Tải bản đầy đủ (.pdf) (11 trang)

Báo cáo khoa học: Brox, a novel farnesylated Bro1 domain-containing protein that associates with charged multivesicular body protein 4 (CHMP4) potx

Bạn đang xem bản rút gọn của tài liệu. Xem và tải ngay bản đầy đủ của tài liệu tại đây (851.32 KB, 11 trang )

Brox, a novel farnesylated Bro1 domain-containing protein
that associates with charged multivesicular body protein 4
(CHMP4)
Fumitaka Ichioka, Ryota Kobayashi, Keiichi Katoh, Hideki Shibata and Masatoshi Maki
Department of Applied Molecular Biosciences, Graduate School of Bioagricultural Sciences, Nagoya University, Japan
Alix (also named AIP1) is an interacting partner of the
penta-EF-hand Ca
2+
-binding protein, ALG-2 [1–5],
and acts as a multifunctional adaptor protein in vari-
ous cellular functions such as cell death, receptor endo-
cytosis, endosomal protein sorting, cell adhesion,
budding of enveloped RNA viruses and development
Keywords
Alix; Bro1; CHMP4; ESCRT-III; farnesylation
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:
Database
The nucleotide sequence of the human Brox
cDNA is available in the DDBJ ⁄ EMBL ⁄ Gen-
Bank database under accession number
AB276123
(Received 11 October 2007, revised 5
December 2007, accepted 10 December


2007)
doi:10.1111/j.1742-4658.2007.06230.x
Human Brox is a newly identified 46 kDa protein that has a Bro1 domain-
like sequence and a C-terminal thioester-linkage site of isoprenoid lipid
(CAAX motif) (C standing for cysteine, A for generally aliphatic amino
acid, and X for any amino acid). Mammalian Alix and its yeast ortholog,
Bro1, are known to associate with charged multivesicular body protein 4
(CHMP4), a component of endosomal sorting complex required for trans-
port III, via their Bro1 domains and to play roles in sorting of ubiquitinat-
ed cargoes. We investigated whether Brox has an authentic Bro1 domain
on the basis of its capacity for interacting with CHMP4s. Both Strep Tac-
tin binding sequence (Strep)-tagged wild-type Brox (Strep–Brox
WT
) and
Strep-tagged farnesylation-defective mutant (Cys fi Ser mutation; Strep–
Brox
C408S
) pulled down FLAG-tagged CHMP4b that was coexpressed in
HEK293 cells. Treatment of cells with a farnesyltransferase inhibitor, FTI-
277, caused an electrophoretic mobility shift of Strep–Brox
WT
, and the
mobility coincided with that of Strep–Brox
C408S
. The inhibitor also caused
a mobility shift of endogenous Brox detected by western blotting using
polyclonal antibodies to Brox, suggesting farnesylation of Brox in vivo.
Fluorescence microscopic analyses revealed that Strep–Brox
WT
exhibited

accumulation in the perinuclear area and caused a punctate pattern of
FLAG–CHMP4b that was constitutively expressed in HEK293 cells. On
the other hand, Strep–Brox
C408S
showed a diffuse pattern throughout the
cell, including the nucleus, and did not cause accumulation of FLAG–
CHMP4b. Fluorescent signals of monomeric green fluorescent protein
(mGFP)-fused Brox
WT
merged partly with those of Golgi markers and with
those of abnormal endosomes induced by overexpression of a dominant
negative mutant of AAA type ATPase SKD1 ⁄ Vps4B in HeLa cells, but
such colocalization was less efficient for mGFP–Brox
C408S
. These results
suggest a physiological significance of farnesylation of Brox in its subcellu-
lar distribution and efficient interaction with CHMP4s in vivo.
Abbreviations
Brox
WT
, wild-type Brox; CAAX motif, a thioester-linkage site of isoprenoid lipid (C standing for cysteine, A for generally aliphatic amino acid,
and X for any amino acid); CHMP, charged multivesicular body protein; E-64, trans-epoxysuccinyl-
L-leucylamido(4-guanidino) butane; ESCRT,
endosomal sorting complex required for transport; FTI-27, farnesyltransferase inhibitor; GAPDH, glyceraldehyde-3-phosphate dehydrogenase;
GFP, green fluorescent protein; GST, glutathione S-transferase; mGFP, monomeric green fluorescent protein; pAb, polyclonal antibody; PTP,
protein tyrosine phosphatase; PVDF, poly(vinylidene difluoride); RabGAPLP, Rab GTPase-activating protein-like protein; Strep, StrepTactin
binding sequence.
682 FEBS Journal 275 (2008) 682–692 ª 2008 The Authors Journal compilation ª 2008 FEBS
[6–9]. Alix associates through its Pro-rich region with
SH3 domains of CIN85 ⁄ Ruk

l
⁄ SETA [10], endophilins
[11] and Src [12] at sites different from the ALG-2
binding site [13]. The PSAP sequence present in the
Pro-rich region of Alix is also recognized by the ubiqu-
itin E2 variant domain of TSG101 [14]. The crystal 3D
structures of the N-terminal domain (designated Bro1
domain) of yeast Bro1 [15] and that of human Alix
[16] have revealed folded cores of  360 residues. The
Bro1 domains are necessary and sufficient for binding
to endosomal sorting complex required for transport
(ESCRT)-III component charged multivesicular body
protein 4 (CHMP4) ⁄ Snf7 [16–18]. ESCRTs and their
associated proteins are conserved from yeast to
humans, and function in the sorting of ubiquitinated
cargoes into intraluminal vesicles that are generated by
inward budding of the endosomal membrane of the
so-called multivesicular bodies (endosomes) [19,20]. A
V domain in the central region of Alix associates with
the YPXnL motif found in late domains of retroviral
Gag proteins [16,21] and it is involved in regulation of
viral budding from cells.
Two Alix homologs containing Bro1 domain-like
sequences are found in the human genome database
(Fig. 1). One is PTPN23, which encodes a putative
protein tyrosine phosphatase, HD-PTP [22]. We pre-
viously demonstrated that, as in the case of Alix,
HD-PTP interacts not only with CHMP4b through its
N-terminal region containing a Bro1 domain-like
sequence but also with TSG101, endophilin A1 and

