Both Strep Tac-tin binding sequence Strep-tagged wild-type Brox Strep–BroxWT and Strep-tagged farnesylation-defective mutant Cysfi Ser mutation; Strep– BroxC408S pulled down FLAG-tagged C
Trang 1that 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 Ca2 +-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: mmaki@agr.nagoya-u.ac.jp
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 investigatubiquitinat-ed 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–BroxWT) and Strep-tagged farnesylation-defective mutant (Cysfi Ser mutation; Strep– BroxC408S) 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–BroxWT, and the mobility coincided with that of Strep–BroxC408S 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–BroxWT 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–BroxC408S 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 BroxWTmerged 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–BroxC408S 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.
Trang 2[6–9] Alix associates through its Pro-rich region with
SH3 domains of CIN85⁄ Rukl⁄ 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 interacmodifica-tion 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.
Trang 3found 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–BroxWT) or Strep-tagged CAAX motif mutant
(substitution of Cys408 with Ser; Strep–BroxC408S) 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–BroxWT
and Strep–BroxC408Sand with the comparable
N-termi-nal regions containing the Bro1 domain of Alix (Strep–
Alix1)423) and HD-PTP (Strep–HD-PTP1)431) (Fig
2A) On the other hand, FLAG–RabGAPLP was
pulled down with Strep–Alix1)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–
BroxC408S 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–BroxWT
to that of Strep–BroxC408S, but the treatment did not
affect that of Strep–BroxC408S (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–BroxWT and Strep–BroxC408S 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–BroxWT (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–BroxWT, pStrep–BroxC408S, pStrep– Alix1)423 or pStrep–HD-PTP1)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).
Trang 4perinuclear area On the other hand, Strep–BroxC408S
(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–BroxWT (Fig 4A), Strep–Alix1)423 (Fig 4E)
and Strep–HD-PTP1)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–Alix1)423 but not with Strep–HD-PTP1)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–BroxWTor pStrep–BroxC408Sand cultured in the
pres-ence (+) or abspres-ence (–) 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 ( )) of 10 l 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.
Fig 4 Subcellular distribution of Brox HEK293 cells were trans-fected with pStrep–Brox WT (A, B), pStrep–Brox C408S (C, D), pStrep– Alix1)423(E, F) or pStrep–HD-PTP1)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.
Trang 5shown in Fig 5, qualitative differences in Bro1 domains
were also observed among the coexpressed Bro1 domain
constructs As in the case of Strep–Alix1)423(Fig 5G–
I), expression of Strep–BroxWTcaused accumulation of
FLAG–CHMP4b and partial colocalization in the
perinuclear area (Fig 5A–C) This accumulation
was not observed in cells expressing Strep–BroxC408S
(Fig 5D–F) or Strep–HD-PTP1)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– BroxWT) As shown in Fig 6, a proportion of the fluorescence signals of mGFP–BroxWT 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 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), SKD1E235Q, is known to induce formation of abnor-mal endosomes As shown in Fig 7, a subset of punc-tate fluorescence signals of mGFP–BroxWT expressed
in HeLa cells merged with those of Myc–SKD1E235Q (Fig 7A–C) A large proportion of mGFP–BroxC408S showed a diffuse pattern, but some fine punctate
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–Alix1)423 (G–I) or
pStrep–HD-PTP1)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.
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), p230 (trans-Golgi), EEA1 (early endosomes) and 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.
Trang 6patterns were also observed (Fig 7D–F) The merging
of signals between mGFP–BroxC408S and Myc–
SKD1E235Q was much less noticeable than in the case
of mGFP–BroxWT
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–BroxWT or FLAG–BroxC408S were
analyzed by western blotting with mAb to FLAG
The intensities of the detected bands were not
signifi-cantly different between FLAG–BroxWT and FLAG–
BroxC408S (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,
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.
Trang 7we 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 (http://mendel.imp.ac.at/sat/PrePS/index2.html),
obtained experimental evidence, by using 3H-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 BroxWT and
BroxC408S 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–BroxWT
(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–BroxWT and Strep–BroxC408S 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–BroxWT, Strep–BroxC408S 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–BroxWTto abnormal endosomes was induced by overexpression of Myc– SKD1E235Q, but this occurred to a lesser degree in the case of mGFP–BroxC408S both in HeLa cells (Fig 7) and in HEK293 cells (supplementary Fig S3) Interest-ingly, mGFP–BroxWT 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
Trang 8Among 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-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 TCA TGA CCC ATT GGT TTC ATA GGA ACC-3¢ and 5¢-GGG AAT TCT 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–BroxWT), pCMV3· FLAG-A [17] (pFLAG–BroxWT
) and pmGFP-C2 (pmGFP–BroxWT), 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–BroxC408S
, which has a point mutation at amino acid 408, was created 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–BroxC408S
and pFLAG– BroxC408S, an EcoRI fragment of BroxC408S derived from pmGFP–BroxC408S 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–Alix1)423, pStrep–HD-PTP1)431, pFLAG– CHMP4b, pFLAG–RabGAPLP and pMyc–SKD1E235Q 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,
Trang 9penicillin (100 unitsÆmL)1) and streptomycin (100 lgÆmL)1)
at 37C under humidified air containing 5% CO2 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⁄ Pi (137 mm
NaCl, 2.7 mm KCl, 8 mm Na2HPO4, 1.5 mm KH2PO4,
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 4C 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⁄ Piand then lysed and subjected to Strep pulldown
as described above, except that cleared lysates were incubated
with Strep-tactin sepharose at 4C 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 MgCl2, 5 mm 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⁄ Pi, fixed in 4% (w⁄ v) paraformaldehyde in NaCl⁄ Pifor 20 min, quenched in 50 mm NH4Cl in NaCl⁄ Pi
for 15 min, and permeabilized in 0.1% (w⁄ v) Triton X-100
in NaCl⁄ Pifor 5 min After blocking with 0.1% (w⁄ v) gela-tin in NaCl⁄ Pi 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-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⁄ Piat 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 4C The pellets were washed by suspending
in NaCl⁄ Pi 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: Ca2 +-binding protein ALG-2 and
Trang 10Alzheimer’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 Ca2 +-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 Ca2+-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