 |
INTRODUCTION |
Apolipoprotein A-I
(apoA-I)1 is the major
protein constituent of high density lipoprotein (HDL) and plays crucial
roles in the synthesis, structure, and functions of HDL (1). ApoA-I undergoes gradual extracellular lipidation by the action of ABCA-1 (2-5), leading to the formation of pre-
1, discoidal pre-2, and
spherical 
-HDL particles (6-9). In either its lipid-free, pre--HDL-associated, -HDL-associated, or reconstituted HDL
(PC/cholesterol/apoA-I disks, rHDL) forms, apoA-I promotes efflux of
cholesterol from cells, thus providing substrate for the
lecithin-cholesterol acyltransferase reaction (10). In HDL, apoA-I is
the principal physiological activator of lecithin-cholesterol
acyltransferase (11-13). ApoA-I also promotes high affinity binding of
HDL and rHDL particles to cells (14-17), and analysis of
apoA-I-deficient mice highlighted the importance of apoA-I for the
delivery of HDL cholesterol to steroidogenic tissues via selective
lipid uptake (18).
A number of the functions of apoA-I appear to be mediated by a cell
surface HDL receptor called scavenger receptor class B type I (SR-BI)
(14, 15, 19, reviewed in Ref. 20). SR-BI binds HDL and rHDL, at least
in part via apoA-I (14, 15). In addition, it can bind other
lipoproteins (21, 22) and exhibits complex binding properties
consistent with multiple classes of binding sites or modes of binding
different ligands (19, 23, 24). On binding lipoproteins, SR-BI mediates
both selective cholesteryl ester uptake from the lipoprotein to the
cells (19, 23-25) and bi-directional unesterified cholesterol movement
(23, 26, 27). In addition to cholesteryl esters (19), SR-BI can mediate cellular uptake from HDL of free cholesterol (27, 28), triglycerides (28, 29), phospholipids (30), and vitamin E (31-35). SR-BI-mediated cholesterol, cholesteryl ester, and triglyceride movement between lipoproteins and cells and the affinity of ligand binding to SR-BI can
be influenced by variations in the size and apolipoprotein and lipid
compositions of the HDL particles (14, 15, 29, 30, 36-40). For
example, when the apparent Kd values (microgram of
protein/ml) of discoidal apoA-I-reconstituted HDLs are compared with
those of native plasma HDL, the rHDL is seen to bind 5-10-fold more
tightly than the spherical HDL (14, 15).
The physiologic importance of the interaction of SR-BI with HDL
(apoA-I) has been established by a variety of in vivo
studies, primarily using rodents (reviewed in Refs. 20 and 41). SR-BI controls the structure and composition of plasma HDL, the cholesterol contents of HDL, the adrenal gland, ovary, and bile (42, 43), and helps
protect against atherosclerosis in murine models (43-46 and reviewed
in Ref. 47). It is clear that the ability of SR-BI to mediate selective
lipid uptake from HDL plays a role in HDL metabolism in rodents.
Although the physiologic significance of SR-BI-mediated efflux of
cholesterol out of cells remains to be determined, analysis of this
efflux has provided important insights into the properties of SR-BI and
the mechanisms underlying its ligand binding and lipid transport activities.
It appears that SR-BI-mediated transport of lipids between cells and
lipoproteins involves two sequential steps as follows: 1) lipoprotein
binding and 2) binding-dependent, yet distinct, SR-BI-mediated lipid transfer (23). Although it has been proposed that
SR-BI-mediated efflux of cholesterol from cells to HDL might occur as a
consequence of SR-BI-mediated changes in the structure of the cell
membrane and be independent of HDL binding to the receptor (48),
compelling evidence that HDL binding to SR-BI is required for
SR-BI-dependent efflux has appeared (23).
The two-step nature of SR-BI activity raised the question of whether
distinct features of the structures of apoA-I or SR-BI were responsible
for each of these steps and if only certain "productive" modes of
apoA-I/HDL binding to SR-BI were compatible with efficient lipid
transfer. Would it be possible using in vitro mutagenesis to
identify residues in apoA-I or SR-BI that, when altered, would block
one step (e.g. lipid transfer) but not the other
(e.g. ligand binding)? In the current study, we have
examined the receptor binding properties and the cholesterol efflux
capacity of discoidal PC/cholesterol/apoA-I rHDL particles prepared
with either unaltered (wild-type, WT) or mutant apoA-I. ApoA-I with
double point mutations in charged residues either in helix 4 (D102A/D103A) or helix 6 (R160V/H162A) were used. These residues were
mutated because they have been proposed to participate in interhelical
interactions of antiparallel apoA-I molecules in discoidal rHDL
particles (49, 50) and are close to the kinks (51) (or -turns
according to Atkinson's model (52)) of apoA-I that precede helix 4 and helix 6 of apoA-I, respectively. We expected that disruption of these
interhelical charge interactions or elimination of the charge of the
surface amino acid exposed to the aqueous phase might alter the nature
of the interactions between apoA-I and SR-BI. In addition to studying
the effects of these mutations in apoA-I on the interactions of rHDL
with wild-type SR-BI, we examined rHDL binding and cholesterol efflux
from transfected cells expressing mutated forms of SR-BI that have
defects in binding and mediating lipid transfer either to both HDL and
LDL (ldlA[M158R]) or to spherical plasma HDL particles but not to plasma LDL (ldlA[Q402R/Q418R]) (23, 24). The results support a model in which efficient SR-BI-mediated lipid transfer depends on productive binding of apoA-I to the receptor. Nonproductive binding interactions are possible, but these do not support efficient lipid transfer. Our findings also provide additional support for the
proposal that efficient cholesterol efflux mediated by SR-BI is
dependent on HDL binding to the receptor.
