We show that, upon CARP over-expression, the transcription factor nuclear factor NF-jB p65 DNA-binding activity decreases.. Interestingly, the expression of the atro-phying protein MURF1
Trang 1identified substrate cardiac ankyrin repeat protein is
involved in the regulation of the nuclear factor-jB pathway
in skeletal muscle
Lydie Laure*, Nathalie Danie`le*, Laurence Suel, Sylvie Marchand, Sophie Aubert,
Nathalie Bourg, Carinne Roudaut, Ste´phanie Duguez, Marc Bartoli and Isabelle Richard
Ge´ne´thon, CNRS UMR8587 LAMBE, Evry, France
Introduction
Calpain 3 is a muscle specific, calcium dependent,
multi-substrate cysteine protease whose mutations are
the cause of limb girdle muscular dystrophy 2A
(LGMD2A, OMIM 253600), a severe muscle disorder leading to selective atrophy and weakness of proximal muscles [1,2] Calpain 3 becomes activated once an
Keywords
calpain 3; cardiac ankyrin repeat protein;
limb girdle muscular dystrophy 2A; NF-jB;
skeletal muscle; titin
Correspondence
I Richard, Ge´ne´thon, CNRS UMR8587
Lambe, 1 bis rue de l’Internationale, 91000
Evry, France
Fax: +33 (0) 1 60 77 86 98
Tel: +33 (0) 1 69 47 29 38
E-mail: richard@genethon.fr
*These authors contributed equally to this
work
(Received 1 June 2010, revised 11 August
2010, accepted 18 August 2010)
doi:10.1111/j.1742-4658.2010.07820.x
A multiprotein complex encompassing a transcription regulator, cardiac ankyrin repeat protein (CARP), and the calpain 3 protease was identified
in the N2A elastic region of the giant sarcomeric protein titin The present study aimed to investigate the function(s) of this complex in the skeletal muscle We demonstrate that CARP subcellular localization is controlled
by the activity of calpain 3: the higher the calpain 3, the more important the sarcomeric retention of CARP This regulation would occur through cleavage of the N-terminal end of CARP by the protease We show that, upon CARP over-expression, the transcription factor nuclear factor NF-jB p65 DNA-binding activity decreases Taken as a whole, CARP and its reg-ulator calpain 3 appear to occupy a central position in the important cell fate-governing NF-jB pathway Interestingly, the expression of the atro-phying protein MURF1, one of NF-jB main targets, remains unchanged
in presence of CARP, suggesting that the pathway encompassing cal-pain3⁄ CARP ⁄ NF-jB does not play a role in muscle atrophy With NF-jB also having anti-apoptotic effects, the inability of calpain 3 to lower CARP-driven inhibition of NF-jB could reduce muscle cell survival, hence partly accounting for the dystrophic pattern observed in limb girdle muscu-lar dystrophy 2A, a pathology resulting from the protease deficiency Structured digital abstract
l MINT-7990388 : Titin (uniprotkb: Q8WZ42 ) physically interacts ( MI:0915 ) with CARP (uni-protkb: Q9CR42 ) by two hybrid ( MI:0018 )
l MINT-7990374 : calpain 3 (uniprotkb: P20807 ) physically interacts ( MI:0915 ) with Titin (uni-protkb: Q8WZ42 ) by two hybrid ( MI:0018 )
l MINT-7990342 : calpain 3 (uniprotkb: P20807 ) physically interacts ( MI:0915 ) with CARP (uni-protkb: Q9CR42 ) by two hybrid ( MI:0018 )
Abbreviations
Ankrd2, ankyrin repeat domain-containing protein 2; CARP, cardiac ankyrin repeat protein; DARP, diabetes-related ankyrin repeat protein; FRAP, fluorescence recovery after photobleaching; GFP, green fluorescent protein; MARP, muscle ankyrin repeat proteins; NF, nuclear factor; NLS, nuclear localization signals; qRT-PCR, quantitative RT-PCR; ROI, region of interest; TA, tibialis anterior; YFP, yellow fluorescent protein.
