| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 277, Issue 22, 19773-19782, May 31, 2002
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Department of Biochemistry and Molecular Biology, Oklahoma
State University, Stillwater, Oklahoma 74078
Received for publication, October 18, 2001, and in revised form, February 21, 2002
The structure of apolipophorin III in the
lipid-bound state and the extent of the conformational change that
takes place when the five-helix bundle apolipoprotein binds to a
lipoprotein lipid surface were investigated by fluorescence resonance
energy transfer in discoidal lipoproteins. Four intramolecular
interhelical distances between helix pairs 1-4, 2-4, 3-4, and 5-4
were estimated by fluorescence resonance energy transfer in both the
lipid-free and the lipid-bound states. Depending on the helices pairs,
the intramolecular interhelical distances increased between 15 and Exchangeable apolipoproteins represent a group of proteins
characterized by their ability to reversibly bind to lipoprotein lipid
surfaces (1). These apolipoproteins play a key role in the regulation
of lipid metabolism (2). The interaction of lipoproteins with receptors
(3-4) and the exchange of lipids between lipoproteins and lipoproteins
and cells (5-7) are among the processes regulated by the structure of
apolipoproteins in both the lipid-free and the lipid-bound states. A
full understanding of the function of exchangeable apolipoproteins in
the homeostasis of lipid metabolism and associated pathologies requires
determining the structure of the protein in the lipid-bound state.
Apolipophorin-III is a 17-kDa exchangeable apolipoprotein that is
involved in the transport of lipids in the hemolymph of insects (8-9).
The structure of apoLp-III1
in the lipid-free state was solved by x-ray crystallography in 1991 (10). ApoLp-III is composed of a bundle of five amphipathic Solving the structure of apolipoproteins in the lipid-bound state
represents a major challenge due to the difficulty of using high
resolution techniques such as NMR or x-ray crystallography to study
lipoprotein complexes. On the other hand, the possibility of expressing
recombinant apolipoproteins such as apoA-I (13), apoE (14), and
apoLp-III (15-16) in bacteria or baculoviruses coupled to the use of
site-directed mutagenesis has opened the door to the use of other
spectroscopic techniques that can render useful structural information.
Although FRET does not have the resolution of x-ray diffraction or
solution NMR, it provides an alternative approach to obtain structural
constraints that can be used to build a low resolution structure of an
apolipoprotein in the lipid-bound state. This approach has been
recently used to study some structural properties of the organization
of human apoA-I molecules in discoidal lipoproteins (19-20). It is
expected that many more studies will apply FRET techniques to
investigate the structure of apolipoproteins in the lipid-bound state.
In this report we present the results of an extensive FRET study that
was directed to investigate structural features of apoLp-III in the
lipid-bound state. Four intramolecular and four intermolecular, interhelical distances were estimated in this study and used to infer
the conformation of individual apoLp-III molecules as well as the
arrangement of different apolipoprotein molecules in discoidal lipoproteins. Moreover, the comparison between the interhelical distances in the lipid-free and in the lipid-bound state demonstrated for the first time that the formation of discoidal lipoproteins is
accompanied by a large increase in the distance of separation between
the helices of the apolipoprotein.
Site-directed Mutagenesis--
Site-directed mutagenesis was
carried out using the commercial kit QuikChange, manufactured by
Stratagene. A pET-32a plasmid (Novagen, Madison, WI) containing the
insert of L. migratoria apoLp-III was used for mutagenesis.
The PCR reaction containing the plasmid and the primers was performed
according to the manufacturer's instructions using the Pfu
DNA polymerase. Competent cells, Epicurian (Stratagene), were
transformed with the PCR reaction. The plasmid DNA of the transformed
bacteria was sequenced from the S-tag and T7-terminator sites. AD494
expression host cells were transformed with the pET-32a-apoIII plasmids
containing the correct sequences. Wild type L. migratoria
apoLp-III contains two Trp residues and no Cys residues. The single Trp
mutant containing a reporter Trp in helix 4 at position 113 was
obtained with a single mutation (W128F) of the wild type apoLp-III.
This mutant was subsequently used as a template to construct four
single Cys mutants. The single Trp mutant has been previously
characterized and shown to have structural and functional properties
nearly identical to those of the wild type apoLp-III in both the
lipid-free (21) and lipid-bound states (22). The location of the Cys
residues was chosen on the basis of the crystal structure of L. migratoria apoLp-III (1aep.pdb). The following
substitutions were carried out: N16C (Cys in helix 1), A51C (Cys in
helix 2), N83C (Cys in helix 3) and A147C (Cys in helix 5); we will
refer to these mutants as Cys-H1, Cys-H2, Cys-H3, and Cys-H5 mutants,
respectively. According to the crystal structure the amino acid
residues replaced by Cys are exposed to the solvent. The location of
the Cys residues and that of the only Trp residue is shown in Fig.
1. A double mutant containing no Trp
residues, W128F/W113F, was also constructed, and we will refer to this
mutant as the Phe mutant. This mutant was needed to correct the
fluorescence spectra for the scattering generated by the lipoproteins
as well as to study possible intermolecular energy transfer
processes.
