Tài liệu Báo cáo khoa học: Autolytic activity of human calpain 7 is enhanced by ESCRT-III-related protein IST1 through MIT–MIM interaction pptx

Autolytic activity of human calpain 7 is enhanced by
ESCRT-III-related protein IST1 through MIT–MIM
interaction
Yohei Osako, Yuki Maemoto, Ryohei Tanaka, Hironori Suzuki, Hideki Shibata and Masatoshi Maki
Department of Applied Molecular Biosciences, Graduate School of Bioagricultural Sciences, Nagoya University, Japan
Keywords
calpain 7; ESCRT-III; IST1; microtubule-
interacting and transport (MIT); proteolysis
Correspondence
M. Maki, Department of Applied Molecular
Biosciences, Graduate School of
Bioagricultural Sciences, Nagoya University,
Furo-cho, Chikusa-ku, Nagoya 464-8601,
Japan
Fax: +81 52 789 5542
Tel: +81 52 789 4088
E-mail:
(Received 10 May 2010, revised 21 July
2010, accepted 20 August 2010)
doi:10.1111/j.1742-4658.2010.07822.x
Calpain 7, a mammalian ortholog of yeast Cpl1 ⁄ Rim13 and fungal PalB, is
an atypical calpain that lacks a penta-EF-hand domain. Previously, we
reported that a region containing a tandem repeat of microtubule-interact-
ing and transport (MIT) domains in calpain 7 interacts with a subset of
endosomal sorting complex required for transport (ESCRT)-III-related
proteins, suggesting involvement of calpain 7 in the ESCRT system.
Although yeast and fungal calpains are thought to be involved in alkaline
adaptation via limited proteolysis of specific transcription factors, proteo-
lytic activity of calpain 7 has not been demonstrated yet. In this study, we
investigated the interaction between calpain 7 and a newly reported ESC-
RT-III family member, increased sodium tolerance-1 (IST1), which pos-

sesses two different types of MIT-interacting motifs (MIM1 and MIM2).
We found that glutathione-S-transferase (GST)-fused tandem MIT
domains of calpain 7 (calpain 7MIT) pulled down FLAG-tagged IST1
expressed in HEK293T cells. Coimmunoprecipitation assays with various
deletion or point mutants of epitope-tagged calpain 7 and IST1 revealed
that both repetitive MIT domains and MIMs are required for efficient
interaction. Direct MIT–MIM binding was confirmed by a pulldown exper-
iment with GST-fused IST1 MIM and purified recombinant calpain 7MIT.
Furthermore, we found that the GST–MIM protein enhances the autolysis
of purified Strep-tagged monomeric green fluorescent protein (mGFP)-
fused calpain 7 (mGFP–calpain 7–Strep). The autolysis was almost com-
pletely abolished by 10 mm N-ethylmaleimide but only partially inhibited
by 1 mm leupeptin or E-64. The putative catalytic Cys290-substituted
mutant (mGFP–calpain 7
C290S
–Strep) showed no autolytic activity. These
results demonstrate for the first time that human calpain 7 is proteolytically
active, and imply that calpain 7 is activated in the ESCRT system.
Structured digital abstract
l
MINT-7990193, MINT-7990213, MINT-7990233: calpain 7 (uniprotkb:Q9Y6W3) physically
interacts (
MI:0915) with IST1 (uniprotkb:P53990)byanti tag coimmunoprecipitation (MI:0007)
l
MINT-7990176: calpain 7 (uniprotkb:Q9Y6W3) physically interacts (MI:0915) with IST1
(uniprotkb:
P53990)bypull down (MI:0096)
l
MINT-7990252: IST1 (uniprotkb:P53990) binds (MI:0407)tocalpain 7 (uniprotkb:Q9Y6W3)
by pull down (

MI:0096)
Abbreviations
ALLNal, N-acetyl-
L-leucyl-L-leucyl-L-norleucinal; CBB, Coomassie Brilliant Blue R-250; CHMP, charged multivesicular body protein; CSD1,
calpastatin domain 1; ESCRT, endosomal sorting complex required for transport; GFP, green fluorescent protein; GST, glutathione-S-transferase;
IST1, increased sodium tolerance-1; mGFP, monomeric green fluorescent protein; MIM, microtubule-interacting and transport-interacting motif;
MIT, microtubule-interacting and transport; pAb, polyclonal antibody; VPS, vacuolar protein sorting; WB, western blot.
4412 FEBS Journal 277 (2010) 4412–4426 ª 2010 The Authors Journal compilation ª 2010 FEBS
Introduction
Calpains are a family of intracellular Ca
2+
-dependent
cysteine proteases [1–3]. Well-studied typical mamma-
lian calpains (l-calpain and m-calpain) are composed
of a catalytic large subunit and a regulatory small sub-
unit. Both subunits have C-terminal penta-EF-hand
domains [4], which contribute to activation of the pro-
tease by Ca
2+
binding, to heterodimerization of each
subunit, and to binding of the endogenous calpain
inhibitor calpastatin [5,6]. Although the detailed
molecular mechanisms are still unknown, ubiquitously
expressed calpains, represented by l-calpain and m-cal-
pain, have been suggested to be involved in fundamen-
tal biological phenomena such as regulation of the cell
cycle and signal transduction [1,3,7–9]. On the other
hand, tissue-specific calpains, such as p94 ⁄ calpain 3
and nCL-2 ⁄ calpain 8, have been suggested to have spe-
cific roles [10–12].

As the mRNA of calpain 7 is expressed ubiquitously
in human tissues, calpain 7 is expected to have
fundamental and essential roles in mammalian cells [13].
Studies on calpain 7 have been preceded by those on
yeast and fungal orthologs (Cpl1 ⁄ Rim13 and PalB,
respectively), and accumulating data indicate that Cpl1
and PalB play critical roles in alkaline adaptation via
processing of transcription factors Rim101 ⁄ PacC [14–
18]. However, the functions of mammalian calpain 7 are
still unknown. It has not even yet been demonstrated
whether calpain 7 has protease activity, and neither
in vivo nor in vitro substrates have been identified.
Although calpain 7 contains a C2-like domain, it lacks a
penta-EF-hand domain and is classified as an atypical
calpain. As one of the significant structural features,
mammalian calpain 7 possesses a tandem repeat of
microtubule-interacting and transport (MIT) domains
[19,20] at the N-terminus (Fig. 1A). Several MIT
domain-containing proteins have been shown to bind
endosomal sorting complex required for transport
(ESCRT)-III proteins and their related proteins [21–23].
Fig. 1. Schematic representations of
calpain 7 and IST1. (A) Calpain 7 possesses
two MIT domains (MITa and MITb) at its
N-terminus, a calpain-like cysteine protease
domain (Cys290, a putative catalytic Cys) in
the middle, and a C2-like domain at its C-ter-
minus. Catalytic triad residues are indicated
by closed triangles. (B) IST1 has a CHMP-
like domain in its N-terminal half, a Pro-rich

linker in the middle, and two different types
of MIMs (from the N-terminal side, MIM2
and MIM1, respectively) at the C-terminus.
Amino acids that are important for binding
to the VPS4 MIT domain are indicated by
open triangles. Wild-type (WT) as well as
deletion and amino acid substituted mutants
of calpain 7 and IST1 used in this study are
schematically represented. The numbers
below the bars indicate positions in amino
acid residues.
Y. Osako et al. Enhancement of calpain 7 autolysis by IST1
FEBS Journal 277 (2010) 4412–4426 ª 2010 The Authors Journal compilation ª 2010 FEBS 4413
The ESCRT system was originally identified as
machinery contributing to multivesicular endosome
(multivesicular body) formation in the endocytic path-
way [24,25]. ESCRT machinery has been proposed to
have additional roles in other membrane deforma-
tion ⁄ fission events, such as retrovirus budding and
membrane fission of daughter cells in cytokinesis [26].
The core ESCRT system is composed of four com-
plexes, termed ESCRT-0, ESCRT-I, ESCRT-II and
ESCRT-III, and associated proteins, including AAA-
type ATPase vacuolar protein sorting (VPS)4. VPS4
interacts with components of ESCRT-III via its MIT
domain, and catalyzes the dissociation of ESCRTs
from membranes [27]. We previously reported that cal-
pain 7 associates with a subset of ESCRT-III and its
related proteins: charged multivesicular body protein
(CHMP)1A CHMP1B, CHMP2A, CHMP4b,

