© 1998 American Heart Association, Inc.
Original Contributions |
Modulation of Arachidonic Acid Release and Membrane Fluidity by Albumin in Vascular Smooth Muscle and Endothelial Cells
From the Department of Pharmacology, UMDS of Guy's and St Thomas's Hospitals, St Thomas's Campus, London, UK.
Correspondence to Dr S.V. Smirnov, Department of Pharmacology, UMDS of Guy's and St Thomas's Hospitals, St Thomas' Campus, Lambeth Palace Road, London SE1 7EH, United Kingdom. E-mail s.smirnov{at}umds.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Abstract—Albumin is the major plasma protein circulating in blood. Albumin potently decreases capillary permeability, although the mechanisms are not understood completely. Albumin also effectively binds arachidonic acid (AA), which increases capillary permeability. To investigate the interactions of BSA and AA with the cell membrane, the effect of these substances on [3H]AA release and membrane fluidity was studied in vascular myocytes and endothelial cells. BSA (0.2 and 1 mg · mL-1) stimulated a significant release of [3H]AA from both intact rat aorta and cultured smooth muscle cells. This effect was not mimicked by
-globulin or myoglobin
(both 1 mg · mL-1) in intact tissue. BSA, but not
-globulin and myoglobin, decreased the membrane fluidity (assessed
as changes in the steady-state fluorescence anisotropy of
1,6-diphenyl-1,3,5-hexatriene) in a concentration-dependent manner with
a half-maximum concentration between 0.007 and 0.4 mg ·
mL-1 in both freshly isolated and cultured rat aortic
myocytes and human umbilical vein endothelial cells. AA
(1 to 200 µmol/L) caused the opposite effect, increasing
membrane fluidity and antagonizing the effect of BSA. BSA modified at
its arginine residues, which are thought to be important in AA binding,
did not stimulate [3H]AA release and was significantly
less potent than native BSA in altering the membrane fluidity. The
effect of BSA can be explained by a high-affinity binding of AA to the
protein and extraction of AA from the cell membrane. The interaction
between BSA and AA could play a role in the regulation of vascular
permeability.
Key Words: vascular smooth muscle • endothelial cell • membrane fluidity • BSA • arachidonic acid
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Albumin is the major blood plasma protein that binds and transports various substances in the blood. Albumin is known to affect capillary permeability in vitro.1 2 Perfusion of vessels with protein-free solution dramatically increased hydraulic permeability in various vascular beds,3 4 5 an effect that was prevented by adding serum albumin to the perfusate at concentrations as low as 1 mg · mL-1 6 or even 0.1 mg · mL-1.7 Although several hypotheses have been proposed to explain the action of albumin on hydraulic permeability in microvessels, the exact mechanism(s) involved remain controversial.1 2
Conversely, arachidonic acid (AA) (2 to 5 mmol/L) has been demonstrated to increase permeability in cerebral microvessels and induced edema in rat brains.8 Unterberg et al9 found in 1987 that superfusion of cat brains with AA (0.03 to 0.3 mmol/L) opened the blood-brain barrier for Na+-fluorescein and, at higher concentrations (3 mmol/L), for fluorescein isothiocyanate-dextran (molecular weight 62 000). The changes in the permeability of cultured cerebromicrovascular endothelial cells for trypan blue by AA (50 to 100 µmol/L) did not depend on arachidonate metabolism, suggesting that events perturbing the cellular membrane may affect the permeability of cerebral capillaries.10 Using a similar cell model, Villacara et al11 in 1989 demonstrated that AA (100 µmol/L) increased both the permeability of endothelial cells for trypan blue albumin and membrane fluidity in isolated endothelial cell membranes loaded with 1,6-diphenyl-1,3,5-hexatriene (DPH). The former effect was also mimicked by other fatty acids.12 The correlation between the increased cell permeability and increased cell membrane fluidity has been demonstrated previously.13 14 Although the ability of albumin to bind long-chain polyunsaturated fatty acids with high affinity is well documented,15 no data are available describing the interactions of albumin, AA, and membrane fluidity in vascular smooth muscle or endothelial cells.
We therefore have investigated whether BSA could affect the AA content in intact rat aorta, cultured smooth muscle, and endothelial cells, by measuring the release of radioactive AA and the steady-state fluorescence anisotropy as an indicator of the cell membrane fluidity. Some of the results have been presented in abstract form.16 17
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Experiments were performed on both freshly isolated rat aorta myocytes (fRAMs) and human mesenteric arterial myocytes (fHMAMs), cultured RAMs (cRAMs), and human umbilical vein endothelial cells (HUVECs). The procedure for obtaining human tissue was approved by the St Thomas's Hospital Ethical Committee.
Cell Isolation
Freshly isolated smooth muscle cells were prepared from either
rat thoracic aorta (collected from 2 to 4 male Wistar rats, 200 to
300 g) or human mesenteric artery. Pieces of arteries (1 to
1.5 mm wide) were preincubated in nominally
Ca2+-free physiological
saline solution (Ca2+-free PSS) for 20 minutes at
36°C and then transferred into Ca2+-free PSS (5
mL) containing collagenase (3 mg ·
mL-1; 230 U/mg; Worthington), papain (1 mg
· mL-1), 1 mmol/L dithiothreitol, and
15 µmol/L Ca2+. After incubation for 30
minutes at 36°C, pieces of vascular tissue were washed twice in an
ice-cold fresh Ca2+-free PSS and then triturated
using a wide-bored glass pipette. Cells were filtered through a 95-µm
nylon mesh and concentrated by centrifugation at
1100g for 5 minutes. Subsequent cycles of digestion (2
mg · mL-1 collagenase and 0.5
mg · mL-1 papain for 15 minutes) were
also performed if required.