| |
EXPERIMENTAL PROCEDURES |
Materials
Vent polymerase, T4 ligase, and restriction enzymes were
purchased from New England Biolabs (Beverly, MA). Other materials for
the PCR were obtained from PerkinElmer Life Sciences. Oligonucleotides for PCR, DH10Bac competent cells, Sf900-II SFM, Ham's F-12
medium, fetal bovine serum, trypsin/EDTA, penicillin/streptomycin,
glutamine, and G418 sulfate and rTEV protease were purchased from
Invitrogen. Bactotryptone and bacto-yeast extract were obtained from
VWR Scientific (Pittsburgh, PA). Other reagents (and sources) are as
follows: sodium [125I]iodide,
[1,2-3H]cholesterol (1 mCi/ml, specific activity of
40-60 Ci/mmol) (PerkinElmer Life Sciences); fatty acid-free bovine
serum albumin, cholesterol, sodium cholate, and
1-palmitoyl-2-oleoyl-L-phosphatidylcholine (POPC);
aprotinin, benzamidine, leupeptin, and phenylmethylsulfonyl fluoride,
o-phenylenediamine dihydrochloride, Triton X-100 from Sigma;
dialysis tubing (Spectrum Medical Industries, Inc., Los Angeles, CA);
immunoblot polyvinylidene difluoride membrane, Bio-Rad Dc Protein Assay
Kit (Bio-Rad); monoclonal anti-human apoA-I antibody 5F-6 (Ottawa
Heart Institute, Ontario, Canada); anti-mouse IgG conjugated
to horseradish peroxidase and 3',3'-diaminobenzidine or enhanced
chemiluminescence reagent (Amersham Biosciences); peroxidase-conjugated
sheep anti-human apoA-I antibody (Biodesign International Inc., Saco,
ME). Nickel-nitrilotriacetic acid resin was purchased from Qiagen Inc.
(Hilden, Germany). All other reagents were purchased from Sigma,
Bio-Rad, or other standard commercial sources as described previously
(14).
Methods
Generation of Expression Plasmids Containing WT and Point
Mutations in apoA-I; Expression of apoA-I in the Baculovirus System and
Protein Purification--
To generate baculoviruses expressing
wild-type and mutant apoA-I forms, a BamHI-SalI
fragment containing human apoA-I cDNA was cloned into the
polylinker region of pFASTBAC donor plasmid that contains the
ampicillin and gentamycin resistance genes (Invitrogen). This
recombinant plasmid also contains a histidine tag and the Tobacco
Etched viral (TEV) protease cleavage site. The apoA-I-containing plasmid was used to transform DH10Bac Escherichia coli cells
(Invitrogen). These cells had been transformed previously with the
baculovirus genome containing the lacZ and kanamycin
resistance genes, along with a helper plasmid containing the
tetracycline resistance gene and the transposase genes. Transposition
of the apoA-I gene into the lacZ gene disrupts the
expression of lacZ and provides a recombinant plasmid
named bacmid.
The generation of the donor plasmid expressing WT apoA-I was described
previously (14). For the generation of AI(D102A/D103A) and
AI(R160V/H162A), the human apoA-I cDNA was mutagenized by PCR,
using a set of specific mutagenesis primers, containing the mutation of
interest, and a set of flanking universal primers containing the
restriction sites BamHI and SalI, using the
pBluescript-AI cDNA plasmid as template. For the mutagenesis of
AI(D102A/D103A), 5' (AIMIV1-5) and 3' (AIMIV1-3) primers were used.
For the mutagenesis of AI(R160V/H162A), 5' (AIMIV4-5) and 3'
(AIMIV4-3) primers were used. The set of universal primers are 5'
(AIWC-5) and 3' (AIWC-3) (see Table I)
(53). The generation of the donor plasmid expressing apoA-I(
(1-59))
was described previously (14). For the generation of
apoA-I((185-243)), the region between amino acids +1 and +184 was
amplified using the 5' (AIC-5) and 3' (AIM3-3) primers, respectively (see Table I). The DNA fragment containing the mutation of interest was
digested with BamHI and SalI and cloned into the
corresponding sites of pFASTBAC donor plasmid. Cells containing
recombinant bacmids were selected by kanamycin, tetracycline, and
gentamycin resistant as white colonies due to the disruption of
lacZ sequence in the recombinant bacmid. Recombinant
bacmid DNA was isolated from minipreps and used to transfect a
monolayer of Sf-9 insect cells (54-56). Recombinant viruses were
isolated, amplified, titrated, and used to infect larger Sf-9 cell
cultures grown in suspension at 27 °C. Sf-9 cells were pelleted and
resuspended in a lysis buffer. The supernatant was used for the
purification of apoA-I fusion proteins using a nickel-nitrilotriacetic
acid resin affinity chromatography (57, 58). The pure apoA-I without
the His tag, when needed, was obtained by cleavage with TEV protease
and purified by a second nickel-nitrilotriacetic acid resin affinity
column.
Preparation of Discoidal Reconstituted HDL Particles (rHDLs) of
Apolipoprotein, POPC, and Cholesterol--
Complexes composing
apolipoprotein, phosphatidylcholine (POPC), and cholesterol were
prepared using the sodium cholate dialysis method (59) with minor
modifications. The complexes were prepared at a apoA-I:POPC:cholesterol
molar ratio of 1:100:10, as reported previously (15). Briefly, POPC and
cholesterol were mixed and dissolved in glass-distilled
chloroform:methanol (2:1). The solvent was evaporated under nitrogen.
After drying, the lipids were suspended in buffer A (10 mM
Tris-HCl, 0.15 M NaCl, and 1% sodium EDTA (w/w), pH 8.0)
by vortexing and held on ice for 1 h. Then sodium cholate (the
final cholate/POPC molar ratio was 1) was added, and the mixture was
incubated for 1 h on ice. The appropriate amount of apolipoprotein
was then added, followed by an hour-long incubation on ice. The sodium
cholate was then removed by extensive dialysis against buffer A at
4 °C using tubing with a molecular mass cut-off of 12-14
kDa. Complexes were stored under nitrogen at 4 °C.
