Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
VIROPORIN-MEDIATED MEMBRANE PERMEABILIZATION: PORE
FORMATION BY NON-STRUCTURAL POLIOVIRUS 2B PROTEIN
Aitziber Agirre‡, Angel Barco¶#, Luis Carrasco¶§, and José L. Nieva‡§
‡Unidad de Biofísica (CSIC-UPV/EHU) and Departamento de Bioquímica,
Universidad del País Vasco, Aptdo. 644, 48080 Bilbao, Spain.
¶Centro de Biología Molecular (CSIC-UAM), Universidad Autónoma de Madrid,
Canto Blanco, 28049 Madrid, Spain.
#Current address: Center for Neurobiology and Behavior. Columbia University. New
York, NY 10027. U.S.A.
§ To whom correspondence may be addressed: Luis Carrasco. Centro de Biología
Molecular (CSIC-UAM), Universidad Autónoma de Madrid, Canto Blanco, 28049
Madrid, Spain. Phone: 34 91 3978450/ Fax: 34 91 3974799 E-mail:
[email protected]; José L. Nieva. Unidad de Biofísica (CSIC-UPV/EHU) and
Departamento de Bioquímica, Universidad del País Vasco, Aptdo. 644, 48080 Bilbao,
Spain. Phone: 34 94 6012615/ Fax: 34 94 4648500 E-mail: [email protected].
Running title: Pore-forming activity of poliovirus 2B
1
Copyright 2002 by The American Society for Biochemistry and Molecular Biology, Inc.
JBC Papers in Press. Published on August 14, 2002 as Manuscript M205393200 by guest on January 12, 2020
http://ww
w.jbc.org/
Dow
nloaded from
Abstract:
Enterovirus non-structural 2B protein is involved in cell membrane permeabilization
during late viral infection. Here we analyze the pore-forming activity of poliovirus
2B and several of its variants. Solubilization of 2B protein was achieved by
generating a fusion protein comprised of poliovirus 2B attached to a maltose-binding
protein (MBP), as an N-terminal solubilization partner. MBP-2B was assayed using
large unilamellar vesicles as target membranes. This fusion protein was able to
assemble into discrete structures that disrupted the permeability barrier of vesicles
composed of anionic phospholipids. The transbilayer aqueous connections generated
by MBP-2B were stable over time, allowing the passage of solutes of MW under
1000. Oligomerization was investigated using fluorescence resonance energy transfer
(FRET). Our data indicate that MBP-2B aggregation occurs at the membrane surface.
Moreover, MBP-2B binding to membranes promoted the formation of SDS-resistant
tetramers. We conclude that MBP-2B forms oligomers capable of generating a
tetrameric aqueous pore in lipid bilayers. These findings are the first evidence of
viroporin activity shown by a protein from a naked animal virus.
2
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
Introduction
Enteroviruses are non-enveloped animal viruses belonging to the
Picornaviridae family. Infection by species of the Enterovirus genus leads to significant
alterations in cell membranes, including morphological changes and functional
modifications (reviewed in: 1, 2). Of the cellular alterations induced by viral
infection, enhancement of plasma membrane permeability is among the most relevant.
Increased membrane permeability occurs at two well-defined times during the viral
infection cycle: at an early stage, when virus particles enter cells, and late during
infection, when most viral products are being synthesized (1-3).
Late membrane leakiness, occurring from the third hour post-infection,
requires viral gene expression. Initially it involves the diffusion of ions and small
molecules, but not macromolecules, through the plasma membrane (4,5). Intense
efforts have been made to unveil the molecular mechanisms underlying this process.
Some years ago, it was argued that the intracellular accumulation of viral particles
and/or massive expression of viral products might be the triggering factors for
enhanced permeability and eventual lysis of the cell. Contrary to this idea, cumulative
evidence suggests that the sole expression of certain enterovirus proteins might be
responsible for the increase in membrane permeability (6-8). Particularly, the
individual expression of 2B and 2BC proteins of coxsackie virus B3 (CVB3) and
poliovirus has been shown to reproduce several late infection effects, such as changes
in plasma membrane permeability (7-9), increased cytoplasmic calcium ion levels (8,
10), inhibition of protein secretion (9, 11, 12) and disassembly of the Golgi complex
(11, 13). Owing to their capacity to alter the permeability barrier of cell membranes,
3
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
these enteroviral proteins have been proposed to represent the first examples of
"viroporins" from naked viruses (7-9, 12, 14, 15).
Enterovirus 2B proteins are hydrophobic overall and rather small
(approximately 100 amino acids). Hydrophobicity within the 2B sequence appears to
be distributed across two main domains (2, 15): a first stretch (amino acids 32-55 in
poliovirus), which is partially amphipathic as a cationic α-helix at the N-terminus
(amino acids 34-49 in poliovirus), and a second hydrophobic region which might
constitute a transmembrane domain1 (TMD) (amino acids 61-81 in poliovirus). This
distribution seems to be functionally meaningful since mutations altering either first
domain amphipathicity or second domain hydrophobicity have been shown to
interfere with the ability of 2B to increase permeability (7, 8, 12, 14, 15) and with
viral RNA replication (12, 16).
In an effort to gain further insight into the cytolytic processes that occur
during enteroviral infection, we explored the molecular mechanism used by
poliovirus 2B protein to disrupt the permeability barrier of phospholipid vesicles.
Large unilamellar vesicles (LUV) encapsulating different solutes in their aqueous
compartments were employed to monitor the kinetics of solute release and entry into
the membrane-bound aqueous compartments. This model system allowed the
definition of several requirements for pore formation by 2B, in terms of individual
lipids, protein doses and limiting size of the solutes able to cross the membrane. The
experimental data presented here, indicate that poliovirus 2B may assemble into
tetrameric aqueous pores of diameter in the range 5-10 Å in anionic bilayers. Our
experimental approach and findings contribute to a better knowledge of the structural
4
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
and functional features of viroporins (1).