Trang 2internal propeptide is removed from its active site by
an auto-proteolytic process [3] Although the large
majority of the substrates identified are structural
pro-teins [3–5], propro-teins involved in cell metabolism [5–7]
and in the regulation of gene and protein expression
[2,7–9] were also suggested to be potential calpain 3
substrates Taken together, the ensuing cleavages were
proposed to play a role in three major physiological
processes: the orchestration of sarcomere remodeling
[10–12], the control of apoptosis [9,13] and the
regula-tion of gene expression [2,7–9]
Calpain 3 is found in several different subcellular
localizations within the muscle fiber, notably in
associa-tion with three regions of titin, a giant structural protein
spanning half the sarcomere [3,14,15] Two of these
regions, the N2A region and the M line, are involved in
the transmission of mechanical signals to signaling
path-ways In the M line, mechanical stimulation activates
the interaction of a protein complex with the kinase
domain of titin, reducing the nuclear translocation of
the transcription factor SRF and impeding gene
tran-scription [16] In the elastic N2A region, mechanical
activity stimulates the expression of the muscle ankyrin
repeat proteins (MARPs), a family of gene expression
regulators [17,18] Passive stretch also induces a
subcel-lular redistribution of the MARPs, suggesting a
titin-N2A-mediated link between stress signals and gene
expression [18] The MARP family is composed of three
proteins, ankyrin repeat domain-containing protein 2
(Ankrd2), cardiac ankyrin repeat protein (CARP) and
diabetes-related ankyrin repeat protein (DARP),
grouped together with respect to their common minimal
structure and their potential role in the control of
tran-scription [17,19–21] Although the three MARPs are
expressed in both heart and skeletal muscle [18,22,23],
Ankrd2 is mainly expressed in skeletal muscle [18,24,25],
CARP in the heart [18,21,26–28] and DARP in
equiva-lent amounts in both tissues [20]
Interestingly, Ankrd2 was previously suggested to be
cleaved by calpain 3 [29] but CARP, which was shown
to be the first MARP whose expression increases in
response to exercise in skeletal muscle [30], was not
assessed as a substrate The structure of CARP
com-prises several ankyrin-like repeats, PEST motifs (i.e
regions of protein instability rich in proline, glutamic
acid, serine and threonine) and putative nuclear
locali-zation signals (NLS) [18,19,26,28] In the heart, CARP
expression increases in remodeling conditions
associ-ated with pathological hypertrophy [31–34] In the
skel-etal muscle, CARP expression is low under basal
conditions but was reported to be induced in several
conditions such as exercise [30,35–38], atrophy [26] and
muscle pathologies [39–43] From a molecular point of
view, CARP is known to act as a transcriptional regula-tor Indeed, CARP can bind to DNA [19] and inhibits the transcription of MLC-2V by association with the transcription factor YB1 in the heart [21]
Considering that (a) a molecular complex encom-passes calpain 3 and CARP in the N2A elastic region [18]; (b) exercise stimulates both calpain 3 activity [44] and CARP expression [30] in skeletal muscle; (c) cal-pain 3 was previously suggested to cleave unidentified regulators of transcription [9]; and (d) a member of the MARP family was previously demonstrated to be cleaved by calpain 3 [29], the present study aimed to identify the possible functional relationship(s) between CARP and calpain 3 and the physiological pathway(s) under control We first showed that calpain 3 cleaves CARP in vitro Once cleaved, the long C-terminal part
of CARP is more efficiently bound to titin, possibly impeding CARP nuclear translocation and any subse-quent gene expression regulation In addition, we dem-onstrated that CARP regulates the transcriptional activities of several transcription factors, including nuclear factor NF-jB p65 Together, CARP and con-sequently calpain 3 appear to have a central role in the regulation of the important cell fate-governing NF-jB pathway
Results
CARP is a substrate of calpain 3 Considering the sarcomeric localization of both CARP and calpain 3 and the fact that calpain 3 cleaves another member of the MARPs family, the possibility that CARP could be processed by active calpain 3 was investigated A direct, in vitro digestion of CARP by calpain 3 using recombinant proteins could not be attempted because calpain 3 is inactivated during purifi-cation [45] In skeletal muscle cells, calpain 3 is consid-ered to be inactive until specific signals trigger its dissociation from a muscle specific inhibitor [46] We therefore tested our hypothesis using ectopic gene expression in non muscular cells, the only system lead-ing to uncontrolled activation of calpain 3 NIH-3T3 fibroblasts were transfected with expression plasmids encoding CARP (pcDNA-CARP-V5) in the presence
of wild-type or catalytically-inactive C129S-mutated calpain 3 (encoded by CFP and pYFP-C3-C129S-CFP, respectively) The activation of the protease was confirmed by the appearance of an autolysis band
at approximately 55 kDa on a calpain-specific western blot (Fig 1A) CARP detection was performed using an antibody raised against the C-terminal V5 epitope In the presence of the protease-dead C129S calpain 3,
Trang 3CARP migrates at an apparent molecular weight of
approximately 40 kDa A lower band is clearly visible
in the presence of wild-type calpain 3 (37 kDa for the
shorter form), demonstrating that CARP is cleaved in
