Báo cáo khoa học: Secretion of the mammalian Sec14p-like phosphoinositide-binding p45 protein pptx

Secretion of the mammalian Sec14p-like
phosphoinositide-binding p45 protein
Maria Merkulova
1
, Huong Huynh
2
, Vitaly Radchenko
1
, Kan Saito
2
, Valery Lipkin
1
,
Tatiana Shuvaeva
1
and Tomas Mustelin
2
1 Shemyakin and Ovchinnikov Institute of Bioorganic Chemistry, Russian, Academy of Sciences, Moscow, Russia
2 Program of Inflammation, Infectious and Inflammatory Disease Center and Program of Signal Transduction, Cancer Center,
The Burnham Institute, La Jolla, CA, USA
Protein–lipid interactions are important for protein
targeting, signal transduction, lipid transport, and the
maintenance of cellular compartments and mem-
branes. The yeast PtdIns transfer protein Sec14p is
the prototype for a large family of protein modules
referred to as the SEC14 domain (Smart entry:
smart00516) or CRAL ⁄ TRIO domain (Pfam entry:
pfam00650) for cellular retinaldehyde binding protein⁄
Trio protein homology. There are presently about
500 proteins with this domain in the conserved
domains database ( and

this number is still growing. It is now apparent that
this is an evolutionary ancient and widespread domain
found in plants, yeast, invertebrates, and mammals [1].
About two thirds of all proteins that contain a Sec14p
homology domain, consist only of this domain, while
the rest are multidomain proteins with additional
protein–protein interaction or catalytic domains. The
prototypical Sec14p from Saccharomyces cerevisiae is
two-lobed globular protein with a large hydrophobic
pocket [2], which binds the entire PtdIns molecule.
While the same overall topology is found in all
Sec14p-like proteins, they differ from each other
mostly in the region predicted to bind the head group
of PtdIns. Other known ligands for these proteins
include phosphatidylcholine [3], trans-retinaldehyde [4],
and PtdIns(3,4,5)P
3
[5]. It seems that Sec14p-like pro-
teins have been adapted during evolution to fulfill a
number of different functions that depend on protein–
lipid interactions.
Keywords
CRAL ⁄ TRIO domain; GOLD domain;
nonclassical protein secretion;
phosphoinositides; Sec14p
Correspondence
T. Mustelin, Program of Signal Transduction,
The Burnham Institute, 10901 N. Torrey
Pines Road, La Jolla, CA 92037, USA
Fax: +1 858 713–6274

Tel: +1 858 713–6270
E-mail:
(Received 11 March 2005, revised 23
August 2005, accepted 2 September 2005)
doi:10.1111/j.1742-4658.2005.04955.x
Protein–lipid interactions are important for protein targeting, signal trans-
duction, lipid transport, and the maintenance of cellular compartments
and membranes. Specific lipid-binding protein domains, such as PH,
FYVE, PX, PHD, C2 and SEC14 homology domains, mediate interactions
between proteins and specific phospholipids. We recently cloned a 45-kDa
protein from rat olfactory epithelium, which is homologous to the yeast
Sec14p phosphatidylinositol (PtdIns) transfer protein and we report here
that this protein binds to PtdIns(3,4,5)P
3
and far weaker to less phosphoryl-
ated derivatives of PtdIns. Expression of the p45 protein in COS-1 cells
resulted in accumulation of the protein in secretory vesicles and in the
extracellular space. The secreted material contained PtdIns(3,4,5)P
3
. Our
findings are the first report of a Sec14p-like protein involved in transport
out of a cell and, to the best of our knowledge, inositol-containing phos-
pholipids have not previously been detected in the extracellular space. Our
findings suggest that p45 and phosphoinositides may participate in the
formation of the protective mucus on nasal epithelium.
Abbreviations
Btk, Bruton’s tyrosine kinase; EEA1, early endosomal antigen-1; FYVE, Fab1, YotB,Vac1, and EEA1 homology; GFP, green fluorescent
protein; GOLD, Golgi dynamics; HA, hemagglutinin; PH, pleckstrin homology; PI3K, phosphatidylinositol 3¢-kinase; PITP, PtdIns transfer
proteins; PtdIns, phosphatidylinositol; PTP, protein tyrosine phosphatase; PX, phox homology; s, secretory; SPF, supernatant protein factor;
TAP, tocopherol-associated protein.

FEBS Journal 272 (2005) 5595–5605 ª 2005 FEBS 5595
The mammalian Sec14p-like protein p45 was origin-
ally thought to be a GTP-binding protein present in
rat olfactory epithelium [6,7]. The determination of its
sequence in 1999 [8] showed that it is homologous to
yeast Sec14p and related lipid-binding proteins. A clo-
sely related bovine protein was reported by Stocker
and coworkers in the same year [9]. A human ortholog
of this protein was later cloned and termed toco-
pherol-associated protein (TAP) [10]. This protein is
identical to supernatant protein factor (SPF), a protein
with squalene transfer activity [11]. The human gen-
ome contains three genes for TAP within 100 kb of
each other on chromosome 22q12.1 [12], designated
hTAP1, hTAP2, and hTAP3. The two latter are 86%
and 80% identical to hTAP1 ⁄ SPF [12]. The rat p45
protein is the ortholog of hTAP2.
The crystal structure of hTAP1 ⁄ SPF [13] revealed a
two-domain topology very similar to yeast Sec14p. In
hTAP1 ⁄ SPF, the 275-residue Sec14p homology domain
has a large horse shoe-shaped ligand-binding pocket
between the central b-sheet and the surrounding a-heli-
ces [13]. The C-terminal 115 residues form a separate
domain consisting of eight b strands organized as jelly
roll barrel [13]. This globular domain was referred to
as a Golgi dynamics (GOLD) domain [12], a putative
protein–protein interaction domain predicted to be
involved in Golgi function or trafficking [14]. GOLD
domains are frequently found in combination with
domains known to interact with lipids, i.e. SEC14,