Experimental procedures
Materials
N-(7-nitro-benz-2-oxa-1,3-diazol-4-yl)phosphatidylethanolamine (N-NBD-PE),
Phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylinositol (PI),
phosphatidylserine (PS), dioleoylphosphatidylglycerol (DOPG), 1-palmitoyl-2-
oleoyl-phosphatidylglycerol (POPG), dipalmitoylphosphatidylglycerol (DPPG) and
1-stearoyl-2-arachidonoyl-phosphatidylglycerol (SAPG), were purchased from
Avanti Polar Lipids (Birmingham, AL, USA). 5-((((2-
iodoacetyl)amino)ethyl)amino)naphthalene-1-sulfonic acid (IAEDANS) and N,N’-
dimethyl-N-(iodoacetyl)-N’-(7-nitrobenz-2-oxa-1, 3-diazol-4-
yl)ethylenediamine (IANBD), 8-aminonaphtalene-1,3,6-trisulfonic acid sodium salt
(ANTS) and p-xylenebis(pyridinium)bromide (DPX) were from Molecular Probes
(Junction City, OR, USA).
2B protein expression and purification
Plasmids encoding maltose binding protein (MBP)-2B fusion proteins were
constructed by sub-cloning poliovirus-2B encoding DNA into pMALc2 vector (New
England Biolabs Inc., Beverly, MA). Mutants were subsequently obtained by the
polymerase chain reaction method as previously described (17). Mutations were
5
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
confirmed by DNA sequencing. All constructs were produced as MBP fusion proteins
from pMALc2 vectors using Escherichia coli BL21 as the host strain. A 20-ml
overnight culture was used to inoculate 1 liter of Luria-Bertani medium supplemented
with ampicilin (100 µg/ml), which was further incubated at 37°C until an A600 value
of 0.5 was achieved. The cells were induced with 1 mM isopropyl- β-D-thio-
galactopyranoside and incubated at 30 °C for an additional 3 h before harvesting by
centrifugation. Cells were then resuspended in a Hepes 5 mM, NaCl 100 mM, DTT 1
mM, EDTA 10 mM (pH, 7.4) buffer solution with protease inhibitors, and disrupted
by sonication at 4°C. The resulting lysate was centrifuged at 9000 x g for 30 min at
4°C to pellet cellular debris. The supernatant was applied to an affinity amylose-
agarose column. MBP-proteins were eluted with Hepes buffer supplemented with
maltose 10 mM. Eluted fractions were subsequently subjected to ultracentrifugation
(100000 x g, 4°C, 30 min). Purified proteins obtained from the supernatants were
characterized by SDS-PAGE in 3–13% gradient gels followed by silver staining
(Pierce, Rockford, IL).
Permeability assays
Large unilamellar vesicles (LUV) were prepared according to the extrusion method of
Hope et al. (18) in 5 mM Hepes, 100 mM NaCl (pH 7.4). The lipid concentrations of
the liposome suspensions were determined by phosphate analysis (19). The mean
diameter of vesicles was 95 nm as estimated by quasielastic light scattering using a
Malvern Zeta-Sizer instrument. The average vesicle size did not change upon
treatment with MBP-2B at any of the tested doses.
6
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
Release of vesicular contents into the medium was routinely monitored by the
ANTS/DPX assay (20). LUV containing 12.5 mM ANTS, 45 mM DPX, 20 mM NaCl
and 5 mM Hepes were obtained by separating the unencapsulated material by gel
filtration in a Sephadex G-75 column eluted with 5 mM Hepes, 100 mM NaCl (pH
7.4). Osmolarities were adjusted to 200 mosm in a cryoscopic osmometer (Osmomat
030, Gonotec, Berlin, Germany). Fluorescence measurements were performed by
setting the ANTS emission at 520 nm and the excitation at 355 nm. A cutoff filter (470
nm) was placed between the sample and the emission monochromator. The absence of
leakage (0%) corresponded to fluorescence of the vesicles at time zero; 100% leakage
was taken as the fluorescence value obtained after addition of Triton X-100 (0.5%
v/v). The degree of permeabilization was then inferred from the following equation:
%Leakage= [(Ff-F0)/ (F100-F0)] x 100 [1]
where Ff is the fluorescence determined after protein addition, F0 is the initial
fluorescence of the intact LUV suspension, and F100 is the fluorescence value after
the addition of Triton X-100. The extent of ANTS release was determined after
fluorescence intensity reached a plateau (routinely> 60 min). Vesicles were stored at
4°C and used within 1–2 days. No spontaneous leakage of entrapped material was
observed for vesicles stored at 4°C for at least 1 week.
Release of aqueous contents induced by MBP-2B was complementarily assessed
after trapping Evans Blue and NADH into the LUV. Leakage of these compounds
into the medium was assayed by absorbance as described elsewhere (21).
7
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
Entry of solutes into vesicles was evaluated through selective reduction of the NBD
probe residing in the inner leaflet of vesicle membranes according to the method
described by McIntyre and Sleight (22). Vesicles symmetrically labeled with N-
NBD-PE (0.6 mole %) were incubated with protein for 30 minutes at 37°C.
Dithionite (10 mM) was then added to the protein-vesicle mixtures. The degree of
vesicle permeabilization was inferred from the percentage of reduced internal NBD.
This value was calculated using the following equation:
%Reduced NBDi = [(F0-Ff)/ (F0-F100)] x 100 [2]
Where F0 and Ff correspond to the NBD fluorescence values after dithionite addition
to sealed and protein-treated vesicles, respectively, and F100 is the fluorescence value
of vesicles solubilized with Triton X-100 (0.5% v/v) and incubated with dithionite.
Binding to membranes and surface oligomerization
MBP-2B protein contains two single Cys residues within the 2B sequence (positions
73 and 86). This allowed the specific labeling of 2B with thiol-reacting NBD (MBP-
2B-NBD) and ANS (MBP-2B-ANS) fluorophores. Briefly, iodoacetamide
derivatives IANBD and IAEDANS were added to DTT-free preparations of 2B (6
µM) at a final 6:1 molar ratio, and incubated overnight in the dark at 4°C. The reaction
was stopped by the addition of DTT (1mM). The unreacted probe fraction was
separated by gel filtration on a Sephadex G-25 column eluted with 5 mM Hepes, 100
mM NaCl (pH 7.4). Eluted protein samples were subsequently concentrated by
filtration-centrifugation (4000 x g) using a Centricon system (Amicon-Millipore,
8
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
Bedford, MA, USA) with a cutoff of 10000 Da. Labeling efficiency was estimated as
described elsewhere (23). For MBP-2B-ANS, excitation and emission maxima were
observed at 339 and 482 nm respectively, whereas for MBP-2B-NBD, the maxima
were 481 (excitation) and 529 nm (emission).