Apolipoprotein-lipid complex formation was verified by analysis with
native polyacrylamide gradient (8-25%) gel electrophoresis (PhastGel
system, Amersham Biosciences). The lipid-free apoA-I that was not
incorporated in the lipid protein complexes was removed from the
preparation either by dialysis using a dialysis membrane with a
molecular mass cut-off of 50 kDa or by gel filtration using an
Amersham Biosciences Superose 6HR column, 10/30 (total bed volume 24 ml). The column was loaded with 200 ml of sample in buffer A and was eluted with buffer A at a rate of 0.4 ml/min, with a 0.5-ml fraction size. rHDL particles were usually recovered in fractions 9-14, as
determined by native gradient gel electrophoresis (8-25%) using the
PhastGel (Amersham Biosciences).
Cell Cultures for Receptor Binding Assays and
[3H]Cholesterol Efflux--
The ldlA-7 cell line is an
LDL receptor-deficient Chinese hamster ovary cell mutant that expresses
very little SR-BI protein or HDL binding/selective uptake activity (21,
60). The ldlA-7 cells were maintained in monolayer culture in Ham's
F-12 medium containing 5% fetal bovine serum, 100 units/ml penicillin,
100 mg/ml streptomycin, and 2 mM glutamine (medium A). The
generation of cell lines expressing the WT and mutant forms of SR-BI
designated as ldlA[mSR-BI], ldlA[Q402R/Q418R], and ldlA[M158R],
respectively, were described previously (19, 23, 24). These cells were maintained as stock cultures in medium A supplemented with 500 µg/ml
G418 (medium B). All incubations with cells were performed at 37 °C
in a humidified 5% CO2, 95% air incubator.
Immunoreceptor Assay by ELISA--
To detect the association of
rHDL particles containing apoA-I with different cells, the
immunoreceptor assay as reported by us previously (14) was modified to
use ELISA, instead of Western blotting. On day 0, cells were plated in
6-well dishes at 3 × 105 cells/well in medium A
(ldlA-7) or medium B (ldlA[mSR-BI], ldlA[M158R], or
ldlA[Q402R/Q418R]). On day 2, the monolayers were washed twice with
Ham's F-12 medium containing 100 units/ml penicillin, 100 mg/ml
streptomycin, and 2 mM glutamine (medium C) and then re-fed with 0.7-1 ml of medium C with the indicated amounts of apoA-I containing rHDL particles or spherical HDL. After a 2-h incubation at
37 °C, the cells were washed twice at 4 °C with buffer B (50 mM Tris-HCl, 0.15 M NaCl, pH 7.4) containing 2 mg/ml fatty acid-free bovine serum albumin, followed by one rapid wash
with buffer B alone. The cells were then solubilized in lysis buffer
(PBS, 1% Triton X-100, 50 mg/ml phenylmethylsulfonyl fluoride, 1 mg/ml pepstatin A, 20 mg/ml aprotinin, and 10 mg/ml leupeptin)
and incubated, with shaking, at 4 °C for 30 min. The cell lysates
were collected by scraping the wells and were clarified by
centrifugation at 14,000 rpm in a Beckman Microfuge for 20 min at
4 °C. The protein concentrations of the cell lysates were determined
by BCA Protein Assay (61). An aliquot of 20~30 µg of cell lysate
was used to determine the receptor association using ELISA.
For receptor association experiments, the Maxisorb 96-well plates were
coated with anti-human apoA-I monoclonal antibody (1 µg/ml) (Ottawa
Heart Institute, Ontario, Canada) diluted 1:800 in PBS at 100 µl/well
and stored at 4 °C overnight. The coating solution was aspirated,
and the well was washed three times with washing buffer (0.05% Tween
20 in PBS) at 200 µl/well. Blocking buffer (10% nonfat dry milk in
washing buffer) was then added at 200 µl/well. The plates were
incubated at room temperature for 1.5 h and washed three times
with washing buffer. Cell lysates were diluted in PBS to a total volume
of 100 µl and were added in each well and then incubated at room
temperature for 1 h. To obtain a standard curve, rHDL particles
containing a different amount of apoA-I, mixed with 20-30 µg of the
lysate of blank cells (ldlA-7, ldlA[mSR-BI], ldlA[Q402R/Q418R], or
ldlA[M158R]) that were not treated with rHDL particles or any
lipoprotein, were adjusted to a total volume of 100 µl and were added
to different wells. After the plates were washed with washing buffer
three times, the secondary antibody (sheep anti-human apoA-I,
peroxidase-conjugated) diluted 1:500 in blocking buffer was added at
100 µl/well. The plates were incubated at room temperature for 1 h and washed with washing buffer three times. An aliquot of 200 µl of
o-phenylenediamine dihydrochloride substrate (0.4 mg/ml;
urea hydrogen peroxide, 0.4 mg/ml; phosphate-citrate buffer 0.05 M) was added to each well. After a 30-min incubation at
room temperature, the reaction was terminated by adding 50 µl of 2 N H2SO4 per well. The absorbance was measured at 490 nm using a microtiter plate reader. A polynomial second order regression was used as the best fit for the standard curve. The specific binding was obtained by subtracting the binding of
the ldlA-7 cells from the binding of the SR-BI expressing cells from
each experimental point, and it was expressed as nanograms of
cell-associated HDL or rHDL particles per mg of total cell protein. The
cell-associated ligands were analyzed as a function of concentration by
nonlinear regression by the program Prism using a one-site binding isotherm.