ALG-2 through its central Pro-rich region, indicating
that HD-PTP is a functional paralog of Alix [23].
The second Alix homolog, designated Brox [a Bro1
domain-containing protein with a thioester-linkage site
of isoprenoid lipid (CAAX motif, C standing for cys-
teine, A for generally aliphatic amino acid, and X for
any amino acid)] in this article, is a 411 amino acid
residue hypothetical protein encoded by C1orf58
(FLJ32421), which was first reported as one of
295 proteins identified in exosomes (extracellular micro-
vesicles) in human urine [24]. Brox lacks a V domain
and Pro-rich region. Multiple sequence alignment and
phylogenetic analysis of Alix orthologs and Bro1
domain-containing proteins have revealed that the
Bro1 domain-like sequence of Brox is less similar to
the Bro1 domains of Alix and HD-PTP and even more
distantly related than yeast Bro1 (supplementary
Fig. S1). It has remained to be established whether
Brox has a functional Bro1 domain, i.e. a capacity
for binding to CHMP4s. A unique feature of Brox is
that it possesses a C-terminal tetrapeptide sequence
Cys-Tyr-Ile-Ser, which conforms to a CAAX motif
[25]. In this study, by expressing in cultured mamma-
lian cells an epitope-tagged wild-type Brox as well as
its amino acid-substituted mutant in the CAAX motif,
we investigated the capacity of Brox to bind to
CHMP4s. We also investigated the subcellular distri-
bution of Brox by confocal fluorescence microscopy
and the extracellular release of Brox into culture
medium. The results suggest that Brox has a functional

Bro1 domain for CHMP4 interaction. Although farn-
esylation is dispensable for in vitro interaction and
extracellular release, post-translational lipid modifica-
tion facilitates interaction with CHMP4s in vivo by
restricting its subcellular localization.
Results
Conservation of Brox in the animal kingdom
Orthologs of human Brox are found in the animal
kingdom, including Caenorhabditis elegans, but there
is no ortholog in the currently available Drosophila
genome database (FlyBase). Brox-like genes are not
Fig. 1. Schematic representations of human
Bro1 domain-containing proteins. Schematic
representations of Alix, HD-PTP and Brox
are shown. Bro1, Bro1 domain; V,
V domain, PRR, Pro-rich region; PTP, protein
tyrosine phosphatase domain; PEST, PEST
motif; CAAX, CAAX motif. The numbers
indicate the amino acid residues. Cys408
was substituted with Ser for creation of a
farnesylation-defective mutant in this study.
F. Ichioka et al. Association of Brox with CHMP4
FEBS Journal 275 (2008) 682–692 ª 2008 The Authors Journal compilation ª 2008 FEBS 683
found in plants, fungi or other lower eukaryotes. The
C-terminal CAAX-motif-like tetrapeptide sequences are
conserved except for the first A (human, Tyr; mouse,
Ser; chicken, Arg; C. elegans, Val) (supplementary
Fig. S2), which is not necessarily aliphatic for the
prenylation motif [25]. A Cys residue in a CAAX motif
is covalently linked with either a 15-carbon isoprenoid,

catalyzed by farnesyltransferase, or a 20-carbon iso-
prenoid, catalyzed by geranylgeranyltransferase I.
Brox has a preferred sequence for farnesyltransferase,
due to the presence of the C-terminal Ser residue
[25,26].
Interaction of Brox with CHMP4b
Whereas CHMP4s interact with both Alix
[14,17,18,27–29] and HD-PTP [29], Rab GTPase-acti-
vating protein-like protein (RabGAPLP) interacts with
Alix but not with HD-PTP [29]. We investigated
whether Brox interacts with CHMP4s and RabGAPLP.
Lysates of HEK293 cells coexpressing either wild-type
StrepTactin binding sequence (Strep)-tagged Brox
(Strep–Brox
WT
) or Strep-tagged CAAX motif mutant
(substitution of Cys408 with Ser; Strep–Brox
C408S
) and
either FLAG–CHMP4b or FLAG–RabGAPLP were
incubated with StrepTactin sepharose beads, and the
complexes were pulled down. Then, the proteins bound
to the beads (pulldown products) were subjected to
western blotting using mAb to FLAG. FLAG–
CHMP4b was pulled down with both Strep–Brox
WT
and Strep–Brox
C408S
and with the comparable N-termi-
nal regions containing the Bro1 domain of Alix (Strep–

Alix
1)423
) and HD-PTP (Strep–HD-PTP
1)431
) (Fig.
2A). On the other hand, FLAG–RabGAPLP was
pulled down with Strep–Alix
1)424
but not with either
Brox or HD-PTP constructs (Fig. 2B).
Effects of farnesylation inhibition on
electrophoretic mobility of Brox
As shown in Fig. 2, the CAAX-motif mutant Strep–
Brox
C408S
exhibited retardation of migration in
SDS ⁄ PAGE. To determine whether the difference in
migration rates between the wild-type and the mutant is
related to the potential farnesylation at Cys408, a farne-
syltransferase inhibitor, FTI-277 [30], was added to the
culture medium during transient overexpression of
Strep-tagged Brox proteins. Treatment of cells with
FTI-277 caused a shift in the mobility of Strep–Brox
WT
to that of Strep–Brox
C408S
, but the treatment did not
affect that of Strep–Brox
C408S
(Fig. 3A). Addition of

the inhibitor to the culture medium also caused an elec-
trophoretic mobility shift of endogenous Brox that was
detected with polyclonal antibody (pAb) to Brox (Fig.
3B). We therefore conclude that Brox is farnesylated at
Cys408 in vivo. Next, we performed biochemical subcel-
lular fractionation of endogenous Brox in HEK293 cells
by differential centrifugation and analysis by western
blotting with pAb to Brox. As shown in Fig. 3C,
although samples of the P1, P2 and P3 fractions were
loaded 20-fold more than those of the total fraction (T)
and cytosolic fraction (S), intensities of the observed
immunoreactive signals were similar between P2 and S,
and the signals for P1 and P3 were weaker than the sig-
nals for S, suggesting that less than  10% of Brox
bound to membranes. There was no significant differ-
ence in the fractionation pattern between farnesylated
and unmodified Brox proteins with treatment with
FTI-277.
Subcellular distribution of Strep–Brox
Subcellular distributions of transiently expressed
proteins of Strep–Brox
WT
and Strep–Brox
C408S
in
HEK293 cells were analyzed by confocal immunofluo-
rescence microscopy, using mAb to Strep, and chro-
mosomal DNAs were stained with TO-PRO-3. As
shown in Fig. 4, Strep–Brox
WT