MBP-2B-NBD was subsequently used to estimate the apparent mole-fraction
partition coefficient, Kx. Partitioning curves were elaborated by monitoring the
fractional change in emitted NBD fluorescence from MBP-2B-NBD titrated with
increasing lipid concentrations. Fluorescence was recorded in a Perkin Elmer MPF-66
spectrofluorimeter with excitation and emission monochromators set at 480 and 530
nm respectively and corresponding slits of 5 nm. Values of Kx were determined by
fitting the experimental values to the hyperbolic function:
[(Fmax/F0)-1] [L]
F/F0 = 1+ ----------------------- [3]
Kd+ [L]
where [L] is the lipid concentration and Kd is the apparent dissociation constant, or
lipid concentration at which the bound protein fraction is 0.5. Thus, Kx=[W]/Kd
where [W] is the molar concentration of water (24).
Protein-protein interactions in solution and at membrane surfaces were monitored
through resonance energy transfer (FRET) from MBP-2B-ANS to MBP-2B-NBD.
FRET was measured by the donor quenching method (23), comparing MBP-2B-
ANS emission spectra in the presence and absence of the acceptor MBP-2B-NBD.
Corrected spectra were recorded in a Perkin Elmer MPF-66 spectrofluorimeter with
9
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
excitation set at 339 nm and slits of 5 nm (excitation) and 10 nm (emission). The
efficiency of energy transfer (E) was determined according to Stryer (25) from the
relative fluorescence intensity of the ANS in the absence (Q0) and presence (QT) of
MBP-2B-NBD in the vesicles as:
E = 1- (QT / Q0) [4]
Oligomerization in the membranes was also evaluated using SDS-PAGE gradient
gels. Samples consisted of 0.1 nM protein incubated with increasing lipid
concentrations for 30 min at 37°C. Silver-stained gels were digitalized and the
relative amount of MBP-2B oligomers produced, quantified using UN-SCAN-IT
software (Silk Scientific, Orem, UTAH).
Results:
Poliovirus 2B is a small, hydrophobic protein. In order to characterize its
porin-like activity, recombinant 2B versions that were suitable for purification and
stable in aqueous solution had to be prepared using several approaches. Solubilization
of 2B alone overexpressed in bacteria requires the presence of high amounts of
detergent in the medium, which interferes with leakage measurements in a vesicular
system (data not shown). From the different possibilities considered for the
production of aqueous soluble 2B fusion proteins, MBP was selected as the best
option on the basis of its weak affinity for anionic membranes (see below). In
10
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
consequence, MBP-2B constructs were expressed in bacteria and the fusion protein
purified as an aqueous soluble species. In a further effort to obtain pure 2B versions
from these fusion constructs, samples were subjected to proteolysis with Factor Xa.
MBP-2B cleavage by Xa to produce free 2B resulted in low yields and reduced
solubility of the 2B product. An additional attempt to generate isolated 2B involved
treating the membrane-bound MBP-2B with the Xa protease. However, this protease
readily induced clotting of the lipid vesicles, preventing the progression of the
reaction. Synthesis of MBP-2B allowed characterization of the porin-like activity of
2B in a vesicular system. Thus, for the experiments described below in which we
evaluated the particular features of MBP-2B interaction with vesicle membranes
(permeabilization, partitioning and surface-oligomerization), it should be noted that,
under similar conditions, control experiments confirmed that MBP alone showed no
detectable effect.
The data shown in Figure 1 reveal the ability of MBP-2B to permeabilize
large unilamellar vesicles composed of anionic phospholipid. Coencapsulation of
ANTS and DPX probes in vesicles extinguishes most of the ANTS fluorescence.
Release of this probe to the medium may be followed by increased fluorescence due
to the relief of DPX quenching upon dilution (20). The size of these solutes (MW:
~425) is in the range of that of the aminoglycoside antibiotic hygromycin B (MW:
527.5) widely used for 2B-induced cell permeability measurements (6, 7, 9, 12, 14).
However, we caution here that experimental evidence for 2B localized at the plasma
membrane has only been obtained for coxsackievirus 2B, in both transfected COS
cells and infected HeLa cells (8). In poliovirus 2B-expressing cells the protein has
11
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
been localized at intracellular membranes (7, 13) and, therefore, it should not be
excluded the possibility that internal membrane permeabilization is indirectly
responsible for enhanced permeability of the plasma membrane.
Figure 1A shows the typical kinetics of the aqueous contents of liposomes
spilling into the medium upon incubation with MBP-2B. It may be seen that a plateau
was reached, depending on the protein-to-lipid mole ratio (panel B). Saturation with
time is consistent with a permeabilization process mediated by structures irreversibly
attached to the lipid bilayers, i.e., not interchanging between the different vesicles in
the samples. 2B-induced permeabilization in PI vesicles started at protein-to-lipid
mole ratios > 1:50000. According to the apparent mole-fraction partition coefficient
estimated for this system (see below), ~70% of the added protein was associated with
lipid vesicles under these measuring conditions. Initiation of leakage at such low
protein doses in the membrane suggests that lysis was mediated by a highly specific
mechanism. Thus, lysis was not a mere consequence of adding massive amounts of
protein to the bounding vesicle membrane.
To validate the previous assay as specifically measuring 2B lytic function, we
next explored the effects of 2B variants (Fig. 2). The changes in the protein sequence
of the mutant MBP-2B proteins generated were previously reported to affect the
ability of 2B to permeabilize cell membranes. The K[39,42,46]E mutation, affecting
the cationic amphipathic α-helix, reduces plasma membrane permeability towards
hygromycin B as well as the capacity of 2B to disrupt intracellular Ca2+ homeostasis
(8, 12). Consistent with these findings, this MBP-2B variant showed severely
impaired induction of the release of the LUV solute contents to the medium. As
12
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
shown in Figure 2 this triple aminoacid substitution in 2B reduced both the extent and
the initial rate of vesicle permeabilization by ~80 and ~90%, respectively.