[3H]Cholesterol Efflux from ldlA-7 Cells and Stably
Transfected ldlA-7 Cells Expressing Wild-type and Mutant Forms of
mSR-BI--
On day 0, cells were plated in 6-well plates as described
above at a density of 2 × 105 cell/well. On day 1, the medium in each well was replaced with 2 ml of labeling medium
(medium C containing 10% heat-inactivated NCLPDS and 1 µCi/ml
[1,2-3H]cholesterol). On day 3, cells were washed once
with medium C and then incubated in 2 ml of equilibration medium
(medium C containing 1% fatty acid-free bovine serum albumin). On day
4, cells were washed once with medium C and then incubated at 37 °C
for 45 min with 1 ml of medium C. After a single washing with medium C,
1 ml of efflux medium (medium C containing 10 mM HEPES, pH
7.0, and the appropriate amounts of lipoproteins) was added to each well. Cells were then incubated at 37 °C. At the indicated times, 65 µl of efflux medium was removed from the wells and clarified by
centrifugation for 1 min in a microcentrifuge. The radioactivity in 50 µl of the supernatant was then measured by liquid scintillation counting. At the end of the incubation, the remaining efflux medium was
discarded, and the cells were lysed in 800 µl of lysis buffer (PBS
and 1% Triton X-100) for 30 min at room temperature. The radioactivity
in 100 µl of each cell lysate was determined by liquid scintillation
counting, and the percent of efflux was calculated. Total cellular
[3H]cholesterol was calculated as the sum of the
radioactivity in the efflux medium plus the radioactivity in the cell
lysate. The percent of [3H]cholesterol efflux represents
cholesterol released into the medium at different times divided by the
total cholesterol and multiplied by 100. To calculate the net
SR-BI-mediated efflux, the cholesterol efflux of the untransfected
ldlA-7 cells was subtracted from the cholesterol efflux of the SR-BI
expressing cells.
| |
RESULTS |
Effects of ApoA-I Mutations on Binding and Cholesterol Efflux
Mediated by Wild-type SR-BI--
To examine the effects of introducing
double point mutations into apoA-I on the ability of the apoA-I to
interact with SR-BI, we prepared PC/cholesterol/apoA-I discoidal rHDL
particles with either native (wild-type, rHDL(WT)) apoA-I, or apo-AI
with mutations in helix 4 (D102A/D103A, rHDL(helix 4 M)) or helix 6 (R160V/H162A, rHDL(helix 6 M)). All three rHDLs had similar particle
sizes consisting primarily of particles with apparent diameters of 96 Å (not shown). Various concentrations of these rHDL preparations were
incubated at 37 °C with control ldlA-7 cells that express almost no
SR-BI (19, 21, 23) or ldlA-7 cells that express high levels of wild-type murine (m) SR-BI as a consequence of stable transfection with
a mSR-BI expression vector (ldlA[mSR-BI]) (19, 23). The mSR-BI-dependent binding of rHDL to the cells as a function
of rHDL concentration (Fig.
1A) was determined using an
ELISA immunoreceptor binding assay (see "Experimental Procedures"),
and the efflux of [3H]cholesterol from cells pre-labeled
with this sterol was measured as a function of time at a fixed
concentration of rHDL (Fig. 1, B-D). Fig. 1A
shows that the high affinity and extent of SR-BI-dependent binding (differences between the binding to ldlA[mSR-BI] and control ldlA-7 cells) for all three rHDL preparations (rHDL(WT), rHDL(helix 4 M), and rHDL(helix 6 M)) were similar (apparent Kd values of 4.4 ± 1.1, 3.9 ± 0.8, and 1.1 ± 0.5 µg
protein/ml, respectively). The similarities of the binding of these
rHDLs to SR-BI were confirmed independently in experiments in which
binding of 125I-HDL to cells was competed by varying
concentrations of the rHDLs (data not shown).
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 1.
rHDL binding to and
rHDL-dependent [3H]cholesterol efflux from
control cells (ldlA-7) and cells expressing wild-type SR-BI
(ldlA[mSR-BI]) A-F. A, concentration
dependence of the binding of rHDL particles prepared with WT apoA-I,
the helix 4 double mutant (D102A/D103A), or the helix 6 double mutant
(R160V/H162A) to ldlA[mSR-BI] cells. On day 2 after plating,
association of the indicated lipoproteins with ldlA[mSR-BI] or ldlA-7
cells was determined using the immunoreceptor assay as described under
"Experimental Procedures." The specific binding shown in
A was determined by subtracting the values of binding to the
ldlA-7 cells from the corresponding values of binding to the
ldlA[mSR-BI] cells. Values shown are representative of three
independent experiments. The apparent Kd values for
association with ldlA[mSR-BI] cells were determined by nonlinear
regression analysis. B-D, ldlA[mSR-BI] and ldlA-7 cells
were plated at 200,000 cells/well in 6-well dishes, labeled with
[3H]cholesterol for 48 h in medium C containing 10%
NCLPDS, washed, incubated in equilibration medium for 24 h,
washed, and incubated for 45 min at 37 °C with 1 ml of medium C. The
medium was then aspirated and replaced with 1 ml of efflux medium
containing 28 µg of protein/ml (1 µM) of rHDL prepared
with the indicated apoA-Is. After incubation at 37 °C for the
indicated times, [3H]cholesterol released to the media
from ldlA-7 cells (B) or ldlA[mSR-BI] cells (C)
was measured as described under "Experimental Procedures."
D shows the differences in cholesterol efflux between
ldlA[mSR-BI] and ldlA-7 cells (SR-BI-mediated efflux). E
shows the relative values (%) of SR-BI-dependent net
cholesterol efflux from the ldlA[mSR-BI] cells to the different
rHDL acceptors. The 100% efflux value was defined as the efflux
observed using ldlA[mSR-BI] cells and rHDL particles containing WT
apoA-I after a 4-h incubation at 37 °C as shown in D. The
value used was the average of three experiments and corresponded to a
net efflux of 23% of 3H-labeled cellular cholesterol.
F shows the relative values (%) of cholesterol efflux from
ldlA[mSR-BI] cells expressing the WT receptor normalized by the
relative amount of SR-BI-dependent rHDL binding to the
ldlA[mSR-BI] cells for the different rHDL acceptors. The 100%
receptor binding value was defined as the amount of rHDL(WT) binding to
ldlA[mSR-BI] cells observed at an rHDL concentration of 28 µg of
protein/ml (1 µM). The acceptor rHDL particles containing
WT or mutant apoA-I forms are indicated as follows: rHDL(WT),
black bar; rHDL(helix 4 M), striped bar;
rHDL(helix 6 M), shaded bar. Values in A-F
represent the averages of 3-5 independent experiments performed in
duplicate.
|
|
Fig. 1B shows that efflux rates of
[3H]cholesterol from control ldlA-7 cells to all of the
rHDLs (28 µg of protein/ml, 1 µM) were similar. In
contrast, efflux rates from ldlA[mSR-BI] cells to rHDL, which
represents the total cholesterol efflux, were all greater than the
rates for the control ldlA-7 cells and depended on the form of apoA-I
in the particles (Fig. 1C). Fig. 1D shows the
differences between the rates measured in the transfected (ldlA[mSR-BI]) and untransfected control (ldlA-7) cells, which represent SR-BI-dependent efflux. The relative efflux rates
were rHDL(WT) > rHDL(helix 6 M) > rHDL(helix 4 M) (Fig.