PSS had the following composition (in mmol/L): 130 NaCl, 5 KCl, 1.5 CaCl2, 1.2 MgCl2, 10 HEPES, and 10 glucose. The pH was adjusted to 7.2 with NaOH. Nominally Ca2+-free PSS was devoid of CaCl2.
Cell Culture
HUVECs were isolated by collagenase perfusion (Type
1A, 1 mg · mL-1) for 25 minutes from
umbilical cords obtained from full-term normal births. Dissociated
cells were transferred to tissue culture flasks containing 12.5 mL of
medium 199 supplemented with FCS (15% v/v), penicillin (100 IU
· mL-1), streptomycin (100 µg ·
mL-1), sodium heparin (90 IU ·
mL-1), and endothelial cell
growth supplement (30 µg · mL-1). RAMs
were explant-cultured from freshly isolated aortas after removal of
endothelium. Small pieces of tissue (
2
mm2) were transferred to a
75-cm2 tissue culture flask containing DMEM
supplemented with FCS (10% vol/vol), penicillin (100 IU ·
mL-1), and streptomycin (100 µg ·
mL-1). Cells were maintained in a humidified
incubator at 37°C/5% CO2, and culture medium
was replaced every 2 or 3 days until confluence was achieved. All cells
were used before their fourth passage.
Isolated cell membranes were prepared from confluent HUVECs (passage 1,
3 to 4 · 106 cells). Cells were
harvested using 0.05% trypsin and 0.02% ethylenediaminetetraacetic
acid solution (10 minutes at 37°C), resuspended in 2 mL PSS, and
homogenized using the Ultra-Turrax T25
homogenizer (Janke & Kunkel, IKA-Labortechnik) at
24 000 rpm for 2 minutes at 4°C. Cell debris and nuclei were removed
by centrifugation at 1000g for 10 minutes at
4°C. The supernatant (predominantly
plasmalemmal/sarcolemmal membranes) was diluted to 6 mL
with PSS and used for measurement of fluorescent anisotropy as
described below. All media, PBS, and FCS were purchased from Gibco BRL
(Paisley, UK). All other chemicals and reagents were purchased from
either Sigma Chemical or BDH Merck.
Measurement of [3H]AA Release
Intact Rat Aorta
Twelve rat aorta rings (1 to 1.5 mm long) were placed in 1
mL of PSS and incubated with 2.5 µCi [3H]AA
(223.4 Ci/mmol, NEN) at a final concentration of 11.2 nmol/L and cold
AA (5 µmol/L) for 24 hours at +4°C. Two rings were transferred
into each superfusion chamber (200 µL), retained between 2 filter
paper discs, and superfused with PSS at a flow rate of 0.7 mL ·
min-1 at room temperature using a Brandel
Superfusion System. Tissue was washed for 20 minutes, and then 2
minutes (1.4 mL) fractions were collected. At the end of each
experiment, the aorta rings and filter papers were removed and
dissolved in 0.5 mL of Soluene 350 and then were neutralized with
glacial acetic acid (200 µL). The radioactivity was measured as
disintegrations per minute after the addition of scintillant (3 mL;
Ultima Gold) using a Wallac 1409 liquid scintillation counter. The
results are presented as a percentage of fractional release (%
FR) of [3H]AA.
Cultured RAMs
Confluent cRAMs (passage 1) were fed with 0.1 µCi ·
mL-1 [3H]AA for 72 hours
at 37°C. The culture medium was then replaced with
[3H]AA-free DMEM, and cells were incubated for
24 additional hours. cRAMs were then harvested as described above and
washed 3 times with fresh PSS. Approximately 1.5 to 2 ·
106 cells were then placed into Brandel
superfusion chambers, and measurement of [3H]AA
release was performed as described above.
Measurement of Cell Membrane Fluidity
Cell membrane fluidity was measured using the
fluorescence probe DPH (Sigma). DPH was dissolved in
tetrahydrofuran to obtain a stock solution of 4 mmol/L. Cell
suspension (4 to 8 · 105 per mL in 3 mL)
was incubated in the presence of 2 µmol/L DPH for 30 minutes at
35°C with constant stirring. The penetration of DPH into the cell
membrane was monitored by measuring the increasing
fluorescence, which reached a steady-state level within 20 to
25 minutes of loading. Cells then were concentrated by
centrifugation at 1100g and resuspended in a
final volume of fresh PSS to give 2 to 3x105
cells per mL. Substances tested were added cumulatively to the cuvette
containing 2 mL of cell suspension. Because the maximal changes in the
total volume of the solution were <10% no correction of concentration
was made.
Steady-state fluorescent polarization of DPH was measured using
a Perkin-Elmer SP-50 spectrophotometer at 35°C. DPH was excited by
vertically polarized light at 360 nm, and its emission intensities were
detected at 430 nm through a polarizer orientated parallel and
perpendicular to the direction of polarization of the excitation beam.