CHMP4c and CHMP7 [28]. We also showed that cal-
pain 7 interacts with CHMP1B via its tandem MIT
domains, and that it partially colocalizes with endocy-
tosed epidermal growth factor, suggesting involvement
of calpain 7 in the ESCRT system [28].
On the basis of the resolved 3D structure, the MIT
domain of VPS4 forms three-helix bundles. ESCRT-III
proteins commonly contain conserved amino acid
sequences for MIT binding, termed MIT-interacting
motif (MIMs), in their C-termini. Two types of MIM
have been identified: MIM1 and MIM2. The former
forms an amphipathic helix that binds to the groove
between VPS4 MIT domain helices 2 and 3 [29,30],
and the latter forms a Pro-rich strand that binds
between helices 1 and 3 [31].
Human increased sodium tolerance-1 (IST1), an or-
tholog of yeast Ist1, possesses both MIM1 and MIM2
at its C-terminus [32,33]. IST1 and Ist1 can bind to
several ESCRT-related proteins, including VPS4 ⁄ Vps4
and CHMP1B ⁄ Did2 [32–35]. Interestingly, the 3D
structure of the IST1 N-terminal domain is very
similar to that of the core domain of CHMP3, an
ESCRT-III component [35,36]. Although small
interfering RNA-mediated knockdown effects on the
endocytic pathway are not evident, IST1 is required
for efficient cytokinesis in HeLa cells [32,33]. IST1 and
Ist1 associate with CHMP1 ⁄ Did2 to regulate the local-
ization and ATPase activity of VPS4 ⁄ Vps4 [32,33,37].
Because of the structural and functional resemblance
to CHMP proteins, IST1 is now regarded as a new

ESCRT-III family member.
The findings described above led us to investigate
whether calpain 7 interacts with IST1 through MIT–
MIM interactions. In this study, we examined cal-
pain 7–IST1 interactions by in vitro and in vivo binding
experiments, using purified recombinant proteins and
cultured mammalian cells expressing epitope-tagged
proteins. We also investigated the effect of this interac-
tion on the autolysis of calpain 7.
Results
Glutathione-S-transferase (GST) pulldown assay
of FLAG–IST1
To investigate whether MIT domains of calpain 7 (cal-
pain 7MIT) interact with IST1, we first performed a
GST pulldown assay (Fig. 2). GST-fused calpain7MIT
(1–165 amino acids) followed by the protease cleavage
site and His
6
-tag (GST–MIT–pHis) was purified with
His-tag affinity resin, immobilized on glutathione–
Sepharose beads, and incubated with cleared lysates of
HEK293T cells expressing FLAG-tagged CHMP1B,
CHMP4b, CHMP6 or IST1. After incubation, the
beads were pelleted by low-speed centrifugation and
washed. Cleared lysates and proteins bound to the
Fig. 2. GST–MIT–pHis pulldown assay of FLAG–IST1. HEK293T
cells were transfected with pFLAG–CHMP1B pFLAG–CHMP4b,
pFLAG–CHMP6 or pFLAG–IST1. At 24 h after transfection, cells
were lysed, and the cleared lysates were incubated with recombi-
nant GST-fused tandem MIT domains of calpain 7 (GST–MIT–pHis)

or GST–pHis (negative control) immobilized on glutathione–Sepha-
rose beads. The beads were then pelleted by low-speed centrifuga-
tion and washed with the lysis buffer. The cleared lysates (Input)
and proteins in the pellets (pulldown product, Pulldown) were sub-
jected to SDS ⁄ PAGE (10% gel) and WB, with mAb against FLAG.
Immunoreactive bands were detected by the chemiluminescence
method. Membranes were also stained with CBB. Bands of GST–
MIT–pHis and GST–pHis in the pulldown products are indicated by
arrows.
Enhancement of calpain 7 autolysis by IST1 Y. Osako et al.
4414 FEBS Journal 277 (2010) 4412–4426 ª 2010 The Authors Journal compilation ª 2010 FEBS
beads (pulldown products) were separated by
SDS ⁄ PAGE and subjected to western blot (WB) analy-
sis with mAb against FLAG or visualized by staining
with Coomassie Brilliant Blue R-250 (CBB). The inten-
sities of the immunoreactive bands for FLAG–IST1 in
the pulldown products of GST–MIT–pHis were much
stronger than those of FLAG–CHMP1B and FLAG–
CHMP4b. For FLAG–CHMP6 (negative control), no
specific immunoreactive band was detected under the
conditions used. No signals were detected in the con-
trol pulldown products of GST–pHis.
Coimmunoprecipitation of FLAG–IST1 with
monomeric green fluorescent protein (mGFP)–
calpain 7 mutants
Next, we investigated the interaction between calpain 7
and IST1 in mammalian cells by the coimmunoprecipi-
tation method. Cleared lysates (Fig. 3, Input) of
HEK293T cells coexpressing mGFP fused with cal-
pain 7 and various mutants (Fig. 1A) and FLAG–IST1

were incubated with anti-green fluorescent protein
(GFP) serum for immunoprecipitation. Clear
immunoreactive bands of FLAG–IST1 were detected
for mGFP–calpain 7, mGFP–calpain 7
C290S
(a mutant
with replacement of the putative catalytic Cys, Cys290,
by Ser), and mGFP–calpain 7MIT by WB analysis with
mAb against FLAG (Fig. 3, IP, lower panel). The
signal was weak but significant for mGFP–cal-
pain 7MITb. Signals were reduced to the background
or control level for mGFP–calpain 7MITa and mGFP–
calpain 7DMIT. The results indicated that tandem MIT
domains are required for efficient calpain 7–IST1 inter-
action. Intriguingly, the degradation bands seen in
mGFP–calpain 7 (Fig. 3, closed and open triangles)
were not detected in the case of mGFP–calpain 7
C290S
,
suggesting that the degradation was caused by proteo-
lytic activity of mGFP–calpain 7 itself. We refer to this
issue later.
Effects of mutations of IST1 MIMs on binding to
mGFP–calpain 7MIT
To investigate whether the MIM1 and ⁄ or MIM2
regions present in IST1 are responsible for interaction
with calpain 7 MIT domains, we performed a similar
coimmunoprecipitation assay with mGFP–cal-
pain 7MIT and various FLAG–IST1
MIM

deletion and
point mutants (L326D, MIM2 Leu326 replaced by
Asp; L353A, MIM1 Leu353 replaced by Ala; Fig. 1B),
which were previously shown to lose the ability to bind
to the VPS4 MIT domain [32]. As shown in Fig.4
(bottom panel), the immunoreactive band for wild-type
FLAG-IST1 (WT) was clearly detected, but for
FLAG–IST1
DMIM1
, FLAG–IST1
DMIM2
, and all MIM
Fig. 3. Coimmunoprecipitation of FLAG–IST1 with mGFP–calpain 7
mutants. HEK293T cells were cotransfected with pFLAG–IST1 and
plasmids expressing calpain 7 mutants fused with mGFP. At 24 h
after transfection, cleared lysates (Input, 10%, upper panel; 1%,
lower panel) were subjected to immunoprecipitation (IP) with anti-
GFP serum followed by WB analysis with mAb against GFP (upper
panel) and mAb against FLAG (lower panel), respectively. Proteoly-
sed fragments of mGFP–calpain 7 in wild-type (WT) and DMIT con-
structs are indicated by closed ( 45 kDa) and open ( 30 kDa)
triangles.
Fig. 4. Effects of mutations of IST1 MIMs on binding to mGFP–cal-
pain 7MIT. mGFP–calpain 7MIT and various FLAG–IST1 mutants
(see Fig. 1B) were independently expressed in HEK293T cells.
Cleared lysate from cells expressing mGFP–calpain 7MIT was
mixed with that expressing each FLAG–IST1 mutant, and each mix-
ture was subjected to coimmunoprecipitation with anti-GFP serum.
The cleared lysates (Input) and immunoprecipitated proteins (IP)
were subjected to WB analysis with mAb against GFP and mAb

against FLAG, respectively.
Y. Osako et al. Enhancement of calpain 7 autolysis by IST1
FEBS Journal 277 (2010) 4412–4426 ª 2010 The Authors Journal compilation ª 2010 FEBS 4415
point mutants, signals were significantly weakened.
The signal for FLAG–IST1
DMIM1,2
decreased to almost
the background or negative control (FLAG–CHMP6)
level.
Direct interaction between recombinant
calpain 7MIT and GST–IST1 proteins
The use of cleared lysates of HEK293T cells for all of
the experiments described above left the possibility
that unknown factors might mediate MIT–MIM inter-
actions. To exclude this possibility, we performed in vi-
tro GST-pulldown assays with purified recombinant
calpain 7MIT, which was obtained by removal of GST
and His
6
-tag by digestion with PreScission protease
followed by ion exchange chromatography. Purified
calpain 7MIT was incubated with GST–IST1 mutants
or GST immobilized on glutathione–Sepharose beads.
Pulldown products were visualized by staining with
CBB. Calpain 7MIT was pulled down by GST–IST1
and GST–MIM (Fig. 5, Pulldown, open triangle) but
not by GST–MIM
L326D,L353A
or GST.
Enhancement of autolytic activity of mGFP–

calpain 7 by calpain 7–IST1 interaction
As shown in Fig. 3, expression of mGFP–calpain 7 in
HEK293T cells generated  45 and 30 kDa fragments
(designated as 45 K and 30 K, respectively, in this arti-
cle), and those bands were not detected in the case of
mGFP–calpain 7
C290S
. A similar result was obtained
when we used HeLa cells (data not shown). mGFP–
calpain 7 was thought to be proteolysed by its own
proteolytic activity, which led us to investigate this
phenomenon further.
Estimation of cleavage sites in mGFP–calpain 7
mGFP–calpain 7, mGFP–calpain 7
C290S
and mGFP–
calpain 7
C290A
(a mutant with the putative catalytic
Cys, Cys290, replaced by Ala) were transiently
expressed in HEK293T cells, and total cell lysates were
analyzed by WB analysis with mAb against GFP or
polyclonal antibody (pAb) against calpain 7 (raised
against recombinant MIT domains [28]). In the case of
WB analysis with mAb against GFP, 45 K and 30 K
were reproducibly detected for mGFP–calpain 7 but
not for mGFP–calpain 7
C290S
and mGFP–cal-
pain 7