1D). To simplify the analysis of the efflux and binding
activities of the different rHDLs at a given rHDL concentration, we
tabulated the extents of SR-BI-dependent efflux after
4 h of incubation at 37 °C and the level of
SR-BI-dependent binding, setting both efflux and binding
values for rHDL(WT) to 100%. Fig. 1E shows the relative [3H]cholesterol efflux values for all three rHDLs, and
Fig. 1F shows those values corrected for the extent of
binding (relative % efflux/relative % binding) at rHDL concentrations
of 28 µg of protein/ml (1 µM). (Similar values were
obtained using the efflux results determined after 2 h of
incubation.) These data clearly show that the mutations in helices 4 and 6 substantially reduced the capacity of the rHDLs to mediate
SR-BI-dependent cholesterol efflux without substantially altering their abilities to bind to SR-BI. Cholesterol efflux from the
WT SR-BI was also reduced by ~40-50% when rHDL particles containing
N- (apoA-I((1-59))) or C-terminal (apoA-I((185-243))) truncated
apoA-I forms were used as a cholesterol acceptor (data not shown),
despite the relatively high affinity of binding of these particles to
SR-BI (14). These results are consistent with the possibility that
rHDL(WT) bound productively to SR-BI and thus promoted lipid transfer,
whereas the binding of rHDLs containing the two mutants was less
productive with respect to lipid transfer.
Effects of ApoA-I Mutations on Binding and Cholesterol Efflux
Mediated by the M158R SR-BI Mutant--
To explore further the effects
of the mutations in helices 4 and 6, we repeated the experiments shown
in Fig. 1 using cells (ldlA[M158R]) expressing SR-BI with a Met 
Arg point mutation at position 158 (23). This mutant receptor is
expressed on the surface of cells and can bind and mediate lipid
transfer to and from chemically modified (acetylated) LDL in a fashion
similar to wild-type SR-BI (23). However it exhibits almost no ability to bind to and thus mediate lipid transfer with spherical
plasma HDL and unmodified LDL (23). Fig.
2A shows that the affinity of
rHDL(WT) binding to SR-BI(M158R) (apparent Kd = 48 ± 12 µg of protein/ml) was 10-fold lower than that for
wild-type SR-BI (apparent Kd of 4.4 ± 1.1 µg
of protein/ml, see Fig. 1A). Thus, both rHDL(WT) and plasma
HDL bind much less tightly to this mutant receptor than to the
wild-type receptor; however, the rHDL(WT) binding to the mutant
receptor was readily detected and quantitated, whereas that of the
spherical plasma HDL was almost undetectable (23). The binding of
rHDL(helix 4 M) to SR-BI(M158R) was also much weaker than its binding
to the wild-type receptor (apparent Kd values of
60 ± 13 versus 3.9 ± 0.8 µg of protein/ml).
Unexpectedly, rHDL(helix 6 M) bound almost as tightly to the mutant
receptor as it bound to the wild-type SR-BI (apparent
Kd values of 7 ± 3 and 1.1 ± 0.5 µg of protein/ml, respectively). This raises the possibility that the altered
structure of the helix 6 mutant might compensate for the altered
structure of the binding site in this mutant receptor.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 2.
rHDL binding to and
rHDL-dependent [3H]cholesterol efflux from
cells expressing a mutant SR-BI (ldlA[M158R]) A-D.
A, concentration dependence of the binding of rHDL particles
prepared with WT apoA-I, the helix 4 double mutant (D102A/D103A), or
the helix 6 double mutant (R160V/H162A) to ldlA[M158R] cells. On day
2 after plating, association of the indicated lipoproteins with
ldlA[M158R] cells was determined using the immunoreceptor assay as
described under "Experimental Procedures." The specific binding
shown in A was determined by subtracting the values of
binding to the ldlA-7 cells from the corresponding values of binding to
the ldlA[M158R] cells. The apparent Kd values for
association with ldlA[M158R] cells were determined by nonlinear
regression analysis. B shows the differences in cholesterol
efflux between ldlA[M158R] and ldlA-7 cells (SR-BI[M158R]-mediated
efflux) determined as described in Fig. 1 and under "Experimental
Procedures." C shows the relative values (%) of
SR-BI-dependent net cholesterol efflux from the
ldlA[M158R] cells to the different rHDL acceptors. The 100% efflux
value was defined as the efflux observed using ldlA[mSR-BI] cells and
rHDL particles containing WT apoA-I after a 4-h incubation at 37 °C
as shown in Fig. 1D. The value used was the average of three experiments and corresponded to a net efflux
of 23% of 3H-labeled cellular cholesterol. D
shows the relative values (%) of cholesterol efflux normalized by the
relative amount of SR-BI-dependent rHDL binding to the
ldlA[M158R] cells for the different rHDL acceptors. The 100%
receptor binding value was defined as the amount of rHDL(WT) binding to
ldlA[mSR-BI] cells observed at an rHDL concentration of 28 µg of
protein/ml (1 µM). The efflux from ldlA[mSR-BI] cells
to rHDL(WT) is indicated in C and D by
black bars. The efflux from ldlA(M158R) cells to rHDL(WT),
rHDL(helix 4 M), and rHDL(helix 6 M) are indicated by
white, striped, and shaded bars,
respectively. Values in A-D represent the averages of 3-5
independent experiments performed in duplicate.
|
|
Fig. 2B shows the SR-BI-dependent efflux of
[3H]cholesterol from ldlA[M158R] cells mediated by the
rHDLs (28 µg of protein/ml, 1 µM) as a function of
time. The relative efflux rates compared with those for rHDL(WT) and
the wild-type receptor (100%) are shown without and with correction
for the extent of binding in Fig. 2, C and D,
respectively. It is clear that the absolute rates of efflux mediated by
rHDL(WT) and rHDL(helix 4 M) were substantially lower when the M158R
mutant replaced the wild-type SR-BI. However, much of the reduced rate
of efflux was a consequence of relatively reduced amounts of binding to
the mutant receptor, rather than intrinsic loss of lipid transfer
activity. When corrected for the extent of binding (Fig.