Protein Expression and Purification--
Expression of the
protein was induced by
isopropyl-1-thio- Protein Labeling--
Acrylodan (AcD),
didansyl-L-cystine (DDC), and
4-acetamide-4'-maleimidylstilbene-2,2-disulfonic acid (AMS) were
purchased from Molecular Probes. AcD was dissolved in methanol (10 mM), DDC was dissolved in dimethylformamide (20 mM), and AMS (20 mM) was dissolved in 0.4 M potassium phosphate buffer, pH 7.0. Before the labeling
reaction, the proteins (20-40 µM) were treated with 2 mM dithiothreitol followed by dialysis against the
appropriate volume of 5 mM potassium phosphate, pH 7.0, to
reduce dithiothreitol concentration to 50 µM. These
protein preparations were incubated with a 20-fold molar excess of
probe (AcD, DDC, or AMS) at room temperature for 12 h. The excess
of probe was removed by gel filtration chromatography on Sephadex G25
(Amersham Biosciences). In the cases of AcD and DDC, due to their low
water solubility part of the probe goes out of solution at this
concentration. These samples were centrifuged before the gel filtration
chromatography. The samples treated with AMS were directly applied to
the column. Unlabeled protein was separated by gel filtration
chromatography in Superdex 75, as explained below. Before the second
chromatography, the samples were dried by speed vacuum to promote the
oxidation and dimerization of the unlabeled protein. When these samples were separated by SDS-PAGE, two bands corresponding to monomer and
dimer were observed; only the band corresponding to the monomer was
fluorescent under the UV light, indicating that the labeling reaction
was specific. The specificity of the labeling was also confirmed using
as negative control wild type apoLp-III, which contains no Cys
residues. Under identical labeling conditions the wild type protein was
not labeled. The samples were dissolved in 1.5 M
guanidinium hydrochloride buffered with 25 mM potassium phosphate, and the dimers were separated from the monomers by fast
protein liquid chromatography in a Superdex 75 HR10/30 column using the
same buffer. The elution was monitored by absorbance at 280 nm as well
as at the absorption maximum of the corresponding probe. The fractions
containing the monomer were collected and dialyzed against 5 mM potassium phosphate, pH 7.0.
The labeling efficiency (fraction of labeled protein) was determined
from the concentration of protein, as determined by reaction with
bicinchoninic acid (BCA kit, Pierce) and from the UV absorption spectra
of the labeled protein using the absorption coefficients (23) of the
corresponding probes (AcD, Preparation of DMPC Discoidal Lipoprotein
Complexes--
Discoidal lipoproteins were prepared by the method
reported by Jonas et al. (35). Proteins were mixed
with DMPC and sodium cholate in a 1/260/715 molar ratio and then
dialyzed exhaustively against 5 mM potassium phosphate, pH
7.0. When indicated, the corresponding proteins were diluted with the
Phe mutant. Analysis of the resulting lipoprotein complexes by gradient
gel electrophoresis (24) indicated one predominant population of
complexes with an apparent diameter of 20.7 nm (Fig.
2).
Fluorescence Spectroscopy--
Steady-state emission and
excitation fluorescence spectra were recorded with an ISS K2
fluorescence spectrophotometer using 0.5-mm slits for excitation and
emission (8-nm bandwidth). The light output of the lamp was recorded
simultaneously with the emission spectrum by recording the intensity in
a reference photomultiplier. Trp emission was recorded with excitation
at both 280 and 285 nm and collected between 295 and 600 nm every 1 nm.
Emission spectra of the acceptor-labeled proteins was recorded with
excitation at both 280 nm and at the wavelength of the acceptor
absorption maximum 340, 380, and 325 nm for DDC, AcD, and AMS, respectively.
All spectra were corrected for the Raman peak of the solvent. The
emission spectra of the lipoprotein samples were also corrected for the
contribution of scattering using as the control lipoproteins prepared
with the mutant containing no Trp residues. All measurements were
carried out at 28 °C in 5 mM potassium phosphate buffer, pH 7.0. The experiments were performed in a 0.4-cm path length cell,
and the absorbance of the samples was kept below 0.025. Donor quantum
yields in the absence of acceptor (QD) were obtained from
the fluorescence of unlabeled mutants using as reference a solution of
tryptophan in water, Qd = 0.13, (25). The quantum yield of
the labeled samples were obtained in the same fashion but were also
corrected to account for the labeling efficiency.
Calculation of Fluorescence Energy Transfer Efficiency, D-A
Förster Distance, and Average Distances of D-A
Separation--
The Förster distances (Ro in
Å) were calculated using the following equation.
The efficiency of energy transfer was calculated from the quantum
yields of Trp fluorescence in the absence (QD) and in the presence (QDA) of acceptor according to Equation 2 below.
D-A distances were estimated from the Ro values, and
the efficiency of FRET was estimated from Equation 3 below.
Circular Dichroism--
CD spectra were acquired at 28 °C in
a Jasco CD spectrophotometer Model 715. The CD spectra of the same
samples used in fluorescence experiments were determined in a 1-cm path
length cell. Spectra were corrected using the buffer as blank and smoothed.
Protein Determinations--
Protein concentrations of the
samples containing lipoproteins or lipid-free proteins were determined
by the bicinchoninic acid method using the BCA kit (Pierce). The
protein concentration of unlabeled and lipid-free apoLp-III samples was
also determined by UV absorption spectroscopy in the presence of 2 M guanidinium hydrochloride using an extinction coefficient
of 5,790 M The structure of apoLp-III in the lipoprotein lipid-bound
state as well as the nature and magnitude of the conformational change
that takes place when lipid-free apoLp-III binds to a lipid surface
were investigated by FRET. Four single-Cys mutants were constructed
using as basis the crystal structure of the lipid-free protein. All
these mutants contain a single Trp residue (Trp-113), which is located
in the nonpolar face and near the middle of helix 4. The Cys residues
were located in the polar face of the helices 1 (Cys-H1 mutant), 2 (Cys-H2), 3 (Cys-H3), and 5 (Cys-H5). Fig. 1 shows the location of the
Trp residue as well as the location of the Cys residues engineered in
helices 1, 2, 3, and 5. The intermolecular distances between the
Trp-113 and probes attached to the helices 1, 2, 3, or 4 were
determined in both the lipid-free and the lipid-bound states. Although
the distances in the lipid-free state are known from the
crystallographic data, the estimation of these distances by FRET and
the use of three different probes provided a mechanism to evaluate the
reliability of the data and detect systematic errors in the estimates
of the intermolecular distances.
Characterization of the Single Cys mutants; Effect of the
Mutations and Cys Labeling in the Structure of the Proteins
A potential influence of the mutations and/or the probes attached
to the Cys residues on the secondary structure of the proteins was
studied by far-UV CD spectroscopy (Fig.
3). All the mutants but one were highly
All the mutants used were also characterized by their ability to
interact with lipids. With no exceptions, all the proteins were active
and able to spontaneously form discoidal lipoprotein particles when
incubated with DMPC (data not shown).
Spectral Properties of Donor and Acceptor Groups and Spectral
Overlaps
Fig. 4 shows the absorption spectra
of the acceptor probes, AcD, DDC, and AMS, covalently bound to the
Cys-H5 mutant. For the sake of clarity and because no major differences
were observed between the absorption spectra of the probes in the
lipid-free and lipid-bound state of the proteins, only one absorption
spectrum per probe is shown in Fig. 4.