C290A
(Fig. 6B, upper panel, closed and open tri-
angles). With pAb against calpain 7, an  45 kDa
fragment was also detected specifically for mGFP–cal-
pain 7 (Fig. 6B, lower panel, gray triangle). These data
indicate that a putative catalytic Cys, Cys290, of cal-
pain 7 has a critical role in the wild-type-specific prote-
olysis. To examine whether the 45 kDa fragment
detected by WB analysis with pAb against calpain 7 is
identical to 45 K, cleared lysates from cells expressing
mGFP–calpain 7 or mGFP–calpain 7
C290S
were sub-
jected to immunoprecipitation with anti-GFP serum or
pAb against calpain 7, followed by WB analysis with
mAb against GFP and mAb against calpain 7 (raised
against calpain 7 MITb [28]), respectively. As shown in
Fig. 6C, 45 kDa fragments were wild-type-specifically
detected in both immunoprecipitation products (upper
Fig. 5. Direct interaction between recombinant calpain 7MIT and GST–IST1. Purified recombinant calpain 7 MIT domain (1–165 amino acids),
calpain 7MIT, was incubated with GST (negative control), GST–IST1, GST–IST1MIM or GST–IST1MIM
L326D,L353A
that had been immobilized
on glutathione–Sepharose beads and subjected to GST-pulldown assay. Purified proteins, initial protein mixtures (Input) and pulldown prod-
ucts (Pulldown) were resolved on a 15% gel by SDS ⁄ PAGE, and subjected to CBB staining. Open triangles and closed triangles indicate
bands of recombinant calpain 7MIT and GST, GST–IST1, GST–IST1MIM, and GST–IST1MIM
L326D,L353A
, respectively.
Enhancement of calpain 7 autolysis by IST1 Y. Osako et al.
4416 FEBS Journal 277 (2010) 4412–4426 ª 2010 The Authors Journal compilation ª 2010 FEBS

panel, closed triangle; lower panel, gray triangle), sug-
gesting that 45 K contains both mGFP and MIT
domains of calpain 7. On the other hand, the anti-
GFP-reacting 30 kDa band was not detected in the
immunoprecipitates of antibody against calpain 7,
indicating a lack of MIT domains in 30 K. Thus,
mGFP–calpain 7 contains at least two cleavage sites:
one lies at the N-terminal boundary of the protease
domain, generating 45 K, and the other lies between
mGFP and MITa, generating 30 K (Fig. 6A). To
roughly estimate the latter cleavage site in mGFP–cal-
pain 7, we used three types of unfused mGFP
Fig. 6. Estimation of cleavage sites in
mGFP–calpain 7. (A) Schematic representa-
tions of mGFP–calpain 7 (fragmentary view)
and unfused mGFP constructs with stop
codons at different positions at their C-ter-
mini. Two estimated cleavage sites generat-
ing 45 and 30 kDa fragments (designated 45
and 30 K, respectively) are indicated by solid
arrows. (B) The putative catalytic residue
Cys290 was replaced by either Ser or Ala,
mGFP–calpain 7 (WT), mGFP–calpain 7
C290S
(C290S) and mGFP–calpain 7
C290A
(C290A)
were transiently expressed in HEK293T
cells, and total cell lysates were then ana-
lyzed by WB with mAb against GFP and

pAb against calpain 7, respectively. Arrows
and closed and open triangles indicate full-
length mGFP–calpain 7 and 45 K and 30 K,
respectively, and the gray triangle indicates
the 45 kDa fragment (45 K) detected by WB
analysis with pAb against calpain 7 [also
shown in (C) and (D)]. (C) Cleared lysates
from cells expressing mGFP–calpain 7 or
mGFP–calpain 7
C290S
were subjected to
immunoprecipitation (IP) with anti-GFP
serum or pAb against calpain 7, followed by
WB analysis with mAb against GFP and
mAb against calpain 7, respectively. (D)
mGFP–calpain 7, mGFP–calpain 7
C290S
and
three types of unfused mGFP constructs
(mGFP
265
, mGFP
259
or mGFP
239
) were tran-
siently expressed in HEK293T cells, and
total cell lysates from those cells and un-
transfected cells were analyzed by WB with
mAb against GFP to compare the electro-

phoretic mobility of 30 K with that of each
mGFP. Asterisks indicate 33 kDa bands that
were detected in both the wild type (WT)
and Cys290-substituted mutants (C290S and
C290A).
Y. Osako et al. Enhancement of calpain 7 autolysis by IST1
FEBS Journal 277 (2010) 4412–4426 ª 2010 The Authors Journal compilation ª 2010 FEBS 4417
constructs that have stop codons at different positions
at their C-termini: mGFP
265
, mGFP
259
and mGFP
239
(see Fig. 6A and Experimental procedures). These
mGFP proteins were transiently expressed in HEK293T
cells, and total cell lysates were analyzed by WB with
mAb against GFP to compare the electrophoretic
mobility of 30 K with that of each mGFP construct. As
shown in Fig. 6D, the electrophoretic mobility of 30 K
was closer to that of mGFP
239
. This result suggested
that mGFP–calpain 7 was cleaved at the point immedi-
ately after or in the vicinity of residue 239 of mGFP.
In order to determine an autolytic cleavage site in
mGFP–calpain 7–Strep, we also attempted to purify a
C-terminal fragment by using Strep-Tactin Sepharose
beads. Although extraneous bands other than intact
expression products were detected, they were common

to both the wild type and the C290S mutant (Fig. 7A).
Moreover, no wild-type-specific bands were detected
by probing with antibody against Strep or Strep-Tac-
tin-conjugated horseradish peroxidase (data not
shown). Thus, it is likely that autolytic cleavage also
occurs near the C-terminus of calpain 7 before or
immediately after N-terminal cleavage. The faint,
 33 kDa, bands detected with mAb against GFP
[indicated by asterisks in Fig. 6: (B), top, lanes 3–5;
(C), top, lanes 1, 2, 4 and 5; (D), last two lanes] were
found not only for the wild type but also for the Cys
mutants (C290S and C290A). Thus, they were proba-
bly derived by limited digestion with other cellular pro-
teases, and not by autolysis of mGFP–calpain 7.
Enhancement of autolysis of mGFP–
calpain 7–Strep by GST–MIM in vitro
As we observed direct MIT–MIM interaction in vitro
(Fig. 5), we speculated that IST1 serves as an activator
for mGFP–calpain 7 via MIT–MIM interaction. To
investigate this possibility, we performed an in vitro
‘autolysis assay’. mGFP–calpain 7–Strep was expressed
in HEK293T cells, and purified by affinity purification
with Strep-Tactin Sepharose beads (Fig. 7A). Purified
mGFP–calpain 7–Strep ( 0.7 lg) was incubated with
1 lg of recombinant GST–IST1, GST–MIM or GST
(negative control) at 30 °C for 20 h. After incubation,
samples were analyzed by WB with mAb against GFP
to detect proteolysed fragments of mGFP–cal-
pain 7–Strep. As expected, addition of GST–IST1 and
GST–MIM enhanced the generation of 30 K, but addi-

tion of GST did not (Fig. S1). As the purified recombi-
nant GST–IST1 contained multiple degraded
fragments, we used GST–MIM for further analyses.
Next, we performed a similar assay with GST–
MIM, GST–MIM
L326D,L353A
or GST-fused CHMP6
N-terminal half (GST–CHMP6NT) as a negative con-
trol. As shown in Fig. 7B, addition of GST–MIM
enhanced the generation of 30 K, but only a marginal
enhancing effect was observed with the addition of
GST–MIM
L326D,L353A
or GST–CHMP6NT. In the
case of mGFP–calpain 7
C290S
–Strep, with or without
any recombinant proteins, no degraded bands were
Fig. 7. Enhancement of autolysis of mGFP–calpain 7–Strep by
GST–MIM in vitro. (A) Purification of mGFP–calpain 7–Strep and
mGFP–calpain 7
C290S
–Strep from HEK293T cells. Cleared lysate of
untransfected HEK293T cells (mock) and those of cells expressing
C-terminally Strep-tagged mGFP–calpain 7 (WT) or mGFP–cal-
pain 7
C290S
(C290S) were incubated with Strep-Tactin Sepharose
beads. After incubation, unbound proteins were removed
(Unbound), and the beads were washed. Proteins bound to the