pleckstrin homology (PH), and Fab1p ⁄ YotB ⁄ Vac1p ⁄
EEA1 (FYVE) domains [14]. Such two-domain pro-
teins may function as adapters that assemble protein
complexes on membranes or aid in packaging of cargo
proteins into membrane vesicles.
The ligand specificity of hTAP1 ⁄ SPF is rather con-
tradictory. First, it was reported to bind tocopherol
[9,10] and squalene [11], then the three-dimensional
crystal structure in complex with a-tocopheryl quinone
was reported [15]. Recent studies with a variety of
hydrophobic ligands showed that recombinant hTAP1 ⁄
SPF can bind a-, b-, c-, and d-tocopherols, a -toco-
pheryl quinone, squalene, phosphatidylcholine, and
PtdIns, with the lowest dissociation constants for the
latter [16]. This absence of a clear ligand preference
leaves the question of the true physiological ligand
unresolved. The ligand(s) for TAP2 (p45) and TAP3
are unknown.
To begin to address the biological function of
p45 ⁄ TAP2 protein we have performed lipid binding
experiments and we studied the subcellular localization
of p45. We report that this protein binds to
PtdIns(3,4,5)P
3
and far weaker to less phosphorylated
derivatives of PtdIns in vitro. When expressed in
COS-1 cells, p45 accumulated in secretory vesicles and
the extracellular space along with PtdIns(3,4,5)P
3
. This

could represent a novel mechanism of export of
PtdIns-derived molecules. The significance of this is
discussed.
Results
P45 binds PtdIns(3,4,5)P
3
in vitro
Several proteins containing the lipid-binding SEC14
domain have now been shown to bind phospholipids
[3,5,12]. For example, the SEC14 domain of PTP-
MEG2 was shown to bind PtdIns(3,4,5)P
3
in vitro and
colocalized with it in intact cells [5]. We therefore deci-
ded to study the binding of the SEC14-containing p45
to phospholipids using a filter-binding assay [17]. Both
natural purified and recombinant p45 proteins were
used to eliminate possible artifacts due to contamin-
ation with bacterial lipids, endogenous lipid occu-
pancy, issues with misfolding, or other potential
problems with either preparation. Both preparations
were of high purity (Fig. 1A). Nitrocellulose filters
with 15 immobilized phospholipids (PIP-Strips
TM
, Ech-
elon) were overlaid with recombinant or natural p45
in solution at a final concentration of 0.5 lgÆmL
)1
(10 nm), washed extensively, and detected by immuno-
blotting. As shown in Fig. 1B, natural p45 bound best

to PtdIns(3,4,5)P
3
and, significantly weaker, to
PtdIns(3)P, PtdIns(4)P, PtdIns(3,5)P
2
, PtdIns(4,5)P
2
,
PtdIns(3,4)P
2
and PA. Other phospholipids did not
bind at all to natural p45 protein. Recombinant p45
protein gave very similar results: it also bound best to
PtdIns(3,4,5)P
3
and far weaker to the same other
phospholipids as natural p45 did (Fig. 1C). We con-
clude from these experiments that p45, like the SEC14
domain PTP-MEG2 [5], binds to phosphoinositides
with highest affinity for PtdIns(3,4,5)P
3
.
Subcellular localization of p45 protein
Previous studies [7] using immunohistochemistry and
electron microscopy indicated that endogenous p45 in
the olfactory epithelium is found in secretory granules
of sustentacular cells and in the adjacent layer of the
surrounding mucus. p45 was also detected in the wash-
out medium of nasal mucosa [6]. However, it was not
clear from these studies if p45 was actively secreted or

only released from disrupted nasal cells [6].
To study the subcellular localization of p45, we
cloned its cDNA into the pEF3HA eukaryotic expres-
sion vector, which adds an HA epitope tag to the
C-terminus of the insert. COS-1 cells were transfected
Secretion of p45 Sec14p-like protein M. Merkulova et al.
5596 FEBS Journal 272 (2005) 5595–5605 ª 2005 FEBS
with either empty pEF3HA vector or the pEF3-
HA_p45 construct, fixed 48 h later, permeabilized,
stained with a TRITC-conjugated 12CA5 anti-HA
mAb, and viewed under a confocal microscope. While
vector-transfected cells did not show any staining
(Fig. 2A, upper panel), the HA-tagged p45 was readily
detected in the cells transfected with pEF3HA_p45 as
a staining of intracellular vesicles, as granular plasma
membrane-associated structures, and as irregular
aggregates in the extracellular space (Fig. 2A, lower
panel). Immunoblotting of the culture medium as well
as lysates of the transfected cells with the 16B12 anti-
HA tag antibody showed that a protein of the correct
M
r
(molecular mass calculated to be 48 192 Da) was
present in both locations (Fig. 2B). Thus, transfected
COS-1 cells produce HA-tagged p45, package it into
vesicles, and secrete much of it into the medium.
To further demonstrate that the intracellular p45
was indeed concentrated in secretory vesicles, we coex-
pressed p45 in COS-1 cells with a fusion protein con-
sisting of GFP plus two tandem FYVE domains from