the presence of calpain 3 in vitro (Fig 1A)
CARP and calpain 3 interaction was tested using
yeast two-hybrid experiments Because the ectopic
expression of wild-type calpain 3 leads to uncontrolled
proteolysis, a construct encoding catalytically inactive
calpain 3 fused to GAL 4 binding domain (pAS-C3)
was used as a bait and a construct encoding CARP
fused with the activation domain (pGAD-CARP) was
used as a prey The yeasts resulting from the mating of
clones transformed with either calpain 3 or CARP
grow on Leu-Trp-His- selection medium, indicating
that calpain 3 and CARP interact (Fig 1B) The fact
that calpain 3 interacts directly with CARP supports
the idea that the cleavage is direct
Efforts to identify CARP cleavage site by protein
sequencing were unsuccessful We therefore
con-structed several N-terminal truncated forms of CARP
(pDNter1, pDNter2, pDNter3 and pDNter4; see
Materi-als and methods) with respect to the different domains
of this protein (Fig 1C) First, after expression in
NIH3T3, their migration patterns were compared by
immunoblotting with the profiles observed upon
cal-pain 3 mediated-CARP cleavage (Fig 1D) The
plas-mid pDNter2 produces a band that matches exactly
CARP cleaved C-terminal fragment (with an apparent
molecular weight of 37 kDa on SDS⁄ PAGE) Second,
the migration patterns of CARP, DNter1 or DNter2 in
presence or absence of calpain 3 were compared
(Fig 1E) Although DNter1 is cleaved when
co-expressed with calpain 3, DNter2 remains unchanged
(Fig 1E), suggesting that the position of the cleavage
site is between amino acids 30 and 71 Interestingly, in this region, three overlapping sequences fit the poten-tial consensus recently reported for calpain 3 cleavage sites almost perfectly (Fig 1C, bottom) [47] Since these sequences are localized between amino acids 65 and 88 and the cleavage site is localized before amino acid 71, the region of cleavage would be between amino acids 65 and 71 Considering the location of the N-terminal extremity of DNter2 on the structure of CARP and the presence of the potential cleavage sites,
we identified the localization of the cleavage site within
a predicted coiled-coil domain (Fig 1B)
Calpain 3-mediated CARP cleavage strengthens its interaction with titin N2A
A core and a bipartite NLS were previously identified around the CARP cleavage site within the coiled-coil region (positions 71–74 and 59–76) [28] We therefore investigated whether calpain 3 activity could influence CARP subcellular localization Plasmids encoding fluo-rescent fusion-proteins corresponding to CARP before and after cleavage by calpain 3 (pYFP-CARP-CFP-HIS, pYFP-DNter2-CFP-HIS and pYFP-Nter-CFP-HIS) were injected and transferred by electroporation
in tibialis anterior (TA) muscles of 129SvPasIco wild-type mice Seven days later, the muscles were exposed and submitted to direct observation using a confocal microscope and an excitation wavelength of 514 nm for yellow fluorescent protein (YFP) emission The set-ting for the CFP emission (excitation wavelength of
457 nm) was also attempted, but the fluorescence was much weaker than YFP and the images were blurry, impeding their analysis We therefore used YFP fluo-rescence only for further analysis
Fig 1 CARP is a substrate of calpain 3 (A) Western blot analysis performed on NIH3T3 extracts over-expressing V5-tagged CARP in the presence of either wild-type or C129S-mutated calpain 3 The appearance of a 37-kDa CARP proteolytic fragment shows that CARP is cleaved in presence of active calpain (V5 specific staining; upper panel) The activation of calpain 3 is verified by the detection of the 58 and
55 kDa autolysis fragments (calpain 3 specific staining; lower panel) (B) Yeast two-hybrid assessment of calpain 3-CARP interaction The calpain 3 construct was mutated on its active site to prevent uncontrolled proteolysis Yeasts resulting from the mating of clones trans-formed with calpain 3 or CARP were grown on Trp-Leu- (control medium; lower panels) or Trp-Leu-His- medium (selective medium; upper panels) As a positive control, an interaction test of calpain 3 and N2A-titin is performed (upper left panel) The yeasts carrying calpain 3 and CARP grow on the selective medium, indicating that CARP and calpain 3 interact (upper middle panel) (C) Schematic representation of CARP structure (top) and sequence (bottom) indicating the presence of two PEST domains (light gray colored box), a coiled-coil region (gray colored box), five ankyrin repeats (five dark gray boxes), two core NLS (in red) and a bipartite NLS (in yellow; the bipartite NLS encompass-ing one of the core NLS) The region of interaction with titin-N2A is highlighted in bold ⁄ blue, as well as by bold ⁄ blue underlined characters
in the sequence The consensus site for calpain 3 cleavage and the positions of the three imperfect cleavage sequences identified in CARP are shown at the bottom The truncated constructs (DNter1-4 and NterCARP) are shown below the CARP structure and the calpain 3 cleav-age site is indicated by an arrow (D) Western blot analysis performed on NIH3T3 extracts over-expressing either the full-length or the trun-cated CARP constructs (DNter1–4) The molecular weight of DNter2 matches the lower band detected when CARP is co-expressed with calpain 3 (37 kDa band; compare the first and the fourth lane) (E) Western blot analysis performed on NIH3T3 extracts over-expressing CARP or the truncated DNter1 or DNter2 CARP constructs, in the presence or absence of calpain 3 DNter1 is cleaved when co-expressed with calpain 3, whereas DNter2 is not.