Calculation of the steady-state fluorescence anisotropy
(r) was performed using Perkin-Elmer computer software
according to the equation:
 | (1) |
Modification of Albumin
Fatty acid–free (fraction V) BSA was modified using a procedure
described in detail elsewhere.18 Briefly, a
solution of 0.825 g BSA in 25 mL of 0.2 mol/L NaOH was mixed with 330
mg 1,2-cyclohexanedione for 5 minutes and then titrated with HCl to
neutral pH. The solution was dialyzed at 4°C for 48 to 72 hours. The
degree of modification of the arginine residues was
>90%18 and was not determined here. Breakdown
of BSA after the modification process was assessed by SDS-PAGE
analysis and was not significant. The final concentration of
protein was estimated using the standard Bradford protein assay
(Sigma). BSA concentration (µmol/L) was calculated assuming a
relative molecular mass of 66 000.
Data are presented as mean±SEM and compared for significance using the 2-tailed Student t test; differences were deemed significant at P<0.05 unless stated otherwise. Data analysis was performed using Origin 4.1 (Microcal Software Inc).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Release of [3H]AA and its Modulation by BSA in Intact Rat Aorta
Initial background [3H]AA release measured in the absence of drug was between 0.2% and 1.7% FR (mean, 0.4±0.04; n=54), which typically decreased gradually by 20% to 30% toward the end of the experiment (Figure 1A
, 
).
This decrease apparently was not due to the slow removal of bound
radioactive label from the extracellular space, because perfusion of
aortic rings with a high concentration of "cold" AA (50
µmol/L) did not affect [3H]AA release (Figure 1B, 
). Stimulation of phospholipase A2
activity with the Ca2+ ionophore ionomycin
(5 µmol/L) for 20 minutes caused a steady-state increase in
release of [3H]AA, which then returned to the
basal level after the removal of the ionophore (Figure 1A, 
). To
estimate the results quantitatively, the maximal FR in the presence of
ionomycin was corrected for the average background signal measured
immediately before drug application and at the end of experiment
(Figure 1) and is presented in Table 1. Statistical analysis showed a
significant increase in [3H]AA release in
comparison to the control data measured in a similar manner.
|
|
Addition of BSA (1 mg · mL-1) to the
perfusate for 8 minutes caused a dramatic initial rise in
release of [3H]AA within the first 2 minutes,
followed by a slow decrease (Figure 2A).
After removal of BSA, the signal returned to the basal level within 6
to 8 minutes. Two subsequent applications of the same concentration of
BSA initiated smaller and similar steady-state increases in the release
of radioactive label (Figure 2A). To compare these data quantitatively,
the area under the curve in the presence of protein (net
[3H]AA FR) corrected for the background FR was
calculated during each application. Figure 2B shows the mean data
obtained in 9 to 14 similar experiments. The net
[3H]AA FR in response to the second and third
applications of BSA (57% and 68% decreases, respectively) fell
significantly in comparison to that during the first addition; however,
there were no significant differences between the last 2
applications.
|
Analogous experiments performed on cRAMs (passage 1) after prolonged
incubation with radiolabeled AA at 37°C (see Materials and Methods),
when the label would be expected to equilibrate within various
intracellular phospholipid pools more completely than in intact tissue,
showed similar results (Figure 2C
and 2D). Initial application of BSA
(1 mg · mL-1) caused a large increase in
[3H]AA release (Figure 2C), which decayed in a
similar manner to that observed in intact tissue. Two subsequent
additions of BSA (applied in 14-minute intervals) resulted in virtually
identical responses (Figure 2C), in contrast to the intact tissue
preparations (Figure 2A). However, after subsequent incubation of cRAMs
without stimulation for 70 minutes, the fourth application of the same
concentration of BSA strongly increased [3H]AA
release (Figure 2C). The net [3H]AA FR in
response to the fourth application of BSA was significantly larger
(P<0.02 and P<0.03; paired t test)
than that recorded during the previous third and second
applications of the protein, respectively (Figure 2D).
The effect of a longer (20 minutes) application of BSA (1 mg ·
mL-1) was studied in intact rat aorta using the
experimental protocol described above for ionomycin. Under these
conditions, the [3H]AA FR remained elevated
over the entire albumin application period, although it had
decreased by 66±2% (n=6) at the end of the application of BSA,
approaching a steady-state level (Figure 3A). Peak responses to this concentration
of BSA in these experiments and experiments described in Figure 2A
(only the first application was used) were corrected for the background
release as described above and averaged and are presented in
Table 1. The maximal [3H]AA FR was smaller when
the concentration of the protein was decreased to 0.2 mg ·
mL-1 (Table 1). The time course of the effect in
0.2 mg · mL-1 BSA resembled that shown in
Figure 3A, although the [3H]AA FR decrease
(51±5%; n=4) in the presence of protein was significantly smaller
(P<0.02) than that in the presence of 1 mg ·
mL-1 BSA (not shown).
|
-Globulin, the second most abundant plasma protein, could not
stimulate significantly [3H]AA release under
similar conditions at a concentration of 1 mg ·
mL-1 (Figure 3B). Myoglobin (1 mg ·
mL-1) was also ineffective (Table 1).