2D), the efficiency of efflux for rHDL(helix 4 M) (107%)
was not significantly different from that for the rHDL(WT) with
wild-type receptor (100%). These data indicate that the M158R receptor
mutant had relaxed structural requirements for productive binding
relative to the wild-type receptor, for rHDL(helix 4 M) the ratios of
% efflux to % binding were 107 versus 23, respectively
(Figs. 2D and 1F). Similar efflux results were obtained when the rHDL concentration was increased to 50 µg of protein/ml (data not shown).
The ability of rHDL(WT) to mediate efflux from the M158R mutant
receptor was somewhat lower (71%) than that from the wild-type receptor (100%) (Fig. 2D). This suggests that the small
reduction in productive binding of rHDL(WT) due to the mutation in
the receptor was compensated for by the helix 4 mutation. Given the
apparently reduced stringency of productive binding of the M158R mutant
receptor and the high affinity of rHDL(helix 6 M) binding to this
mutant (Fig. 2A), it was not surprising to observe a high
absolute (98%) and binding-corrected relative (111%) efflux of
cholesterol from ldlA[M158R] cells to rHDL(helix 6 M) (Fig. 2,
C and D). Both the binding and efflux
observed for the rHDL(helix 6 M)/ldlA[M158R] pair of mutant ligand
and mutant receptor were similar to those activities in the
corresponding pair of wild-type control molecules (compare Fig. 1,
A and D, with Fig. 2, A and
B).
Effects of ApoA-I Mutations on Binding and Cholesterol Efflux
Mediated by the Q402R/Q418R SR-BI Double Mutant--
To determine
whether the properties exhibited by the M158R mutant receptor, relaxed
structural requirements for productive binding and restoration of high
affinity rHDL binding by the helix 6 mutant (Fig. 2), were common to
another SR-BI mutant with reduced plasma HDL binding, we examined
cholesterol efflux and ligand binding by ldlA[Q402R/Q418R] cells. The
Q402R/Q418R mutations in SR-BI dramatically reduce the binding of
plasma HDL and the consequent lipid transfer, without substantially
altering the binding of plasma LDL and its corresponding lipid transfer
(23, 24). Analysis of the saturation binding curves (not shown) showed that all three rHDL preparations (WT, helix 4, and helix 6) bound less
tightly to the Q402R/Q418R mutant receptor (apparent
Kd values of 10 ± 3, 9 ± 3.5, and
27 ± 10 µg of protein/ml, respectively) than to the wild-type
receptor (apparent Kd values of 4.4 ± 1.1, 3.9 ± 0.8, and 1.1 ± 0.5 µg of protein/ml, respectively). Nevertheless, all three preparations of rHDL appeared to bind well to
this mutant receptor and certainly more tightly than spherical plasma
HDL, whose binding was so low that we were unable to determine accurately the corresponding apparent Kd (23,
24).
Fig. 3 shows the
SR-BI(Q402R/Q418R)-dependent efflux of
[3H]cholesterol to the rHDLs (28 µg of protein/ml, 1 µM) as a function of time (Fig. 3A) and the
relative efflux rates compared with those for rHDL(WT) and the
wild-type receptor (100%), without or with correction for the extent
of binding (Fig. 3, B and C, respectively). There
was a striking reduction in both the absolute and binding-corrected
efflux rates for all three rHDLs compared with the 100% controls.
These data support the conclusions drawn from the analysis of binding
and efflux mediated by the wild-type receptor; the structure of the
ligand-receptor complex can substantially affect the efficiency of
lipid transfer independently of effects on binding affinity. In
contrast to the results with the M158R receptor mutant, there was
little evidence for reduction in the stringency of productive binding
by the Q402R/Q418R mutant.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 3.
rHDL-dependent
[3H]cholesterol efflux from cells expressing a mutant
SR-BI (ldlA[Q402R/Q418R]) A-C. A shows the
differences in cholesterol efflux between ldlA[Q402Q/Q418R] and
ldlA-7 cells (SR-BI(Q402R/Q418R)-mediated efflux) determined as
described in Fig. 1 and under "Experimental Procedures."
B shows the relative values (%) of
SR-BI-dependent net cholesterol efflux from the
ldlA[Q402R/Q418R] cells to the different rHDL acceptors. The 100%
efflux value was defined as the efflux observed using ldlA[mSR-BI]
cells and rHDL particles containing WT apoA-I after a 4-h incubation at
37 °C as shown in Fig. 1D. The value used was the average
of three experiments and corresponded to a net efflux of 23% of
3H-labeled cellular cholesterol. C shows the
relative values (%) of cholesterol efflux normalized by the relative
amount of SR-BI-dependent rHDL binding to the
ldlA[Q402R/Q418R] cells for the different rHDL acceptors. The 100%
of receptor binding value was defined as the amount of rHDL(WT) binding
to ldlA(mSR-BI) cells observed at an rHDL concentration of 28 µg of
protein/ml (1 µM). The efflux from ldlA(mSR-BI) cells to
rHDL(WT) is indicated in B and C by black
bars. The efflux from ldlA[Q402R/Q418R] cells to rHDL(WT),
rHDL(helix 4 M), and rHDL(helix 6 M) are indicated
white, striped, and shaded bars,
respectively. Values in A-C represent the averages of 3-5
independent experiments performed in duplicate.
|
|
Comparison of Binding-dependent and Binding-independent
Mechanisms of SR-BI-mediated Cholesterol Efflux--
The results in
Figs. 1-3 provide additional insight into the question of whether or
not HDL binding to SR-BI is required for SR-BI-mediated efflux of
cellular cholesterol to HDL. Compelling evidence for a direct
binding-dependent mechanism of efflux has been reported
(23). Nevertheless the observation that SR-BI can apparently alter the
distribution of cholesterol in the plasma membrane of cells in which it
is expressed and other data led to the suggestion of the possibility of
binding-independent, SR-BI-mediated efflux of cholesterol to HDL and
other acceptors (48). If SR-BI-mediated efflux to HDL (or rHDL) were
binding-independent, differences in the efficiencies of efflux to the
three rHDL preparations used here would not depend on the nature of the
cellular SR-BI (mutant or wild-type). Similarly, differences in the
efficiencies of efflux mediated by wild-type or mutant (M158R or
Q402R/Q418R) receptors would not depend on the nature of the rHDL
acceptor in the extracellular medium. However, this was not the case.