Trang 4The staining for pYFP-CARP-CFP-HIS clearly
demonstrated a mixed nuclear and cytoskeletal pattern
(Fig 2A; note the striated pattern of fluorescence at
higher magnification, lower left panel), as previously
found in heart cells, where it was initially identified as
a partner of skeletal muscle titin-N2A [18] An analysis
of CARP expression in the subcellular compartments
obtained from the muscle of the mice injected with pYFP-CARP-CFP-HIS confirms the presence of the protein in the nucleus and on the cytoskeletal fraction (Fig 2B) Because we also confirmed using a two-hybrid assay that CARP is able to interact with titin-N2A (Fig 2D, top left panel), the fluorescent cytoskel-etal staining most likely corresponds to the sarcomeric
Trang 5location of titin-N2A In the nucleus, a spotted pattern
is clearly distinguished at very high magnification,
sug-gesting that CARP is localized into a very peculiar, yet
unidentified, nuclear subcompartment (Fig 2A, lower
right panel) This pattern is reminiscent of the PML
bodies, a compartment in which Ankrd2 has previously
been observed [17] The localization of
pYFP-DNter2-CFP-HIS is undistinguishable from the
pYFP-CARP-CFP-HIS localization (Fig 2C, upper panel) By
contrast, the short fragment CARP-Nter has a very
different localization pattern Indeed, its expression is
scattered throughout the fiber without any sarcomeric
pattern, and it does not translocate into the nucleus
(Fig 2C, lower panel) This undefined localization,
taken together with the absence of physiologically
rele-vant protein domains in this region, suggests that this
fragment is probably devoid of biological activity
To further investigate the possibility of translocation
in between various cell compartments, the strength of
the interaction between CARP and the muscle
sarco-mere was assessed in vitro using a two-hybrid assay and
in vivousing fluorescence recovery after photobleaching
(FRAP) Two-hybrid experiments were carried out
between yeast competent cells transformed either with
pAS-N2A-titin fused with GAL4-binding domain or
pGAD-CARP or DNter2 fused with GAL4-activation
domain The growth of the clones is more important
when titin-N2A is expressed with DNter2, suggesting
that the weak interaction detected between CARP and
titin-N2A is reinforced after CARP cleavage (Fig 2D)
FRAP analysis is commonly used to quantify the
mobility of a fluorescent molecule in a cell
compart-ment of interest [48] FRAP expericompart-ments were carried
out after injection of CARP-CFP-HIS,
pYFP-DNter2-CFP-HIS or pYFP-Nter-CFP-HIS in the TA
of 129SvPasIco mice The fluorescence recovery speed
observed in the presence of the Nter protein is so rapid
that we could not even bleach a region of interest
(ROI) efficiently, impeding FRAP measurement (data
not shown) This result suggests that the short
N-ter-minal CARP fragment is freed from the sarcomere,
which is consistent with the results obtained by direct
localization of the fluorescence and with the fact that
this fragment does not bear the binding site for
titin-N2A (Fig 1C) After photobleaching, the recovery
speed of DNter2 is significantly slower than the
recov-ery speed of CARP, suggesting that the long
C-termi-nal CARP fragment is more efficiently bound to titin
after cleavage by calpain 3 (P < 0.01; Fig 2E) It is
worth noting that, in skeletal muscle, an endogenous
inhibitor maintains calpain 3 in an inactive state until
a signal, such as eccentric exercise [44], activates its
proteolytic functions As a result, after 7 days of
expression in 129SvPasIco mice, only a minor propor-tion of the CARP substrate is cleaved, as indicated by the clear sarcomeric pattern (Fig 2A, upper left panel) Considering that the weak proportion of YFP-Nter protein resulting from the cleavage cannot be bleached, the comparison of the FRAP results obtained with CARP or DNter2 in wild-type animals does not take into account anything else other than the motilities of these proteins These results suggest that, once cleaved
by calpain 3, the C-terminal region of CARP binds more efficiently to the sarcomeric N2A region, possibly reducing CARP nuclear translocation
Because CARP was previously suggested to be able
to form a dimer [49], we aimed to determine whether the cleavage of a molecule of CARP could affect the sarcomeric binding of another uncleaved CARP mole-cule Accordingly, we compared CARP subcellular localization in the presence or absence of calpain 3 using a new calpain 3 knockout mouse model (C3-null) generated by disruption of the calpain 3 gene using homologous recombination (Figs S1 and S2 and Doc S1) Although a weak quantity of calpain 3 mutated mRNA is still expressed (< 20% of the wild-type level) (Fig S1B), western blot analysis confirmed the complete knockout of the protein in this murine model (Fig S1C) CARP subcellular localization and mobility were assessed after injection of a plasmid encoding pYFP-CARP-CFP-HIS in the TA muscles of C3-null and 129SvPasIco strains Since CARP will not