(Fig. 4A) exhibited
a diffuse pattern but some accumulation in the
A
B
Fig. 2. Brox interacts with CHMP4b but not with RabGAPLP.
HEK293 cells were cotransfected with pFLAG–CHMP4b (A) or
pFLAG–RabGAPLP (B) and either an empty vector of Strep-tag
(pEXPR-IBA105) (Ctrl), pStrep–Brox
WT
, pStrep–Brox
C408S
, pStrep–
Alix
1)423
or pStrep–HD-PTP
1)431
. The cleared cell lysates (Input)
were incubated with StrepTactin sepharose beads at 4 °C over-
night. The pellets (Pulldown) were subjected to SDS ⁄ PAGE and
western blotting. PVDF membranes were probed with mAb to
FLAG (A and B, upper panels) and mAb to Strep-tag II (A and B,
lower panels).
Association of Brox with CHMP4 F. Ichioka et al.
684 FEBS Journal 275 (2008) 682–692 ª 2008 The Authors Journal compilation ª 2008 FEBS
perinuclear area. On the other hand, Strep–Brox
C408S
(Fig. 4C) exhibited a diffuse pattern throughout the
cytoplasm and nucleus. This diffuse pattern is in con-
trast to the absence of conspicuous fluorescence signals
in nuclei of cells transfected with expression plasmids

of Strep–Brox
WT
(Fig. 4A), Strep–Alix
1)423
(Fig. 4E)
and Strep–HD-PTP
1)431
(Fig. 4G).
We previously showed that the distribution of
FLAG–CHMP4b that was constitutively expressed in
HEK293 cells was diffuse, but that coexpression
with C-terminally deleted Alix (AlixDC-V5) caused
accumulation of FLAG–CHMP4b in the perinuclear
area [17], and this effect was also observed after coex-
pression with Strep–Alix
1)423
but not with Strep–HD-
PTP
1)431
, suggesting a qualitative difference between
the Bro1 domains of Alix and HD-PTP in vivo [23]. As
A
B
C
Fig. 3. Farnesylation of Brox. (A) HEK293 cells were transfected
with pStrep–Brox
WT
or pStrep–Brox
C408S
and cultured in the pres-

ence (+) or absence (–) of 10 l
M FTI-277 for 24 h. The cleared cell
lysates were incubated with Strep-tactin sepharose beads at 4 °C
for 3 h. The pellets (Pulldown) were subjected to SDS ⁄ PAGE and
western blotting. PVDF membranes were probed with mAb to
Strep-tag II. (B) HEK293 cells were cultured in the presence (+) or
absence ())of10l
M FTI-277 for 24 h. The total cell lysates were
subjected to SDS ⁄ PAGE and western blotting. PVDF membranes
were probed with pAb to Brox. Arrowheads indicate non-specific
bands. (C) Effects of treatment with FTI-277 on the subcellular dis-
tribution of Brox were investigated by subcellular fractionation as
described in Experimental procedures. T, total lysate; P1, nuclear
and cell debris (pellet, 1000 g for 5 min); P2, crude mitochondrial
and organelle-enriched fraction (pellet, 10 000 g for 15 min); P3,
microsomal fraction (pellet, 100 000 g for 30 min); S, soluble frac-
tion (supernatant, 100 000 g for 30 min). Different relative amounts
of fractionated samples (% corresponding to harvested cells: T and
S, 0.1%; P1, P2 and P3, 2%) were loaded and analyzed by
SDS ⁄ PAGE and immunoblotting analysis with antibody against
Brox. Arrowheads indicate nonspecific crossreacting bands.
AB
CD
EF
GH
Fig. 4. Subcellular distribution of Brox. HEK293 cells were trans-
fected with pStrep–Brox
WT
(A, B), pStrep–Brox
C408S

(C, D), pStrep–
Alix
1)423
(E, F) or pStrep–HD-PTP
1)431
(G, H). After 24 h, the cells
were fixed and stained with mAb to Strep-tag II and Alexa
Fluor 488-conjugated goat anti-(mouse IgG) (A, C, E, G) and with
TO-PRO-3 for chromosomal DNA (B, D, F, H). The fluorescence sig-
nals of Alexa Fluor 488 and TO-PRO-3 were analyzed with a confo-
cal laser-scanning microscope. Bars, 10 lm. Asterisks indicate
transfected cells.
F. Ichioka et al. Association of Brox with CHMP4
FEBS Journal 275 (2008) 682–692 ª 2008 The Authors Journal compilation ª 2008 FEBS 685
shown in Fig. 5, qualitative differences in Bro1 domains
were also observed among the coexpressed Bro1 domain
constructs. As in the case of Strep–Alix
1)423
(Fig. 5G–
I), expression of Strep–Brox
WT
caused accumulation of
FLAG–CHMP4b and partial colocalization in the
perinuclear area (Fig. 5A–C). This accumulation
was not observed in cells expressing Strep–Brox
C408S
(Fig. 5D–F) or Strep–HD-PTP
1)431
(Fig. 5J–L) or in
untransfected cells in the same microscopic fields.

Analyses of monomeric green fluorescent protein
(mGFP)–Brox distribution with organelle markers
Next, for further analyses of the perinuclear distribu-
tion of Brox by immunostaining with commercially
available mAbs against organelle markers, we
expressed a Brox protein fused with mGFP (mGFP–
Brox
WT
). As shown in Fig. 6, a proportion of the
fluorescence signals of mGFP–Brox
WT
expressed in
HEK293 cells merged well with those of Golgi marker
proteins GM130 (Fig. 6A–C) and p230 (Fig. 6D–F).
No merged signals were observed for an early endo-
some marker (EEA1) (Fig. 6G–I) or a late endo-
some ⁄ lysosome marker (LAMP-1) (Fig. 6J–L).
Overexpression of an ATPase-defective mutant of
AAA type ATPase SKD1 (also named Vps4B),
SKD1
E235Q
, is known to induce formation of abnor-
mal endosomes. As shown in Fig. 7, a subset of punc-
tate fluorescence signals of mGFP–Brox
WT
expressed
in HeLa cells merged with those of Myc–SKD1
E235Q
(Fig. 7A–C). A large proportion of mGFP–Brox
C408S

showed a diffuse pattern, but some fine punctate
ABC
DE F
GH I
JKL
Fig. 5. Formation of CHMP4b puncta by Brox overexpression.
FLAG–CHMP4b ⁄ HEK293 cells were transfected with pStrep–-
Brox
WT
(A–C), pStrep–Brox
C408S
(D–F), pStrep–Alix
1)423
(G–I) or
pStrep–HD-PTP
1)431
(J–L). After 24 h, the cells were fixed and
stained with primary antibodies (mAb to Strep-tag II and pAb to
FLAG) and secondary antibodies [Alexa Fluor 488-conjugated goat
anti-(mouse IgG) and Cy3-labeled goat anti-(rabbit IgG)]. The fluores-
cence signals of Alexa Fluor 488 (A, D, G, J) (green) and Cy3 (B, E,
H, K) (red) were analyzed with a confocal laser-scanning micro-
scope and are represented in black and white. The merged images
are shown in (C), (F), (I) and (L), respectively, in color. Bars, 10 lm.
ABC
DEF
GHI
JKL
Fig. 6. Subcellular distribution of Brox. HEK293 cells were trans-
fected with pmGFP–Brox