early endosomal antigen-1 (EEA1). This GFP-EEA1-
(FYVE)
2
constructs binds to PtdIns(3)P [18,19], a
phospholipid most abundant on endosomes and secre-
tory vesicles, but also found on Golgi and post-Golgi
membranes and in small quantities at the plasma
membrane. In these transfected cells, the intense red
immunofluorescence staining of intracellular p45 was
surrounded by rings of green fluorescence (Fig. 2C).
There were also smaller vesicles marked by GFP-
EEA1-(FYVE)
2
, without p45 inside, as well as a strong
green fluorescence in the nucleus. While the former
probably represent lysosomes and ⁄ or endosomes, the
nuclear staining is presumably nonspecific as GFP
alone also tends to accumulate in the nucleus [5]. We
conclude from these experiments that p45, which is
synthesized in these cells, translocates into the Golgi
system (perhaps directly from the rough endoplasmic
reticulum through cis-Golgi) and is subsequently pack-
aged in the trans-Golgi network into secretory vesicles.
Thus, it appears that many of the vesicles with high
p45 content are cytosolic secretory vesicles destined for
exocytosis (‘secretory vesicle’ in Fig. 2C). Indeed,
many cells contained vesicles in the process of being
emptied to the extracellular environment (‘exocytosis’
in Fig. 2C). Extracellular aggregates of p45 were detec-
ted in every transfection experiment (‘extracellular’ in

Fig. 2C). In contrast, numerous other proteins with a
HA tags remain completely intracellular [5,20,21].
The SEC14 domain, but not GOLD domain,
is sufficient for secretion of p45
Secreted proteins are usually synthesized as precursors
with a cleavable N-terminal signal peptide, composed
of a positively charged region, a hydrophobic core,
and a proteolytic cleavage site [22]. No such cleavable
signal sequence was found in p45 using the hidden
Markov model (HMM)-based signalip software,
which is available on the internet [23]. However, many
AB C
Fig. 1. Phospholipid binding by p45 protein. (A) SDS ⁄ PAGE gel stained with Coomassie blue of purified recombinant (lane 1) and natural
(lane 2) p45 proteins. (B, C) Protein lipid overlay assay using PIP Strips
TM
from Echelon and the indicated concentrations of natural (B) and
recombinant (C) p45 proteins. PE, phosphatidylethanolamine, PC, phosphatidylcholine, PS, phosphatidylserine, LPA, lysophosphatidic acid,
LPC, lysophosphatidylcholine, S1P, sphingosine-1-phosphate, PA, phosphatidic acid.
M. Merkulova et al. Secretion of p45 Sec14p-like protein
FEBS Journal 272 (2005) 5595–5605 ª 2005 FEBS 5597
other well documented secretory proteins lack signal
sequence and cannot currently be identified computa-
tionally as secreted proteins. These proteins have been
proposed to contain nonlinear or unrecognized
signal regions, which can be N-terminal, internal, or
C-terminal [24]. In fibroblast growth factor-16, a
A
C
B
Secretion of p45 Sec14p-like protein M. Merkulova et al.

5598 FEBS Journal 272 (2005) 5595–5605 ª 2005 FEBS
hydrophobic motif was shown to be important for
translocation of the protein into the endoplasmic reti-
culum membrane and mutation of this region abro-
gated secretion [25]. In order to determine if p45
contains such hydrophobic motifs, we analyzed its
hydrophilicity ⁄ hydrophobicity using the method of
Kyte and Doolittle [26]. The resulting plot (Fig. 3A)
showed that p45 contains several strongly hydrophobic
motifs, which could be signaling for secretion. How-
ever, these motifs were located within the SEC14 and
GOLD domains of p45 (Fig. 3A), which probably fold
as independent units. Thus, simply mutating the
hydrophobic segments would likely impair the proper
folding of these domains. Therefore, we instead chose
to make deletion mutants of p45. In the first deletion
mutant, we cloned amino acid residues 76–246 of p45
comprising the SEC14 domain into the pEF3HA vec-
tor. The second mutant consisted of the GOLD
domain, amino acid residues 265–381. When these con-
structs were expressed in COS-1 cells and visualized by
anti-HA staining, it was clear that the SEC14 construct
behaved like full-length p45 (Fig. 3B) and was found
both in cytoplasmic vesicles and in the extracellular
space (Fig. 3C), while the GOLD domain construct
was only seen inside cells (Fig. 3D). By immunoblot-
ting, the 19-kDa SEC14 protein was readily detectable
in the culture supernatant (Fig, 3E, lane 4), while the
13-kDa GOLD protein was present only in trace
amounts (lane 3) despite being well expressed by the

cells (lane 1). Thus, all the information necessary for
secretion of p45 appears to be contained within the
SEC14 domain.
Co-localization of p45 with PtdIns(3,4,5)P
3
in
intact cells
Based on the finding that p45 bound to
PtdIns(3,4,5)P
3
in vitro we decided to examine if p45
was associated with this phospholipid in intact cells.
COS-1 cells were transfected with the HA-
tagged p45 plasmid and then immunostained with an
anti-PtdIns(3,4,5)P
3
mAb, Alexa FluorÒ 488 goat
anti-(mouse Ig) Ig, and last with the TRITC-conju-
gated anti-HA mAb. As shown in Fig. 4, the resulting
green fluorescence was found to colocalize with p45
inside secretory vesicles and to a lesser extent at the
plasma membrane and in the extracellular space. The
weak staining of extracellular material is not surprising
given that detergents are included in the staining pro-
tocol. Thus, the staining for PtdIns(3,4,5)P
3
is partly
overlapping with p45, particularly in the secretory vesi-
cles, which usually do not stain specifically with this
mAb [5].