With wild-type SR-BI, rHDL(WT) was a "good" acceptor (100%) of
cellular cholesterol, whereas rHDL(helix 6 M) was a "poor" acceptor
(49%) (Fig. 1E). Yet with the M158R mutant form of SR-BI,
rHDL(WT) was a poor acceptor (32%), whereas rHDL(helix 6 M) was a good
acceptor (98%) (Fig. 2C). If rHDL(helix 6 M) was an
intrinsically poor cholesterol acceptor compared with rHDL(WT), and if
the M158R mutant receptor were intrinsically less efficient than the
wild-type receptor at mediating efflux (based on the data with
rHDL(WT)), perhaps because it was not able to alter appropriately the
cellular distribution of cholesterol, then one would expect that efflux
mediated by the M158R receptor to rHDL(helix 6 M) would have been
especially poor (32 × 49% = 16%, compared with 100% for
rHDL(WT) with wild-type SR-BI) (Fig.
4A). Similarly, it would be
expected that the efflux mediated by the M158R receptor to rHDL(helix 4 M) should be 32 × 21% = 7% compared with 100% for rHDL(WT)
with wild-type SR-BI (Fig. 4A). Thus, the observed relative
efflux rates of 98% for rHDL(helix 6 M) and 44% for rHDL(helix 4 M)
with the M158 mutant receptor were not compatible with a
binding-independent model, in which the expected efflux rates would
have been 16 and 7%, respectively (Fig. 4A). Fig. 4,
B and C, shows a comparison of the observed
relative efflux rates (not corrected for binding) with the expected
efflux rates calculated with the assumption that efflux was
binding-independent. The values in Fig. 4, B and C, are based on the data in Figs. 1-3 and additional data
in which binding (14) and efflux (not shown) experiments using rHDLs prepared with N- or C-terminal deletion mutants of apo-AI. The N-terminal mutant lacks residues 1-59, and the C-terminal mutant lacks
residues 185-243. The predominant sizes of the rHDL particles containing the N- and C-terminal truncated mutants of apoA-I were 96 and 78 Å, respectively (14). rHDL particles prepared with these
truncated mutants bind tightly to SR-BI (14). In six of the eight
comparisons between observed and expected relative efflux rates, the
observed rates were substantially greater than the rates expected from
the binding-independent model. Taken together, these data and
previously published studies (23) provide very strong support for a
binding-dependent mechanism of SR-BI-mediated cholesterol efflux to HDL or rHDL.
View larger version (29K):
[in this window]
[in a new window]
|
Fig. 4.
Observed and calculated cholesterol efflux
values from cells expressing mutant receptors by rHDL particles
containing mutant forms of apoA-I A-C. A,
schematic representation of predictions based on a binding independent
mechanism of cholesterol efflux. The figure defines the receptor
activity and the efflux capacity of the acceptor. The model assumes
that (a) the efflux capacity of an acceptor is a property of
the acceptor and is influenced by structural mutations in apoA-I;
(b) the receptor activity in cholesterol efflux is a
property of a receptor and is influenced by structural mutations in the
SR-BI; and (c) there is no receptor-acceptor interaction and
thus the receptor cannot alter the properties of the acceptor and vice
versa. Based on these assumptions, the "expected cholesterol
efflux" was defined as the product of the receptor activity times the
efflux capacity of the acceptor. B and C,
observed and expected values of relative cholesterol efflux promoted by
rHDL particles containing different mutant apoA-I forms from cells
expressing the mutant receptor SR-BI(M158R) (B) or
SR-BI(Q402R/Q418R) (C). Observed values indicated
experimentally determined values of cholesterol efflux (open
bars). The expected cholesterol efflux was calculated and
represented the product of the % efflux promoted by rHDL(WT) from
cells expressing the mutant receptor (receptor activity) times the
relative efflux promoted by rHDL particles containing a particular
apoA-I mutant (i.e. rHDL(helix 4 M), etc.) from cells
expressing the WT SR-BI (efflux capacity) (closed
bars).
|
|
| |
DISCUSSION |
SR-BI-mediated transfer of lipid between cells and lipoproteins is
a two-step process: 1) lipoprotein binding and 2)
binding-dependent, yet distinct, SR-BI-mediated lipid
transfer (23, 25, 62). The results from the current and previous (23)
studies very strongly support the proposal that plasma HDL or rHDL must
bind to SR-BI if this receptor is to mediate lipid transfer between the
lipoprotein and the cells. To determine if high affinity lipoprotein binding to SR-BI were sufficient to ensure efficient lipid transfer, we
prepared a series of discoidal, PC/cholesterol/apoA-I rHDL particles
with wild-type or mutant apoA-Is and examined their binding to and
mediation of [3H]cholesterol efflux from cells expressing
wild-type or mutant forms of SR-BI. Two different sets of double
mutations in apoA-I could dramatically reduce cholesterol efflux to the
rHDLs without substantially changing their binding affinities
(apparent Kd values) to wild-type SR-BI. Thus, tight
binding is not sufficient to ensure efficient lipid transfer.