WT
. After 24 h, the cells were fixed and
stained with mAbs of organelle markers: anti-GM130 (cis-Golgi),
anti-p230 (trans-Golgi), anti-EEA1 (early endosomes) and anti-
LAMP-1 (late endosomes and lysosomes). Cy3-labeled goat anti-
(mouse IgG) was used as a secondary antibody. The fluorescence
signals of mGFP (A, D, G, J) (green) and Cy3 (B, E, H, K) (red) were
analyzed with a confocal laser-scanning microscope and repre-
sented in black and white. The merged images are shown in (C),
(F), (I) and (L), respectively, in color. The small boxed areas are
magnified in the respective large boxed areas. Bars, 10 lm.
Association of Brox with CHMP4 F. Ichioka et al.
686 FEBS Journal 275 (2008) 682–692 ª 2008 The Authors Journal compilation ª 2008 FEBS
patterns were also observed (Fig. 7D–F). The merging
of signals between mGFP–Brox
C408S
and Myc–
SKD1
E235Q
was much less noticeable than in the case
of mGFP–Brox
WT
.
Detection of Brox in extracellular vesicles
Pisitkun et al. [24] performed proteomic profiling of
extracellular microvesicles (exosomes) in human urine
by MS analysis and identified 295 proteins, including a
hypothetical protein FLJ32421 (designated Brox in this
study). To gain more insights into the release of Brox
in exosomes in conjunction with farnesylation, we per-

formed western blotting of vesicles released from
HEK293T cells into culture medium. As shown in Fig.
8A, Brox, Alix and TSG101 (a positive control) were
detected with their respective antibodies in 100 000 g
pellets (vesicular fraction) of cultured medium that had
been precleared by centrifugation at 10 000 g (culture
supernatant) but not in the case of the medium with-
out culture (medium). In contrast, glyceraldehyde-
3-phosphate dehydrogenase (GAPDH), a cytosolic
protein used as a negative control, was not detected in
the vesicular fraction. Similarly, 100 000 g pellets of
the culture supernatant from HEK293T cells express-
ing either FLAG–Brox
WT
or FLAG–Brox
C408S
were
analyzed by western blotting with mAb to FLAG.
The intensities of the detected bands were not signifi-
cantly different between FLAG–Brox
WT
and FLAG–
Brox
C408S
(Fig. 8B). We therefore conclude that Brox
is packaged into exosomal vesicles and released to the
extracellular milieu but that this process does not
depend on its farnesylation.
Discussion
In the human genome database, there are three genes

that encode Bro1 domains: Alix, HD-PTP and Brox
(Fig. 1). Alix and HD-PTP, possessing similarities in a
wider region, have been shown to share several associ-
ated proteins, such as CHMP4b, TSG101, endophi-
lin A1 and ALG-2 [23]. On the other hand, Brox lacks
a V domain and Pro-rich region. In the present study,
we demonstrated for the first time that Brox also asso-
ciates with CHMP4b (Fig. 2A). In our previous study,
we found that the Bro1 domain of Alix directly bound
the three CHMP4 isoforms (CHMP4a, CHMP4b and
CHMP4c), by in vitro GST-pulldown assays using each
recombinant protein, and that CHMP4b is a major
Alix-interacting isoform, based on their expression
levels and binding capacities [18]. In the present study,
ABC
DEF
Fig. 7. Colocalization of Brox with SKD1
E235Q
. HeLa cells were co-
transfected with pMyc–SKD1
E235Q
and either with pmGFP–Brox
WT
(A–C) or with pmGFP–Brox
C408S
(D–F). After 24 h, cells were fixed
and stained with mAb to c-Myc and Cy3-labeled goat anti-(mouse
IgG). The fluorescence signals of mGFP (A and D, green) and Cy3
(B and E, red) were analyzed with a confocal laser-scanning micro-
scope and are represented in black and white. The merged images

are shown in (C) and (F), respectively. The small boxed areas are
magnified in the respective large boxed areas. Bars, 10 lm.
A
B
Fig. 8. Release of Brox from cells. (A) HEK293T cells were incu-
bated at 37 °C for 48 h in culture medium containing 10% fetal
bovine serum, and vesicles released into the medium were col-
lected by ultracentrifugation as described in Experimental proce-
dures. The total cell lysate (lysate) and vesicular fractions were
analyzed by western blotting with antibodies against Brox,
Alix, TSG101 and GAPDH. Cult Sup, cultured medium supernatant;
medium, control medium that was incubated without cells. (B)
HEK293T cells were transfected with pFLAG–Brox
WT
or pFLAG–
Brox
C408S
. After 48 h, transfectants were harvested, and vesicles
released from transfectants were collected by ultracentrifugation.
The total cell lysates (Lysate) and vesicle fractions (vesicles) were
analyzed by western blotting with mAb to FLAG.
F. Ichioka et al. Association of Brox with CHMP4
FEBS Journal 275 (2008) 682–692 ª 2008 The Authors Journal compilation ª 2008 FEBS 687
we similarly prepared thioredoxin-fused CHMP4s and
glutathione–sepharose beads immobilizing glutathione
S-transferase (GST)-fused Brox protein that was puri-
fied from Escherichia coli. GST–Brox beads pulled
down all thioredoxin–CHMP4s (data not shown), indi-
cating direct physical interaction between Brox and
CHMP4s. Although the physiological significance

remains to be established, we reported that the
N-terminal domain of Alix associated with Rab-
GAPLP [29], which was later reported to be identical
to a Rab5-specific GAP (RabGAP-5) [31]. No binding
of either Brox or HD–PTP to RabGAPLP was
detected, and interaction with RabGAPLP was found
to be specific for Alix among the three human Bro1
domain-containing proteins (Fig. 2B).
Brox contains a C-terminal tetrapeptide motif
known as a CAAX motif, which is a site for post-
translational modification with isoprenoids (prenyla-
tion) [25]. Recently, Maurer-Stroh et al. [32], who
constructed PRENbase-Database of Prenylated Pro-
teins ( />obtained experimental evidence, by using
3
H-labeled
prenyl precursors in an in vitro transcription–transla-
tion system, that GST–FLJ32421 is farnesylated but
not geranylgeranylated. We observed differences in the
migration rates in SDS ⁄ PAGE between Brox
WT
and
Brox
C408S
mutant (Figs 2 and 3A) and between endo-
genous Brox proteins from farnesyltransferase inhibi-
tor (FTI-277)-treated and untreated cells (Fig. 3B).
The effect of FTI-277 is smaller on endogenous Brox
(Fig. 3B) than on transiently expressed Strep–Brox
WT