Modulation of Cell Membrane Fluidity by BSA
Figure 4A shows that
the application of BSA (15 µmol/L; 1 mg ·
mL-1) caused a large increase in steady-state
fluorescence anisotropy (decrease in cell membrane fluidity) in
cRAMs (passage 2) preloaded with DPH, as described in Materials and
Methods. This effect was reversible and could be obtained repeatedly
after washing cells with fresh PSS. The addition of BSA did not affect
cell viability (assessed by trypan blue exclusion), although 16% to
20% of cells were lost during repeated washing and
centrifugation procedures (Figure 4A). Cumulative
addition of BSA to a cuvette containing HUVECs (passage 2) caused a
progressive step-like increase in r. The effect of BSA
become noticeable at BSA concentrations near 0.5 µmol/L, and
r reached a maximal value of 0.29 to 0.3 when the
concentration of protein exceeded 10 µmol/L (Figure 4B, ).
Under control conditions, however, no change in anisotropy occurred,
and a small decrease in r usually was observed (Figure 4B, ). To compare the effect of BSA on the steady-state anisotropy in
various cell types, r was calculated as the mean value when
the effect stabilized at each protein concentration and is plotted
against BSA concentration in Figure 4C. This dependence was sigmoidal
and was fitted to the following equation:
 | (2) |
.
Significant differences in the relative potency of BSA in increasing
the membrane fluidity neither between different cultured and freshly
isolated cell types nor between different passages
(K1/2, 0.1 and 6 µmol/L or 0.007 and 0.4
mg · mL-1, respectively) within the same
cell type were found. In 2 experiments with fHMAMs, qualitatively
similar results were obtained, with K1/2=10 and
7.8 µmol/L and rmax and
rmin=0.3 and 0.31, and 0.18 and 0.24,
respectively.
|
|
Qualitatively similar results of the effect of BSA on membrane fluidity also were obtained on isolated plasmalemmal/sarcolemmal membranes prepared from HUVECs. K1/2 was 9.6±3.9 µmol/L, and rmax and rmin were 0.31±0.02 and 0.2±0.003, respectively, in 3 preparations tested.
-Globulin exerted no significant effect on r over a
similar range of concentrations in HUVECs (passage 1) (Figure 4D, )
and other cell types (Table 2). Myoglobin also was ineffective up to a
concentration of 100 µmol/L (molecular mass, 17 500; Table 2).
Effect of AA on Cell Membrane Fluidity
Cumulative addition of various doses of AA to the suspension of
fRAMs produced the opposite effect on r to that seen on
addition of BSA, decreasing its value (Figure 5A, ). Changes in r caused
by AA also had a sigmoidal dependency on the log of the concentration,
and data were analyzed using the same equation described in the
legend of Figure 4C (Figure 5B). No significant differences in
K1/2 for AA were found between fRAMs, cRAMs, and
HUVECs (Table 2), although large variations in K1/2
values (5 to 129 µmol/L) were observed. The AA solvent DMSO did
not affect cell membrane fluidity (Figure 5A and 5B, ).
|
The interaction of BSA and AA in their capacity to modulate cell
membrane fluidity was studied by applying mixtures of these substances
to fRAMs. Figure 5C illustrates the results of such an experiment.
Cells were split into 3 cuvettes, and BSA (10 µmol/L) and AA
(100 µmol/L) were added to 2 cuvettes. As expected, the addition
of these substances caused contrasting changes in r that
quickly reached steady-state levels of 0.3 and 0.1 for BSA () and
AA (), respectively. Mixtures of BSA (10 µmol/L) and various
concentrations of AA were introduced to the third cuvette (Figure 5C, 
). When 10 µmol/L BSA and 100 µmol/L AA (1:10 ratio)
were added to the suspension of fRAMs, r increased, but only
by 50% compared with when the same concentration of BSA alone was
added. An increase in the BSA:AA molar ratio to 1:20 because of further
addition of AA caused a greater decrease in r, and only at a
concentration of AA 40 times greater than that of BSA (400
µmol/L) did r reach the level measured in the presence of
100 µmol/L AA alone (Figure 5C). Qualitatively similar results
were obtained in another experiment.
Comparison of the Effect of Native and Modified BSA on Cell
Membrane Fluidity and [3H]AA Release
It has been proposed previously that positively charged arginine
residues in the albumin molecule are at or near binding sites
for long-chain fatty acids.19 To disrupt these
putative binding sites in the BSA molecule, the protein was modified on
arginine residues (see Materials and Methods). Figure 6A and 6B illustrates the comparison of
the effect of native () and modified () BSAs on r in
cRAMs (passage 1). Figure 6 shows that modified BSA was 100 times
less potent than the native protein in its ability to decrease cell
membrane fluidity (Figure 6A and 6B). In general, modified BSA was
significantly less potent (K1/2, 10 to 63
µmol/L) than native protein in all cell types studied (Table 2).
|
A comparison of the effect of native and modified BSA (0.2 mg ·
mL-1) on [3H]AA release
showed that modified BSA did not mimic the effect of the native protein
in increasing the release of AA in intact rat aorta preparations
(Figure 6C). Increasing the concentration of modified BSA up to 1
mg · mL-1 also did not potentiate
[3H]AA release significantly (Table 1).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The major findings described in the present study are: (1) BSA was able to increase release of AA from both intact rat aorta and cultured smooth muscle cells; (2) BSA increased steady-state anisotropy (decreased cell membrane fluidity) in DPH-labeled vascular smooth muscle and endothelial cells; (3) both effects were not mimicked by other plasma proteins such as
-globulin and myoglobin;
(4) AA caused the opposite effect of that of BSA, increasing cell
membrane fluidity; and (5) the effect of BSA on both cell membrane
fluidity and AA release was significantly decreased when the protein
was modified at its arginine residues.