Structure of Apolipophorin-III in Discoidal Lipoproteins
INTERHELICAL DISTANCES IN THE LIPID-BOUND STATE AND
CONFORMATIONAL CHANGE UPON BINDING TO LIPID*
ABSTRACT
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
20 Å upon binding of the apolipoprotein to lipid, demonstrating for
the first time that binding to lipid is accompanied by a major change
in interhelical distances. Using discoidal lipoproteins made with a
combination of apolipophorin III molecules containing donor and
acceptor groups and apolipophorin III molecules containing neither
donor nor acceptor groups, it was possible to obtain information about
intermolecular interhelical distances between the helix 4 of one
apolipoprotein and the helices 1, 2, 3, and 5 of a second
apolipoprotein residing in the same discoidal lipoprotein. Altogether,
the estimated intermolecular and intramolecular interhelical distances
suggest a model in which the apolipoprotein arranges in pairs of
antiparallel and fully extended polypeptide chains surrounding the
periphery of the bilayer disc.
INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
-helices
connected by short loops (10). The study of the structure-function relationship of Locusta migratoria apoLp-III is of interest
because the knowledge of the crystal structure of the full-length
apoLp-III molecule offers the possibility of relating the structure of
an apolipoprotein to its function. Because apoLp-III shares many physical-chemical properties with exchangeable apolipoproteins of
humans and other vertebrates (1, 11-12), it is likely that elucidation
of the mechanism of interaction of apoLp-III with lipids would be
useful to the understanding of the function of human apolipoproteins.
EXPERIMENTAL PROCEDURES
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
View larger version (63K):
[in a new window]
Fig. 1.
Location of the energy donor and acceptors in
the apoLp-III molecule. The location of the Cys residues
engineered to construct the mutants Cys-H1 (Cys-16), Cys-H2 (Cys-51),
Cys-H3 (Cys-83), and Cys-H5 (Cys-147) is indicated in the figure
together with the position of the single Trp residue (energy donor),
which is common to all the mutants and located in helix 4.
-D-galactopyranoside, and the protein
isolated from the bacteria and purified as previously reported (21).
Briefly, the fusion protein was purified by nickel affinity
chromatography. Recombinant locust apoLp-III was cleaved from the rest
of the fusion protein with enterokinase (Novagen). ApoLp-III does not
have any cleavage sites for enterokinase. ApoLp-III was purified from
the cleavage reaction by nickel affinity chromatography and ion
exchange chromatography in DE52.
386 = 19,200 M
1 cm1; AMS, 328 = 31,500 M1 cm1; DDC,
334 = 8,300 M1
cm1). UV-visible absorption spectra were
recorded with a HP 8453 diode array spectrophotometer. The labeling
efficiencies obtained with each of Cys mutants and probes are indicated
in Table I. The fraction of labeled protein obtained for a given probe
was compared with the intensity of the emission spectra of the protein, obtained by direct excitation of the probe. This comparison showed a
very good correlation of the data of the efficiency of labeling with
the intensities of the emission spectra of the probes.
View larger version (20K):
[in a new window]
Fig. 2.
Non-denaturing PAGE of discoidal
lipoproteins. Left panel, Coomassie Blue-stained 4-20%
acrylamide gradient gel of discoidal lipoproteins containing 260 molecules of phospholipid per molecule of apoLp-III. Lane 1 shows the high molecular weight markers (thyroglobulin, ferritin,
catalase, and lactate dehydrogenase). The diameters of the standards
are indicated (in nm) on the left side of the picture.
Lanes 2-6 show the discoidal lipoproteins made with DMPC
and the wild type apoLp-III, Cys-H1, Cys-H2, Cys-H3, and Cys-H5
mutants, respectively. The lipoprotein samples contained 0.5 mM dithiothreitol to prevent the formation of
inter-discoidal disulfide bonds. Right panel, the
calibration curve was constructed as indicated under "Experimental
Procedures."
The spectral overlap integral, J (in units of
M
(Eq. 1)
1
cm1nm4), was calculated from the
absorption spectra of the acceptors and the normalized fluorescence
spectra of the donor in the absence of acceptor. A 
2
value of 0.67 was assumed for all the samples and conditions.
(Eq. 2)
(Eq. 3)
1
cm1.
RESULTS
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
-helical in both the lipid-free and the lipid-bound states (Table
I). The only exception was the Cys-H5
mutant labeled with AMS that showed a lower -helical content in the
folded lipid-free state. However, the CD spectrum of the AMS-labeled
Cys-H5 mutant in the lipid-bound state was indistinguishable to the
spectra of other mutants. Therefore, it was considered that the
AMS-labeled Cys-H5 mutant could still provide useful information on the
interhelical distances in the lipid-bound state, and consequently, it
was included in the studies.
View larger version (30K):
[in a new window]
Fig. 3.
Secondary structure of Cys-H1 and Cys-H5 in
the lipid free and lipid-bound states. Far-UV CD spectra of the
unlabeled and acceptor-labeled Cys-H1 mutant in the lipid-free state
(panel A) and in discoidal lipoproteins of DMPC (panel
B); the symbols corresponding to different acceptor
probes are included in the figure. Panels C and D
show the CD spectra obtained with the mutant Cys-H5 in the lipid-free
and lipid-bound states, respectively. The samples were dissolved in 5 mM potassium phosphate buffer, pH 7.4, and the CD spectra
were acquired at 28 °C. MRE, mean residue ellipticity.
Spectral overlap integral J(
) and Förster distances
(Ro) for the energy transfer from tryptophan in helix 4 (Trp-113) to different acceptors in helices 1, 2, 3, and 5
View larger version (24K):
[in a new window]
Fig. 4.
Overlaps between the emission spectra of Trp
and the absorption spectra of AcD, DDC, and AMS. Panel A
shows the overlaps between the absorption spectrum of acrylodan and the
emission spectra of the Trp in Cys-H5 mutant in the folded state
(Pi, 5 mM phosphate buffer, pH 7) and
lipid-bound state (discoidal lipoproteins in 5 mM phosphate
buffer). Panels B and C show the overlaps for AMS
and DDC, respectively. The corresponding symbols are shown in the
figure.