To account for these findings, we propose that efficient SR-BI-mediated
cholesterol efflux to lipoproteins may require not only binding, but
also the formation of a "productive complex." This is also probably
the case for selective lipid uptake. Whereas other explanations for our
data are possible, we suggest that for productive complexes to form and
thus facilitate efficient lipid transport, both the lipoprotein and the
receptor should be precisely aligned, have the capacity to undergo
appropriate conformational changes, or both. The concept of precise
orientation of the apoA-I-containing lipoprotein bound to the receptor
resulting in productive binding for lipid transfer is analogous to that of correct substrate orientation in the active site of an enzyme. In
the context of this model, the binding to wild-type SR-BI of rHDLs
comprising mutant apoA-Is with either helix 4 (D102A/D103A, rHDL(helix
4 M)) or helix 6 (R160V/H162A, rHDL(helix 6 M)) mutations was not
productive. This is manifested by poor lipid transfer despite tight
binding. It has been reported that the presence of apoA-II in HDL
particles or rHDL particles containing apoE results in tight binding to
the wild-type SR-BI, but transfer of cholesteryl esters from these
particles to cells was less efficient than that from apoA-I only HDL or
rHDL particles reconstituted with apoA-I (28, 38, 39). The efficiency
of lipid transfer from apolipoprotein-free lipid particles to cells via
SR-BI has also been reported to be less efficient than from apoA-I
containing rHDL particles (28). The results of these independent
studies are consistent with the productive complex model for
SR-BI-mediated lipid transfer. In this regard it is noteworthy that
high affinity binding of HDL to a surface receptor called CD36, which
in many ways resembles SR-BI (41), is not sufficient to ensure
efficient cellular selective lipid uptake or unesterified cholesterol
efflux. Thus, CD36, which shares many structural and functional
similarities to SR-BI (41), was initially shown to bind HDL tightly
(22, 25) but not mediate efficient selective lipid uptake (25). These
findings were independently confirmed (62) and extended by the analysis
of cholesterol efflux (48). Thus, HDL binding to CD36 does not appear
to be associated with the formation of a productive complex.
Remarkably, efficient receptor-mediated cholesterol efflux to
rHDL(helix 6 M) or rHDL(helix 4 M) was restored when the receptor was
mutated (SR-BI(M158R)). The M158R mutation disrupts both spherical plasma HDL and LDL binding to SR-BI (23). One possible explanation is
that the M158R mutation may have relaxed the stringency of the
conformational requirements for productive complex formation between
SR-BI and rHDL particles. The affinity of the rHDL binding clearly was
not a determining factor for productive complex formation, the apparent
Kd values of rHDL(helix 4 M) and rHDL(helix 6 M)
binding to this mutant receptor were 60 ± 13 (weak binding) and
7 ± 3 µg protein/ml, respectively. These values should be compared to apparent Kd values for binding to native SR-BI of ~16 and 4 µg of protein/ml for spherical plasma HDL and rHDL with wild-type apoA-I (rHDL(WT)) (14, 15, 19). Not surprisingly, a
different mutation in SR-BI(Q402R/Q418R), which results in the loss of
plasma HDL (24) but not plasma LDL, binding, and lipid transfer, did
not restore efficient lipid transfer to rHDLs with these mutant
apoA-Is. Additional studies will be necessary to determine if only a
limited number of mutations in SR-BI can have the effect of relaxing
the conformational requirements for productive complex formation
between SR-BI and its ligands.
The apoA-I helix 4 (D102A/D103A) and helix 6 (R160V/H162A) mutations
used in this study were chosen after screening the ability of rHDLs
prepared with several apoA-I mutants to promote SR-BI-mediated efflux
of cellular cholesterol. Two models have been proposed to account for
the structure of apoA-I on rHDL particles: the "belt" model and the
"picket fence" model (49, 50, 63). In the belt model, two
apolipoprotein A-I molecules are wrapped together in an anti-parallel
belt around a small discoidal bilayer patch of phospholipid and
cholesterol (160 phospholipid molecules/particle) (49, 50). Each
apoA-I molecule is composed primarily of a set of 10 amphipathic -helices whose hydrophobic surfaces face inward toward
the fatty acid chains of the disc. In the belt model (49, 50), the side
chains of Asp-103 in one apoA-I molecule and His-162 in its
anti-parallel mate form a salt bridge, whereas the side chains of
residues Asp-102 and Arg-160 extend out into the aqueous medium (49,
50). It is possible that the loss of the salt bridge as a result of the
mutations introduced might alter the conformation of the rHDL and its
ability to productively bind to wild-type SR-BI. It is also possible
that the mutations of Asp-102 and Arg-160 either altered the
conformation of the rHDLs or perhaps disrupted direct contacts between
the rHDL and SR-BI. In the picket fence model, the 22-mer amphipathic
-helical repeats of apolipoprotein A-I form tandem antiparallel
helices that are perpendicular to the plane of the disc (63). In this configuration, most of the charged amino acids of the amphipathic helices that are close to the edges of the disc are exposed to the
aqueous phase and may likewise be available to interact with SR-BI.
Neither rHDL(WT) nor rHDL(helix 4 M) bind as tightly as rHDL(helix 6 M)
to the M158R mutant receptor (apparent Kd values of
48 ± 12, 60 ± 13, and 7 ± 3 µg of protein/ml,
respectively). Perhaps the R160V mutation in the helix 6 apoA-I mutant
reduces repulsive charge interactions between the arginine side chain at position 160 and the arginine in the mutant receptor at position 158. Direct physical interactions between SR-BI and the amphipathic helices of apoA-I in rHDL particles were also suggested by
cross-linking experiments (64). Further studies, including those using
single rather than double point mutations in apoA-I, should help
identify the key structural determinants underlying apoA-I/SR-BI
interactions and productive complex formation. Furthermore, structural
analysis may provide in the near future insight into the precise nature of the proposed productive complex that allows efflux of cellular cholesterol.
Regardless of the mechanisms underlying the results described here
(productive binding or some other model), it should be possible to
express in animals dominant negative apoA-I mutants that would permit
the formation of HDL particles that could bind to SR-BI but not
participate in efficient SR-BI-mediated lipid transfer. The presence of
such defective particles in the circulation of animals might be useful
in examining the physiologic role of SR-BI-mediated lipid transport.
,
TOP
ABSTRACT
INTRODUCTION