be processed by calpain 3 in C3 deficient animals and will only be slightly processed in wild-type animals, the full-length CARP protein is therefore the main YFP-tagged protein present in both cases With respect to CARP localization, no significant difference was noted: in both models, CARP is localized in the nucleus, as well as on the fiber sarcomere, easily recog-nizable by the striated pattern of the fluorescence (Fig 3A) In FRAP experiments (Fig 3B), the fluores-cence recovery speed is significantly slower in wild-type muscles than in calpain 3 deficient muscles (Fig 3C), suggesting that the interaction between CARP and titin is reinforced in the presence of cal-pain 3 These results suggest that the calcal-pain 3-medi-ated cleavage of some molecules of CARP reinforces the interaction of other unprocessed CARP molecules with the sarcomere
Taken together, the results obtained in the present study appear to corroborate that, once cleaved by cal-pain 3, the C-terminal part of CARP, as well as unprocessed CARP molecules, bind more efficiently to titin N2A, which is consistent with the fact that CARP nuclear translocation could be controlled by calpain 3 activity
Trang 6In vitro, CARP can act as a regulator of
transcription factors activity in the nucleus
Considering that CARP is a known regulator of
tran-scription in the heart, we investigated the possibility
that CARP might play a similar role on gene
regula-tion in skeletal muscle The DNA binding activities
of nuclear proteins from C2 myotubes transfected either with pcDNA3-CARP or with a mock vector (pcDNA3-lacZ) were compared using a membrane-based analysis (protein⁄ DNA array) of a set of 345 pre-selected transcription factors Sixty-eight transcrip-tion factors appear to be regulated by CARP (cut-off ratio CARP⁄ lacZ < 0.6 or > 1.3) Apart from pax5,
Fig 2 CARP cleavage strengthens its interaction with the titin N2A region (A) Localization of YFP-CARP in the mouse TA after electrotrans-fer CARP is localized both on the sarcomere (lower left panel) and in the nucleus (lower right panel) of the fibers Scale bars = 20 lm (B) Analysis of CARP expression in subcellular compartments of TA transduced by YFP-CARP (detection by GFP-specific western blot) confirms that CARP is present on the cytoskeleton and the nucleus in skeletal muscle (C) Localization of YFP-DNter2 and YFP-Nter in the mouse TA after electrotransfer Similar to CARP, DNter2 is localized on the sarcomere and in the nucleus (upper panel), whereas the Nter-CARP fluo-rescence (lower panel) is scattered throughout the fiber (D) Yeast two-hybrid assay of the interaction of titin with CARP or DNter2 Yeasts resulting from the mating of clones transformed with titin or CARP were grown on Trp-Leu- (lower panels, control medium) or Trp-Leu-His-medium (upper panels, selective Trp-Leu-His-medium) On the selective Trp-Leu-His-medium, the yeasts carrying titin and DNter2 grow more than the yeasts carry-ing titin and CARP, suggestcarry-ing that, once cleaved, the association of the long C-terminal fragment of CARP and titin is reinforced (upper panels) (E) Quantification of FRAP experiments FRAP was measured in several ROI after photobleaching at 514 nm in mouse TA trans-duced with YFP-tagged CARP or DNter2 The fluorescence recovery speed is slower when CARP is truncated (i.e slower in the presence of the DNter2 construct compared to the CARP construct) (**P < 0.01, n = 12).
Trang 7whose DNA-binding activity is slightly elevated in the
presence of CARP (on one out of the two consensus
sequences considered), all the other transcription
fac-tors are inhibited (Fig 4A) The 32 factors most
signif-icantly inhibited (cut-off ratio CARP⁄ lacZ < 0.5) are
presented schematically in Fig S3 Although its
inhibi-tion does not reach this level with this quantificainhibi-tion
method, the activity of the transcription factor NF-jB,
as measured with three different consensus sequences,
is consistently repressed This transcription factor
appeared to be particularly interesting considering that
it was previously described as having a role in muscle
atrophy [50] and is abnormally distributed subsequent
to calpain 3 deficiency [8]
Using a quantitative ELISA-based method, we
con-firmed that, when CARP is significantly over-expressed
by two-fold, NF-jB p65 transcriptional activity is
sig-nificantly decreased two-fold (P < 0.05; Fig 4B)
Using the quantification of its messenger level on
RNA extracts of the same cells, we confirmed that this
transcription factor is not transcriptionally regulated
(Fig 4C) and concluded that its nuclear translocation
or its activity is modulated by CARP We also
performed real-time quantitative RT-PCR (qRT-PCR)
of MuRF1, an E3 ubiquitin ligase whose transcription