V52D/I54K polar substitution within the hydrophobic residue cluster at the C-
terminal end of the amphipathic helix, has been shown to interfere with 2B ability to
enhance membrane permeability in HeLa cells (7). This non-conservative mutation
reduced both the extent and the initial rates of vesicle permeabilization by ~40 and
~65%, respectively. In comparison, a conservative/non-conservative A66G/L68R
double substitution within the second hydrophobic domain of 2B (14) had no effect
on the permeabilization levels attained, while the initial rate of vesicle
permeabilization was reduced by ~30%. A further double A69E/I71E polar
substitution that decreases second domain hydrophobicity in the 2B protein (16)
reduced the extent of permeabilization by ~50% and slowed the process by ~70%. In
summary, the data derived from the set of MBP-2B variants tested appear to
reinforce the functional significance of the hydrophobic domains within 2B
sequences, highlighting the physiological relevance of the in vitro vesicle assay.
The permeabilization phenomenon described above showed a strict
requirement for anionic phospholipids in the membrane composition (Table 1).
Vesicles made of electrically neutral PC were not efficiently permeabilized by MBP-
2B. Mixing PC with phosphatidylethanolamine and/or cholesterol did not result in
efficient permeabilization either (data not shown). Among the negatively charged PI
and PS species present in cell membranes, highest efficiency was observed for
vesicles made of pure PI, while PS was a poor substrate for the lytic activity of the
protein. To ascertain whether the PI effect was due to the specific recognition of
13
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
inositol by MBP-2B versus the serine polar head group, or to different acyl-chains
predominantly present in both lipids (see Table 1), 2B lytic activity was assayed in
negatively charged vesicles made from synthetic PG-s containing acyl-chains of
different length and degree of saturation. The data shown in Table 1 indicate that
MBP-2B was unable to permeabilize PG with acyl chains comparable to those
mainly present in PS. In contrast, vesicles made of synthetic SAPG with a fatty acid
composition similar to that of PI, were readily permeabilized by the protein. It is of
note that most of PC lipid’s acyl chains were also similar to those present in PI. Thus,
according to these results it seems that the combination of negatively charged polar
head groups with a minimal amount of long unsaturated acyl chains was required to
sustain an optimal 2B lytic activity in vitro.
The experiments described in Figure 3 were designed to assess the stability
and size of the permeabilizing structures assembled by MBP-2B. Several
prerequisites need to be satisfied to be able to conclude that leakage takes place via
the irreversible assembly of discrete lytic pores: i) the complex should not interchange
from one vesicle to another once inserted in the lipid bilayer, ii) the pore should allow
continuous passage of solutes, i.e., membrane integrity is not to be transiently
perturbed as a consequence of the mere insertion of the protein, and iii) there has to
be a size limit for molecules able to leak out from the vesicles.
MBP-2B exchange between vesicles was analyzed in the experiments shown
in Figure 3A. The results show that the ability of MBP-2B to permeabilize ANTS-
containing PI vesicles was reduced with the time of pre-incubation with empty PI
vesicles (Figure 3A). No inhibition was detected using empty PC vesicles, indicating
14
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
that the blocking effect is probably induced by the irreversible attachment of MBP-
2B to the anionic PI membranes.
The stability of the MBP-2B lytic structure was evaluated by combining
solute leakage-out with solute entry assays (Figure 3B). The rationale behind these
experiments was that if leakage-out reflects the formation of discrete pores, they
should also allow the entry of solutes once the pores have formed. Permeable vesicles
were produced by pre-incubating PI LUV with increasing amounts of MBP-2B.
Entry of solutes into these pre-formed protein-vesicle complexes was inferred from
the percentage of NBD reduced by externally added dithionite. In symmetrically
labeled vesicles, this compound caused an approximate 55% reduction in NBD –
judged by the decrease in fluorescence intensity (data not shown) – indicating specific
reduction of the probe residing in the external membrane leaflet (22, 26). We thus
took this value to represent a zero level of solute entry and assigned complete NBD
reduction in detergent-solubilized vesicles the value of 100% entry. The data in
Figure 3B indicate that 2B-permeabilized vesicles allowed the entry of solutes.
Moreover, when solute entry and leakage-out were compared as a function of the
protein dose, both processes were shown to occur to a similar extent. These findings
are consistent with the presence in the vesicle membranes of stably-open aqueous
connections that allowed free diffusion of the tested solutes.
To calculate the molecular weight cutoff of the released solutes induced by
MBP-2B, several compounds were encapsulated in vesicles, including ANTS (MW
427), DPX (MW 422), NADH (MW 660) and Evans Blue (MW 960) (Figure 3C).
After incubation of vesicles with MBP-2B, determination was made of the substance
15
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
leaking out. It was found that compared to NADH, there was very scarce leakage of
Evans blue. The Stokes radii of the permeating compounds is of the order of 6 Å (27).
As mentioned above, this pore aperture size also lies within the range of that
corresponding to hygromycin B, an agent broadly used in cell permeability
determinations. In conclusion, the stability of the permeabilizing structures and the
size dependence of leakage observed also support the notion of MBP-2B forming an
aqueous pore in the vesicle membrane.
Possible interactions between MBP-2B and membranes, and among MBP-2B
protein monomers were subsequently analyzed by means of the fluorescently labeled
derivatives MBP-2B-ANS and MBP-2B-NBD (Figure 4). Derivatization processes
did not appreciably affect 2B pore-forming activity. This was inferred by the similar
degrees of vesicle permeabilization induced by labeled and unlabeled MBP-2B
versions (data not shown). The MBP-2B fusion protein contains two reacting Cys
residues in the 2B sequence, C73 and C86. These are found within the second
hydrophobic domain and the adjacent interfacial region, respectively. The sensitivity
of the NBD probe to changes in the dielectric constant provides a suitable means of
investigating the partitioning of 2B into the apolar milieu of the membrane. The
fluorescence emission spectrum of MBP-2B-NBD was measured in aqueous solution
and in the presence of PI vesicles (Figure 4A). Addition of PI vesicles resulted in
significant enhancement of the intensity. The increased fluorescence observed in PI
samples was due to relocation of the NBD moiety to the more hydrophobic
environment of the bilayer. This is consistent with the insertion into the PI membrane
of at least one of the labeled Cys residues, most likely Cys-73 located within the
16
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
second hydrophobic domain. Of note under these measuring conditions, is that the
spectral contribution of unbound MBP-2B-NBD species in the PI sample was
negligible (see below).