Effect of BSA on AA Release
To the best of our knowledge, the release of labeled AA from
intact blood vessels has not been demonstrated previously. To confirm
whether labeled AA is incorporated into some intracellular phospholipid
pools, we examined the effect of ionomycin, which has been shown to
enhance the activity of phospholipase A2 by
elevating the intracellular Ca2+
concentration.20 Under our conditions, ionomycin
caused a significant and sustained [3H]AA
release. This suggests that the radioactive label was incorporated at
least into a phospholipase A2-accessible
phospholipid pool.
We found that BSA significantly stimulated AA release in a
concentration-dependent manner in rat aortic rings (Figures 2A, 2B, and 3A). This effect was not due to removal of extracellular bound
radioactive AA by BSA, because first, perfusion of tissue with a high
concentration (50 µmol/L) of "cold" AA, which would be
expected to displace extracellular bound radioactive AA, did not cause
significant changes in [3H]AA release (Figure 1B). Secondly, both repeated (Figure 2A and 2B) and continuous (Figure 3A) applications of 1 mg · mL-1 BSA
resulted in sustained AA release in intact tissue. A similar effect
also was observed during a continuous perfusion of tissue with a 5-fold
lower concentration of BSA (Figure 6C, ). In addition, we were able
to obtain similar multiple responses to BSA in cultured RAMs (Figure 2C and 2D) under conditions in which the label presumably was distributed
more equally within intracellular phospholipid
pools.21
The small background release of [3H]AA and the initial rapid release in response to BSA followed by a sustained release in the presence of protein are consistent with the possibility that there is a small pool of a BSA-accessible endogenous [3H]AA. Because it seems unlikely that BSA readily would cross the membrane because of its large molecular mass and lipid insolubility, it is probable that the protein can interact with the extracellular surface of both endothelial and smooth muscle cells, thus removing AA from the membrane. Although BSA is able to bind other phospholipids, such as phosphatidylcholine and lysophosphatidylcholine, which also could contain labeled AA, this seems to be an unlikely explanation for our results, because it has been shown that the release of lysophosphatidylcholine by BSA only occurred after a lag phase of 30 minutes and then increased slowly over 4 hours in cultured rat hepatocytes.22 Additionally, BSA-induced release of phosphatidylcholine was found to be very small and slow in the same cells.22
Additionally, the data shown in Figure 2C and 2D indicate that the
fraction of BSA-accessible [3H]AA recovered
completely after 70 minutes in cultured RAMs. These results could
suggest that there is a turnover of AA in vascular smooth muscle and
endothelial cells, yielding free AA that was then
released from intact tissue or isolated cells by BSA.
Modulation of Cell Membrane Fluidity by BSA and AA
To evaluate the possibility that BSA could affect the physical
properties of the cell membrane, the effect of BSA, as well as that of
-globulin, myoglobin, and AA (cell membrane fluidity measured by DPH
fluorescence anisotropy) was compared in
endothelial and smooth muscle cells. DPH is the most
frequently and successfully used fluorescence probe in lipid
bilayers, isolated membranes, and living cells.23
The main disadvantages of using living cells in measurements of
fluorescence polarization or anisotropy are
heterogeneity of the intact cell membrane and slow
passive diffusion of DPH into intracellular membrane
structures.23 Both factors make a quantitative
interpretation of the results difficult. Partition of DPH could be
estimated by monitoring the degree of fluorescence
polarization,24 which would be expected to
decrease with time during loading of DPH.23 Under
our experimental conditions, no changes in the fluorescence
anisotropy were found during a 30-minute loading procedure (not shown).
In addition, the reversibility of the effect of BSA (Figure 4A) also
suggests that the phenomenon was associated preferentially with the
extracellular cell membrane.
BSA (1 mg · mL-1) caused a marked and
reversible increase in r (Figure 4A). The effect on
r appeared at concentrations of BSA between 0.007 and 0.04
mg · mL-1 (0.1 to 0.5 µmol/L), and
r was saturated at concentrations >0.7 mg ·
mL-1 (10 µmol/L; Figure 4B and 4C). This
effect appeared to be specific to BSA, because it was not mimicked by
-globulin or myoglobin. The potency of the effect of BSA was similar
in freshly isolated smooth muscle and cultured smooth muscle and
endothelial cells (Table 2). An equivalent effect of
BSA was also recorded with isolated HUVEC membrane preparations.
Our preliminary results also indicated that BSA caused a qualitatively
similar effect in the DPH-loaded 3T3 fibroblast cell line
(K1/2=1.2±0.4 µmol/L; n=3). These results
suggest that the effect of BSA on the dynamic properties of the cell
membrane is not limited to a particular cell type and may be a general
phenomenon.