beads were eluted with a buffer containing 5 m
MD-desthiobiotin
(Purified proteins). Samples were separated by SDS ⁄ PAGE fol-
lowed by CBB staining. The arrow and asterisk indicate bands of
mGFP–calpain 7–Strep and Strep-Tactin detached from beads,
respectively. (B) Purified mGFP–calpain 7–Strep (WT and C290S)
proteins ( 0.7 lg) were incubated at 30 °C for 20 h with either
GST–MIM, GST–MIM
L326D,L353A
or GST–CHMP6NT (1 lg) or with-
out additional proteins ()). Samples without incubation (time 0)
were also analyzed. After incubation, samples were subjected to
SDS ⁄ PAGE (15% gel) and analyzed by WB with mAb against GFP
to detect proteolysed fragments of mGFP–calpain 7–Strep. Bands
of full-length WT and C290S are indicated by the arrow, and those
of 30 K are indicated by the open triangle.
Enhancement of calpain 7 autolysis by IST1 Y. Osako et al.
4418 FEBS Journal 277 (2010) 4412–4426 ª 2010 The Authors Journal compilation ª 2010 FEBS
detected. This result strongly suggests that 30 K is
generated by proteolytic activity of mGFP–cal-
pain 7–Strep itself, not by potentially contaminating
proteases in the preparations, and that MIT–MIM
interaction enhances the autolytic activity of mGFP–
calpain 7–Strep in vitro.
Autolytic properties of mGFP–calpain 7–Strep
We further characterized the autolytic activity of
mGFP–calpain 7–Strep. Purified mGFP–cal-
pain 7–Strep was incubated with GST–MIM in a buf-
fer containing 2 mm CaCl
2

or EGTA. As compared
with the control (a sample without addition of CaCl
2
or EGTA), neither enhancing nor inhibitory effects on
the generation of 30 K were observed with the addi-
tion of 2 mm CaCl
2
or EGTA (data not shown). On
the other hand, when purified mGFP–calpain 7–Strep
was incubated with GST–MIM in the presence of vari-
ous protease inhibitors or N-ethylmaleimide, a sulfhy-
dryl modification reagent, the generation of 30 K was
almost completely abolished by 10 mm N-ethylmalei-
mide (Fig. 8A) and partially inhibited by 1 mm leupep-
tin (inhibitor of trypsin-type serine proteases and
cysteine proteases) or 1 mm E-64 (cysteine protease
inhibitor) (Fig. 8B). Obvious effects of other protease
inhibitors were not observed with the use of 3 lm
recombinant human calpastatin domain 1 (CSD1, cal-
pain inhibitor protein), 0.5 lm ovocystatin (cysteine
protease inhibitor protein), 20 lm MG-132 (protea-
some inhibitor), 20 lm antipain (cysteine protease
inhibitor), 20 lm N-acetyl-l-leucyl-l-leucyl-l-norleuci-
nal (ALLNal) (calpain inhibitor) or 2 mm pefabloc
(serine protease inhibitor).
Effects of ESCRT-related proteins on autolysis of
mGFP–calpain 7 in vivo
Next, we examined whether IST1 affects the generation
of 30 K in vivo. mGFP–calpain 7 was coexpressed with
either FLAG–IST1 or FLAG–IST1

DMIM1,2
in
HEK293T cells, and total cell lysates were analyzed by
WB with mAb against GFP. As shown in Fig. 9A, the
effect of coexpression with FLAG–IST1 on the genera-
tion of 30 K was not so obvious regarding the ratio of
precursor (arrow) and 30 K (open triangle). On the
other hand, coexpression with FLAG–IST1
DMIM1,2
reduced 30 K generation. Overexpression of VPS4B
E235Q
(a VPS4B mutant with replacement of Glu235 by Gln,
lacking ATPase activity) is known to cause accumula-
tion of ESCRTs on the endosomal membrane to form
aberrant multivesicular bodies MVB [27]. As shown in
Fig. 9B, coexpression with FLAG–VPS4B
E235Q
signifi-
cantly reduced the generation of 30 K as compared
with the control vector.
Discussion
IST1 is a newly reported ESCRT-III (or CHMP) fam-
ily member, and interacts with the MIT domain of
VPS4 [32,33]. In this study, we showed for the first
time that a tandem repeat unit of MIT domains of
human calpain 7 directly interacts with the C-terminal
region of IST1 (Fig. 5). We previously reported an
interaction between calpain 7 and CHMP1B [28], but
this interaction seems to be much weaker than that
between calpain 7 and IST1 under the conditions

employed (Fig. 2). As shown by mutational analyses
(Fig. 4), the observed stronger interaction may be
attributable to the presence of two potential binding
sites in the IST1 C-terminal region, which contains
Fig. 8. Autolytic properties of mGFP–calpain 7–Strep. Effects of
protease inhibitors and sulfhydryl modification reagent on autolysis
of mGFP–calpain 7–Strep were investigated. Purified mGFP–cal-
pain 7–Strep was incubated at 30 °C for 20 h with GST–MIM in a
buffer containing protease inhibitors as indicated (A). As a control,
the same volume of a vehicle used for dissolving reagents was
added to the reaction mixture in place of inhibitors. Bands of full-
length mGFP–calpain 7–Strep and those of 30 K are shown in the
upper and lower panels, respectively. Additionally, leupeptin, E-64
and pefabloc were tested at higher concentrations (B).
Y. Osako et al. Enhancement of calpain 7 autolysis by IST1
FEBS Journal 277 (2010) 4412–4426 ª 2010 The Authors Journal compilation ª 2010 FEBS 4419
two types of MIM motif (MIM1 and MIM2) for bind-
ing to VPS4 MIT [32,33]. MIM1 and MIM2 were orig-
inally defined by differences in binding to different
grooves formed by a three-helix bundle of the MIT
domain of mammalian VPS4 or yeast Vps4 [29–31].
MIM1 of CHMP1A or Vps2 binds to the groove
between helices 2 and 3, and MIM2 of CHMP6 binds
to that between helices 1 and 3. Bajorek et al. [32]
suggested that MIM1 and MIM2 of IST1 also bind to
the different grooves of VPS4 MIT, on the basis of
NMR chemical shift mapping. Their mutational analy-
ses revealed that MIM1 and MIM2 have a synergistic
effect on binding to MIT, suggesting that the two
grooves in the three-helix bundle of VPS4 MIT accept

MIM1 and MIM2 simultaneously [32]. In analogy to
those findings, either one of the MIT domains of cal-
pain 7 seems to be sufficient for binding to MIMs of
IST1. However, our data indicated that both MIT
domains are required for efficient interaction (Fig. 3).
One conceivable explanation for this observation is
that tandem MIT domains may act as a single inte-
grated module. The yeast ESCRT-related protein Vta1
also has tandem MIT domains, and the 3D structures
showed that they are closely associated by extensive
hydrophobic interactions, which make two MIT
domains an apparent single module [38]. As the linker
region between the MIT domains of calpain 7 is much
shorter than that of Vta1 (five residues versus 21 resi-
dues), it is not certain whether the same theory
applies to calpain 7, but the idea that tandem MIT
domains of calpain 7 affect each other to maintain an
interacting interface seems attractive. However, at
present, we have no clue as to whether MIM1 and
MIM2 bind to only one MIT domain or bind to each
of the two MIT domains of calpain 7, and it is not
known why interaction between calpain 7 and the
MIM2-containing protein CHMP6 was not observed
(Fig. 2) [28]. Structural studies, such as cocrystalliza-
tion and X-ray analysis of tandem MIT domains of
calpain 7 and IST1 MIM elements, should clarify
these issues in the future.
Although the physiological role of human calpain 7
is still unknown, the presence of tandem MIT domains
might contribute to its role being different from that in

