ISSN 0006-2979, Biochemistry (Moscow), 2024, Vol. 89, No. 11, pp. 2028-2036 The Author(s) 2024. This article is an open access publication.
2028
Vimentin and Desmin Intermediate Filaments
Maintain Mitochondrial Membrane Potential
Alexander A. Dayal
1
, Olga I. Parfenteva
1
, Wang Huiying
1
, Anton S. Shakhov
1,2
,
Irina B. Alieva
1,2
, and Alexander A. Minin
1,a
*
1
Institute of Protein Research, Russian Academy of Sciences, 119334, Moscow, Russia
2
Belozersky Institute of Physico-Chemical Biology, Lomonosov Moscow State University, 119992, Moscow, Russia
a
e-mail: alexminin@gmail.com
Received August 28, 2024
Revised October 4, 2024
Accepted October 9, 2024
Abstract— Intermediate filaments (IFs) represented by a diverse range of proteins, are one of the three main
cytoskeleton components in different types of animal cells. IFs provide mechanical strength to cells and help
position the nucleus and organelles in the cell. Desmin is an IF protein typical of muscle cells, while vimentin,
which has a similar structure, is expressed in many mesenchymal cells. Both proteins are synthesized during
myogenesis and regeneration of damaged muscle tissue and form a mixed IF network. Both desmin and vi-
mentin regulate mitochondrial activity, including mitochondrial localization and maintenance of mitochondrial
membrane potential, in the corresponding cells, but the role of mixed IFs in the control of mitochondrial
functions remains unclear. To investigate how a simultaneous presence of these proteins affects mitochondrial
membrane potential, we used BHK21 cells expressing both vimentin and desmin IFs. Expression of vimentin
or desmin individually or both proteins simultaneously was suppressed using gene knockout and/or RNA
interference. It was found that disruption of biosynthesis of either vimentin or desmin did not affect the
mitochondrial membrane potential, which remained unchanged compared to cells expressing both proteins.
Simultaneous abolishment of both proteins resulted in a 20% reduction in the mitochondrial membrane po-
tential, indicating that both vimentin and desmin play an equally important role in its maintenance.
DOI: 10.1134/S0006297924110154
Keywords: mitochondria, vimentin, desmin, mitochondrial membrane potential
Abbreviations: IF, intermediate filament; MMP, mitochon-
drial membrane potential.
* To whom correspondence should be addressed.
INTRODUCTION
Mitochondria are eukaryotic cell organelles that
play a critical role in cell functioning under both nor-
mal and pathological conditions. They are essential
for ATP synthesis, generation of reactive oxygen spe-
cies, regulation of apoptosis and intracellular calcium
concentration, and numerous other cellular functions
[1-4]. At the organismal level, mitochondria are in-
volved in the immune response, stress response, and
adaptation to exercise, as well as play an important
role in the development of various pathologies, includ-
ing cancer and aging [5-10]. Most mitochondrial func-
tions directly depend on the mitochondrial membrane
potential (MMP), which is maintained by the respirato-
ry chain complexes located in the inner mitochondrial
membrane [11]. Electron transport and proton trans-
fer across the membrane during respiration result in
the establishment of the electrical potential of about
–180 mV between the cytosol and the mitochondrial
matrix [11]. The importance of maintaining the MMP
is emphasized by the fact that depolarized mitochon-
dria, i.e., those with unacceptably low membrane
potential, are selectively eliminated through mitopha-
gy [12]. Extensive depolarization of mitochondria can
activate apoptosis, resulting in cell death [13].
Numerous intracellular factors involved in the
MMP regulation have been identified [14], including
intermediate filament(IF) proteins that play a signifi-
cant role in this process [15]. For example, mutations
INTERMEDIATE FILAMENTS MAINTAIN MITOCHONDRIAL MEMBRANE POTENTIAL 2029
BIOCHEMISTRY (Moscow) Vol. 89 No. 11 2024
in keratins have been shown to cause mitochondri-
al fragmentation in hepatocytes, leading to the ag-
gregation of mitochondria and impaired respiratory
function [16-18]. Vimentin (type III IF protein) regu-
lates MMP in fibroblasts; its removal decreases MMP
by 20% [19]. Desmin, an IF protein specific for mus-
cle cells, has also been demonstrated to participate
in the mitochondrial respiratory function [20-24].