Based on the NBD spectral changes, binding experiments were performed to
establish the mole fraction coefficient (Kx) for the partitioning of MBP-2B into PI
membranes (Figure 4B). From the partitioning curve, an apparent Kx value of 2.85 x
106 was obtained, indicating that under the conditions of Figure 4A, >90% of added
MBP-2B-NBD was associated with PI vesicles. In contrast, no change in NBD
fractional fluorescence was observed at lipid concentrations of up to 500 µM in the
presence of PC membranes. These findings suggest that the protein did not bind to
these vesicles and also explain the lack of lytic activity detected in PC vesicles (Table
1).
The affinity evaluation in the previous experiments allowed monitoring of the
aggregation state of membrane-bound MBP-2B by means of resonance energy
transfer measurements (Figure 4C). To this end, we selected a lipid concentration at
which the amount of free protein was negligible. MBP-2B-ANS served as the energy
donor, while MBP-2B-NBD was used as the fluorescence acceptor. In the presence
of PI vesicles quenching of donor emission, consistent with an efficient energy
transfer E = 0.4, was observed. In contrast, no energy transfer was observed in
absence of vesicles or when proteins were incubated with PC vesicles. At the 1:1000
protein-to-lipid mole ratio used in the PI experiment, the low surface density would
preclude energy transfer between unassociated protein monomers. This may be
17
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
inferred from the following calculations. The R0 distance computed for the ANS-
NBD couple is around 27 Å (28). Approximating monomer surface distribution to
circular areas, the protein-to-lipid ratio required in a vesicle for an E = 0.05 would be
1:20. Or, in other words, at the experimentally used 1:1000 protein-to-lipid ratio, E
would be H 0 for randomly distributed MBP-2B. Thus, the significant levels of
energy transfer for the ANS-donor/NBD-acceptor pair detected suggest that 2B
multimers associate within the membrane.
In order to determine the nature of MBP-2B self-association within
membranes, the oligomerization state of the protein was next analyzed, in the
presence and absence of vesicles (Figure 5). As shown in Figure 5A, the purified
protein in solution showed a single SDS-PAGE band with a molecular weight
consistent with that of the MBP-2B monomer (H51 kDa). The presence of PI vesicles
induced the formation of two prominent higher molecular weight bands of
approximately 100 and 200 kDa, corresponding to dimers and tetramers of MBP-2B,
respectively. In the presence of PC vesicles, no MBP-2B tetramers were observed
and only minor traces of dimers could be detected. Bands corresponding to higher
order complexes were no observed either in control experiments when MBP alone
was incubated with PI vesicles (data not shown). Quantification of the oligomeric
MBP-2B species formed as a function of increasing PI concentrations (Figure 5B and
5C) gave rise to curves reproducing the saturation behavior of the partitioning curve
in Figure 4B. Apparent Kd values were ~20 µM for both NBD binding and oligomer
formation, indicating a similar dependence of both processes on lipid concentration.
Taken together, these data suggest that binding to membranes induced the formation
18
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
of MBP-2B oligomers, tetramers being the highest MW species.
Discussion
Viroporins are a family of viral proteins whose common functional
characteristic is their capacity to compromise cell membrane integrity during virus
infection. In this manner, viroporins favor the release of viral progeny (1-3). Thus
far, there were no functional assays that could quantitatively evaluate viroporin lytic
activity in vitro. Viroporin sequences are short and extremely hydrophobic and,
consequently, difficult to manipulate in aqueous solution. Among our many attempts
to produce aqueous soluble 2B, i.e., using pure 2B or several derived fusion proteins,
we found that the best option was to generate MBP-2B constructs to obtain pure
soluble protein. This MBP-2B complex proved suitable for characterizing the pore-
forming activity of 2B in a vesicular system.
Two putative models had been previously proposed to account for the
increased membrane permeability developed by cells upon 2B expression (2, 12). One
such model is based on the structural strategies described for several pore-forming
cytolytic toxins (29, 30). According to this model, the Lys-rich amphipathic helices
of 2B (residues 32-55) might form transbilayer oligomers and constitute an aqueous
pore, with their hydrophobic sides facing the lipid bilayer and their hydrophilic
residues orientated towards the water-filled lumen of the pore. An alternative, but not
exclusive, model postulated that the amphipathic helices, lying parallel to the
membrane plane, might partially penetrate into a single monolayer interface in a
19
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
"carpet-like" fashion (30, 31). The first model predicts the formation of membrane-
embedded oligomers of a discrete size. These complexes should be able to assemble
aqueous pores with a molecular weight cutoff for the solutes capable of passing
through, i.e., they should be of a defined diameter. In the second model, oligomers, if
any, may be of heterogeneous sizes and all types of solutes may eventually permeate
across the membranes, whose general architecture is disturbed. A further requirement
is that there be relatively high protein doses in the membranes, such that they may be
permeabilized by a mechanism like the one predicted by the second model. The
experimental results presented here, illustrate the ability of poliovirus 2B to
permeabilize vesicles at low protein-to-lipid ratios (Figures 1-2) such that stable
discrete permeating structures are created (Figure 3) and oligomerized in membranes
giving rise to SDS-resistant tetramers (Figures 4-5). Thus, our findings would appear
to be consistent with the first model, i.e., the formation by viroporin 2B of a
transmembrane aqueous pore that allows free diffusion of small solutes.
The production of soluble MBP-2B allowed characterization of the pore-
formation requirements in terms of lipid composition (Table 1). Negatively charged
polar head groups were required for 2B activity. The presence of negative charge in
the membrane was necessary for initial binding of MBP-2B (Figure 4). This finding
confirmed the involvement of electrostatic interactions in the early stages of the 2B-
membrane interaction process. Currently proposed mechanisms for the formation of
“barrel-stave”-like pores by amphipathic sequences assume that, following rapid
binding to membranes, aggregation occurs at their surface (27, 32-34). When the
protein aggregates attain a critical size, a pore will be formed. This basic mechanism
20
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
seems to be functional in the case of 2B. According to the FRET assays, the MBP-2B
protein aggregated only upon binding to membranes. These data suggest that the
cationic amphipathic helix drives initial partitioning, thereby allowing further
surface-assembly of the oligomeric MBP-2B structure giving rise to membrane
permeabilization.
