
WT
clu RNAi
clu OE
0%
50%
100%
150%
200%
250%
Relative mito size
(EM)
P
= 0.0060
P
= 0.0009
WT
clu RNAi
clu OE
0%
50%
100%
150%
200%
Relative mito size
(MitoGFP)
P
= 0.0015
P
= 0.0042
gh
WT
clu
RNAi
clu
OE
MitoGFP
a
b
c
de f
d’
e’
f’
EM
Fig. 2
clu
regulates mitochondrial morphology in
Drosophila
.a
c
Confocal microscopy images of the thoracic muscle labeled with mitoGFP.
clu
RNAi leads
to mitochondrial elongation (
b
), and
clu
overexpression (OE) results in mitochondrial fragmentation (
c
), as compared with the wild type (WT,
a
). White
dashed lines mark the boundaries of representative mitochondria. Scale bar: 5
μ
M.
d
f
,
d
’–
f
TEM ultrastructural images show that mitochondria in wild-
type muscle are well aligned in between myo
fi
brils (
d
,
d
). Muscle with
clu
RNAi exhibits dramatically elongated mitochondria (
e
,
e
), and muscle with
clu
overexpression shows fragmented mitochondria as well as cristae with uneven density (
f
,
f
).
d
’–
f
Enlarged images of the boxed regions in (
d
f
). White
dashed lines illustrate the boundaries of representative mitochondria. Scale bars: 1
μ
M. Expression of UAS-mitoGFP (
a
c
), UAS-
clu
RNAi (
b
,
e
,
e'
), and
UAS-
clu
(
c
,
f
,
f
) were driven by the indirect
fl
ight muscle (IFM)-speci
fi
c driver IFM-GAL4
24
. Expression of UAS-
clu
RNAi driven by Mef2-GAL4 results in
adult lethality and severe mitochondrial defects (data not shown), which are similar to
clu
f04554
mutants (see Fig.
3
).
g
,
h
Quanti
fi
cation of the relative sizes
of mitochondria labeled with mitoGFP (
g
) or visualized under TEM (
h
). For each genotype, 3 different male
fl
ies were analyzed (mean ± SEM,
n
=
3). For
each individual
fl
y, 30
50 individual mitochondria were analyzed using Fiji/ImageJ and the average mitochondrial size was calculated (one-way ANOVA
with post hoc Tukey
s HSD test).
ARTICLE
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-022-29071-4
4
NATURE COMMUNICATIONS
| (2022) 13:1582 | https://doi.org/10.1038/s41467-022-29071-4 | www.nature.com/naturecommunications