Previously, we found that recombinant vimentin and
desmin expressed and isolated from bacteria, bound
to mitochondria in vitro [25, 26], suggesting that they
may be directly involved in regulating mitochondrial
properties.
Although vimentin is typically found in various
mesenchymal cells, while desmin IFs are characteris-
tic of muscle cells, these proteins can form mixed IF
networks under certain conditions. For instance, vi-
mentin and desmin are expressed simultaneously at
the early stages of muscle fiber differentiation and
during regeneration [27]. Therefore, both vimentin
and desmin can be present in the same cell, as well
as interact with mitochondria and influence their
properties.
The role of vimentin in regulating MMP in the
presence of desmin has been investigated insufficient-
ly. Although several studies have provided compelling
evidence that desmin is involved in the distribution,
morphology, and respiratory function of mitochon-
dria [20-23], its role in the MMP regulation remains
poorly understood. Here, we studied the role of vi-
mentin and desmin in this process in BHK21 cells
expressing both proteins. By selectively suppressing
expression of either desmin or vimentin using RNA
interference (RNAi) and/or the CRISPR-Cas9 system,
we found that each protein could independently
maintain the MMP.
MATERIALS AND METHODS
Cell culture. BHK21 cells and two derivative
cell lines generated using the CRISPR/Cas9 system,
BHK21(Vim
–/–
) and BHK21(Des
–/–
), were cultured in
DMEM (PanEco, Russia) supplemented with 10% fetal
bovine serum (Biolot, Russia), penicillin (100 µg/ml),
and streptomycin (100 µg/ml) (Sigma-Aldrich, USA) at
37°C in a humidified atmosphere with 5% CO
2
. For mi-
croscopy, the cells were seeded on sterile coverslips
and incubated for 16-20 h.
RNAi. To deplete desmin in BHK21 cells via
RNAi, we used the pG-SHIN2-des plasmid encoding
shRNA 5′-AAGCAGGAGAUGAUGGAGU-3′ [28] and GFP
as a reporter. Vimentin was knocked down using
the pG-SHIN2-vim plasmid encoding shRNA 5′-CAGA-
CAGGAUGUUGACAAU-3′ [29, 30] (kindly provided by
Prof. R. Goldman; Northwestern University, Chicago).
Control cells were transfected with the pG-SHIN2-scr
plasmid encoding scrambled shRNA sequence of the
same length (5′-AUGUACUGCGCGUGGAGA-3′).
Vimentin and desmin knockouts. To knock out
the vimentin gene, BHK21 cells, were transfected
with the pSpCas9n(BB)-Puro-(1+2)Vim plasmid en-
coding two guide RNAs: 5′-CACCGAACTCGGTGTTGAT-
GGCGT-3′ and 5′-CACCGAACACCCGCACCAACGAGA-3′
[31]. The gene for desmin was knocked out using the
pSpCas9(BB)-Puro-Des plasmid encoding the guide
RNA 5′-CACCGCGGCGACCCGGGUCGGCUCG-3′. Cell were
transfected using Transfectin reagent (Evrogen, Rus-
sia) in complete DMEM medium. Briefly, 1 µg of plas-
mid DNA was mixed with 1 µl of Transfectin in 0.1 ml
of serum-free DMEM and added to cells in 1 ml of
complete DMEM. BHK21(Vim–/–) and BHK21(Des–/–)
cells were selected in DMEM containing 2 µg/ml pu-
romycin and 1 µg/ml verapamil.
Fluorescent microscopy of live cells. Mito-
chondria were stained by incubating cells with 5 nM
tetramethylrhodamine (TMRM; Molecular Probes,
USA) in the presence of 2.2 µM verapamil for 30 min
at 37°C. Following incubation, the coverslips with
the cells were placed in a sealed chamber contain-
ing DMEM and imaged with a Keyence BZ-9000 mi-
croscope (USA) equipped with an incubator for live-
cell imaging. The temperature in the incubator was
maintained at 36 ± 2°C. The cells were imaged with a
PlanApo 63× objective and a 12-bit digital CCD camera.
The images were transferred to a computer using the
BZ II Viewer software (Keyence, USA) and saved as
12-bit graphic files for further analysis.