is up-regulated through NF-jB activation in atrophic muscle fibers [51–53] Interestingly, MuRF1 expression remains constant in this experiment, strongly suggest-ing that the correspondsuggest-ing signalsuggest-ing pathway regulated
by NF-jB does not involve this protein (Fig 4D)
Discussion
In the present study, we provide new insights into the regulation and function of the molecular complex encompassing CARP and calpain 3 in the N2A region
We identified CARP as a new calpain 3 substrate Interestingly, this cleavage regulates CARP subcellular localization by increasing the strength of its interaction with the sarcomere In addition, we investigated the modification of transcription factor activities induced
by CARP over-expression and demonstrated CARP-induced regulation of NF-jB activity
Even though CARP bears two potential PEST insta-bility regions, its cleavage does not occur in any of these sequences but takes place in a strongly structured coiled-coil region [49] A core and a bipartite NLS were previously predicted to be encoded in this region, the bipartite NLS encompassing the core NLS (Fig 1B) [28] It should be noted that an additional core NLS is present downstream of this region (posi-tion 94–98; Fig 1B) [19,28,54] Cleavage by calpain 3 would theoretically disrupt the bipartite NLS and leave the two core NLS intact on CARP C-terminal region The N-terminal region liberated by the cleavage has no NLS left and is consistently never observed inside the nucleus However, it is possible that the loss of one NLS in the C-terminal fragment affects the nuclear transport of this form of CARP, although the sensitiv-ity of the methods we used could not confirm this hypothesis
The results obtained in the present study strongly suggest that, once cleaved, CARP interaction with the region N2A is reinforced CARP interacts with titin-N2A using a region situated in its second ankyrin repeat (Fig 1B) [18] Bio-informatics analysis indicated that this region remains structurally unaffected by the cleavage, whereas the coiled-coil region appears to be destructed (for methodology, see Materials and meth-ods) In addition to carrying NLS, this region was previously proposed to be involved in the homodi-merization of CARP [49] We therefore propose that the loss of CARP dimerization promotes the binding to titin by improving the accessibility of the titin-binding domain Interestingly, the importance of CARP inter-action for its function was recently demonstrated in a
Fig 3 Calpain 3 produces a reinforcement of CARP interaction
with titin (A) CARP localization after electrotransfer of YFP tagged
CARP in TA muscles from wild-type (left) and C3-null (right) mice.
In both models, CARP is localized in the nucleus and on the
sarco-mere of the fibers Scale bars = 20 lm (B) Quantification of FRAP
experiments FRAP was measured for several ROI after
photoble-aching at 514 nm in TA muscles from wild-type and C3-null mice
transduced with YFP-tagged CARP The fluorescence recovery
speed is slower in muscles of animals bearing functional calpain 3
(**P < 0.01, n = 10).
Trang 8pathophysiological context since pathogenic mutations
result in the loss of CARP binding to talin and FHL2
and, consequently, in the perturbation of its function
[55]
Regulation of function through the control of
sub-cellular localization represents novel information for
a member of the calpain family, although it was
previously described for other proteases such as casp-ases [56] Several different subcellular traffic mechanisms controlling transcription are known to be controlled by protein cleavage In particular, the liberation and nuclear translocation of transcription factors can be triggered by proteasome-mediated degradation of a partner, as exemplified by the prototypical regulation of
Fig 4 In vitro, CARP can act as a regulator of transcription factors activity in the nucleus (A) Effect of CARP on the DNA-binding activities
of 345 transcription factors DNA-binding activities were measured on nuclear extracts of C2 myotubes over-expressing either CARP or lacZ (control) Black boxes highlight the 32 transcription factors that were most significantly inhibited (cut-off ratio CARP ⁄ lacZ < 0.5) Although less severely, the DNA-binding activity of NF-jB is also inhibited (single white boxes indicate NF-jB DNA-binding activities measured on three different consensus sequences) Pax5 is the only factor whose DNA-binding activity is slightly elevated (double white boxes) (B) Quan-tification of NF-jB DNA-binding activities in C2 myotubes over-expressing CARP The DNA-binding activity of the NF-jB isoform p65 is sig-nificantly inhibited when CARP is over-expressed (*P < 0.05, n = 3) (C) Real-time quantification of the mRNA level of NF-jB p65 in myotubes over-expressing CARP The gene expression of NF-jB p65 does not vary with CARP over-expression (n = 3) (D) Real-time quanti-fication of the mRNA level of MuRF1 in myotubes expressing CARP The gene expression of MURF1 does not vary with CARP over-expression (n = 3).