Since p45 appears to reside mostly inside a post-
Golgi vesicular compartment, we decided to test if a
cytosolic marker for PtdIns(3,4,5)P
3
, a fusion between
GFP and the pleckstrin homology (PH) domain of the
Btk kinase (GFP-Btk-PH) [27], would colocalize with
p45. When COS-1 cells cotransfected with GFP-Btk-
PH and HA-p45 were stained for p45 and viewed
under a confocal microscope, it was clear that the
green and red fluorescence did not colocalize at all.
While p45 was confined to discrete spots, GFP-Btk-PH
excluded these spots and was instead diffusely cyto-
plasmic and accumulated at parts of the plasma mem-
brane (Fig. 4B). Thus, unlike PTP-MEG2 [5] which
also binds PtdIns(3,4,5)P
3
via its SEC14 domain [5],
p45 is inaccessible to a cytoplasmic marker for this
lipid. This is in agreement with the notion that p45
with bound PtdIns(3,4,5)P
3
resides in the secretory
vesicle compartment. Finally, we stained cells for the
secretory vesicle marker carboxypeptidase E, which
largely colocalized with p45 (Fig. 5).
Discussion
Taken together, our findings indicate that the mam-
malian Sec14p-like protein p45 (TAP2) is a secreted
protein that binds PtdIns(3,4,5)P

3
and perhaps other
highly phosphorylated inositol phospholipids. Al-
though this transport function is not too different
from that of S. cerevisiae Sec14p, which transports
PtdIns between cellular membranes, our findings are
the first to report that phosphoinositides can be trans-
ported out by TAP2 into the extracellular environment.
Because p45 was purified from olfactory epithelium
Fig. 2. Endogenous p45 is located in granular cytoplasmic structures and extracellular space in COS-1 cells. (A) Confocal microscopy COS-1
cells transfected with empty pEF3HA vector (upper panels) or pEF3HA_p45 (lower panels) and then immunostained for HA-tagged p45 with
TRITC-conjugated anti-HA mAb (red). Right hand panels are Nomarski differential interference contrast images of the same field. The shown
cells are representative of the majority of stained cells. (B) Immunoblot with anti-HA mAb of conditioned medium (lanes 1 and 2) and cell ly-
sates (lanes 3 and 4) from COS-1 cells transfected with pEF3HA_p45 (lanes 1 and 3) or empty vector pEF3HA (lanes 2 and 4) for 48 h. p45
is indicated by an arrow. (C) Confocal microscopy of a COS-1 cells transfected with cotransfected with GFP-EEA1-(FYVE)
2
(green) and pEF3-
HA_p45 and stained as in (A). Note that overexpressed p45 protein is localized inside secretory vesicles and in extracellular space, while
GFP-EEA1-(FYVE)
2
is enriched around secretory vesicles and other organelles. There is no direct colocalization of GFP-EEA1-(FYVE)
2
with
p45.
M. Merkulova et al. Secretion of p45 Sec14p-like protein
FEBS Journal 272 (2005) 5595–5605 ª 2005 FEBS 5599
A
B
E
C

D
Secretion of p45 Sec14p-like protein M. Merkulova et al.
5600 FEBS Journal 272 (2005) 5595–5605 ª 2005 FEBS
and can be detected in secretory vesicles of sustentacu-
lar cells in the nasal mucosa, as well as in the mucus
produced by these cells, it seems that the secretion of
p45 we observe represents the physiological behavior of
this protein. The mRNA for p45 was also detected in
the surface layer of epithelial tissues [8]. p45 and the
B
A
Fig. 4. p45 colocalizes with PtdIns(3,4,5)P
3
, but not with GFP-Btk-PH, inside secretory vesicles. (A) Confocal microscopy of COS-1 cells
transfected with pEF3HA_p45 and double stained for PtdIns(3,4,5)P
3
with the specific mAb plus Alexa FluorÒ 488 goat anti-(mouse Ig) Ig
(green), and then for p45 with the TRITC-conjugated anti-HA mAb (red). (B) Confocal microscopy of COS-1 cells cotransfected with GFP-Btk-
PH construct (green) plus pEF3HA_p45 and stained with the TRITC-conjugated anti-HA mAb (red). Differential interference contrast images
of the same cells are shown in the far right panels. The shown cells are representative of the majority of stained cells. Note that
PtdIns(3,4,5)P
3
colocalizes with p45 whereas the GFP-Btk-PH cannot penetrate the secretory vesicles and therefore cannot colocalize with
p45.
Fig. 3. Secretion of the SEC14 domain of p45. (A) The hydrophilicity plot of p45 protein and location of the SEC14 and GOLD domains. Note
that the strongly hydrophobic motifs in p45 lie within these two domains. (B–D) Confocal microscopy of COS-1 cells transfected with pEF3-
HA_p45, pEF3HA_SEC14, and pEF3HA_GOLD constructs, as indicated, and stained with the TRITC-conjugated anti-HA mAb (red). Two rep-
resentative fields with accompanying Nomarski differential interference contrast images are shown. (E) Anti-HA immunoblot of cell lysates
(lanes 1 and 2) or conditioned medium (lanes 3 and 4) of COS-1 cells transfected with pEF3HA_GOLD (lanes 1 and 3) or pEF3HA_SEC14
(lanes 2 and 4). Note that only the SEC14 protein is present at high levels in the medium.