Immunofluorescence and immunoblotting. For
IF staining, the cells were fixed with methanol at –20°C
for 10min and incubated with mouse monoclonal an-
ti-vimentin antibodies V9 (Sigma-Aldrich) and mouse
monoclonal anti-desmin antibodies DE-U-10 Sigma-
Aldrich). FITC- and TRITC-conjugated anti-mouse sec-
ondary antibodies (Jackson, USA) were used for pro-
tein detection. Microphotographs were acquired using
a Keyence BZ-9000 microscope (USA) with a PlanApo
63× objective and a 12-bit digital CCD camera.
Super-resolution structured illumination micros-
copy (SR-SIM) was performed with a Nikon N-SIM mi-
croscope (Nikon, Japan) with a ×100/1.49 NA oil im-
mersion objective and a 561 nm diode laser; Z-stacks
were acquired at 0.12-µm intervals with an EMCCD
camera (iXon 897, Andor, Japan). Exposure was opti-
mized to maintain an average brightness of ~5000 a.u.
to minimize photobleaching. Images were acquired
with the NIS-Elements 5.1 software (Nikon).
SDS-PAGE was conducted according to Laemmli’s
method [32], followed by immunoblotting as previous-
ly described [26]. The membranes were stained with
V9 antibodies (vimentin), DE-U-10 antibodies (desmin),
and DM1A monoclonal antibodies (tubulin) and then
DAYAL et al.2030
BIOCHEMISTRY (Moscow) Vol. 89 No. 11 2024
Fig. 1. SR-SIM imaging of vimentin (green) and desmin (red) IFs in BHK21 cells (immunostaining). Scale bar: 10 µm.
with secondary anti-mouse antibodies conjugated to
horseradish peroxidase (Jackson). Protein bands were
visualized using hydrogen peroxide and diaminoben-
zidine as peroxidase substrates.
MMP measurements. MMP was evaluated by
measuring TMRM fluorescence as reported in [19]. Mi-
tochondrial contours were defined using the “analyze
particles” plugin of the ImageJ software in a region
with a single layer of cells, and the average fluores-
cence intensity of all pixels within each contour was
calculated. For each experiment, 10-15 regions each
containing 15-40 mitochondria were analyzed. The
data are presented as the mean fluorescence intensity
for all mitochondria ± standard error. The normality
of data distribution was assessed with the Shapiro–
Wilk test; the homogeneity of variance was checked
with the F-test. The significance of differences was
estimated with the paired Student’s t-test.
RESULTS
Fibroblast-like BHK21 cells contain IFs com-
posed of vimentin and desmin (type III proteins),
with small amounts of nestin (type VI protein) [33].
As demonstrated by the high-resolution immunofluo-
rescence microscopy, these filaments were formed by
co-polymerization of vimentin and desmin. Figure 1
shows that the distribution of these proteins was not
perfectly uniform, suggesting different proportions
of vimentin and desmin in individual filaments. IFs
in the perinuclear region were enriched in desmin,
while vimentin predominantly localized to the cell pe-
riphery. The overall organization of IFs was a radial
network, which implies participation of the microtu-
bule transport system in the intracellular distribution
ofIFs. To investigate the role of desmin and vimentin
in regulating mitochondrial function, we decided to
selectively delete each protein from the cells.
First, we assessed how disruption of vimentin or
desmin biosynthesis would affect the IF network in
BHK21 cells. Using RNAi with the plasmids based on
the pG-SHIN2 carrying the GFP reporter gene [29, 30],
we generated BHK21 cells with IFs composed solely
of vimentin or desmin. As shown in Fig. 2, expres-
sion of the corresponding interfering RNAs resulted in
almost complete absence of desmin (Fig. 2, a and b)
or vimentin (Fig. 2, c and d) in the transfected cells
(identified by co-expression of GFP). In contrast, con-
trol cells transfected with pG-SHIN2-scr displayed no
changes in the levels of either desmin (Fig.2,e andf)
or vimentin (not shown).
Previously, we found that suppression of vimen-
tin biosynthesis in fibroblasts led to a decrease in the
MMP [16]. Here, we investigated the effect of vimentin
knockdown in BHK21 cells also expressing desmin and
found that the absence of vimentin did not affect the
MMP in these cells. (Fig. 3). This suggests that desmin
was able to maintain the MMP at a high level, similar
to vimentin. The knockdown of desmin did not lead
to the decrease in the MMP either (Fig. 3), likely due
to the presence of vimentin in the cells. These data
suggest that both vimentin and desmin independently
regulate MMP in BHK21 cells.