unicellular organisms. Whereas Cpl1 (yeast calpain 7)
does not possess an apparent MIT domain, PalB (fun-
gal calpain 7) has only a single MIT domain. In accor-
dance with this difference, reported binding partners
are not identical among calpain 7, Cpl1 and PalB.
Cpl1 and PalB were shown to interact with the ESC-
RT-III core proteins Snf7 ⁄ Vps32 (CHMP4) and Vps24
(CHMP3), respectively [39,40], but interaction between
Cpl1 ⁄ PalB and CHMP1 orthologs (Did2 ⁄ DidB) has
not been reported. Thus, the N-terminal regions of
calpain 7 might have evolved to acquire different strat-
egies for involvement in the ESCRT system, and the
tandem MIT domains may govern interacting features
unique to human calpain 7, enabling it to execute its
physiological roles differently from lower eukaryotic
calpains.
In yeast and fungi, the transcription factor Rim101 ⁄
PacC is thought to be a substrate of Cpl1 ⁄ PalB, and it
Fig. 9. Effects of ESCRT-related proteins on autolysis of mGFP–cal-
pain 7 in vivo. (A) mGFP–calpain 7 of either the wild type (WT) or
C290S mutant was coexpressed with either FLAG–IST1 or FLAG–
IST1
DMIM1,2
in HEK293T cells, and total cell lysates were analyzed
by WB with mAb against GFP and mAb against FLAG, respectively.
Bands of 45 kDa (45 K) and 30 kDa (30 K) are indicated by closed
and open triangles, respectively. Cotransfection with a blank vector
instead of IST1 expression plasmids was performed for control
experiments. (B) The effect of coexpression of FLAG–VPS4B
E235Q

on mGFP–calpain 7 autolysis was investigated as shown in (A).
Enhancement of calpain 7 autolysis by IST1 Y. Osako et al.
4420 FEBS Journal 277 (2010) 4412–4426 ª 2010 The Authors Journal compilation ª 2010 FEBS
has been proposed that Rim101 ⁄ PacC is also recruited
around the ESCRTs on the endosomal membranes by
binding to Snf7 ⁄ Vps32-interacting factor Rim20 ⁄ PalA
[16,18]. On the other hand, a human homolog of
Rim101 ⁄ PacC has not been identified. Futai et al.
showed that His-tagged calpain 7 purified from COS
cells does not proteolyse typical calpain substrates
in vitro [13]. In this study, we found that GST-fused
MIM of IST1 enhances the autolysis of purified
mGFP–calpain7–Strep in vitro (Fig. 7B), demonstrat-
ing the protease activity of calpain 7 for the first time.
This finding suggests that calpain 7 also functions as a
protease rather than as a structural protein in mamma-
lian cells, and that MIT domains are involved in cal-
pain 7 activation. This notion leads us to suggest two
possible activation mechanisms of calpain 7 in vitro:
(a) by binding of MIM, MIT domains dissociate from
the protease domain to expose the catalytic core; and
(b) binding of MIM causes a conformational change
of calpain 7 to create an active catalytic core. We
observed that an mGFP-fused calpain 7 mutant lack-
ing tandem MIT domains (mGFP–calpain 7DMIT) is
still proteolysed to generate 30 K in cultured cells
(Fig. 3), apparently supporting the former possibility.
However, it is also possible that IST1 acts on the sub-
strate rather than on the protease. To investigate fur-
ther whether the autolysis is an intermolecular or

intramolecular reaction, we purified N-terminally
Strep-tagged calpain 7 as a protease source, and incu-
bated it with either mGFP–calpain 7
C290S
–Strep or
mGFP–calpain 7DMIT
C290S
–Strep in the presence of
GST–MIM. As a result, proteolysed mGFP–cal-
pain 7
C290S
–Strep fragment (30 K) was detected
(Fig. S2), suggesting that autolysis of calpain 7 is inter-
molecular. As the degree of degradation of mGFP–cal-
pain 7DMIT
C290S
–Strep was slightly decreased, it is
likely that IST1 acts on MIT of the substrate and
influences the accessibility of the substrate to the
enzyme. However, there remains a possibility that a
gross conformational change induced by deletion of
MIT from mGFP–calpain 7
C290S
–Strep itself made the
substrate more resistant to the protease. Moreover, the
efficiency of generation of 30 K in the intermolecular
reaction experiment seems less than that observed in
the experiment in which mGFP–calpain 7 was incu-
bated, and we cannot exclude the possibility that both
an intramolecular reaction and an intermolecular reac-

tion occur in the autolysis. Therefore, it is premature
to draw conclusions regarding the mechanism of the
enhancing effects of IST1 on mGFP–calpain 7–Strep
autolysis in vitro.
Both mGFP–calpain 7 and IST1 have been reported
to accumulate on aberrant endosomes when an
ATPase-defective VPS4 mutant (VPS4B
E235Q
, used in
this study) is expressed in HeLa cells [28,33]. However,
overexpression of FLAG–VPS4B
E235Q
reduced 30 K
generation (Fig. 9B), suggesting that proper recruit-
ment of calpain 7 is important for its activation. In the
case of conventional calpains, a C2-like domain has
been suggested to partially contribute to Ca
2+
-depen-
dent membrane binding [41]. However, we previously
reported that the subcellular distribution of calpain 7
is not significantly affected by Ca
2+
, and that mGFP–
calpain 7DMIT coexpressed with monomeric red fluo-
rescent protein–VPS4B
E235Q
does not accumulate on
aberrant endosomes [28]. These observations strongly
suggest that MIT domains are responsible for regulat-

ing the subcellular localization of calpain 7. As shown
in Fig. 9A, overexpression of FLAG–IST1 did not
enhance the autolysis of mGFP–calpain 7 in cultured
cells. On the other hand, overexpression of FLAG–
IST1
DMIM1,2
suppressed the autolysis. This observation
might be explained by regarding IST1 as a regulator of
the intracellular localization of calpain 7, because IST1
was previously reported to contribute to recruitment of
VPS4 to an ESCRT-III-accumulated region in the cell
[32,33]. Given that endogenous IST1 is sufficient for
the recruitment of calpain 7 around ESCRTs, overex-
pression of FLAG–IST1 would have no additive
effects. On the other hand, overexpressed FLAG–
IST1
DMIM1,2
would occupy the ESCRT surface, and
hamper binding of endogenous IST1, resulting in fail-
ure of calpain 7 recruitment and exhibiting a domi-
nant-negative effect. To test this hypothesis, we
performed fluorescence microscopic analyses, and
investigated the subcellular localization of overexpres-
sed FLAG–IST1 ⁄ IST1
DMIM1,2
and mGFP–calpain 7 in
HeLa cells. These proteins displayed diffuse or par-
tially colocalized punctate distribution around nucleus.
There were no significant differences in the punctate
distribution of mGFP–calpain 7 between cells coex-

pressing FLAG–IST1 and and those coexpressing
FLAG-IST1
DMIM1,2
(data not shown). Thus, it is not
clear why FLAG–IST1 had no enhancing effects on
autolysis and FLAG–IST1
DMIM1,2
inhibited the autoly-
sis of mGFP–calpain 7. Other unknown cytosolic fac-
tors that physically associate with IST1 but whose
amounts are limited might be involved in enhancing
the autolysis of mGFP–calpain 7.
When fungal calpain 7 (PalB) cleaves PacC (a tran-
scription factor acting on alkaline adaptation), PalA
functions as a scaffold to recruit PacC to endosomal
membranes by recognizing two YPXL motifs present
in the C-terminal half of PacC [16]. A human ortholog
of PalA, ALIX (also known as AIP1), functions in the
budding of enveloped RNA viruses from plasma
Y. Osako et al. Enhancement of calpain 7 autolysis by IST1
FEBS Journal 277 (2010) 4412–4426 ª 2010 The Authors Journal compilation ª 2010 FEBS 4421
membranes [42]. ALIX is recruited to plasma mem-
branes by Gag proteins of HIV-1 and equine infectious
anemia virus through binding of the V domain of
ALIX to YPX(n)L late-domain motifs (n = 1–3)
[42,43]. As virus-encoded aspartyl proteases are already
well known to process Gag precursor proteins, cal-
pain 7 may not be involved in virus budding. The con-
servation of the YPX(n)L motif for binding to
ALIX ⁄ PalA, however, hints an approaching way to

search for potential calpain 7 substrates. As IST1 is
involved in cytokinesis rather than endosomal sorting
[32,33], calpain might process factors that work in cell
division. Studies are in progress to search for
YPX(n)L-containing ALIX-interacting proteins for the
investigation of potential calpain 7 substrates.
Experimental procedures
Antibodies and reagents
The following mouse mAbs were used: mAb against
FLAG (clone M2; Sigma, St Louis, MO, USA), and mAb
against GFP (clone B-2; Santa Cruz Biotechnology, Santa
Cruz, CA, USA). Anti-GFP serum (A6455) was obtained
from Invitrogen ⁄ Molecular Probes (Carlsbad, CA, USA).
Rabbit pAb against recombinant human calpain 7 and
mouse mAb against human calpain 7 were raised as
described previously [28]. Peroxidase-conjugated goat anti-
(mouse IgG) and anti-(rabbit IgG) were obtained from
Jackson Immunoresearch Laboratories (West Grove, PA,
USA). N-ethylmaleimide, leupeptin, pepstatin A (pepsta-
tin), phenylmethanesulfonyl fluoride and calpain inhibitor I
(ALLNal) were obtained from Nacalai Tesque (Kyoto,
Japan). Antipain and E-64 were obtained from the Peptide
Institute (Osaka, Japan). Pefabloc and ovocystatin were
obtained from Calbiochem (San Diego, CA, USA). MG-132
was obtained from Wako Pure Chemical Industries (Osaka,
Japan). d-Desthiobiotin was purchased from IBA GmbH
(Go
¨
ttingen, Germany).
Construction of plasmids