M. Merkulova et al. Secretion of p45 Sec14p-like protein
FEBS Journal 272 (2005) 5595–5605 ª 2005 FEBS 5601
bound PtdIns(3,4,5)P
3
may play important roles in
the protective nasal mucosa, perhaps as components
of the extracellular matrix. p45 may also release
PtdIns(3,4,5)P
3
to serve a function of its own in the
mucosa. Interestingly, plants also use a similar exocyto-
sis of lipid transfer proteins as part of the formation
of a protective surface wax [28,29].
We have found that secretion of p45 does not occur
via the conventional targeting mechanism of N-ter-
minal signal sequence recognition and cleavage. Rather,
p45 secretion is determined by the SEC14 domain of
p45, which is well secreted as an isolated protein
(Fig. 3). The SEC14 domain is an ancient lipid-binding
domain found in both plants and animals [1]. While the
prototypical yeast Sec14p binds and transports PtdIns
[2], this function has been taken over in higher eukary-
otes by structurally unrelated PtdIns transfer proteins
(PITPs) [30]. Dictyostelium discoideum represents a
transition stage in evolution with orthologs of both
S. cerevisiae Sec14p and mammalian PITPs involved in
PtdIns transport in the same cell [31]. In higher eukary-
otes, SEC14 domain-containing proteins have appar-
ently evolved to carry out other functions, such as
transport of tocopherol [10] or retinaldehyde [4] and

allosteric regulation of enzymes like the PTP-MEG2
tyrosine phosphatase [5,32] or Dbl family regulators of
small Ras-related G-proteins [1].
The metabolism of inositol phospholipids has been
intensely investigated during the last two decades [33].
The D3 hydroxyl group of the inositol ring is phos-
phorylated by a family of phosphoinositide 3-kinases
(PI3Ks), which can be subdivided into three classes
[34]. While class I PI3Ks mainly phosphorylate
PtdIns(4,5)P
3
to produce PtdIns(3,4,5)P
3
in response
to activation of many cell-surface receptors [35], class
III PI3Ks phosphorylate only PtdIns to produce
PtdIns(3)P [36]. Mammalian class III PI3K are ortho-
logs of the yeast vacuolar sorting protein Vps34p [37].
They are located primarily in the Golgi and other
internal membranes, where they produce PtdIns(3)P
and are thought to play role in protein sorting and
vesicle transport. Their function is mediated through
a number of effector proteins, containing FYVE
domains, which specifically bind PtdIns(3)P [18,19],
such as EEA1 [38,39], Hrs [40], PIKfyve (a vesicle-
bound PtdIns-5-kinase) [41]. In addition, phosphoryla-
tion of the D4 and D5 positions of the inositol ring
also play important roles in a variety of site-specific
recruitment and activation events in membrane traf-
ficking [42,43]. PtdIns(3,4,5)P

3
was also detected
recently on the cytosolic face of the enclosing mem-
brane of secretory vesicles [5]. However, it remains
unclear how PtdIns(3,4,5)P
3
gains access to the lumen
of secretory vesicles. Is it synthesized in this location
or is it transported there by p45? It also remains
unknown if PtdIns(3,4,5) P
3
bound to p45 was synthes-
ized using the class III PI3K (plus D4 and D5 kinases)
pathway on intracellular membranes or if it was syn-
thesized by class I PI3K and then transported to the
Golgi and ⁄ or secretory pathway from plasma mem-
brane. In the latter case, p45 could be involved in the
transport process. These questions, as well as the role
of p45-mediated export of phosphoinositides, will
require further studies.
Experimental procedures
Antibodies and reagents
The anti-influenza hemagglutinin (HA) tag epitope mAb
12CA5 conjugated to tetramethyl rhodamine isothiocyanate
(TRITC) was from Roche Molecular Biochemicals (Indi-
anapolis, IN, USA). The 16B12 anti-HA from Covance
(Richmond, CA, USA), was used for immunoblotting.
Alexa FluorÒ 594 goat anti-(mouse IgG H + L) Ig and
Alexa FluorÒ 488 goat anti-(mouse IgG H + L) Ig were
Fig. 5. p45 colocalizes with carboxypeptidase E, a secretory vesicle

marker. Confocal microscopy of COS-1 cells transfected with pEF3-
HA_p45 and double stained for carboxypeptidase E with a specific
mAb plus quantum dot-605-conjugated anti-(mouse Ig) Ig (green),
and then for p45 with rat anti-HA mAb plus quantum dot-655-conju-
gated anti-(rat Ig) Ig (red). DNA was stained with DAPI and the
fourth panel is a merge of the three colors. Note that carboxypepti-
dase E and p45 colocalize extensively.
Secretion of p45 Sec14p-like protein M. Merkulova et al.
5602 FEBS Journal 272 (2005) 5595–5605 ª 2005 FEBS
from Molecular Probes (Eugene, OR, USA). Rabbit poly-
clonal antiserum against native rat p45 protein was raised
as described previously [7]. All phospholipids, the anti-
PtdIns(3,4,5)P
2
mAb, and commercial PIP Strips
TM
with 15
different phospholipids were from Echelon (Salt Lake City,
UT, USA). Each PIP Strip
TM
had 100 pmol of PtdIns,
PtdIns(3)P, PtdIns(4)P, PtdIns(5)P, PtdIns(3,4)P
2
,
PtdIns(3,5)P
2
, PtdIns(4,5)P
2
, PtdIns(3,4,5)P
3