(Fig. 3A). In these experiments, HEK293 cells were
treated with FTI-277 for 24 h. However, this length of
treatment may not be long enough to replace all
pre-existing farnesylated Brox with de novo synthesized
unfarnesylated Brox (Fig. 3B). On the other hand, the
majority of Strep–Brox was synthesized during the
period of this inhibitor treatment, and thus only
unfarnesylated protein could be detected (Fig. 3A).
Faster migration of farnesylated proteins than that
of unmodified proteins has been reported [33]. The elec-
trophoretic mobility shift is explained by sequential
enzymatic processing of CAAX proteins after prenyla-
tion [25,34]: (a) proteolytic cleavage of the C-terminal
AAX residues by Ras-converting enzyme 1; and (b)
methylesterification of prenylated Cys by isoprenyl-
cysteine carboxymethyltransferase. In addition to the
CAAX motif, a second signal is known to be required
for Ras proteins to be stably anchored to plasma mem-
branes: palmitoylation of Cys immediately upstream
(2–6 residues) of farnesylated Cys of H-Ras, N-Ras
and K-Ras4A, and a stretch of basic residues upstream
of the farnesylated Cys186 of K-Ras4B [34–36]. Brox
does not possess immediate upstream Cys residues or a
typical basic stretch upstream of the farnesylated
Cys408. This fact may explain why only  10% of
Brox was recovered in the particulate fractions of
HEK293 cells by the biochemical subcellular fraction-
ation, which may release loosely bound Brox from
membranes. As no significant difference was observed
in the distribution pattern between the cells treated and

untreated with the farnesyltransferase inhibitor FTI-
277 (Fig. 3C), unmodified Brox may also bind to mem-
branes indirectly by interacting with other proteins.
Although both Strep–Brox
WT
and Strep–Brox
C408S
mutant bound to CHMP4b in the pulldown assay
(Fig. 2A), a significant difference was observed in the
subcellular localization of transiently overexpressed
tagged proteins (Fig. 4). In contrast to the distribution
of Strep–Brox
WT
, Strep–Brox
C408S
exhibited a diffuse
pattern throughout the cell, including the nucleus. This
mutational effect of Cys408 agrees with the results of
a study by Maurer-Stroh et al. showing that the
perinuclear condensed distribution of wild-type GFP–
FLJ32421 was changed to a diffuse pattern throughout
the cell, including the nucleus, by substituting Cys408
with Ala [32]. In addition, we observed a difference
between the wild-type and the C408S mutant in the
capacity to induce accumulation of constitutively
expressed FLAG–CHMP4b in HEK293 cells (Fig. 5).
Moreover, localization of mGFP–Brox
WT
to abnormal
endosomes was induced by overexpression of Myc–

SKD1
E235Q
, but this occurred to a lesser degree in the
case of mGFP–Brox
C408S
both in HeLa cells (Fig. 7)
and in HEK293 cells (supplementary Fig. S3). Interest-
ingly, mGFP–Brox
WT
showed partial colocalization
with cis-Golgi marker protein GM130 (Fig. 6A–C)
and trans-Golgi marker protein p230 (Fig. 6D–F).
Whereas farnesylation is catalyzed by cytosolic
farnesyltransferase, both post-prenylation processing
enzymes, Ras-converting enzyme 1 and isoprenylcyste-
ine carboxymethyltransferase, are endoplasmic reti-
culum-integral membrane proteins, and proteolytic
cleavage and methylesterification are catalyzed on the
cytosolic side of the endoplasmic reticulum membrane
[34–36]. Some processed proteins, either further palmi-
toylated or not, are transported to the Golgi apparatus
and then to the plasma membrane or to other
intracellular membranes [35]. Thus, it remains to be
established whether the observed localization of
mGFP-fused Brox to Golgi represents a reservoir of
the transiently overexpressed protein that awaits trans-
portation to its final destination, e.g. to multivesicular
endosomes. It would be interesting to investigate
whether there exist cytosolic factors that regulate the
distribution of farnesylated Brox.

Association of Brox with CHMP4 F. Ichioka et al.
688 FEBS Journal 275 (2008) 682–692 ª 2008 The Authors Journal compilation ª 2008 FEBS
Among the CHMP family members (ESCRT-III
components and related proteins containing SNF7
domains), Alix is known to interact with only
CHMP4s. We investigated whether Brox has specifici-
ties different from those of Alix toward CHMPs, each
of which has a unique feature, e.g. myristoylation
(CHMP6) [37], tandem repeat of SNF7-like domain
(CHMP7) [38], and binding to phosphatidylinositol
3,5-bisphosphate (CHMP3) [39]. We performed a yeast
two-hybrid assay using GAL4-DNA-binding domain-
fused Brox as bait and GAL4-activation domain-fused
CHMPs (CHMP1A, CHMP1B, CHMP2A, CHMP2B,
CHMP3, CHMP4a, CHMP4b, CHMP4c, CHMP6,
CHMP7) as prey. Positive interaction was observed
only for CHMP4s (data not shown). Other ESCRT-
related proteins, including TSG101, Vps28, Vps37A,
EAP20, EAP30, EAP45 and Alix, showed no interac-
tions. At present, the physiological function of Brox is
not known. Secretion of this protein to the extracellu-
lar space by exosomes may not represent a bona fide
role, because a large number of multivesicular body-
related proteins are also secreted [24]. We expected
that inhibition of farnesylation would reduce secretion
of Brox, but there was no significant difference in the
amount of secreted Brox proteins between the wild-
type and the farnesylation-defective C408S mutant of
FLAG-tagged Brox (Fig. 8B). This may be explained
by their similar capacities for binding to CHMP4b