However, the transition from a surface aggregate to a pore requires the
insertion into the lipid bilayer of protein portions to act as integral anchors, a process
that might be dependent on phospholipid acyl chain composition (33-35). The
analysis of lipid requirement (Table 1) suggested that the permeabilization process
depends on acyl chain composition rather than on polar head group identity. 2B
contains two putative helical transmembrane domains, approximately spanning
residues 32-55 (amphipathic helix) and 61-81 (TMD), respectively. Notably, the 2B
variant K[39,42,46]E, in which positively charged Lys residues in the amphipathic
helix were substituted by negatively charged Glu, showed impaired permeability,
while polar substitutions directed towards the putative TMD, also interfered with 2B
lytic activity (Figure 2). In an aqueous pore, these domains might combine in each
monomer to form a hairpin "α-loop-α" motif that might span the bilayer (Figure 6).
The degree of hydrophobic mismatch between the length of this element and the
thickness of the bilayer core might regulate spontaneous translocation and pore
assembly. However, bilayer thicknesses for the phosphorylglycerol-based DPPG,
DOPG, POPG, SAPG series are very similar, the phosphate spacing ranging between
37 and 38 Å (36, 37). Thus it seems more plausible that the different levels of
permeability observed for this series are related to the different acyl-chain
21
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
unsaturation degrees. It is likely that single units can better become surface-adsorbed
and/or inserted perpendicularly into the membrane in the presence of unsaturated acyl
chains (35).
Analysis of the oligomeric state of the protein in membranes indicated that the
tetramer might represent the pore-forming structure of 2B. Homomultimerization
reactions of the poliovirus 2B protein had been previously inferred using a yeast two-
hybrid system (38). In a more recent work, multimerization reactions of the coxsackie
virus 2B protein have been further analyzed using a mammalian two-hybrid system
(39). Here we present compelling evidence to support the formation of membrane-
bound 2B oligomers. Formation of SDS-resistant oligomers is a common theme
among membrane-disrupting proteins. This phenomenon has been observed for
channel-forming (40) and pore-forming toxins (41, 42) as well as for fusion-
inducing envelope glycoproteins (43). In this regard membrane insertion might lead
to the establishment of hydrophobic interactions between 2B monomers that cannot
be disrupted within the nonpolar bilayer milieu, a requirement that must be fulfilled in
order to assemble stable and functional pores.
According to our estimates, the molecular weight cutoff for the release of
solutes was approximately that of NADH (MW 660), whose Stokes radius is about 6
Å (27). Thus, our data suggest a tetrameric aqueous pore with a diameter in the range
of 6 Å as the lytic 2B unit (Fig. 6). The model in Figure 6 highlights the different
structural roles played by both transbilayer helices in establishing the 2B membrane
pore. The amphipathic helix in each monomer serves to establish the interface
between the lumen of the aqueous pore and the hydrophobic milieu of the membrane,
22
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
and the second helix constitutes an integral transmembrane anchor. While the former
assembles the permeating pore, the latter reverses the chain direction in the
monomers, so that both the N- and the C-terminal ends of the protein remain at the
same side of the membrane.
Remarkably, a membrane-embedded hairpin organization, oligomeric state
and pore diameter, comparable to those described above for poliovirus 2B, have been
proposed for the membrane-inserting domain of Bacillus thuringiensis Cry1A toxin
(44-46). It is tempting to speculate that structural responses in proteins of diverse
origin have converged to produce integral products that are capable to perturb the
membrane permeability barrier. In summary, poliovirus 2B protein might use
common structural themes, such as those found in several bacterial toxin families, to
assemble tetrameric aqueous pores in membranes. The development of in vitro assays
designed to measure the lytic interaction of 2B with membranes, as well as its degree
of oligomerization therein, will lead to a better understanding of the membrane
permeabilization mechanisms that act during enterovirus infection.
ACKNOWLEDGMENTS
We acknowledge the financial support of the DGICYT project numbers
PM99-0002 (AB and LC) and BIO2000-0929 (AA and JLN). Further support to JLN
was obtained from the Basque Government (PI-1998-32) and the University of the
Basque Country (UPV 042.310-G03/98). A.A. was a recipient of a predoctoral
fellowship of the Basque Government. CBM was awarded with an institutional grant
23
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
by the Fundación Ramón Areces.
24
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
References
1. Carrasco, L. (1995) Adv. Virus Res. 45, 61-112.
2. Irurzun, A., Guinea, R., Barco, A., and Carrasco, L. (2002) in The Picornaviruses.
Eds. Semler B. and Wimmer E.,. (ASM Press). In press.
3. Carrasco, L., Perez, L., Irurzun, A., Lama, J., Martinez-Abarca, F., Rodriguez, P.,
Guinea, R., Castrillo, J. L., Sanz, M. A., and Ayala, M. J. (1993) in Regulation
of Gene Expression in Animal Viruses, eds. Carrasco, L., Sonenberg, N. and
Wimmer, E. (Plenum Press, London, New York). pp. 283-305.
4. Carrasco, L. and Smith A. E. (1976) Nature 264, 807-809.
5. Carrasco, L. (1978) Nature 272, 694-699.
6. Lama, J., and Carrasco, L. (1992) J. Biol. Chem. 267, 15932-15937.
7. Aldabe, R., Barco, A., and Carrasco, L. (1996) J. Biol. Chem. 271, 23134-23137.
8. Van Kuppeveld, F. J. M., Hoenderop, J. G. J., Smeets, R. L. L., Willems, P. H. G.
M., Kijkman, H. B. P. M., Galama, J. M. D., and Melchers W. J. G. (1997)
EMBO J. 16, 3519-3532.
9. Doedens, J.R., and Kirkegaard, K. (1995) EMBO J. 14, 894-907.
10. Aldabe, R., Irurzun, A., and Carrasco, L. (1997) J. Virol. 71, 6214-6217.
11. Barco, A., and Carrasco, L. (1995) EMBO J. 14, 3349-3364.
25
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
12. Van Kuppeveld, F. J. M., Melchers, W. J. G., Kirkegaard, K., and Doedens J. R.
(1997) Virology 227, 111-118.