, phosphatidyl-
ethanolamine, phosphatidylcholine, phosphatidylserine,
lysophosphatidic acid, lysophosphatidylcholine, sphingo-
sine-1-phosphate, and phosphatidic acid spotted onto a
nitrocellulose membrane.
Preparation of water-soluble extract and
purification of natural p45 protein
Preparation of water-soluble extract of rat olfactory epithe-
lium was done as before [7]. Briefly, rat olfactory epithe-
lium was homogenized in 10 mm sodium phosphate buffer,
containing 150 mm NaCl and 0.5 mm phenylmethanesulfo-
nyl fluoride, pH 7.4. The homogenates were centrifuged at
20 000 g for 1 h, and supernatants (total protein concentra-
tion approximately 1 mgÆmL
)1
) were subjected consequently
to DEAE-Sepharose and Sephacryl S-200 chromatography
as described [7]. Purity of the protein was about 90% as
judged by SDS ⁄ PAGE followed by Coomassie blue stain-
ing. Immunoblots with rabbit polyclonal antiserum showed
a single band of the expected size ( 45 kDa). This protein
preparation will be referred here to as ‘natural p45 protein’.
Prokaryotic expression and purification of
recombinant p45 protein
For protein overexpression p45 cDNA was subcloned into
the prokaryotic expression vector pET-11a (Novagen,
Darmstadt, Germany) and then transformed into Escheri-
chia coli strain Rosetta
TM
(Novagen) which was designed

to enhance the expression of eukaryotic proteins that con-
tain codons rarely used in E. coli [44]. Expression procedure
was done according to conventional technique [45]. The
recombinant protein was purified by a two-step chromato-
graphic procedure using DEAE-Sepharose and Sephacryl
S-200 as described earlier [7]. Purity of the protein was
about 90% as judged by SDS ⁄ PAGE followed by Coomas-
sie blue staining. Immunoblot with rabbit polyclonal anti-
serum showed a single band of expected size ( 45 kDa).
Protein lipid overlay Assay
The lipid binding specificity of the p45 protein was assayed
as described [5,17]. Nitrocellulose filters spotted with
100 pmol of phospholipids (PIP Strips
TM
) were blocked in
3% (w ⁄ v) fatty acid-free BSA in 10 mm Tris ⁄ HCl, pH 8.0,
150 mm NaCl, and 0.1% (v ⁄ v) Tween-20 for 1 h and incu-
bated with 0.5 lg ÆmL
)1
(10 nm) of the p45 protein, natural
or recombinant, overnight at 4.C. The membrane was
washed three times for 10 min with 3% fatty acid-free BSA
in 10 m m Tris ⁄ HCl, pH 8.0, 150 mm NaCl, and 0.1%
Tween-20, and then incubated for 1 h with a 1 : 750-diluted
anti-p45 rabbit polyclonal antiserum at 37.C. The mem-
brane was washed as before, and incubated for 1 h with
1 : 2,000-diluted anti-rabbit ⁄ HRP conjugate. Finally, the
membranes were washed and p45 protein bound to the
membrane by virtue of its interaction with phospholipids
was detected by enhanced chemiluminescence using a kit

from Amersham (Arlington Heights, IL, USA).
Immunoblots
Immunoblotting was performed as before [5]. All immuno-
blots were developed by the enhanced chemiluminescence
technique according to the manufacturer’s instructions.
Transient transfection, immunofluorescence, and
confocal microscopy
For transient expression, p45 cDNA was subcloned into
the mammalian expression vector pEF3HA [20], in-frame
with a carboxy terminal HA tag. COS-1 cells seeded on
100-mm Petri dishes (for immunoblotting) or glass cover
slips (for microscopy) were transfected with a pEF3HA_p45
construct in the presence of the Lipofectamine
TM
transfec-
tion reagent (Invitrogen Corporation, Carlsbad, CA, USA)
according to the manufacturer’s instructions. 48 h after
transfection, cells were washed in phosphate-buffered saline
and fixed in freshly made 4% (v ⁄ v) formaldehyde in phos-
phate-buffered saline. Fixed cells were permeabilized with
0.1% (w ⁄ v) saponin in phosphate-buffered saline, then
blocked in 2.5% (v ⁄ v) normal goat serum in 0.1% (w ⁄ v)
saponin in phosphate-buffered saline for 30 min at room
temperature, and then incubated with primary and secon-
dary Ab diluted in the same buffer for 1 h each at room
temperature. For double immunostaining, the permeabilized
cells were first incubated with anti-PtdIns(3,4,5)P
3
mAb,
then with an Alexa FluorÒ 488 goat anti-(mouse IgG

H + L) Ig and then with the TRITC- conjugated anti-HA
mAb. After three washes with phosphate-buffered saline,
cover slips with cells were mounted onto glass slides and
viewed under a confocal laser scanning microscope (MRC-
1024; Bio-Rad, Hercules, CA, USA) with a 60· oil immers-
ion objective. A differential interference contrast image was
also taken of most cells.
Acknowledgements
We are grateful to Lewis C. Cantley for GFP con-
structs. This work was supported by grants from the
U.S. Civilian Research & Development Foundation
M. Merkulova et al. Secretion of p45 Sec14p-like protein
FEBS Journal 272 (2005) 5595–5605 ª 2005 FEBS 5603
(RB1-2338-MO-02, to TS and TM), the Russian Foun-
dation of Basic Research (02-04-48364, to MM), the
Scientific School (312-2003-4) and the Molecular and
Cellular Biology (no.200101, to VL), grants AG00252
(to HH), AI55741 (to TM), and CA96949 (to TM)
from the National Institutes of Health.
References
1 Aravind L, Neuwald AF & Ponting CP (1999)
Sec14p-like domains in NF1 and Dbl-like proteins
indicate lipid regulation of Ras and Rho signaling.
Curr Biol 9, 195–197.
2 Sha B, Phillips SE, Bankaitis VA & Luo M (1998) Crys-
tal structure of the Saccharomyces cerevisiae phosphati-
dylinositol-transfer protein. Nature 391, 506–510.
3 Bankaitis VA, Aitken JR, Cleves AE & Dowhan W
(1990) An essential role for a phospholipid transfer pro-
tein in yeast Golgi function. Nature 347, 561–562.