To illustrate the extent of the donor-acceptor spectral overlap in different conformational states, Fig. 4 includes the normalized tryptophan fluorescence emission spectra of Cys-H5 in the lipid-free and lipid-bound states. Binding to lipid promotes a small red shift of the emission of Trp-113, from 317 to 323 nm. A similar red shift was observed and previously reported for the single Trp mutant containing no Cys (17). The overlap-integral values (J) and Förster distances (Ro) for each donor/acceptor pair in different conformational states are given in Table I. In the lipid-free and lipid-bound proteins AMS has the largest Ro (mean Ro ~ 31 Å), followed by AcD (mean Ro ~ 26 Å) and DDC (mean Ro ~ 22 Å). Minor differences are also found between the Ro values estimated for different mutants when the comparison is made between proteins in equivalent conformational states and labeled with the same probe (Table I).
Efficiency of FRET and D-A Distances from Trp-113 (helix 4) to AcD, DDC, and AMS acceptors bound to Cys-16 (helix 1) or Cys-147 (helix 5)
Folded Lipid-free State--
FRET efficiencies were calculated
from the intensities of Trp emission with excitation at 280 nm as
indicated under "Experimental Procedures." Fig.
5, top panel, shows the
corrected fluorescence emission spectra of the folded and unlabeled
lipid-free Cys-H1 mutant together with the spectra obtained with the
folded lipid-free protein labeled with AcD, DDC, or AMS. The equivalent
spectral data obtained with the Cys-H5 mutant is shown in Fig.
6, top panel. The FRET
efficiencies observed in the folded lipid-free state for all the
mutants are shown in Table II. As
expected from the Ro values of the probes and the
crystallographic distances, a high level of quenching of the Trp
fluorescence by the three acceptor probes is observed for all Cys
mutants. The donor/acceptor distances, which were calculated using
these efficiency values and the individual Ro values
for each D-A pair, are also shown in Table II. These distances are
larger than the crystallographic distances of separation between the
Trp residue (Trp-113) and the Cys residues (Table II). However, the
difference between the estimated distance and the crystallographic
distance could be simply attributed to the size and orientation of the
probes. The distances of separation between Trp-113 and the Cys
residues estimated with DDC are significantly shorter than those
estimated with AcD or AMS (Table II). Even though the DDC-derived
distances correspond very well to the crystallographic distances
between the Trp residue and the Cys residues, they are likely to be
underestimated. This is so because of the size of the probes, where
longer separation distances are expected between D-A pairs than between
the corresponding Trp-Cys pairs. The shorter D-A distances estimated
with DDC could be due to the presence of a disulfide group in the
proteins labeled with this probe. This disulfide bond, which links the
probe to the Cys residue, could quench the Trp fluorescence by an extra mechanism (26), leading to an overestimation of the actual FRET to the
dansyl group. Overall and considering the potential effect of the size
of the probes and the orientation factors, the three probes provided
reasonable D-A distance estimates in the folded state of the
protein.
|
|
|
Lipid-bound State-- FRET between the Trp residue and the acceptors located in helices 1, 2, 3, and 5 was studied in discoidal lipoprotein particles prepared by the method of cholate dialysis using a 1:260 protein to DMPC molar ratio. The lipoprotein particles obtained under these conditions are nearly homogeneous (Fig. 2) and have an apparent diameter of 20.7 nm, as determined by gradient gel electrophoresis under non-denaturing conditions.
The spectra corresponding to lipoproteins made with unlabeled and AcD-,
AMS-, and DDC-labeled Cys-H1 and Cys-H5 mutants are shown in the
middle panels of Figs. 5 and 6, respectively. The efficiencies of energy transfer calculated from the spectra (including those obtained with the Cys-H2 and Cys-H3 mutants) are shown in Table
II. For all the mutants and with any of the three probes, the FRET
efficiencies observed are much lower than the efficiencies obtained in
the folded lipid-free state. The D-A distances, uncorrected for
intermolecular energy transfer (see below), calculated for all
the sets of protein and probes are also shown in Table II. A
significant increase of the donor/acceptor distance is observed in the
lipid-bound state with respect to the folded state for all the mutants.
These results indicate that the distances of separation between the
helix 4 and the helices 1, 2, 3, and 5 increase significantly in the
lipid-bound state. However, due to the small size of the discoidal
lipoproteins and the presence of five or six molecules of
apolipoprotein per lipoprotein particle (27-28), a potential
intermolecular energy transfer between Trp and probes located in
different apoLp-III molecules was considered and investigated. This
type of energy transfer, which could occur between donor (Trp) and
acceptors located in different apolipoprotein molecules, could
potentially lead to an underestimation of the distances of separation
between donor and acceptor. To estimate the importance of this
mechanism of energy transfer and remove its impact in the estimate of
distances, we prepared discoidal lipoproteins in which half of the
acceptor-labeled Cys mutant was replaced by the Phe mutant
(W128F/W113F), which cannot act as donor or acceptor. The
spectra obtained when the lipoproteins were prepared with 50% labeled
apoLp-III and 50% of the Phe mutant are shown in the bottom
panels of Figs. 5 and 6 for the mutants Cys-H1 and Cys-H5,
respectively. The corresponding efficiencies of FRET obtained with all
four Cys mutants are shown in Table II. Taking the average of the data
obtained with the three probes, a significant additional D-A distance
increase is estimated in lipoproteins made with the Cys-H1 (
= 7 Å) and Cys-H3 ( = 5 Å) mutants diluted with the Phe mutant.
The values are the average difference (three probes) between the
distance estimated in discoidal lipoproteins containing 50% of the Phe
mutant (distance less affected by intermolecular FRET) and the distance
estimated when the lipoproteins were prepared with 100% of the Cys
mutant (distance affected by intermolecular FRET). These observations
indicate that in discs of undiluted Cys-H1 and Cys-H3 mutants a
considerable intermolecular FRET occurs, suggesting a relatively short
distance of separation between helices 1 and 4 and 3 and 4 of different
apoLp-III molecules. This effect was not observed in discs made with
the Cys-H2 and Cys-H5 mutants.