The effect of BSA on the fluorescence anisotropy was
antagonized by AA, which caused the opposite effect on cell membrane
fluidity (Figure 5). The apparent potency of the effect of AA on cell
membrane fluidity was significantly lower than that of BSA
(P<0.02 to 0.008) in all cell types studied (Table 2). It
is well known that serum albumin possesses multiple
high-affinity binding sites for fatty acids25 ; it
is able to bind 
20 M of fatty acids per mole
protein.26 These facts concur with our findings,
which showed that the effect of BSA on r could be completely
reversed only at AA concentrations 40 times greater than that of the
protein. This suggests that BSA alters the cell membrane fluidity
because of its high-affinity binding of AA and after the
"extraction" of the fatty acid from the cell membrane. Our other
experimental observations also support this idea. First, the
experiments with BSA modified on arginine residues that were proposed
to be associated with the putative fatty acid binding
sites19 showed that the modified protein was
significantly less potent in modulation of the cell membrane fluidity
and did not stimulate release of AA from intact tissue at
concentrations between 0.2 and 1 mg ·
mL-1 (Figure 6; Tables 1 and 2). Also, the
presence of bilirubin (1 µmol/L), which has a higher affinity
for albumin than AA but binds to a different binding site in
the protein molecule,26 did not prevent the
increase in r by BSA (1 µmol/L; data not shown).
What Is the Physiological Significance of the
Effects of BSA?
Previous reports indicate that albumin alters several
processes in both intact vasculature and single cells. The mechanisms
underlying its actions remain incompletely understood. A widely studied
observation is the effect of albumin on capillary
permeability,1 2 which it dramatically decreases
if the vessel first has been superfused with protein-free solution. The
characteristic features of this effect of albumin on capillary
permeability were: (1) it occurred at concentrations of 
1 mg ·
mL-1 7 27 ; (2) it was not reproduced by
myoglobin and hemoglobin in similar concentration
ranges28 ; and (3) BSA, modified in a manner
similar to that employed in the present article, did not affect the
capillary permeability.18 Taking into account
reports that showed the opposite effect of AA on capillary
permeability,9 10 11 there is a positive
correlation between the effect of BSA and AA on the membrane fluidity
demonstrated in the present article and corresponding capillary
permeability changes. This suggests that the modulation of capillary
permeability by albumin could be mediated, at least partly, via
the affect of protein on the cell membrane fluidity. This is also
supported by the direct correlation between membrane fluidity and
membrane permeability to water and small solutes, which previously has
been demonstrated in artificial membrane
vesicles,29 bovine tracheal epithelial cell
apical membranes,13 and microvillus membrane
vesicles prepared from rat small intestine.14
The contribution of the cell membrane fluidity to the transport of macromolecules across the capillary is less understood, and the binding of albumin to the glycocalyx is thought to play an important role in this process.1 2 Specific albumin-binding proteins of both low (18 and 31 kDa)30 and high (60 kDa)31 molecular weight have been identified in endothelial cells, which could play the role of the anchor protein for albumin on the surface of the cell membrane. The possibility that an alteration in membrane fluidity could affect the structure of the extracellular glycocalyx matrix cannot be excluded entirely, and further experimental work is required to clarify such a possibility.
In 1993, He and Curry32 found that removal of BSA from the perfusate reversibly elevated intracellular [Ca2+] ([Ca2+]i) in intact frog mesenteric microvessels. This effect was associated with a transient increase in capillary permeability and did not occur in the absence of extracellular Ca2+. Recently, it has been shown that plasma albumin caused a decrease in [Ca2+]i in HUVECs33 and astrocytes34 ; however, serum albumin, which was associated with an attached polar lipid, caused a transient increase in [Ca2+]i.33 It was suggested by these workers that albumin might modulate capillary permeability by the removal of active lipids that raise [Ca2+]i levels.33 One possible candidate for the putative lipid is thought to be lysophosphatidic acid35 ; however, in 1995, Nadal et al34 suggested that it also could be a different lipid. Free AA is known to increase [Ca2+]i directly in various cell types.36 37 AA also activates the Ca2+-activated K+ current.38 39 In endothelial cells, this will lead to hyperpolarization of the cell membrane and stimulation of Ca2+ influx because of an increased electrochemical gradient for Ca2+.40 An increase in membrane fluidity by AA could also affect the Ca2+ leak through the endothelial cell membrane, which was shown to be important in the regulation of Ca2+ homeostasis in these cells.41 Therefore, the decrease in [Ca2+]i by albumin in part might be due to a reduction of the concentration of membrane AA by the protein. Our preliminary data also showed that lysophosphatidylcholine potently increased the cell membrane fluidity in vascular and endothelial cells,16 which may suggest that lysophospholipids can share a mechanism similar to the mechanism we described for AA, although this requires further investigation.
|
Acknowledgments |
|---|
This work was supported by the British Heart Foundation (grants BS/95001, PG/96151, and PG/96044). We would like to thank Dr Jo Cunningham for help and advice in measurement of release of radioactive AA and also for useful comments on the manuscript.
Received May 26, 1998; accepted July 29, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1. Michel CC. Capillary permeability and how it may change. J Physiol. 1988;404:1–29.
2.
Curry FE. Determinants of capillary permeability: a
review of mechanisms based on single capillary studies in the frog.