Trang 9the factor NF-jB by the protein IjBa [57]
Addition-ally, the transmembrane receptor Notch is the target
of ligand-dependent proteolysis and one of the
frag-ments released migrate into the nucleus to regulate
gene expression [58] The results reported in the
pres-ent study illustrate a novel mechanism of gene
regula-tion through nuclear translocation inhibition,
combining the destruction of an NLS with an increase
in the affinity of the targeted gene regulator for one of
its partners
The sarcomeric sequestration consecutive to calpain
3 activation might, as a consequence, control
CARP-dependent gene expression Indeed, MARPs are
con-sidered to be involved in gene transcription because (a)
Ankrd2 localizes in euchromatin, the region of
chro-matin where active gene transcription occurs [59], is
able to bind to three transcription factors, YB-1, PML
and p53, and enhances the up-regulation of the
p21(WAFI⁄ CIPI) promoter by p53 [17] and (b) CARP
can bind to DNA [19] and is a negative regulator of
the transcription factor YB1 in the heart [21]
Interest-ingly, calpain 3 was also reported to participate in the
control of gene expression [2,7–9], suggesting that the
complex calpain 3⁄ CARP might comprise an axis for
gene regulation Amongst the possible CARP targets
identified in the present study, NF-jB p65 DNA
bind-ing activity was confirmed to be inhibited by CARP
over-expression Interestingly, we previously
demon-strated that calpain 3 possibly participates in the
control of the NF-jB pathway because calpain 3
deficiency is associated with an altered distribution of
both NF-jB and of its regulator IjBa [8], as well as
with blockade of the induction of specific
anti-apopto-tic NF-jB target genes such as c-FLIP [9] From a
mechanistic point of view, it could be postulated that a
direct interaction between CARP ankyrin repeats and
NF-jB p65 is the cause of a cytoplasmic sequestration
(and hence inhibition) of NF-jB, similar to IjB which
associates through its ankyrin repeats with NF-jB
[60] However, CARP could also act upstream of a
sig-naling cascade controlling directly NF-jB activity as
the transcription factor is now known to be regulated
by both phosphorylation and acetylation [61]
Interest-ingly, it was previously reported that the inhibition of
the NF-jB pathway during the induction of apoptosis
induces CARP upregulation, suggesting that a positive
feedback mechanism could exacerbate this
phenome-non [62] On the other hand, a recent study shows that
the stimulation of NF-jB activity by skeletal muscle
longitudinal stretch up-regulates Ankrd2 expression
through direct stimulation of its promoter [63] Taken
together, these studies suggest that the NF-jB pathway
might be a key differential regulator of the expression
of the MARPs This differential regulation might be necessary to tune muscle signaling pathways with pre-cision in response to various physiological stimuli Under which conditions the pathway identified in the present study is physiologically relevant and how its dysfunction participates in the pathogenesis of LGMD2A represent two important issues that remain
to be addressed The NF-jB pathway is a key regula-tor of numerous cellular events, such as proliferation and differentiation, and catabolic or apoptotic path-ways, in many organs In particular, it was previously established that NF-jB is a major inducer of muscle atrophy in the skeletal muscle Indeed, NF-jB inhibi-tion in mice models invariably protects against muscle atrophy, whereas NF-jB activation promotes proteoly-sis in vivo [51,64–66] However, in our hands, although the over-expression of CARP in muscle cells results in NF-jB p65 inhibition, it does not affect the expression
of MURF1, which is one of the main mediators of NF-jB-dependent muscle atrophy [51] It was also pre-viously suggested that p65 is not the member of the NF-jB family involved in the induction of atrophy [64] Taken together, CARP-dependent NF-jB inhibi-tion therefore appears unlikely to play a role in muscle atrophy On the other hand, several studies have sug-gested a possible involvement of NF-jB in muscle cell survival through induction of anti-apoptotic factors [8,9,67] Calpain 3 deficiency was previously reported
to be associated with a deregulation of the NF-jB pathway and an increase in muscle fiber apoptosis [8] The participation of NF-jB signaling in the pathogen-esis of LGMD2A is therefore an interesting possibility The findings obtained in the present study lead to a proposed working hypothesis: in the absence of calpain
3, CARP nuclear activities would be exacerbated, which would lead to a decrease in NF-jB activity (Fig S4) NF-jB inhibition would impede the protec-tion of muscle from apoptosis, an event leading to progressive muscle destruction In line with this hypothesis, CARP ectopic expression was previously reported to be able to induce apoptotic cell death in hepatoma cells [62] In conclusion, calpain 3, through its action on CARP, appears to have a central role in regulating important cell fate-governing pathways
Materials and methods
Plasmid constructions and antibodies