Overall these results indicate that compared with the lipid-free state the intramolecular distance between the Trp in helix 4 and the Cys residues located in helices 1, 2, and 5 increases by at least 20 Å in the lipid-bound state. A significant interhelical distance increase is also observed between helix 3 and helix 4 (~14 Å). These results also indicate that in discoidal lipoproteins there is a relatively short distance between the residue 113 of an apoLp-III molecule and the residues 16 and 83 of a second neighboring apoLp-III molecule.
It must be noted that, assuming a FRET efficiency of 10%, which is of
the order of the experimental error, the maximum distance that could be
estimated with the probes used in this study would range between 37 and
46 Å. This is so because the Ro values of the
protein-probe pairs used in this study range between 26 and 32 Å.
Because the FRET efficiency values obtained in the lipid-bound state
are very small and often negative (Table II), the estimates of the D-A
distances in the lipid-bound state could be highly underestimated
(i.e. an efficiency of 0% would indicate an infinite D-A distance).
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
|
|---|
The main goals of this study were obtaining information about the structure of apoLp-III in the lipid-bound state and inferring the nature of the conformational change required to achieve the final lipid-bound state found in discoidal lipoproteins. FRET between probes attached to Cys residues and a single Trp residue located in helix 4 allowed us to determine the intramolecular distances separating the centers of helices 1 (N-terminal helix), helix 2, helix 3, and helix 5 (C-terminal helix) from the center of the helix 4. The experimental data obtained with lipoproteins prepared with a mixture of two types apoLp-III molecules, one containing donor and acceptors and the other containing neither energy donor nor energy acceptor, also allowed us to infer intermolecular interhelical distances. Altogether, the intermolecular and intramolecular distances were used to build a model including the organization of different apoLp-III molecules in a discoidal lipoprotein.
Quality of Samples and Data-- Cys-directed labeling is a broadly used technique to study the properties of protein domains. Even though labeling of solvent-exposed Cys residues is one of the less perturbing methods for the incorporation of spectroscopic probes in proteins, the main limitation of this approach resides in the potentially perturbing effect of the mutations required to introduce the Cys residues as well as on the potential effect of the probes on the structure and properties of the protein. Because of this potential problem, the secondary structure of the Cys mutants used in this study was carefully studied by circular dichroism and the position of the fluorescence maximum. Among the 32 proteins involved and investigated in our study, 16 proteins in the lipid-free state, and the same number in the lipid-bound state, only the AMS-labeled Cys-H5 mutant showed an altered structure in the lipid-free state. Moreover, confirming the lack of major structural perturbations, the position of the fluorescence spectra of the Trp residue, which provides a highly sensitive parameter to determine changes in the tertiary structure of the protein, also indicated that the mutations and/or the probes did not affect the structure of the proteins in the lipid-bound state. In all the cases the position of maximum was at or near 323 nm.
FRET is a very useful and widely used technique, but its application to
determine distances has some limitations. The major limitation of FRET
is associated to the uncertainty in the value of the orientation
factor, 2 (29-31), which is needed to estimate the
distance between emission and absorption dipoles of donor and acceptor
groups (Equation 2). Because the experimental estimation of the
orientation factor is difficult, in most studies it is assumed that the
donor and acceptor transition dipoles are characterized by a random
distribution of orientations, and a 2 value of 0.67 is
adopted (29, 32). To evaluate the potential impact of the value in
the distance estimates, we performed the FRET study using three
different probes. Because different probes are unlikely to have the
same average orientation of the transition dipole moments, the
comparison of the results obtained with different probes was used as a
means to detect inconsistent results. For any given type of sample, the
comparison of the D-A distance estimates obtained with AcD and AMS
showed high similarities, whereas the distances estimated with DDC in
the native lipid-free state were shorter than those obtained with AcD
and AMS. On the other hand, the comparison of the changes in
interhelical distances (D-A distance in lipid-bound state discs
(1/2) minus the distance in folded lipid-free state) shows that
the estimates obtained with all the probes are highly similar and of
the order of 20 Å for the distances between the helices pairs 1-4,
2-4, and 5-4 and slightly shorter, 14 Å, for the distance between
helices 3 and 4 (Table II). The consistency of the data suggests that
the results obtained were not significantly influenced by the nature of
the probe. Overall, the comparison of the data indicates that the
differences between the distance estimates obtained with different probes are reasonable, and within the limitations of FRET, the results
and trends observed are reliable. It must be noted, however, that
because the efficiencies of FRET are very small in discoidal lipoproteins, the distances determined in the lipid-bound state could
be underestimated. The Ro values of the
protein-probe pairs used in this study range between 22 and 32 Å.
Thus, assuming a FRET efficiency of 10%, which is of the order of the
experimental error, the maximum distance that could be estimated with
the probes used in this study would range between 32 and 46 Å. Because
an efficiency of 0% would indicate an infinite D-A distance, and the
FRET efficiency values obtained in the lipid-bound state are very
small, the interhelical distances measured in discoidal lipoproteins could be underestimated. It must be noted that a FRET efficiency of
20% would only represent a change in the distance estimate that,
depending on the Ro value, would be between of 4 and
6 Å.
Structure of ApoLp-III in the Lipid-bound State-- The study of the structural organization of apolipoproteins in discoidal lipoproteins can be arbitrarily separated into four areas, apolipoprotein topology, identification of the lipid interacting regions of the apolipoprotein, conformation of the apolipoprotein in the lipid-bound state, and organization of different apolipoprotein molecules in the lipoprotein particle. The present study concerns the last two areas mentioned.
The topology of apolipophorin-III in discoidal lipoprotein has been
defined in two recent studies. The study of apoLp-III·DMPC complexes by Fourier transform-attenuated total reflection infrared spectroscopy provided experimental support to an arrangement of the
-helices predominantly perpendicular to the phospholipid acyl chains
(33). Additional experimental evidence about the topology of apoLp-III
in discoidal lipoproteins was recently presented in a
fluorescence-quenching study. The quenching pattern of single-Trp mutants obtained with phospholipids containing quenchers at different depths was shown to be consistent with a location of apoLp-III around
the periphery of the disc bilayer (18).