(Fig. 2A). The finding of the ability of Brox to bind to
a specific ESCRT-III component extends our under-
standing of the molecular mechanism underlying the
recognition of CHMP4 by Bro1 domains.
Experimental procedures
Antibodies and reagents
Mouse mAbs were Strep-tag II (IBA GmbH, Go
¨
ttingen,
Germany), FLAG-tag (M2), c-Myc-tag (9E10) (Sigma,
St Louis, MO, USA), GM130, p230 trans-Golgi, CD107a ⁄
Lamp-1, EEA1 (BD Biosciences, San Jose, CA, USA),
GAPDH (Chemicon ⁄ Millipore, Billerica, MA, USA), and
TSG101 (4A10) (GeneTex, San Antonio, TX, USA). Rabbit
pAbs of FLAG-tag and GST were purchased from Sigma
and Santa Cruz Biotechnology (Santa Cruz, CA, USA),
respectively. Rabbit pAbs against Brox were raised by the
conventional method using a GST-fused Brox protein as an
antigen and affinity-purified by a column immobilizing malt-
ose-binding protein (MBP)-fused Brox. Peroxidase-conju-
gated goat anti-rabbit IgG and goat anti-mouse IgG were
obtained from Wako (Osaka, Japan). Preparation of rabbit
pAbs against Alix has been described previously [40].
Cy3-labeled anti-mouse or rabbit IgG and Alexa Fluor
488-conjugated anti-mouse IgG used for indirect immuno-
fluorescence analyses were obtained from Amersham (Little
Chalfont, UK) and BD Biosciences, respectively. The fol-
lowing reagents were purchased: farnesyltransferase inhibi-
tor FTI-277 (EMD ⁄ Calbiochem, San Diego, CA, USA),
poly(l-lysine) (Sigma) and TO-PRO-3 (Invitrogen ⁄ Molecu-

lar Probes, Carlsbad, CA, USA).
Construction of plasmids
A Brox cDNA (DDBJ accession number: AB276123) was
cloned from an HeLa cDNA library by the PCR method,
using a pair of primers based on the registered sequence
NM_144695 for C1orf58:5¢-GGG AAT TC
A TGA CCC
ATT GGT TTC ATA GGA ACC-3¢ and 5¢-GGG AAT
TC
T TAG GAG ATG TAG CAC CCA GTG TC-3 ¢ (nu-
cleotides corresponding to a cDNA of the hypothetical pro-
tein C1orf58 are underlined), and an EcoRI fragment was
inserted into pEXPR-IBA105-A [38] (pStrep–Brox
WT
),
pCMV3 · FLAG-A [17] (pFLAG–Brox
WT
) and pmGFP-
C2 (pmGFP–Brox
WT
), respectively. pmGFP-C2, a mamma-
lian expression vector for monomeric enhanced GFP fusion
protein [41], was created from pEGFP-C2 (Clontech) by
PCR-based site-directed mutagenesis according to the
instructions provided with a QuikChange Site-Directed
Mutagenesis Kit from Stratagene using two complementary
primers (5¢-CAG TCC AAG CTG AGC AAA GAC CCC
AAC GAG AAG CGC GAT CAC-3¢ and 5¢-GTG ATC
GCG CTT CTC GTT GGG GTC TTT GCT CAG CTT
GGA CTG- 3¢). pmGFP–Brox

C408S
, which has a point
mutation at amino acid 408, was c reated by PCR-based
site-directed mutagenesis using pmGFP–Brox as a
template and complementary primers (5¢-CAA AAG
GAC ACT GGG TCC TAC ATC TCC TAA G-3¢ and
5¢-CTT AGG AGA TGT AGG ACC CAG TGT CCT
TTT G-3¢). To create pStrep–Brox
C408S
and pFLAG–
Brox
C408S
,anEcoRI fragment of Brox
C408S
derived from
pmGFP–Brox
C408S
was inserted into the EcoRI site of
pEXPR-IBA105-A and pCMV3 · FLAG-A, respectively.
pEXPR-IBA105 was purchased from IBA GmbH. Construc-
tions of pStrep–Alix
1)423
, pStrep–HD-PTP
1)431
, pFLAG–
CHMP4b, pFLAG–RabGAPLP and pMyc–SKD1
E235Q
have been described previously [17,18,23,29].
Cell culture and transfection
HEK293 cells were subjected to limiting dilution cloning,

and one of the isolated cell lines, designated YS14, was
used in this study. FLAG–CHMP4b ⁄ HEK293 cells,
HEK293 cells constitutively expressing FLAG–CHMP4b,
were established as described previously [17]. HEK293
YS14, FLAG–CHMP4b ⁄ HEK293 and HeLa cells were
cultured in DMEM supplemented with 5% (HEK293 cells)
or 10% (HeLa cells) heat-inactivated fetal bovine serum,
F. Ichioka et al. Association of Brox with CHMP4
FEBS Journal 275 (2008) 682–692 ª 2008 The Authors Journal compilation ª 2008 FEBS 689
penicillin (100 unitsÆmL
)1
) and streptomycin (100 lgÆmL
)1
)
at 37 °C under humidified air containing 5% CO
2
. One day
after cells had been seeded, the cells were transfected with
the expression plasmid DNAs by the conventional calcium
phosphate precipitation method or by using FuGENE6
(Roche, Basel, Switzerland).
Strep-pulldown assay
At 24 h after transfection with expression vectors, HEK293
cells were washed and harvested with NaCl ⁄ P
i
(137 mm
NaCl, 2.7 mm KCl, 8 mm Na
2
HPO
4

, 1.5 mm KH
2
PO
4
,
pH 7.3) and then lysed in lysis buffer A (10 mm
Hepes ⁄ NaOH, pH 7.4, 142.5 mm KCl, 0.2% Nonidet P-40,
0.1 mm pefabloc, 25 lgÆmL
)1
leupeptin, 1 lm pepstatin
and 1 lm E-64). Supernatants after centrifugation at
14 000 r.p.m. were incubated with Strep-tactin sepharose
(IBA GmbH) at 4 °C overnight with gentle mixing. After
the beads had been recovered by low-speed centrifugation
and washed three times with the lysis buffer without prote-
ase inhibitors, the bead-bound proteins (pulldown products)
were subjected to SDS ⁄ PAGE followed by western blotting
using poly(vinylidene difluoride) (PVDF) membranes
(Immobilon-P) (Millipore). The membranes were then blot-
ted either with mAb to Strep-tag II or mAb to FLAG, and
then with a horseradish peroxidase-conjugated secondary
antibody. Signals were detected by the chemiluminescence
method using Super Signal West Pico Chemiluminescent
Substrate (PIERCE, Rockford, IL, USA).
Treatment with farnesyltransferase inhibitor
At 4 h after HEK293 cells had been transfected with expres-
sion vectors, the culture medium was changed to DMEM
containing 5% fetal bovine serum supplemented with FTI-
277 to a final concentration of 10 lm or vehicle (0.1% di-
methylsulfoxide). After 24 h, cells were washed and harvested