13. Sandoval, I. V., and Carrasco, L. (1997) J. Virol. 71, 4679-4693.
14. Barco, A., and Carrasco, L. (1998) J. Virol. 72, 3560-3570.
15. Van Kuppeveld, F. J. M., Galama, J. M. D., Zoll, J., Van den Hurk, P. J. J. C., and
Melchers W. J. G. (1996) J. Virol. 70, 3876-3886.
16. Van Kuppeveld, F. J. M., Galama, J. M. D., Zoll, J., and Melchers W. J. G. (1995)
J. Virol. 69, 7782-7790.
17. Higuchi, R., Krummel, B. and Saiki, R. K. (1988) Nucleic Acids Res. 16, 7351-
67.
18. Hope, M. J., Bally, M. B., Webb, G., and Cullis, P. R. (1985) Biochim. Biophys.
Acta 812, 55-65.
19. Böttcher, C.S.F., van Gent, C. M., and Fries, C. (1961) Anal. Chim. Acta 24,
203-204.
20. Ellens, H., Bentz, J., and Szoka, F. C. (1985) Biochemistry 24, 3099-3106.
21. De los Rios, V., Mancheño, J.M., Lanio, M.E., Oñaderra, M., and Gavilanes, J. G.
(1998) Eur. J. Biochem. 252, 284-289.
22. McIntyre, J. C., and Sleight, R. (1991) Biochemistry 30, 11819-11827.
23. Ubarretxena-Belandia, I., Hozeman, L., van-der-Brink, E., Pap, E.H., Egmond,
26
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
M. R., Verheij, H., and Dekker, N. (1999) Biochemistry 38, 7398-7405.
24. White, S., Wimley, W. C., Ladokhin, A. S., and Hristova, K. (1998) Methods
Enzymol. 295, 62-87.
25. Stryer L. (1978) Annu. Rev. Biochem. 47, 819-846.
26. Agirre, A., Nir, S., Nieva, J.L., and Dijkstra, J. (2000) J. Lipid Res. 41, 621-628.
27. Parente, R. A., Nir, S., and Szoka, F. C. Jr. (1990) Biochemistry 29, 8720-8728.
28. Franzen, J.S., Marchetti, P.S., and Feingold, D.S. (1980) Biochemistry 19, 6080-
6089.
29. Cramer, W.A., Heymann, J. B., Schendel, S. L., Deriy, B. N., Cohen, F. S.,
Elkins, P. A., and Stauffacher, C.V. (1995) Annu. Rev. Biophys. Biomol.
Struct. 24, 611-641.
30. Shai, Y. (1995) Trends Biochem. Sci. 20, 460-464.
31. Bechinger, B. (1997) J. Membrane Biol. 156, 197-211.
32. Matsuzaki, K. (1998) Biochim. Biophys. Acta 1376, 391-400.
33. Nir, S., and Nieva, J.L. (2000) Prog. Lipid Res. 39, 181-206.
34. Huang, H.W. (2000) Biochemistry 39, 8347-8352.
35. Nicol, F., Nir, S., and Szoka, FC Jr. (2000) Biophys. J. 78, 818-829.
36. Rawicz, W., Olbrich, K.C., McIntosh, T., Needmham, D., and Evans, E. (2000)
27
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
Biophys. J. 79, 328-339.
37. Nagle, J.F., and Tristram-Nagle, S. (2000) Biochim. Biophys. Acta 1469, 159-
195.
38. Cuconati, A., Xiang, W. K., Lahser, F., Pfister, T., and Wimmer E. (1998) J.
Virol. 72, 1297-1307.
39. De Jong, A. S., Schrama, I. W. J., Willems, P.H.G.M., Galama, J. M. D.,
Melchers, W. J. G., and van Kuppeveld, F. J. M. (2002) J. Gen. Virol. 83, 783-
793.
40. Buckley, J.T., Wilmsen, H.U., Lesieur, C., Schulze, A., Pattus, F., Parker, M. W.,
and van der Goot, F. G. (1995) Biochemistry 34, 16450-16455.
41. Zitzer, A., Zitzer, O., Bhakdi, S., and Palmer, M. (1999) J. Biol. Chem. 274,
1375-1380.
42. Shepard, L. A., Shatursky, O., Johnson, A. E., and Tweten, R. K. (2000)
Biochemistry 39, 10284-10293.
43. Bron, R., Wahlberg, J. W., Garoff, H., and Wilschut, J. (1994) EMBO J. 12, 693-
701.
44. Schwartz, J. L., Juteau, M., Grochulski, P., Cygler, M., Prefontaine, G.,
Brousseau, R., and Masson, L. (1997) FEBS Lett. 410, 397–402.
45. Masson, L., Tabashnik, B. E., Liu, Y. B., Brousseau, R., and Schwartz, J. L.
(1999) J. Biol. Chem. 274, 31996–32000.
28
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
46. Gerber, D., and Shai, Y. (2000) J. Biol. Chem. 275, 23602-23607.
47. Rost, B., and Sander, C. (1993) Proc. Natl. Acad. Sci. U S A 90, 7558-7562.
48. Rost, B., Casadio, R., Fariselli, P., and Sander, C. (1995) Protein Sci. 4, 521-533.
29
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
Footnotes:
1ANTS, 8-aminonaphtalene-1,3,6-trisulfonic acid sodium salt; DOPG,
dioleoylphosphatidylglycerol; DPPG, dipalmitoylphosphatidylglycerol; DPX
p-xylenebis(pyridinium)bromide; FRET, fluorescence resonance energy
transfer; IAEDANS, 5-((((2-iodoacetyl)amino)ethyl)amino)naphthalene-1-
sulfonic acid; IANBD N,N’-dimethyl-N-(iodoacetyl)-N’-(7-nitrobenz-2-
oxa-1, 3-diazol-4-yl)ethylenediamine; LUV, large unilamellar vesicles; MBP,
maltose-binding protein; N-NBD-PE, N-(7-nitro-benz-2-oxa-1,3-diazol-
4-yl)phosphatidylethanolamine; PC, phosphatidylcholine; POPG, 1-palmitoyl-
2-oleoylphosphatidylglycerol; PI, phosphatidylinositol; PS, phosphatidylserine;
SAPG, 1-stearoyl-2-arachidonoyl-phosphatidylglycerol; TMD,
transmembrane domain.