Circ Res. 1986;59:367–380.
3. Rippe B, Folkow B. Capillary permeability to albumin in normotensive and spontaneously hypertensive rats. Acta Physiol Scand. 1977;101:72–83.[Medline] [Order article via Infotrieve]
4. McDonagh PF. Both protein and blood cells reduce coronary microvascular permeability to macromolecules. Am J Physiol. 1983;245:H698–H706.
5. Watson PD. Effects of blood-free and protein-free perfusion on CFC in the isolated cat hindlimb. Am J Physiol. 1983;245:H911–H919.
6. Mason JC, Curry FE, Michel CC. The effects of proteins upon the filtration coefficient of individually perfused frog mesenteric capillaries. Microvasc Res. 1977;13:185–202.[Medline] [Order article via Infotrieve]
7.
Huxley VH, Curry FE. Albumin modulation of
capillary permeability: test of an adsorption mechanism. Am
J Physiol. 1985;248:H264–H273.
8. Ohnishi T, Posner JB, Shapiro WR. Vasogenic brain edema induced by arachidonic acid: role of extracellular arachidonic acid in blood-brain barrier dysfunction. Neurosurgery. 1992;30:545–551.[Medline] [Order article via Infotrieve]
9. Unterberg A, Wahl M, Hammersen F, Baethmann A. Permeability and vasomotor response of cerebral vessels during exposure to arachidonic acid. Acta Neuropathol. 1987;73:209–219.
10. Kempski O, Villacara A, Spatz M, Dodson RF, Corn C, Merkel N, Bembry J. Cerebromicrovascular endothelial permeability: in-vitro studies. Acta Neuropathol. 1987;74:329–334.[Medline] [Order article via Infotrieve]
11. Villacara A, Spatz M, Dodson RF, Corn C, Bembry J. Effect of arachidonic acid on cultured cerebromicrovascular endothelium: permeability, lipid peroxidation and membrane "fluidity." Acta Neuropathol. 1989;78:310–316.[Medline] [Order article via Infotrieve]
12.
Hennig B, Shasby DM, Fulton AB, Spector AA. Exposure to
free fatty acid increases the transfer of albumin across
cultured endothelial monolayers.
Arteriosclerosis. 1984;4:489–497.
13. Worman HJ, Brasitus TA, Dudeja PK, Fozzard HA, Field M. Relationship between lipid fluidity and water permeability of bovine tracheal epithelial cell apical membranes. Biochemistry. 1986;25:1549–1555.[Medline] [Order article via Infotrieve]
14. Meddings JB. Lipid permeability of the intestinal microvillus membrane may be modulated by membrane fluidity in the rat. Biochim Biophys Acta. 1989;984:158–166.[Medline] [Order article via Infotrieve]
15. Thompson GA Jr. Lipid molecular species retailoring and membrane fluidity. Biochem Soc Trans. 1989;17:286–289.[Medline] [Order article via Infotrieve]
16. Smirnov SV, Beck R. Modulation of membrane fluidity in vascular smooth muscle cells: the role of phospholipase A2 (PLA2). Biophys J. 1998;74:A197. Abstract.
17. Beck R, Abbot SE, Aaronson PI, Smirnov SV. Effect of temperature and albumin on membrane fluidity in vascular smooth muscle and endothelial cells. J Physiol. 1998;507:71P. Abstract.
18.
Michel CC, Phillips ME, Turner MR. The effects of
native and modified bovine serum albumin on the permeability of
frog mesenteric capillaries. J Physiol.. 1985;360:333–346.
19. Jonas A, Weber G. Presence of arginine residues at the strong, hydrophobic anion binding sites of bovine serum albumin. Biochemistry. 1971;10:1335–1339.[Medline] [Order article via Infotrieve]
20. Lustig KD, Erb L, Landis DM, Hicks-Taylor CS, Zhang X, Sportiello MG, Weisman GA. Mechanisms by which extracellular ATP and UTP stimulate the release of prostacyclin from bovine pulmonary artery endothelial cells. Biochim Biophys Acta. 1992;1134:61–72.[Medline] [Order article via Infotrieve]
21. Nollert MU, Hall ER, Eskin SG, McIntire LV. The effect of shear stress on the uptake and metabolism of arachidonic acid by human endothelial cells. Biochim Biophys Acta. 1989;1005:72–78.[Medline] [Order article via Infotrieve]
22. Baisted DJ, Robinson BS, Vance DE. Albumin stimulates the release of lysophosphatidylcholine from cultured rat hepatocytes. Biochem J. 1988;253:693–701.[Medline] [Order article via Infotrieve]
23. Shinitzky M, Barenholz Y. Fluidity parameters of lipid regions determined by fluorescence polarization. Biochim Biophys Acta. 1978;515:367–394.[Medline] [Order article via Infotrieve]
24.
de Laat SW, van der Saag PT, Shinitzky M.
Microviscosity modulation during the cell cycle of neuroblastoma cells.
Proc Natl Acad Sci U S A.. 1977;74:4458–4461.
25.
Ashbrook JD, Spector AA, Santos EC, Fletcher JE. Long
chain fatty acid binding to human plasma albumin. J
Biol Chem. 1975;250:2333–2338.
26. Kragh-Hansen U. Molecular aspects of ligand binding to serum albumin. Pharmacol Rev. 1981;33:17–53.[Medline] [Order article via Infotrieve]
27.