Two recent studies (17-18) of single-Trp mutants of apoLp-III have
provided evidence about the helical regions of the apolipoproteins that
interact with the lipid surface. The first of these studies, which used
water-soluble quenchers, suggested that at least three of the five
-helices (helices 1, 4, and 5) interact with the lipid surface (17).
The second study, which used lipid quenchers, suggested that in
discoidal lipoproteins all -helices of apoLp-III interact with the
phospholipid acyl chains (18).
Conformation of ApoLp-III in the Lipid-bound State Inferred from
the Interhelical Distances H1-H4, H2-H4, H3-H4, and H5-H4--
The
distance between the centers of helices 5 and 4 (
40 Å) estimated
in this study is consistent with a collinear arrangement of these two
helices in the lipid-bound state. If we assume an average contribution
to the helix length of 1.5 Å per residue and a fully extended
arrangement of helices 4 and 5, the distance of separation between the
Trp 113 and the Cys-147 should be of 55 Å. Therefore, the distance
estimate obtained in this study ( 40 Å) suggests a nearly aligned
arrangement of these two helices.
The average distance of separation between the centers of helices 3 (Cys-83) and 4 (Trp-113) was estimated to be 35.8 Å (calculated from the data obtained with lipoproteins containing the Phe mutant). This distance is slightly shorter than the 45 Å that would separate the Cys and Trp residues assuming a collinear arrangement of the helices 3 and 4 and 1.5 Å per amino acid residue. However shorter, this distance also suggests a nearly collinear arrangement of helices 3 and 4. We have recently observed that a disulfide mutant linking helices 3 and 4 is unable to spontaneously form discoidal lipoproteins when it is incubated with liposomes of DMPC. This observation would support further, although indirectly, the fact that helices 3 and 4 spread apart upon binding to a lipid surface.2
Because the Ro values of the D-A pairs used limit
the maximum distance estimates to 40-48 Å, the arrangement of helices 1 and 2 relative to helix 4 cannot be fully inferred from the distances
estimated in this study. This limitation arises from the fact that the
linear distances between the Trp-113 and the Cys residues at positions
16 and 51 are very long, ~145 and 93 Å, respectively. Despite this
limitation, the separation distance between the center of helices 1 and
4, which is 40 Å, still indicates a major rearrangement of these
two helices when the protein binds to a lipid surface. Similarly, the
estimated separation distance between the helices 2 and 4 in the
lipid-bound state is also greater than 40 Å (AcD) or 48 Å (AMS),
indicating a major rearrangement of the helices 2 and 4 upon binding to
lipid. Because the estimated distances for the helix1-helix 4 and helix
2-helix 4 pairs are very close to the maximum distances that could be determined by FRET, the results obtained do not discard that longer separation distances between the center of helix 4 and the centers of
helices 1 and 2 could actually occur. Thus, a fully extended polypeptide chain would produce essentially the same results that were
observed in this study. In fact, on the basis of the results of this
study as well as on the conclusions of recent studies reporting on the
topology and lipid-interacting regions of the apolipoprotein (17, 18),
a structural model of apoLp-III in discoidal lipoproteins is proposed
below. In this model it is assumed that the apolipoprotein adopts a
fully extended helical conformation. The extended conformation is
directly supported by the long distances separating the centers of the
pairs of helices 3-4 and 4-5 as well as by the fact that the
estimates for the separation distance between the helix pairs 1-4 and
2-4 are at the limit of the maximum distances that can be estimated by
FRET, suggesting than even longer distances may be possible. The
extended conformation is also compatible with recent studies indicating that in discoidal lipoproteins all helices of apoLp-III appear to
interact with the phospholipid acyl chains (18). Moreover, the extended
conformation is consistent with a previous report showing that a
disulfide mutant tethering the helices 1 and 5 (22) is unable to
interact with liposomes of phosphatidylcholine, leading to the
spontaneous formation of discoidal lipoproteins. The new structural
model is described below after discussing the arrangement of different
apolipoprotein molecules in the lipoprotein particle.
Spatial Arrangement of Different Apolipoproteins Molecules in Each Lipoprotein Particle-- Information about the organization of different apoLp-III molecules can be derived from the effect on the efficiency of energy transfer of the dilution of the Cys mutants with the Phe mutant. The fact that the efficiency of FRET between the acceptor probes in helix 1 and the Trp in helix 4 was significantly reduced when the lipoproteins were prepared with a 1:1 mixture of the Cys-H1 mutant and the Phe mutant suggests that in discoidal lipoproteins the helix 1 of one apoLp-III molecule resides close to the helix 4 of a second apoLp-III molecule. A similar situation was observed in lipoproteins made with the Cys-H3 mutant. The efficiencies of FRET decreased with all three acceptors when the lipoproteins included the Phe mutant. Thus, this result will also indicate a close distance of approach between the helix 3 of one apoLp-III molecule and the Trp in helix 4 of a second apoLp-III molecule present in the disc. Contrary to the case of the helices 1 and 3, the results of FRET obtained with the Cys-H2 and Cys-5 mutants do not show differences, suggesting that in discoidal lipoproteins the helices 2 and 5 of one apoLp-III molecule reside at least 40 Å apart from the helix 4 of a second apoLp-III molecule.
To explain these results, we searched for an arrangement of apolipoproteins that could be consistent with a close distance of separation between the helices 2 and 3 of different apolipoproteins as well as between helices 1 and 4 of different apolipoprotein molecules. We were unable to find such kind of apolipoprotein arrangement without compromising the data obtained for the conformation of individual apolipoproteins. Because of this it is hypothesized that there are two potential arrangements of apoLp-III molecules. In one of these arrangements the helices 1 and 4 of different apoLp-III molecules would reside at a distance shorter than 40 Å. In the other potential arrangement the helices 2 and 3 of different apoLp-III molecules would reside at a distance shorter than 40 Å. The model that is consistent with this data is described below. Because the discoidal lipoprotein contains between five and six molecules of apoLp-III (27-28), these two arrangements could be present in each individual lipoprotein.