The next step in testing our hypothesis was to as-
sess the MMP in cells lacking both proteins. To achieve
this, we used a two-step approach. First, BHK21 cells
expressing vimentin or desmin only were generated
using the CRISPR-Cas9 system. Next, the remaining
protein was knocked down by RNAi. The data of
Western blotting (Fig. 4) showed that the obtained
BHK21(Des
–/–
) and BHK21(Vim
–/–
) cell lines completely
lacked the knocked-out proteins and, according to im-
munofluorescence microscopy analysis, contained IFs
composed solely of vimentin (Fig. 5, a, b) or desmin
(Fig.5,c,d), respectively. Vimentin IFs in BHK21(Des
–/–
)
cells form a normal, uniformly distributed radial net-
work. In contrast, desmin IFs in BHK21(Vim
–/–
) cells
INTERMEDIATE FILAMENTS MAINTAIN MITOCHONDRIAL MEMBRANE POTENTIAL 2031
BIOCHEMISTRY (Moscow) Vol. 89 No. 11 2024
Fig. 2. Disruption of desmin (a, b) and vimentin (c, d) IFs by RNAi in BHK21 cells transfected with pG-SHIN2-des and
pG-SHIN2-vim plasmids, respectively. Control cells(e,f) were transfected with pG-SHIN2-scr. IFs were stained with antibodies
against desmin (Des) and vimentin (Vim). Transfected cells were identified by GFP expression. Scale bar: 10 µm.
were partially aggregated, suggesting that the proper
distribution of desmin IFs depends on the presence
of vimentin.
Transfection of BHK21(Des
–/–
) cells with the plas-
mid encoding interfering RNA against vimentin al-
lowed us to significantly reduce the content of this
protein and to obtain the cells lacking vimentin IFs
(Fig. 6, a and b). However, complete removal of de-
smin IFs in BHK21(Vim
–/–
) cells proved to be more
challenging. As shown in Fig. 6, c and d, the trans-
fected cells contained desmin aggregates, while the
filament network at the cell periphery was entirely
absent. In contrast, cells transfected with the control
plasmid retained the IF network (data not shown).
Disruption of vimentin biosynthesis in
BHK21(Des
–/–
) cells decreased the MMP compared
to the control cells (Fig. 7a), indicating involvement
of vimentin IFs in the MMP maintenance. A similar
effect was observed when desmin expression was
suppressed in BHK21(Vim
–/–
) cells, even though de-
smin was not completely eliminated from the cells.
Presumably, the remaining desmin aggregates did
not impact the MMP. Interestingly, the reduction in
the MMP (~20%) was similar in BHK21(Des
–/–
) and
DAYAL et al.2032
BIOCHEMISTRY (Moscow) Vol. 89 No. 11 2024
Fig. 3. RNAi-mediated knockdown of desmin (shRNA-des) and
vimentin (shRNA-vim) in BHK21 cells did not affect the MMP.
Control cells were transfected with pG-SHIN2-scr. The data
are presented as mean fluorescence intensity of mitochon-
dria in the indicated number of cells ± standard error, ex-
pressed as a percentage of the average fluorescence intensity
in non-transfected cells (none); p > 0.10.
Fig. 4. Western blot analysis of cell homogenates from the
original BHK21 cells and BHK21(Des
–/–
) and BHK21(Vim
–/–
)
cell lines generated by the knock-out of desmin and vimen-
tin genes, respectively. Alpha-tubulin was used as a loading
control.
BHK21(Vim
–/–
) cells, suggesting that both IF proteins
independently contribute to the MPP maintenance
in BHK21 cells.
DISCUSSION
Various IF proteins play a critical role in the mi-
tochondrial motility, shape, and functions in different
types of cells [34, 35]. However, the molecular mech-
anisms governing the interaction between IF proteins
and mitochondria remain largely unexplored. Analysis
of amino acid sequences of IF proteins has identified
in some of them regions that could function as mi-
tochondrial localization signals [19, 25, 26]. Hence, at
least some IF proteins can bind directly to the mi-
tochondria, as was demonstrated in in vitro experi-
ments for vimentin and desmin [25, 26]. In this study,
Fig. 5. IFs in BHK21(Des
–/–
) (a, b) and BHK21(Vim
–/–
) (c, d) cells. Immunofluorescent staining with anti-vimentin and anti-
desmin antibodies. Scale bar: 10 µm.