Cloning of human calpain 7 cDNA and construction of
mammalian expression plasmids for various mGFP-fused
calpain 7 mutants (FLAG–CHMP1B, FLAG–CHMP4b,
FLAG–CHMP6 and FLAG–VPS4B
E235Q
) and the bacterial
expression plasmid for GST–CHMP6NT was performed as
described previously [13,28,44]. A mammalian expression
plasmid for mGFP–calpain 7–Strep was constructed as fol-
lows. The DNA fragment encoding Strep-tag II was ampli-
fied by PCR, with pEXPR-IBA105-C [45] as a template
and a pair of primers (forward, 5¢-CCG
CTCGAG
GCTAGCTGGAGCCACCCG-3¢, containing an additional
XhoI site, underlined; and reverse, 5¢-TAGAAGGCACAG
TCGAGGCTG-3¢). The PCR product was digested with
XhoI, and then ligated into the XhoI site of the vec-
tor downstream of the stop-codon-mutated calpain 7
cDNA (AAGCTTGGTGGAAGCGGTGGTTCT
CTCGAG;
mutated stop codon italicized and XhoI site underlined).
From that vector, a DNA fragment corresponding to a part
of calpain 7 (amino acids 390–813) followed by Strep-tag II
was isolated by BamHI digestion and inserted into the
BamHI site of pmGFP–calpain 7.
To construct pCMV3xFLAG–IST1 and pGEX–IST1, an
IST1 cDNA fragment was amplified by PCR, using a
cDNA clone KIAA0174 (GenBank ID: D79996.1) encoding
364 amino acids containing four tandem MP repeats,
obtained from Kazusa DNA Research Institute (Chiba,

Japan), using a pair of primers (forward, 5¢-CTA
GAATT
CAACAGCACAGCATGCTGG-3¢; reverse, 5¢-AGAGAA
TTCTGCCTGGTTTAAGAGACC-3¢; restriction sites
underlined). The amplified cDNA fragment was first
inserted into the Zero Blunt TOPO PCR Cloning vector
(Invitrogen), and the EcoRI fragment was then inserted
into the EcoRI site of pCMV3xFLAG-B [46]. Expression
vectors for IST1 mutants of MIM were obtained by PCR-
based site-directed mutagenesis with a Quik-Change Site-
Directed Mutagenesis Kit (Stratagene, Cedar Creek, TX,
USA), using specific primers (Table S1) and either
pCMV3xFLAG–IST1 or pCMV3xFLAG–IST1
L326D
as a
template. The mutations were confirmed by nucleotide
sequencing. For bacterial expression of GST–IST1MIM, a
cDNA fragment encoding amino acids 300–364 was ampli-
fied by PCR, with a pair of primers (forward, 5¢-TA
GGA
TCCCCTGGACCCAAGCCAGAAG-3¢; reverse, 5¢-AGA
GAATTCTGCCTGGTTTAAGAGACC-3¢; restriction sites
underlined) and pCMV3xFLAG–IST1 as a template. The
amplified cDNA fragment was first inserted into the Zero
Blunt TOPO PCR Cloning vector (Invitrogen), and the
BamHI–EcoRI fragment was then inserted into the
BamHI–EcoRI site of pGEX4T-2. The mutant of pGST–
IST1MIM
L326D,L353A
was obtained by the same method,

with pCMV3xFLAG–IST1
L326D,L353A
as a template.
A pair of oligonucleotides encoding the His
6
sequence
(forward, 5¢-TCGACCACCA TCACCATCACCATTGACA-3¢;
reverse, 5¢-GGCCTGTCAA TGGTGATGGTGATGGTGG-3¢)
and those encoding the PreScission Protease recognition
sequence (forward, 5¢-AATTCCTGGAAGTTCTGTTCCA
GGGTCCAA-3¢; reverse, 5¢-TCGATTGGACCCTGGAAC
AGAACTTCCAGG-3¢) were inserted into the SalI–NotI site
and the EcoRI–SalI site of pGEX-6p-3 (GE Healthcare,
Amersham Place, Little Chalfont, UK), and the resultant
plasmid was named pGST–pHis. After a pair of oligonucleo-
tides including the BglII site (forward, 5¢-GATCCA
AGAT
CTCTG-3¢; reverse, 5¢-AATTCAGAGATCTTG-3¢; BglII
site underlined) had been inserted into the BamHI–EcoRI
sites of pGST–pHis, a cDNA fragment encoding amino
acids 1–165 of calpain 7 (calpain 7MIT) was amplified by
using a pair of primers (forward, 5¢-GAG
AGATCT
Enhancement of calpain 7 autolysis by IST1 Y. Osako et al.
4422 FEBS Journal 277 (2010) 4412–4426 ª 2010 The Authors Journal compilation ª 2010 FEBS
CTATGGACGCCACAGCACTGGAGC-3¢; r everse, 5¢-GAG
AG
AGATCTTTGGCTTAACACTTGTTGAACTG-3¢; BglII
site underlined), and inserted.
Mammalian expression vectors for mGFP

259
and mGFP
239
were obtained by PCR-based site-directed mutagenesis with
a Quik-Change Site-Directed Mutagenesis Kit (Stratagene),
using specific primers and templates (Table S1). In this
study, pmGFP-C1 [28] was used as an expression vector for
mGFP
265
.
Cell culture
HEK293T cells were cultured in DMEM supplemented
with 5% heat-inactivated fetal bovine serum, 100 UÆmL
)1
penicillin and 100 lgÆmL
)1
streptomycin at 37 °C in humid-
ified air containing 5% CO
2
.
Expression and purification of recombinant
proteins
Escherichia coli BL21 cells were transformed with each
expression plasmid for GST and GST-fused proteins (GST–
IST1, GST–IST1MIM, GST–IST1MIM
L326D,L353A
, GST–
CHMP6NT, GST–pHis and GST–MIT–pHis). Expression
of GST–IST1, GST–IST1MIM and GST–IST1
L326D,L353A

was induced with 0.5 mm isopropyl thio-b-d-galactoside for
3 h at 30 °C, and the proteins were purified by binding to
glutathione–Sepharose 4B beads according to the manufac-
turer’s instructions. GST–CHMP6NT was expressed and
purified essentially in the same way as described above,
except for the use of elution buffer containing 10 mm
reduced glutathione. Purified proteins were dialyzed against
NaCl ⁄ P
i
(137 mm NaCl, 2.7 mm KCl, 8 mm Na
2
HPO
4
and
1.5 mm KH
2
PO
4
, pH 7.3), and stored at 4 °C until use.
Expression of GST–pHis and GST–MIT–pHis was
induced with 0.5 mm isopropyl thio-b-d-galactoside over-
night at 16 °C, and the proteins were purified by binding to
TALON metal affinity resin (Clontech, Palo Alto, CA,
USA), according to the manufacturer’s instructions. To
obtain recombinant calpain 7 MIT domains, GST–MIT–
pHis was purified by HisTrap HP (GE Healthcare), fol-
lowed by GSTrap HP (GE Healthcare), according to the
manufacturer’s instructions. The eluate was incubated with
PreScission protease (GE Healthcare) at 4 °C overnight to
remove the N-terminal GST tag and the C-terminal His

6
-
tag. After being dialyzed against HiTrap Q HP starting
buffer (50 mm phosphate buffer, pH 6.0, 50 mm NaCl,
0.5% Tween-20), the processed protein was applied to a Hi-
Trap Q HP (GE Healthcare), washed with starting buffer
and low-salt washing buffer (50 mm phosphate buffer,
pH 6.0, 100 mm NaCl, 0.5% Tween-20), and then eluted
with elution buffer (50 mm phosphate buffer, pH 6.0,
200 mm NaCl, 0.5% Tween-20). Purified proteins were
stored at 4 °C until use.
Purification of recombinant human CSD1 was performed
as described previously [47].
Pulldown assay of GST–MIT–pHis binding to
FLAG-tagged CHMPs and IST1
At 24 h after transfection with expression vectors by the
conventional calcium phosphate precipitation method,
HEK293T cells were washed with NaCl ⁄ P
i
, and harvested
cells were lysed in buffer A (10 mm Tris ⁄ HCl, pH 7.4,
142.5 mm KCl, 0.2% NP-40) supplemented with protease
inhibitors (0.4 mm phenylmethanesulfonyl fluoride, 0.2 mm
pefabloc, 6 lgÆmL
)1
leupeptin, 2 lm E-64, 2 lm pepstatin)
and 5 mm b-mercaptoethanol. Supernatants (cleared
lysates) obtained by centrifugation at 15 000 g were incu-
bated with glutathione–Sepharose beads immobilizing
GST–pHis (negative control) or GST–MIT–pHis for 2 h at