4 Crabb JW, Goldflam S, Harris SE & Saari JC (1988)
Cloning of the cDNAs encoding the cellular retinalde-
hyde-binding protein from bovine and human retina
and comparison of the protein structures. J Biol Chem
263, 18688–18692.
5 Huynh H, Wang X, Li W, Bottini N, Williams S, Nika
K, Ishihara H, Godzik A & Mustelin T (2003) Homo-
typic secretory vesicle fusion induced by the protein
tyrosine phosphatase MEG2 depends on polyphospho-
inositides in T cells. J Immunol 171, 6661–6671.
6 Novoselov VI, Peshenko IV, Evdokimov VJ, Nikolaev
JV, Matveeva EA & Fesenko EE (1994) Water-soluble
GTP-binding protein from rat olfactory epithelium.
FEBS Lett 353, 286–288.
7 Novoselov VI, Peshenko IV, Evdokimov VA, Nikolaev
JV, Matveeva EA & Fesenko EE (1996) 45-kDa GTP-
binding protein from rat olfactory epithelium: purifica-
tion, characterization and localization. Chem Senses 21,
181–188.
8 Merkulova MI, Andreeva SG, Shuvaeva TM, Novose-
lov SV, Peshenko IV, Bystrova MF, Novoselov VI,
Fesenko EE & Lipkin VM (1999) A novel 45 kDa
secretory protein from rat olfactory epithelium:
primary structure and localisation. FEBS Lett 450,
126–130.
9 Stocker A, Zimmer S, Spycher SE & Azzi A (1999)
Identification of a novel cytosolic tocopherol-binding
protein: structure, specificity, and tissue distribution.
IUBMB Life 48, 49–55.
10 Zimmer S, Stocker A, Sarbolouki MN, Spycher SE,

Sassoon J & Azzi A (2000) A novel human tocopherol-
associated protein: cloning, in vitro expression, and
characterization. J Biol Chem 275, 25672–25680.
11 Shibata N, Arita M, Misaki Y, Dohmae N, Takio K,
Ono T, Inoue K & Arai H (2001) Supernatant protein
factor, which stimulates the conversion of squalene to
lanosterol, is a cytosolic squalene transfer protein and
enhances cholesterol biosynthesis. Proc Natl Acad Sci
USA 98, 2244–2249.
12 Kempna P, Zingg JM, Ricciarelli R, Hierl M, Saxena S
& Azzi A (2003) Cloning of novel human SEC14p-like
proteins: ligand binding and functional properties. Free
Radic Biol Med 34, 1458–1472.
13 Stocker A, Tomizaki T, Schulze-Briese C & Baumann
U (2002) Crystal structure of the human supernatant
protein factor. Structure 10, 1533–1540.
14 Anantharaman V & Aravind L (2002) The GOLD
domain, a novel protein module involved in Golgi func-
tion and secretion. Genome Biol 3, research0023.
15 Stocker A & Baumann U (2003) Supernatant protein
factor in complex with RRR-alphatocopherylquinone: a
link between oxidized vitamin E and cholesterol bio-
synthesis. J Mol Biol 332, 759–765.
16 Panagabko C, Morley S, Hernandez M, Cassolato P,
Gordon H, Parsons R, Manor D & Atkinson J (2003)
Ligand specificity in the CRAL-TRIO protein family.
Biochemistry 42, 6467–6474.
17 Deak M, Casamayor A, Currie RA, Downes CP &
Alessi DR (1999) Characterisation of a plant 3-phos-
phoinositide-dependent protein kinase-1 homologue

which contains a pleckstrin homology domain. FEBS
Lett 451, 220–226.
18 Patki V, Lawe DC, Corvera S, Virbasius JV & Chawla
A (1998) A functional PtdIns(3)P binding motif. Nature
394, 433–434.
19 Gaullier JM, Simonsen A, D’Arrigo A, Bremnes B,
Stenmark H & Aasland R (1998) FYVE fingers bind
PtdIns(3)P. Nature 394, 432–433.
20 von Willebrand M, Jascur T, Bonnefoy-Berard N, Yano
H, Altman A, Matsuda Y & Mustelin T (1996) Inhibi-
tion of phosphatidylinositol 3-kinase blocks TCR ⁄ CD3-
induced activation of the mitogen-activated kinase Erk2.
Eur J Biochem 235, 828–835.
21 Gjorloff-Wingren A, Saxena M, Han S, Wang X,
Alonso A, Renedo M, Oh P, Williams S, Schnitzer J &
Mustelin T (2000) Subcellular localization of intracellu-
lar protein tyrosine phosphatases in T cells. Eur J
Immunol 30, 2412–2421.
22 von Heijne G (1985) Signal sequences. The limits of
variation. J Mol Biol 184, 99–105.
23 Nielsen H, Engelbrecht J, Brunak S & von Heijne G
(1997) Identification of prokaryotic and eukaryotic
signal peptides and prediction of their cleavage sites.
Protein Eng 10, 1–6.
24 Blobel G (1980) Intracellular protein topogenesis. Proc
Natl Acad Sci USA 77, 1496–1500.
25 Miyakawa K & Imamura T (2003) Secretion of FGF-16
requires an uncleaved bipartite signal sequence. J Biol
Chem 278, 35718–35724.
Secretion of p45 Sec14p-like protein M. Merkulova et al.