Organization of Apolipophorin-III in Discoidal Lipoproteins; a
Structural Model--
The results obtained in this study allowed us to
propose a model of the discoidal lipoprotein including the conformation
of individual apoLp-III molecules as well as the organization of different apoLp-III molecules in the discoidal particle. To build this
model we have also considered previous studies that addressed the
topology of apoLp-III in discoidal lipoproteins as well as those
studies that investigated the lipid-interacting domains of the
apoLp-III molecule. The model proposed is represented in Fig.
7. The main features of the model are 1)
each apolipoprotein molecule adopts a fully extended conformation; this
extended conformation would support a belt-like model for the
arrangement of the helices of apoLp-III in discoidal lipoproteins, such
as that proposed for apoA-I (34); 2) the thickness of the disc bilayer
can only accommodate two -helices, which are normal to the
phospholipid acyl chains; 3) the apolipoproteins arrange in pairs of
antiparallel extended polypeptide chains; 4) two types of pairs are
possible and are constructed by the linear displacement of one of the
apolipoprotein toward either the N terminus or the C-terminus of the
second apolipoprotein molecule such that the helix 1 or the helix 5 of
both apolipoproteins will constitute hanging ends.
|
There are two studies on the composition of discoidal lipoprotein of apoLp-III. One of the studies concluded that the lipoproteins contain six molecules of apoLp-III (27), whereas the second report indicated the presence of five apolipoprotein molecules per discoidal particle (28). Our model would not be fully correct if the discoidal lipoproteins contain five molecules of apoLp-III. However, it would be fully consistent with a lipoprotein containing six molecules of apoLp-III.
Conformational Change Upon Binding to Lipid--
When exchangeable
apolipoproteins bind to a lipid surface it is common to observe changes
in the spectroscopic properties of the aromatic residues of the
proteins. These changes have been interpreted as due to the
conformational change that takes place when the protein binds to a
lipid surface. It must be noted, however, that in none of the previous
studies has there been a direct estimation of intramolecular distances.
The comparison of the distances estimated in this study for the
lipid-bound and lipid-free state of apoLp-III shows that upon binding
to lipid there is a major conformational change in which the distance
of separation between the helix 4 and the centers of the remaining
helices increases. In this sense, this study represents the first
quantitative estimation of the magnitude of the conformational change
that takes place when an apolipoprotein binds to a lipid surface. On
the other hand, the magnitude of the separation between the helices of
apoLp-III that takes place upon binding to lipid ( 20 Å) may
represent the largest conformational change ever recorded. Consistent
with these results we have observed that two disulfide mutants
tethering either the helices 1 and 5 (22) or the helices 3 and
42 are unable to promote the spontaneous formation of
discoidal lipoproteins when incubated with liposomes of DMPC.
Concluding Remarks-- Elucidation of the lipid-bound structure of exchangeable apolipoproteins represents a major challenge that must be faced with biophysical techniques of lower resolution than x-ray crystallography or NMR. Despite this limitation, the possibility of constructing mutants that can be specifically labeled allows the design and performance of FRET studies, which will certainly provide enough structural information to build low resolution models of lipoproteins.
The present study represents the first attempt to investigate the
arrangement of the helices in the lipid-bound state. We have estimated
four intramolecular and four intermolecular interhelical distances in
the lipid-bound state. These distances provided important structural
constrains for the conformation and arrangement of apolipoprotein
molecules in discoidal lipoproteins leading to a model reasonably
supported by experimental data. Given the limitation in the magnitude
of the separation distances that can be estimated by FRET, additional
studies will be needed to test the validity of certain aspects of the
proposed model. For instance, it will be valuable in determining
interhelical distances relative to a helix other than helix 4. These
types of studies are in progress in our laboratory and will help to
elucidate the relative arrangement of the helices 1, 2, and 3.
| |
FOOTNOTES |
|---|
To whom correspondence should be addressed: Dept. of Biochemistry
and Molecular Biology, 355 Noble Research Center, Oklahoma State
University, Stillwater, OK 74078. Tel.: 405-744-6212; Fax: 405-744-7799; E-mail: jose@biochem.okstate.edu.
Published, JBC Papers in Press, March 14, 2002, DOI 10.1074/jbc.M110089200
This research was supported by an Atorvastatin Research Award (to J. L. S.), National Institutes of Health Grant GM 55622, and the Oklahoma State University Agricultural Experiment Station.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
2 P. S. Chetty and J. L. Soulages, unpublished results.
| |
ABBREVIATIONS |
|---|
The abbreviations used are: apoLp-III, apolipophorin-III; AcD, acrylodan, 6-acryloyl-2-dimethylaminonaphtalene; AMS, 4-acetamido-4'-maleimidylstilbene-2-2'-disulfonic acid, disodium salt; DDC, N,N'-didansyl-L-cystine; DMPC, dimyristoylphosphatidylcholine; Cys-H1, mutant (W128F/N16C) containing a Cys residue in helix 1; Cys-H2, mutant (W128F/A51C) containing a Cys residue in helix 2; Cys-H3, mutant (W128F/N83C) containing a Cys residue in helix 3; Cys-H5, mutant (W128F/A147C) containing a Cys residue in helix 5; Phe mutant, apoLp-III mutant (W113F/W128F) that contains no Trp residues; D-A, donor-acceptor; FRET, fluorescence resonance energy transfer.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
|
|---|
| 1. | Segrest, J. P., Jones, M. K., de Loof, H., Brouillette, C. G., Venkatachalapathi, Y. V., and Anantharamaiah, G. M. (1992) J. Lipid Res. 33, 141-166[Abstract] |
| 2. | Fielding, P. E., and Fielding, C. J. (1996) in Biochemistry of Lipids, Lipoproteins, and Membranes (Vance, D. E. , and Vance, J. E., eds) , pp. 495-516, Elsevier Science Publishers B. V., Amsterdam |
| 3. | Weisgraber, K. H. (1994) Adv. Protein Chem. 45, 249-302[Medline] [Order article via Infotrieve] |
| 4. | Xu, S., Lacctripe, M., Huang, X., Rigotti, A., Zannis, V. I., and Krieger, M. (1997) J. Lipid Res. 38, 1289-1298[Abstract] |
| 5. |
Wang, N.,
Silver, D. L.,
Costet, P.,
and Tall, A. R.