4 °C with gentle mixing. After Sepharose beads had been
recovered by low-speed centrifugation (700 g) for 1 min and
washed three times with buffer A, proteins bound to the
beads (pulldown products) were subjected to SDS ⁄ PAGE
followed by WB analyses. Proteins transferred to poly(vinyl-
idene difluoride) membranes (Immobilon-P; Millipore, Bed-
ford, MA, USA) were probed with appropriate antibodies.
WB chemiluminescent signals were detected with a LAS-
3000mini lumino-image analyzer (Fujifilm, Tokyo, Japan),
using Super Signal West Pico Chemiluminescent Substrate
(Pierce, Rockford, IL, USA). Bands of GST-fusion proteins
were detected by staining the PVDF membranes with CBB.
In vitro binding assay using recombinant
proteins
Ten micrograms of GST (negative control) or GST–IST1 pro-
teins was immobilized on glutathione–Sepharose beads and
mixed with 10 lg of recombinant calpain 7 MIT domains
diluted in buffer B (50 mm Tris ⁄ HCl, pH 8.0, 350 mm NaCl,
0.2% NP-40, 1 mm dithiothreitol) for 1 h at 4 °C. After
Sepharose beads had been pelleted by brief centrifugation
(1 000 g, 1 min) and washed three times with buffer B, bound
protein complexes were separated on a 15% gel by SDS ⁄
PAGE. Protein bands were detected by CBB staining.
Coimmunoprecipitation assay
One day after HEK293T cells had been seeded, they were
transfected with 5 lg of expression plasmid DNA. After
24 h, cells were harvested in NaCl ⁄ P
i
and lysed in buffer A
containing protease inhibitors, as described above. Cleared

lysates of cells were incubated with anti-GFP serum for 2 h
and protein G–Sepharose 4 Fast Flow (GE Healthcare) for
2 h at 4 °C, the beads were washed three times with buffer A,
and bound proteins were subjected to WB analysis with
appropriate antibodies as described above.
Y. Osako et al. Enhancement of calpain 7 autolysis by IST1
FEBS Journal 277 (2010) 4412–4426 ª 2010 The Authors Journal compilation ª 2010 FEBS 4423
Purification of mGFP–calpain 7–Strep from
HEK293T cells
At 24 h after transfection with expression vectors by the
conventional calcium phosphate precipitation method,
HEK293T cells were washed with NaCl ⁄ P
i
, and harvested
cells were lysed in buffer C (20 mm Hepes ⁄ NaOH, pH 7.4,
150 mm NaCl, 1 mm dithiothreitol, 1 mm pefabloc) sup-
plemented with 0.1% Triton X-100. Cleared lysates
obtained by centrifugation at 10 000 g were incubated
with Strep-Tactin Sepharose beads (IBA GmbH) overnight
at 4 °C with gentle mixing. The beads were recovered by
low-speed centrifugation (600 g) for 1 min, and washed
five times with buffer C supplemented with 0.1% Triton
X-100 and once with buffer C. Proteins were eluted with
buffer D (buffer C containing 5 mmd-desthiobiotin and
5mm b-mercaptoethanol in place of dithiothreitol). After
purification, proteins were immediately used for autolysis
assays.
Autolysis assay
In each experiment, purified mGFP–calpain 7–Strep
( 0.7 lg) in buffer D was incubated at 30 °C for 20 h.

The reaction was stopped by adding 5 · SDS sample buffer
and boiling at 95 °C for 3 min. To examine the autolysis-
enhancing effect, mGFP–calpain 7–Strep was incubated
with one of the following proteins: GST–MIM, GST–
MIM
L326D,L353A
or GST–CHMP6NT (1 lg). To examine
the effect of protease inhibitors and the thiol-reactive com-
pound, mGFP–calpain 7–Strep was incubated with GST–
MIM in buffer D supplemented with one of the following
proteins ⁄ chemicals: 3 lm recombinant human CSD1,
0.5 lm ovocystatin, 20 lm MG-132, 20 lm antipain, 20 lm
ALLNal, 20 l m or 1 mm leupeptin, 10 lm or 1 mm E-64,
2mm pefabloc, or 10 mm N-ethylmaleimide. After incuba-
tion, all samples were analyzed by SDS ⁄ PAGE and WB
with mAb against GFP to detect autolyzed mGFP–cal-
pain 7–Strep.
Acknowledgements
We thank E. Goto for technical assistance. We also
thank K. Hitomi and all members of the Laboratory
of Molecular and Cellular Regulation for valuable sug-
gestions and discussion. This work was supported by a
Grant-in-Aid for Scientific Research on Priority Areas
(to M. Maki) and a Grant-in-Aid for JSPS Fellows (to
Y. Osako).
References
1 Goll DE, Thompson VF, Li H, Wei W & Cong J
(2003) The calpain system. Physiol Rev 83, 731–801.
2 Sorimachi H & Suzuki K (2001) The structure of cal-
pain. J Biochem 129, 653–664.

3 Suzuki K, Hata S, Kawabata Y & Sorimachi H (2004)
Structure, activation, and biology of calpain. Diabetes
53, S12–S18.
4 Maki M, Kitaura Y, Satoh H, Ohkouchi S & Shibata
H (2002) Structures, functions and molecular evolution
of the penta-EF-hand Ca
2+
-binding proteins. Biochim
Biophys Acta 1600, 51–60.
5 Moldoveanu T, Gehring K & Green DR (2008) Con-
certed multi-pronged attack by calpastatin to occlude
the catalytic cleft of heterodimeric calpains. Nature 456,
404–408.
6 Hanna RA, Campbell RL & Davies PL (2008)
Calcium-bound structure of calpain and its mechanism
of inhibition by calpastatin. Nature 456, 409–412.
7 Arthur JS, Elce JS, Hegadorn C, Williams K & Greer
PA (2000) Disruption of the murine calpain small sub-
unit gene, Capn4: calpain is essential for embryonic
development but not for cell growth and division. Mol
Cell Biol 20, 4474–4481.
8 Huang Y & Wang KKW (2001) The calpain family and
human disease. Trends Mol Med 7, 355–362.
9 Dutt P, Croall DE, Arthur JS, Veyra TD, Williams K,
Elce JS & Greer PA (2006) m-Calpain is required for
preimplantation embryonic development in mice. BMC
Dev Biol 6 (doi:10.1186/1471-213X-6-3).
10 Richard I, Broux O, Allamand V, Fougerousse F,
Chiannilkulchai N, Bourg N, Brenguier L, Devaud C,
Pasturaud P, Roudaut C et al. (1995) Mutations in the

proteolytic enzyme calpain 3 cause limb-girdle muscular
dystrophy type 2A. Cell 81, 27–40.
11 Ono Y, Shimada H, Sorimachi H, Richard I, Saido TC,
Beckmann JS, Ishiura S & Suzuki K (1998) Functional
defects of a muscle-specific calpain, p94, caused by
mutations associated with limb-girdle muscular dystro-
phy type 2A. J Biol Chem 273, 17073–17078.
12 Hata S, Koyama S, Kawahara H, Doi N, Maeda T,
Toyama-Sorimachi N, Abe K, Suzuki K & Sorimachi
H (2006) Stomach-specific calpain, nCL-2, localizes in
mucus cells and proteolyzes the beta-subunit of
coatomer complex, beta-COP. J Biol Chem 281, 11214–
11224.
13 Futai E, Kubo T, Sorimachi H, Suzuki K & Maeda T
(2001) Molecular cloning of PalBH, a mammalian
homologue of the Aspergillus atypical calpain PalB.
Biochim Biophys Acta 1517, 316–319.
14 Denison SH, Orejas M & Arst HN Jr (1995) Signaling
of ambient pH in Aspergillus involves a cysteine prote-
ase. J Biol Chem 270, 28519–28522.
15 Pen
˜
alva MA & Arst HN Jr (2004) Recent advances in
the characterization of ambient pH regulation of gene
expression in filamentous fungi and yeasts. Annu Rev
Microbiol 58, 425–451.
Enhancement of calpain 7 autolysis by IST1 Y. Osako et al.
4424 FEBS Journal 277 (2010) 4412–4426 ª 2010 The Authors Journal compilation ª 2010 FEBS
16 Pen
˜