The sequences encoding full-length DNter1-4 and Nter CARP were amplified by PCR on a random primed cDNA library obtained by reverse transcription of murine 129SVter skeletal muscle RNA (primers indicated in
Trang 10Table 1) PCR products were cloned in pcDNA3.1D⁄
V5-His-Topo using the TOPOcloning technology (Invitrogen,
Carlsbad, CA, USA) After digestion by XhoI and BamHI,
the inserts of the resulting plasmids were subcloned into
pYFP-CFP-HIS, a plasmid carrying the enhanced YFP at
the 5¢ end of the cloning site The plasmids pYFP-C3-CFP
and pYFP-C3-C129S-CFP were previously described and
bear the murine calpain 3 coding sequence (wild-type or
C129S protease-dead mutant respectively) between
enhanced YFP in 5¢ and eCFP in 3¢ [3] The plasmid
pcDNA3-lacZ was obtained from Invitrogen Every
ampli-fied sequence was confirmed by automated sequencing
For two-hybrid experiments, the cloning of the N2A
region of titin in the pGAD vector (Clontech, Mountain
View, CA, USA) and of human calpain 3 in the pAS vector
(Clontech) were described previously [68] The calpain 3
construct carries the C129S mutation, which invalidates
the protease activity of calpain 3 The N2A region of titin
(exons 101–110) was PCR amplified (see primers in Table 1)
from a random primed cDNA library obtained by reverse
transcription of an adult human skeletal muscles poly(A)
RNA library (Ambion AM7983; Ambion, Austin, TX,
USA) The PCR product was digested by XmaI and NcoI
and cloned in fusion with the GAL4 DNA-binding domain
in pAS CARP and DNter2 cDNA were fused to GAL4
activation domain in pGAD Briefly, CARP and DNter2
were PCR amplified with primers containing the restriction
sites NcoI in the 5¢ primer and XmaI in the 3¢ primer
(Table 1) The digested fragments were ligated in EcoRI⁄
BamHI linearized pGAD Every amplified sequence was
validated by automated sequencing
Rabbit polyclonal antibody directed against the epitope
QESEEQQQFRNIFKQ in exon 17 of the calpain 3 (B3)
was kindly provided by Dr Ahmed Ouali (INRA UR 370,
Saint Genes Champanelle, France) and has been described previously [8] NF-jB-specific rabbit polyclonal antibody was obtained from Chemicon Mouse monoclonal antibody specific for the V5 epitope was purchased from Invitrogen Horseradish peroxidase linked donkey anti-rabbit IgG and sheep anti-mouse IgG antibodies were obtained from GE Healthcare (Piscataway, NJ, USA)
Cell culture and transfection
The NIH3T3 cell line was obtained from the American Type Culture Collection (Rockville, MD, USA) and the C2 mouse myoblasts from the ATCC [69] Fibroblasts and myoblasts were cultured in DMEM containing gentamicin (10 lgÆmL)1) and supplemented with 10% or 20% fetal bovine serum (HyClone Thermo Scientific, Hudson, NH, USA), respectively Myogenic differentiation of C2 cells was initiated by replacement of the growth medium with DMEM containing 5% horse serum (Gibco Invitrogen, Carlsbad, CA, USA) and the subsequent maintenance of the cells in this medium for 4–10 days
For plasmid transfections, cells were plated (300 000 C2 cells or 1 000 000 NIH3T3 cells per 100 mm dish) and allowed to grow for 24 h Transfections were performed with 6 lg of plasmid and 30 lL of FuGENE 6 transfection reagent (Roche Applied Science, Indianapolis, IN, USA)
In case of co-transfections, plasmids were mixed at equimo-lar concentrations To increase C2 transfection efficiency, the same transfection method was used a second time before the start of myogenic differentiation
Preparation of protein samples and immunoblotting
Cells were washed with NaCl⁄ Pi and proteins were extracted using a buffer containing 20 mm Tris (pH 7.5),
150 mm NaCl, 2 mm EGTA, 1% Triton X-100, 2 lm E64 and protease inhibitors (Complete mini protease inhibitor cocktail; Roche Applied Sciences) After centrifugation at
10 000 g for 10 min at 4C, the supernatants were recov-ered for western blot analysis
The muscle proteins of the different sub-cellular compart-ments were extracted using the ProteoExtractSubcellular Proteome Extraction Kit (S-PEK; Calbiochem Merck KGaA, Darmstadt, Germany) Briefly, TA muscles were homogenized in 1 mL of lysis buffer with a Fast-Prep instru-ment (MP-Biomedicals, Solon, OH, USA), and proteins of the cytosol, membranes, nucleus and cytoskeleton were extracted in accordance with the manufacturer’s instructions The samples were denatured before SDS⁄ PAGE using LDS NuPage buffer (Invitrogen) supplemented with 100 mm dithiothreitol Sample protein concentrations were deter-mined by the BCA methodology (Thermo Scientific, Rock-ford, IL, USA) Protein samples were submitted to SDS⁄ PAGE in precast 4–12% acrylamide gradient gels
(Nu-Table 1 Primers used for cloning.
Upper primer Lower primer pGAD-CARP Full-length
CARP
CGCCATGGCAATGATGGTACTGAAAGTAGAGG CGGCCCGGGAACTGATTAAGAGTCTGTCG pGAD-DNter2 CARP from
71 to 319
GAGCCATGGAACAACGGAAAAGCGAGAAAC CGGCCCGGGAACTGATTAAGAGTCTGTCG
pYFP-CARP-CFP-HIS
Full-length
CARP
1–319
CACCATGATGGTACTGAGAG GAATGTAGCTATGCGAGAGTTC
30 to 319
CACCATGGCCGAGTTCAGAAATGGAGAAG GAATGTAGCTATGCGAGAGTTC
71 to 319
CACCATGCTGAAGACACTTCCGGCCAACAG GAATGTAGCTATGCGAGAGTTC
102 to 319
CACCATGCTGAAAGCTGCGCTGGAGAAC GAATGTAGCTATGCGAGAGTTC
124 to 319
CACCATGACCAAAGTTCCAGTTGTGAAGG GAATGTAGCTATGCGAGAGTTC
1 to 70
CACCATGATGGTACTGAGAG GAATGTAGCTATGCGAGAGTTC