with NaCl ⁄ P
i
and then lysed and subjected to Strep pulldown
as described above, except that cleared lysates were incubated
with Strep-tactin sepharose at 4 °C for 3 h. For analysis of
endogenous Brox, the total cell lysates of HEK293 cells cul-
tured in DMEM containing 5% fetal bovine serum supple-
mented with FTI-277 to a final concentration of 10 lm were
analyzed by western blotting using pAb to Brox.
Subcellular fractionation
After treatment with 5 lm FTI-277 for 48 h, HEK293 cells
were suspended in buffer B (10 mm Hepes ⁄ KOH, pH 7.6,
10 mm KCl, 1.5 mm MgCl
2
,5mm 2-mercaptoethanol,
0.1 mm pefabloc, 25 lgÆmL
)1
leupeptin, 1 lm E-64, 1 lm
pepstatin), freeze–thawed twice, and lysed by passing them
20 times through a 26-gauge needle. After addition of NaCl
(final concentration 150 mm) to the lysates, subcellular frac-
tions were obtained by differential centrifugation (see Fig. 3
legend for centrifugation details), essentially as described
previously [40].
Immunofluorescence microscopic analyses
One day after cells had been seeded on coverslips that had
been precoated with poly(l-lysine) (HEK293 cells) or
uncoated (HeLa cells), they were transfected with the
expression plasmid DNAs. After 24 h, cells were washed
with NaCl ⁄ P

i
, fixed in 4% (w ⁄ v) paraformaldehyde in
NaCl ⁄ P
i
for 20 min, quenched in 50 mm NH
4
Cl in NaCl ⁄ P
i
for 15 min, and permeabilized in 0.1% (w ⁄ v) Triton X-100
in NaCl ⁄ P
i
for 5 min. After blocking with 0.1% (w ⁄ v) gela-
tin in NaCl ⁄ P
i
for 30 min, the cells were incubated with
primary antibodies (mAb to Strep-tag II, pAb to FLAG,
mAb to GM130, mAb to p230 trans-Golgi, and mAb to
c-Myc-tag) at room temperature for 1 h and then with sec-
ondary antibodies [Alexa Fluor 488-conjugated goat anti-
(mouse IgG) and Cy3-labeled goat anti-(mouse IgG) or
Cy3-labeled goat anti-(rabbit IgG)] at room temperature
for 1 h. For chromosomal DNA staining, cells were incu-
bated with 0.2 lm TO-PRO-3 in NaCl ⁄ P
i
at room tempera-
ture for 15 min. Finally, they were mounted with antifading
solution [25 mm Tris ⁄ HCl (pH 8.7), 10% polyvinyl alcohol,
5% glycerol, 2.5% 1,4-diazobicyclo(2,2,2)-octane], and ana-
lyzed under a confocal laser-scanning microscope (LSM5
PASCAL; Carl Zeiss, Oberkochen, Germany).

Analyses of extracellular vesicles (exosomes)
The transfected or untransfected HEK293T cells were incu-
bated in 10% fetal bovine serum ⁄ DMEM for 48 h at 37 °C.
Then, the culture media were collected and centrifuged at
1000 g and 10 000 g for 5 min each to remove cell debris,
and the supernatants were further centrifuged at 100 000 g
for 30 min at 4 °C. The pellets were washed by suspending
in NaCl ⁄ P
i
and recentrifuging under the same conditions.
The collected vesicles were subjected to western blotting.
Acknowledgements
We thank Ms H. Yoshida for subcloning of HEK293
cells and Dr K. Hitomi for valuable suggestions. This
work was supported by a Grant-in-Aid for Scientific
Research B (to M. Maki), a Grant-in-Aid for Young
Scientists B (to H. Shibata) and Fellowships for Young
Scientists (to F. Ichioka) from JSPS.
References
1 Vito P, Lacana
`
E & D’Adamio L (1996) Interfering
with apoptosis: Ca
2+
-binding protein ALG-2 and
Association of Brox with CHMP4 F. Ichioka et al.
690 FEBS Journal 275 (2008) 682–692 ª 2008 The Authors Journal compilation ª 2008 FEBS
Alzheimer’s disease gene ALG-3. Science 271, 521–
525.
2 Vito P, Pellegrini L, Guiet C & D’Adamio L (1999)

Cloning of AIP1, a novel protein that associates with
the apoptosis-linked gene ALG-2 in a Ca
2+
-dependent
reaction. J Biol Chem 274, 1533–1540.
3 Missotten M, Nichols A, Rieger K & Sadoul R (1999)
Alix, a novel mouse protein undergoing calcium-
dependent interaction with the apoptosis-linked-gene 2
(ALG-2) protein. Cell Death Differ 6, 124–129.
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 Maki M & Shibata H (2007) The penta-EF-hand
protein ALG-2 and its interacting ptoteins. Calcium
Binding Proteins 2, 4–10.
6 Tarabykina S, Mollerup J, Winding P & Berchtold MW
(2004) ALG-2, a multifunctional calcium binding
protein? Front Biosci 9, 1817–1832.
7 Odorizzi G (2006) The multiple personalities of Alix.
J Cell Sci 119, 3025–3032.
8 Sadoul R (2006) Do Alix and ALG-2 really control
endosomes for better or for worse? Biol Cell 98,
69–77.
9 Mattei S, Klein G, Satre M & Aubry L (2006)
Trafficking and developmental signaling: Alix at the
crossroads. Eur J Cell Biol 85, 925–936.
10 Chen B, Borinstein SC, Gillis J, Sykes VW & Bo

¨
gler O
(2000) The glioma-associated protein SETA interacts
with AIP1 ⁄ Alix and ALG-2 and modulates apoptosis in
astrocytes. J Biol Chem 275, 19275–19281.
11 Chatellard-Causse C, Blot B, Cristina N, Torch S,
Missotten M & Sadoul R (2002) Alix (ALG-2-interact-
ing protein X), a protein involved in apoptosis, binds to
endophilins and induces cytoplasmic vacuolization.
J Biol Chem 277, 29108–29115.
12 Schmidt MH, Dikic I & Bogler O (2005) Src phosphor-
ylation of Alix ⁄ AIP1 modulates its interaction with
binding partners and antagonizes its activities. J Biol
Chem 280, 3414–3425.
13 Shibata H, Yamada K, Mizuno T, Yorikawa C,
Takahashi H, Satoh H, Kitaura Y & Maki M (2004)
The penta-EF-hand protein ALG-2 interacts with a
region containing PxY repeats in Alix ⁄ AIP1, which is
required for the subcellular punctate distribution of the
amino-terminal truncation form of Alix ⁄ AIP1.
J Biochem 135, 117–128.
14 von Schwedler UK, Stuchell M, Muller B, Ward DM,
Chung HY, Morita E, Wang HE, Davis T, He GP,
Cimbora DM et al. (2003) The protein network of HIV
budding. Cell 114, 701–713.
15 Kim J, Sitaraman S, Hierro A, Beach BM, Odorizzi G
& Hurley JH (2005) Structural basis for endosomal
targeting by the Bro1 domain. Dev Cell 8, 937–947.
16 Fisher RD, Chung HY, Zhai Q, Robinson H, Sund-
quist WI & Hill CP (2007) Structural and biochemical