alva MA, Tilburn J, Bignell E & Arst HN Jr (2008)
Ambient pH gene regulation in fungi: making connec-
tions. Trends Microbiol 16, 291–300.
17 Futai E, Maeda T, Sorimachi H, Kitamoto K, Ishiura
S & Suzuki K (1999) The protease activity of a calpain-
like cysteine protease in Saccharomyces cerevisiae is
required for alkaline adaptation and sporulation. Mol
Gen Genet 260, 559–568.
18 Hayashi M, Fukuzawa T, Sorimachi H & Maeda T
(2005) Constitutive activation of the pH-responsive
Rim101 pathway in yeast mutants defective in late steps
of the MVB ⁄ ESCRT pathway. Mol Cell Biol 25, 9478–
9490.
19 Ciccarelli FD, Proukakis C, Patel H, Cross H, Azam S,
Patton MA, Bork P & Crosby AH (2003) The identifi-
cation of a conserved domain in both spartin and spa-
stin, mutated in hereditary spastic paraplegia. Genomics
81, 437–441.
20 Hurley JH & Yang D (2008) MIT domainia. Dev Cell
14, 6–8.
21 Scott A, Gaspar J, Stuchell-Brereton MD, Alam SL,
Skalicky JJ & Sundquist WI (2005) Structure and
ESCRT-III protein interactions of the MIT domain
of human VPS4A. Proc Natl Acad Sci USA 102,
13813–13818.
22 Row PE, Prior IA, McCullough J, Clague MJ & Urbe
´
S (2006) The ubiquitin isopeptidase UBPY regulates
endosomal ubiquitin dynamics and is essential for
receptor down-regulation. J Biol Chem 281, 12618–

12624.
23 Shim S, Merrill SA & Hanson PI (2008) Novel interac-
tions of ESCRT-III with LIP5 and VPS4 and their
implications for ESCRT-III disassembly. Mol Biol Cell
19, 2661–2672.
24 Saksena S, Sun J, Chu T & Emr SD (2007) ESCRTing
proteins in the endocytic pathway. Trends Biochem Sci
32, 561–573.
25 Williams RL & Urbe
´
S (2007) The emerging shape of the
ESCRT machinery. Nat Rev Mol Cell Biol 8, 355–368.
26 McDonald B & Martin-Serrano J (2009) No strings
attached: the ESCRT machinery in viral budding and
cytokinesis. J Cell Sci 122, 2167–2177.
27 Scott A, Chung HY, Gonciarz-Swiatek M, Hill GC,
Whitby FG, Gaspar J, Holton JM, Viswanathan R,
Ghaffarian S, Hill CP et al. (2005) Structural and mech-
anistic studies of VPS4 proteins. EMBO J 24, 3658–
3669.
28 Yorikawa C, Takaya E, Osako Y, Tanaka R, Terasawa
Y, Hamakubo T, Mochizuki Y, Iwanari H, Kodama T,
Maeda T et al. (2008) Human calpain 7 ⁄ PalBH associ-
ates with a subset of ESCRT-III-related proteins in its
N-terminal region and partly localizes to endocytic
membrane compartments. J Biochem 143, 731–745.
29 Obita T, Saksena S, Ghazi-Tabatabai S, Gill DJ, Perisic
O, Emr SD & Williams RL (2007) Structural basis for
selective recognition of ESCRT-III by the AAA ATPase
Vps4. Nature 449, 735–739.

30 Stuchell-Brereton MD, Skalicky JJ, Kieffer C, Karren
MA, Ghaffarian S & Sundquist WI (2007) ESCRT-III
recognition by VPS4 ATPases. Nature 449
, 740–744.
31 Kieffer C, Skalicky JJ, Morita E, De Domenico I, Ward
DM, Kaplan J & Sundquist WI (2008) Two distinct
modes of ESCRT-III recognition are required for VPS4
functions in lysosomal protein targeting and HIV-1
budding. Dev Cell 15, 62–73.
32 Bajorek M, Morita E, Skalicky JJ, Morham SG, Babst
M & Sundquist WI (2009) Biochemical analyses of
human IST1 and its function in cytokinesis. Mol Biol
Cell 20, 1360–1373.
33 Agromayor M, Carlton JG, Phelan JP, Matthews DR,
Carlin LM, Ameer-Beg S, Bowers K & Martin-Serrano
J (2009) Essential role of hIST1 in cytokinesis. Mol Biol
Cell 20, 1374–1387.
34 Rue SM, Mattei S, Saksena S & Emr SD (2008) Novel
Ist1–Did2 complex functions at a late step in multivesic-
ular body sorting. Mol Biol Cell 19, 475–484.
35 Xiao J, Chen XW, Davies BA, Saltiel AR, Katzmann
DJ & Xu Z (2009) Structural basis of Ist1 function and
Ist1–Did2 interaction in the multivesicular body path-
way and cytokinesis. Mol Biol Cell 20, 3514–3524.
36 Bajorek M, Schubert HL, McCullough J, Langelier C,
Eckert DM, Stubblefield WM, Uter NT, Myszka DG,
Hill CP & Sundquist WI (2009) Structural basis for
ESCRT-III protein autoinhibition. Nat Struct Mol Biol
16, 754–762.
37 Dimaano C, Jones CB, Hanono A, Curtiss M & Babst

M (2007) Ist1 regulates Vps4 localization and assembly.
Mol Biol Cell 19, 465–474.
38 Xiao J, Xia H, Zhou J, Azmi IF, Davies BA,
Katzmann DJ & Xu Z (2008) Structural basis of Vta1
function in the multivesicular body sorting pathway.
Dev Cell 14, 37–49.
39 Ito T, Tashiro K, Muta S, Ozawa R, Chiba T, Nishiza-
wa M, Yamamoto K, Kuhara S & Sakaki Y (2000)
Toward a protein–protein interaction map of the bud-
ding yeast: a comprehensive system to examine two-
hybrid interactions in all possible combinations between
the yeast proteins. Proc Natl Acad Sci USA 97, 1143–
1147.
40 Rodrı
´
guez-Gala
´
n O, Galindo A, Herva
´
s-Aguilar A,
Arst HN Jr & Pen
˜
alva MA (2009) Physiological
involvement in pH signaling of Vps24-mediated recruit-
ment of Aspergillus PalB cysteine protease to ESCRT-
III. J Biol Chem 284, 4404–4412.
41 Tompa P, Emori Y, Sorimachi H, Suzuki K &
Friedrich P (2001) Domain III of calpain is a Ca
2+
-reg-

ulated phospholipid-binding domain. Biochem Biophys
Res Commun 280, 1333–1339.
42 Morita E & Sundquist WI (2004) Retrovirus budding.
Annu Rev Cell Dev Biol 20, 395–425.
Y. Osako et al. Enhancement of calpain 7 autolysis by IST1
FEBS Journal 277 (2010) 4412–4426 ª 2010 The Authors Journal compilation ª 2010 FEBS 4425
43 Bieniasz PD (2009) The cell biology of HIV-1 virion
genesis. Cell Host Microbe 5, 550–558.
44 Yorikawa C, Shibata H, Waguri S, Hatta K, Horii M,
Katoh K, Kobayashi T, Uchiyama Y & Maki M (2005)
Human CHMP6, a myristoylated ESCRT-III protein,
interacts directly with an ESCRT-II component EAP20
and regulates endosomal cargo sorting. Biochem J 387,
17–26.
45 Horii M, Shibata H, Kobayashi R, Katoh K, Yorikawa
C, Yasuda J & Maki M (2006) CHMP7, a novel
ESCRT-III-related protein, associates with CHMP4b
and functions in the endosomal sorting pathway.
Biochem J 400, 23–32.
46 Katoh K, Shibata H, Suzuki H, Nara A, Ishidoh K,
Kominami E, Yoshimori T & Maki M (2003) The
ALG-2-interacting protein Alix associates with
CHMP4b, a human homologue of yeast Snf7 that is
involved in multivesicular body sorting. J Biol Chem
278, 39104–39113.
47 Maki M & Hitomi K (2000) Purification of recombi-
nant calpastatin expressed in Escherichia coli. Methods
Mol Biol 144, 85–94.
Supporting information
The following supplementary material is available:

Fig. S1. Enhancement of autolysis of mGFP–cal-
pain 7–Strep by GST–IST1 and GST–MIM in vitro.
Fig. S2. Intermolecular proteolysis of mGFP–cal-
pain 7
C290S
–Strep and mGFP–calpain 7 DMIT
C290S
–
Strep by Strep–calpain 7 in vitro.
Table S1. Primers used for site-directed mutagenesis
performed in this study.
This supplementary material can be found in the
online version of this article.
Please note: As a service to our authors and readers,
this journal provides supporting information supplied
by the authors. Such materials are peer-reviewed and
may be re-organized for online delivery, but are not
copy-edited or typeset. Technical support issues arising
from supporting information (other than missing files)
should be addressed to the authors.
Enhancement of calpain 7 autolysis by IST1 Y. Osako et al.
4426 FEBS Journal 277 (2010) 4412–4426 ª 2010 The Authors Journal compilation ª 2010 FEBS