30
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
Table 1: Effect of lipid composition on MBP-2B-induced LUV permeabilization
LUV composition Head group charge Hydrocarbon chaina % Leakageb
PCPIPSDPPGDOPGPOPGSAPG
+/-------
18:0-20:418:0-20:418:0-18:116:0-16:018:1-18:116:0-18:118:0-20:4
7100157
14 2575
a: Prevalent acyl chains according to the supplier specifications.
b: Percent changes in extents of ANTS/DPX leakage induced at a protein-to-lipid ratio of 1:8000 (mol:mol) are given, relative to pure PI. Mean values of two independent measurements are indicated.
31
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
Figure legends
Fig 1: Permeabilization of large unilamellar PI vesicles by 2B (ANTS/DPX assay). A:
Leakage of internal contents as a function of time. The curves were obtained at
different protein-to-lipid molar ratios as indicated. B: Extent of aqueous content
release (percentage of maximum leakage at time= 60 min) induced by increasing
protein-to-lipid molar ratios. In all cases the lipid concentration used was 50 µM.
Data points correspond to mean values of two independent determinations in a
representative experiment.
Fig. 2: Effect of mutations on the permeabilizing activity of 2B. A: Characteristic
kinetic traces of aqueous content release (ANTS/DPX assay) at a protein-to-lipid
ratio of 1:8000 (mol:mol). B: Final extents (black bars) and initial rates (white bars)
of leakage (protein-to-lipid molar ratio, 1:8000); percent changes of ANTS/DPX
leakage (mean values of two independent determinations in a representative
experiment) are given relative to wild-type MBP-2B. Final leakage extents were
measured at time = 60 min, while initial rates were obtained from the maximum
slopes of kinetic curves such as those in panel A. Conditions were otherwise as in Fig.
1.
Fig. 3: Stability and size of 2B pores formed in PI vesicles. A: 2B exchange between
vesicles. MBP-2B was pre-incubated with empty PI (filled circles) or PC (empty
32
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
circles) vesicles for different time periods at a protein-to-lipid molar ratio, 1:8000.
The leakage activity fraction remaining at the different incubation times (%L/%L0)
was subsequently determined by supplementing the mixtures with ANTS/DPX-
containing PI vesicles. B: Extent of aqueous content release from vesicles
(ANTS/DPX assay, filled circles), compared to the extent of entry of solutes into
them (dithionite assay, empty circles) at different protein-to-lipid molar ratios. For
the "entry" assay, vesicles were preincubated with protein for 1 hour, and then treated
with dithionite. Data points represent mean values of two independent determinations
in a representative experiment. C: Size dependence of solute retention in vesicles.
Vesicles were incubated with protein (protein-to-lipid molar ratio, 1:8000) for 30
minutes prior to determination of the released material. Calculated leakage extents are
mean values ± standard errors of 4 independent determinations. Solutes encapsulated
in vesicles: ANTS (MW 427)/DPX (MW 422), NADH (MW 660) and Evans Blue
(MW 960). In all cases, the lipid concentration was 50 µM.
Fig. 4: Binding to membranes and surface aggregation of 2B protein. A: Fluorescence
emission spectra of MBP-2B-NBD in buffer (dotted lines), and incubated with PI
LUV (continuous lines). Protein concentration was 0.2 µM and, in the presence of
vesicles, the protein-to-lipid molar ratio was 1:1000. B: Partitioning curve as
calculated from the fractional change in MBP-2B-NBD fluorescence upon titration
with PI vesicles (filled circles). The solid line corresponds to best fit of the
experimental values to the function in equation [3]. Empty circles correspond to a
titration experiment with PC vesicles. Data points correspond to mean values of two
independent determinations in a representative experiment. C: Fluorescence
33
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
resonance energy transfer of labeled MBP-2B in PI LUV and in solution. Continuous
lines: Emission spectra of an MBP-2B-ANS:MBP-2B (1:3 mol:mol) mixture.
Dotted lines: Emission spectra of an MBP-2B-ANS:MBP-2B-NBD (1:3 mol:mol)
mixture. The total protein concentration was 0.4 µM. In the vesicle samples, the
protein-to-lipid molar ratio was 1:1000.
Fig. 5: Oligomerization of 2B induced by its association with membranes. A: MBP-
2B in solution (sol.) or mixed with either PI or PC liposomes (protein-to-lipid molar
ratio, 1:1000) was incubated for 30 minutes and solubilized in SDS-PAGE sample
buffer. Proteins were subsequently resolved on 3-13 % polyacrylamide gradient gels.
In the first lane, the molecular weight markers were run as indicated. B: Oligomer
formation with increasing lipid concentrations. MBP-2B was mixed with increasing
quantities of PI vesicles and subjected to SDS-PAGE analysis as indicated in the
previous panel. C: Percentage of MBP-2B converted into oligomers with increasing
lipid concentration. The relative amount of converted protein was estimated from
band intensities in the previous experiment. The solid lines correspond to best fitting
to hyperbolic functions such as that in equation [3].
Fig. 6: Structural model of the poliovirus 2B pore. A: Amino acid sequence of
poliovirus 2B protein. White and light gray cylinders represent the amphipathic and
TMD regions, respectively. The bar diagram corresponds to the probability (scaled
for the range 0-9) of assigning a helical trans-membrane region (white bars) and an
interconnecting loop (black bars), as predicted by the PHD neural network (47, 48). A
detailed description of this procedure may also be found at
http://cubic.bioc.columbia.edu/predictprotein. B. Schematic and structural models for
34
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
2B transmembrane pores. The structure was rendered using the Swiss-PDBviewer
program. Side-chains correspond to positively charged and polar residues facing the
lumen of the aqueous pore. Side chain in the reverting loop indicates the aromatic Tyr
residue, which is embedded in the interfacial membrane region.
35
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from
Aitziber Agirre, Angel Barco, Luis Carrasco and Jose L. Nievapoliovirus 2B protein
Viroporin-mediated membrane permeabilizatin: pore formation by non-structural
published online August 14, 2002J. Biol. Chem.
10.1074/jbc.M205393200Access the most updated version of this article at doi:
Alerts:
When a correction for this article is posted•
When this article is cited•
to choose from all of JBC's e-mail alertsClick here
by guest on January 12, 2020http://w
ww
.jbc.org/D
ownloaded from