5604 FEBS Journal 272 (2005) 5595–5605 ª 2005 FEBS
26 Kyte J & Doolittle RF (1982) A simple method for dis-
playing the hydropathic character of a protein. J Mol
Biol 157, 105–132.
27 Salim K, Bottomley MJ, Querfurth E, Zvelebil MJ,
Gout I, Scaife R, Margolis RL, Gigg R, Smith CI, Dris-
coll PC, Waterfield MD & Panayotou G (1996) Distinct
specificity in the recognition of phosphoinositides by the
pleckstrin homology domains of dynamin and Bruton’s
tyrosine kinase. EMBO J 15, 6241–6250.
28 Sterk P, Booij H, Schellekens GA, Van Kammen A &
De Vries SC (1991) Cell-specific expression of the carrot
EP2 lipid transfer protein gene. Plant Cell 3, 907–921.
29 Pyee J, Yu H & Kolattukudy PE (1994) Identification
of a lipid transfer protein as the major protein in the
surface wax of broccoli (Brassica oleracea) Leaves Arch
Biochem Biophys 311, 460–468.
30 Cockcroft S (1998) Phosphatidylinositol transfer pro-
teins: a requirement in signal transduction and vesicle
traffic. Bioessays 20, 423–432.
31 Swigart P, Insall R, Wilkins A & Cockcroft S (2000)
Purification and cloning of phosphatidylinositol transfer
proteins from Dictyostelium discoideum: homologues of
both mammalian PITPs and Saccharomyces cerevisiae
sec14p are found in the same cell. Biochem J 347, 837–
843.
32 Wang X, Huynh H, Gjorloff-Wingren A, Monosov E,
Stridsberg M, Fukuda M & Mustelin T (2002) Enlarge-
ment of secretory vesicles by protein tyrosine phospha-
tase PTP-MEG2 in rat basophilic leukemia mast cells

and Jurkat T cells. J Immunol 168 , 4612–4619.
33 Vanhaesebroeck B, Leevers SJ, Ahmadi K, Timms J,
Katso R, Driscoll PC, Woscholski R, Parker PJ & Water-
field MD (2001) Synthesis and function of 3-phosphory-
lated inositol lipids. Annu Rev Biochem 70, 535–602.
34 Rameh LE & Cantley LC (1999) The role of phospho-
inositide 3-kinase lipid products in cell function. J Biol
Chem 274, 8347–8350.
35 Fruman DA, Meyers RE & Cantley LC (1998) Phos-
phoinositide kinases. Annu Rev Biochem 67, 481–507.
36 Schu PV, Takegawa K, Fry MJ, Stack JH, Waterfield
MD & Emr SD (1993) Phosphatidylinositol 3-kinase
encoded by yeast VPS34 gene essential for protein sort-
ing. Science 260, 88–91.
37 Volinia S, Dhand R, Vanhaesebroeck B, MacDougall
LK, Stein R, Zvelebil MJ, Domin J, Panaretou C &
Waterfield MD (1995) A human phosphatidylinositol
3-kinase complex related to the yeast Vps34p-Vps15p
protein sorting system. EMBO J 14, 3339–
3348.
38 Mu FT, Callaghan JM, Steele-Mortimer O, Stenmark
H, Parton RG, Campbell PL, McCluskey J, Yeo JP,
Tock EP & Toh BH (1995) EEA1, an early endosome-
associated protein. EEA1 is a conserved alpha-helical
peripheral membrane protein flanked by cysteine
‘fingers’ and contains a calmodulin-binding IQ motif.
J Biol Chem 270, 13503–13511.
39 Stenmark H, Aasland R, Toh BH & D’Arrigo A (1996)
Endosomal localization of the autoantigen EEA1 is
mediated by a zinc-binding FYVE finger. J Biol Chem

271, 24048–24054.
40 Komada M, Masaki R, Yamamoto A & Kitamurai N
(1997) Hrs, a tyrosine kinase substrate with a conserved
double zinc finger domain, is localized to the cytoplas-
mic surface of early endosomes. J Biol Chem 272,
20538–20544.
41 Sbrissa D, Ikonomov OC & Shisheva A (1999) PIKfyve,
a mammalian ortholog of yeast Fab1p lipid kinase,
synthesizes 5-phosphoinositides. J Biol Chem 274,
21589–21597.
42 Rameh LE & Cantley LC (1999) The role of phospho-
inositide 3-kinase lipid products in cell function. J Biol
Chem 274, 8347–8350.
43 Corvera S, D’Arrigo A & Stenmark H (1999) Phospho-
inositides in membrane traffic. Curr Opin Cell Biol 11,
460–465.
44 Saxena P & Walker JR (1992) Expression of argU, the
Escherichia coli gene coding for a rare arginine tRNA.
J Bacteriol 174, 1956–1964.
45 Sambrook J, Fritsch EF & and Maniatis T (1989) Mole-
cular Cloning: a Laboratory Manual, 2nd edn. Cold
Spring. Harbor Laboratory Press, Cold Spring Harbor,
NY, USA.
M. Merkulova et al. Secretion of p45 Sec14p-like protein
FEBS Journal 272 (2005) 5595–5605 ª 2005 FEBS 5605