Paulmichl M, Nasmith P, Hellmiss R, Reed K, Boyle WA,
Nerbonne JM, Clapham DE. Cloning and expression of a rat cardiac
delayed rectifier potassium channel. Proc Natl Acad Sci
U S A.. 1991;88:7892–7895.
28. Phillips ME, Turner MR. The effects of haemoglobin and myoglobin on the filtration coefficient of single frog mesenteric capillaries. J Physiol. 1981;320:39P–40P. Abstract.
29.
Lande MB, Donovan JM, Zeidel ML. The relationship
between membrane fluidity and permeabilities to water, solutes,
ammonia, and protons. J Gen Physiol. 1995;106:67–84.
30.
Ghinea N, Fixman A, Alexandru D, Popov D, Hasu M,
Ghitescu L, Eskenasy M, Simionescu M, Simionescu N. Identification of
albumin-binding proteins in capillary
endothelial cells. J Cell Biol. 1988;107:231–239.
31.
Schnitzer JE, Carley WW, Palade GE. Albumin
interacts specifically with a 60-kDa microvascular
endothelial glycoprotein. Proc Natl
Acad Sci U S A.. 1988;85:6773–6777.
32.
He P, Curry FE. Albumin modulation of capillary
permeability: role of endothelial cell
[Ca2+]i. Am J
Physiol. 1993;265:H74–H82.
33.
Fuentes E, Nadal A, Jacob R, McNaughton P. Actions of
serum and plasma albumin on intracellular
Ca2+ in human endothelial cells.
J Physiol. 1997;504:315–326. Abstract.
34.
Nadal A, Fuentes E, Pastor J, McNaughton PA. Plasma
albumin is a potent trigger of calcium signals and DNA
synthesis in astrocytes. Proc Natl Acad Sci U S A.. 1995;92:1426–1430.
35.
Tigyi G, Miledi R. Lysophosphatidates bound to serum
albumin activate membrane currents in
Xenopus oocytes and neurite retraction in PC12
pheochromocytoma cells. J Biol Chem. 1992;267:21360–21367.
36. Ramanadham S, Gross R, Turk J. Arachidonic acid induces an increase in the cytosolic calcium concentration in single pancreatic islet beta cells. Biochem Biophys Res Commun. 1992;184:647–653.[Medline] [Order article via Infotrieve]
37.
Oike M, Droogmans G, Nilius B. Mechanosensitive
Ca2+ transients in endothelial
cells from human umbilical vein. Proc Natl Acad Sci U S A.. 1994;91:2940–2944.
38. Kirber MT, Ordway RW, Clapp LH, Walsh JV Jr, Singer JJ. Both membrane stretch and fatty acids directly activate large conductance Ca2+-activated K+ channels in vascular smooth muscle cells. FEBS Lett. 1992;297:24–28.[Medline] [Order article via Infotrieve]
39. Smirnov SV, Knock GA, Aaronson PI. Effects of 5-lipoxygenase activating protein inhibitor MK886 on voltage-gated and Ca2+-activated K+ currents in rat arterial myocytes. Br J Pharmacol.. 1998;124:572–578.[Medline] [Order article via Infotrieve]
40.
Chen GF, Cheung DW. Characterization of
acetylcholine-induced membrane hyperpolarization in
endothelial cells. Circ Res. 1992;70:257–263.
41. Demirel E, Laskey RE, Purkerson S, van Breemen C. The passive calcium leak in cultured porcine aortic endothelial cells. Biochem Biophys Res Commun. 1993;191:1197–1203.[Medline] [Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  S. J. Liu Inhibition of L-type Ca2+ channel current and negative inotropy induced by arachidonic acid in adult rat ventricular myocytes Am J Physiol Cell Physiol, November 1, 2007; 293(5): C1594 - C1604. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Berrougui, M. Isabelle, M. Cloutier, G. Grenier, and A. Khalil Age-related impairment of HDL-mediated cholesterol efflux J. Lipid Res., February 1, 2007; 48(2): 328 - 336. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Evidence-based Colloid Use in the Critically Ill: American Thoracic Society Consensus Statement Am. J. Respir. Crit. Care Med., December 1, 2004; 170(11): 1247 - 1259. [Full Text] [PDF]
|
|
|
|
 |
|
 A. Larbi, N. Douziech, G. Dupuis, A. Khalil, H. Pelletier, K.-P. Guerard, and T. Fulop Jr Age-associated alterations in the recruitment of signal-transduction proteins to lipid rafts in human T lymphocytes J. Leukoc. Biol., February 1, 2004; 75(2): 373 - 381. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Fiorini, M. E. Ferretti, C. Biondi, B. Pavan, L. Lunghi, G. Paganetto, and L. Abelli 17{beta}-Estradiol Stimulates Arachidonate Release from Human Amnion-Like WISH Cells through a Rapid Mechanism Involving a Membrane Receptor Endocrinology, August 1, 2003; 144(8): 3359 - 3367. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W.-J. Zhang and B. Frei Albumin selectively inhibits TNF{alpha}-induced expression of vascular cell adhesion molecule-1 in human aortic endothelial cells Cardiovasc Res, September 1, 2002; 55(4): 820 - 829. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||