studies of ALIX ⁄ AIP1 and its role in retrovirus bud-
ding. Cell 128, 841–852.
17 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.
18 Katoh K, Shibata H, Hatta K & Maki M (2004)
CHMP4b is a major binding partner of the ALG-2-
interacting protein Alix among the three CHMP4 iso-
forms. Arch Biochem Biophys 421, 159–165.
19 Slagsvold T, Pattni K, Malerød L & Stenmark H (2006)
Endosomal and non-endosomal functions of ESCRT
proteins. Trends Cell Biol
16, 317–326.
20 Williams RL & Urbe
´
S (2007) The emerging shape of the
ESCRT machinery. Nat Rev Mol Cell Biol 8, 355–368.
21 Lee S, Joshi A, Nagashima K, Freed EO & Hurley JH
(2007) Structural basis for viral late-domain binding to
Alix. Nat Struct Mol Biol 14, 194–199.
22 Toyooka S, Ouchida M, Jitsumori Y, Tsukuda K, Sakai
A, Nakamura A, Shimizu N & Shimizu K (2000) HD-
PTP: a novel protein tyrosine phosphatase gene on
human chromosome 3p21.3. Biochem Biophys Res
Commun 278, 671–678.
23 Ichioka F, Takaya E, Suzuki H, Kajigaya S, Buchman
VL, Shibata H & Maki M (2007) HD-PTP and Alix

share some membrane-traffic related proteins that inter-
act with their Bro1 domains or proline-rich regions.
Arch Biochem Biophys 457, 142–149.
24 Pisitkun T, Shen RF & Knepper MA (2004) Identi-
fication and proteomic profiling of exosomes in
human urine. Proc Natl Acad Sci USA 101, 13368–
13373.
25 Clarke S (1992) Protein isoprenylation and methyla-
tion at carboxyl-terminal cysteine residues. Annu Rev
Biochem 61, 355–386.
26 Moores SL, Schaber MD, Mosser SD, Rands E,
O’Hara MB, Garsky VM, Marshall MS, Pompliano
DL & Gibbs JB (1991) Sequence dependence of protein
isoprenylation. J Biol Chem 266, 14603–14610.
27 Strack B, Calistri A, Craig S, Popova E & Gottlinger
HG (2003) AIP1 ⁄ ALIX is a binding partner for HIV-1
p6 and EIAV p9 functioning in virus budding. Cell 114,
689–699.
28 Martin-Serrano J, Yarovoy A, Perez-Caballero D &
Bieniasz PD (2003) Divergent retroviral late-budding
domains recruit vacuolar protein sorting factors by
using alternative adaptor proteins. Proc Natl Acad Sci
USA 100, 12414–12419.
29 Ichioka F, Horii M, Katoh K, Terasawa Y, Shibata H
& Maki M (2005) Identification of Rab GTPase-activat-
ing protein-like protein (RabGAPLP) as a novel
F. Ichioka et al. Association of Brox with CHMP4
FEBS Journal 275 (2008) 682–692 ª 2008 The Authors Journal compilation ª 2008 FEBS 691
Alix ⁄ AIP1-interacting protein. Biosci Biotechnol Bio-
chem 69, 861–865.

30 Lerner EC, Qian Y, Blaskovich MA, Fossum RD, Vogt
A, Sun J, Cox AD, Der CJ, Hamilton AD & Sebti SM
(1995) Ras CAAX peptidomimetic FTI-277 selectively
blocks oncogenic Ras signaling by inducing cytoplasmic
accumulation of inactive Ras–Raf complexes. J Biol
Chem 270, 26802–26806.
31 Haas AK, Fuchs E, Kopajtich R & Barr FA (2005) A
GTPase-activating protein controls Rab5 function in
endocytic trafficking. Nat Cell Biol 7, 887–893.
32 Maurer-Stroh S, Koranda M, Benetka W, Schneider G,
Sirota FL & Eisenhaber F (2007) Towards complete
sets of farnesylated and geranylgeranylated proteins.
PLoS Comput Biol 3, doi:10.1371/journal.pcbi.0030066
33 Benetka W, Koranda M, Maurer-Stroh S, Pittner F &
Eisenhaber F (2006) Farnesylation or geranylgeranyla-
tion? Efficient assays for testing protein prenylation
in vitro and in vivo. BMC Biochem 7, doi:10.1186/1471-
2091-7-6
34 Winter-Vann AM & Casey PJ (2005) Post-prenylation-
processing enzymes as new targets in oncogenesis. Nat
Rev Cancer 5, 405–412.
35 Hancock JF (2003) Ras proteins: different signals from
different locations. Nat Rev Mol Cell Biol 4, 73–84.
36 Plowman SJ & Hancock JF (2005) Ras signaling from
plasma membrane and endomembrane microdomains.
Biochim Biophys Acta 1746, 274–283.
37 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.
38 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.
39 Whitley P, Reaves BJ, Hashimoto M, Riley AM, Potter
BV & Holman GD (2003) Identification of mammalian
Vps24p as an effector of phosphatidylinositol 3,5-bis-
phosphate-dependent endosome compartmentalization.
J Biol Chem 278, 8786–8795.
40 Shibata H, Suzuki H, Yoshida H & Maki M (2007)
ALG-2 directly binds Sec31A and localizes at endoplas-
mic reticulum exit sites in a Ca
2+
-dependent manner.
Biochem Biophys Res Commun 353, 756–763.
41 Zacharias DA, Violin JD, Newton AC & Tsien RY
(2002) Partitioning of lipid-modified monomeric GFPs
into membrane microdomains of live cells. Science 296,
913–916.
Supplementary material
The following supplementary material is available
online:
Fig. S1. Multiple sequence alignment and phylogenetic
tree of Bro1-domain-containing proteins.
Fig. S2. Multiple sequence alignment of Brox orthologs.
Fig. S3. Colocalization of Brox with SKD1E235Q in
HEK293 cells.

This material is available as part of the online article
from
Please note: Blackwell Publishing are not responsible
for the content or functionality of any supplementary
materials supplied by the authors. Any queries (other
than missing material) should be directed to the corre-
sponding author for the article.
Association of Brox with CHMP4 F. Ichioka et al.
692 FEBS Journal 275 (2008) 682–692 ª 2008 The Authors Journal compilation ª 2008 FEBS

×