(2000)
J. Biol. Chem.
275,
33053-33058 |
| 6. | Rothblat, G. H., Mahlberg, F. H., Johnson, W. J., and Phillips, M. C. (1992) J. Lipid Res. 33, 1091-1097[Abstract] |
| 7. | Jonas, A. (2000) Biochim. Biophys. Acta 1529, 245-256[Medline] [Order article via Infotrieve] |
| 8. | Ryan, R. O., and van der Horst, D. J. (2000) Annu. Rev. Entomol. 45, 233-260[CrossRef][Medline] [Order article via Infotrieve] |
| 9. | Soulages, J. L., and Wells, M. A. (1994) Adv. Protein Chem. 45, 371-415[Medline] [Order article via Infotrieve] |
| 10. | Breiter, D. R., Kanost, MR., Benning, M. M., Wesemberg, G., Law, J. H., Wells, M. A., Rayment, I., and Holden, M. (1991) Biochemistry 30, 603-608[CrossRef][Medline] [Order article via Infotrieve] |
| 11. | Smith, A. F., Owen, L. M., Strobel, L. M., Chen, H. D. R., Kanost, M. R., Hanneman, E., and Wells, M. A. (1994) J. Lipid Res. 35, 1976-1984[Abstract] |
| 12. | Narayanaswami, V., and Ryan, R. O. (2000) Biochim. Biophys. Acta 1483, 15-36[Medline] [Order article via Infotrieve] |
| 13. | McGuire, K. A., Davidson, S., and Jonas, A. (1996) J. Lipid Res. 37, 1519-1528[Abstract] |
| 14. | Morrow, J. A., Arnold, K. S., and Weisgraber, K. H. (1999) Protein Expression Purif. 16, 224-230[CrossRef][Medline] [Order article via Infotrieve] |
| 15. | Soulages, J. L., Pennington, J., Bendavid, O., and Wells, M. A. (1998) Biochem. Biophys. Res. Commun. 243, 272-276 |
| 16. | Ryan, R. O., Schieve, D., Wientzek, M., Narayanaswami, V., Oikawa, K., Kay, C. M., and Agellon, L. B. (1995) J. Lipid Res. 36, 1066-1072[Abstract] |
| 17. | Soulages, J. L., and Arrese, E. L. (2000) Biochemistry 39, 10574-10580[CrossRef][Medline] [Order article via Infotrieve] |
| 18. | Soulages, J. L., and Arrese, E. L. (2001) Biochemistry. 40, 14279-14290[CrossRef][Medline] [Order article via Infotrieve] |
| 19. |
Li, H-h,
Lyles, D. S.,
Thomas, M. J.,
Pan, W.,
and Sorci-Thomas, M. G.
(2000)
J. Biol. Chem.
275,
37048-37054 |
| 20. | Tricerri, M. A., Behling Agree, A. K., Sanchez, S. A., Bronski, J., and Jonas, A. (2001) Biochemistry 40, 5065-5074[CrossRef][Medline] [Order article via Infotrieve] |
| 21. |
Soulages, J. L.,
and Arrese, E. L.
(2000)
J. Biol. Chem.
275,
17501-17509 |
| 22. |
Soulages, J. L.,
Arrese, E. L.,
Chetty, P. S.,
and Rodriguez, V.
(2001)
J. Biol. Chem.
276,
34162-34166 |
| 23. | Haughland, R. P. (1996) Handbook of Fluorescent Probes and Research Chemicals , 6th Ed. , p. 23, Molecular Probes, Eugene, OR |
| 24. | Nichols, A. V., Krauss, R. M., and Musliner, T. A. (1986) Methods Enzymol. 128, 417-431[Medline] [Order article via Infotrieve] |
| 25. |
Griep, M. A.,
and McHenry, C. S.
(1990)
J. Biol. Chem.
265,
20356-20363 |
| 26. | Chen, Y, and Barkley, M. D. (1998) Biochemistry 37, 9976-9982[CrossRef][Medline] [Order article via Infotrieve] |
| 27. |
Wientzek, M.,
Kay, C. M.,
Oikawa, K.,
and Ryan, R. O.
(1994)
J. Biol. Chem.
269,
4605-4612 |
| 28. | Dettloff, M., Weers, P. M. M., Niere, M., Kay, C. M., Ryan, R. O., and Wiesner, A. (2001) Biochemistry 40, 3150-3157[CrossRef][Medline] [Order article via Infotrieve] |
| 29. | Cheung, H. (1991) in Topics in Fluorescence Spectroscopy (Lakowicz, J. R., ed), Vol. 2 , pp. 55-83, Plenum Press, New York |
| 30. |
Dale, R. E,
and Eisinger, J.
(1976)
Proc. Natl. Acad. Sci. U. S. A.
73,
271-273 |
| 31. | Hass, E., Katchalski-Katzir, E., and Steinberg, I. Z. (1978) Biochemistry 17, 5064-5070[CrossRef][Medline] [Order article via Infotrieve] |
| 32. | Lakowicz, J. R. (1999) Principles of Fluorescence Spectroscopy , pp. 374-376, Academic Publishers, Norwell, MA, / Plenum Publishing Corp., New York |
| 33. |
Raussens, V.,
Narayanaswami, V.,
Goormaghtigh, E.,
Ryan, R. O.,
and Ruysschaert, J-M.
(1995)
J. Biol. Chem.
270,
12542-12547 |
| 34. |
Segrest, J. P.,
Jones, M. K.,
Klon, A. E.,
Sheldahl, C. J.,
Hellinger, M.,
DeLoof, H.,
and Harvey, S. C.
(1999)
J. Biol. Chem.
274,
31755-31758 |
| 35. |
Jonas, A.,
Kezdy, K. E.,
and Wald, J. H.
(1989)
J. Biol. Chem.
264,
4818-4824 |
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C. D. Blanchette, R. Law, W. H. Benner, J. B. Pesavento, J. A. Cappuccio, V. Walsworth, E. A. Kuhn, M. Corzett, B. A. Chromy, B. W. Segelke, et al. Quantifying size distributions of nanolipoprotein particles with single-particle analysis and molecular dynamic simulations J. Lipid Res., July 1, 2008; 49(7): 1420 - 1430. [Abstract] |
