(Received for publication, March 12, 1997, and in revised form, May 21, 1997)
From the Biophysics Research Institute, Medical College of Wisconsin, Milwaukee, Wisconsin 53226
1-Substituted diazen-1-ium-1,2-diolates, a class
of nitric oxide (·NO) donor compounds that spontaneously release
·NO at different rates, were used to investigate the effect of ·NO release rate upon the oxidation of low density lipoprotein (LDL). All donor compounds conferred an inhibitory effect upon the
oxidation of LDL; however, the effect exhibited a biphasic dependence
upon the rate of ·NO release. The ·NO release rate that
maximally inhibited oxidation was dependent upon the rate of oxidation.
When LDL was rapidly oxidized by copper(II) sulfate, a faster release
rate was more effective. In contrast, when LDL was oxidized slowly by
2,2-azobis-2-amidinopropane hydrochloride, a slower release rate was
most effective. This biphasic relationship between ·NO release
rate and the duration of inhibition was also demonstrated when LDL
oxidation was initiated with
5-amino-3-(4-morpholinyl)-1,2,3-oxadiazolium, a peroxynitrite
generator. We conclude that the antioxidant ability of ·NO is
dependent not only upon the rate of its release from ·NO donors,
but also upon the rate of oxidation. This conclusion is supported by a
kinetic model of LDL oxidation in the presence of ·NO.
The chronic inhibition of nitric oxide synthase (NOS)1 accelerates atherosclerosis in cholesterol-fed rabbits (1, 2). Supplementation of L-arginine, the substrate for NOS, has been shown to decrease fatty streak formation in rabbits and correct endothelial dysfunction in hypercholesterolemic humans (3-5). These data suggest that basal production of ·NO from the vascular endothelium can suppress the changes associated with early atherosclerotic lesion formation. Moreover, augmentation of ·NO production can partially counteract the effects of elevated cholesterol levels.
The mechanisms by which ·NO affects atherosclerotic lesion formation are unclear; however, there are many possibilities. ·NO has been shown to inhibit inflammatory cell adhesion to the endothelium (6) and suppress smooth muscle cell proliferation (7). In later stages of atherosclerosis, the anti-thrombolytic properties of ·NO (8) may also be relevant.
It has frequently been proposed that the initiation of the atherosclerotic process is oxidative in nature. Free radical-mediated oxidative damage to both the vascular endothelium and low density lipoprotein (LDL) has been proposed to be a vital component of the atherosclerotic mechanism (9, 10). We have recently shown that ·NO suppresses the oxidative modification of LDL (11-13) and inhibits the toxicity of oxidized LDL to endothelial cells (14). We have proposed that the way in which ·NO inhibits oxidative processes is by scavenging propagatory lipid peroxyl radicals, thus suppressing the lipid peroxidation chain reaction (11-14). This mechanism will generate lipid/NO adducts that have recently been identified (15, 16).
Many biological effects of ·NO appear to be related to the rate of ·NO formation (15, 17, 18). Bolus delivery of high concentrations of ·NO may result in limited LDL oxidation that can be augmented by Cu2+ (12). Slow release of ·NO from the donor spermine NONOate, however, suppresses LDL oxidation at concentrations of 2-10 µM (12).
In this investigation, we have examined the effect of ·NO
release rate on the oxidative modification of LDL using four
structurally related ·NO donor compounds (NONOates) with a wide
range of ·NO release rates (see Fig.
1 for structures). We conclude that an
optimal rate of ·NO release is required for maximal suppression
of LDL oxidation, and that this optimal rate depends upon the rate and
the mechanism of LDL oxidation. This conclusion is supported by a
kinetic model of LDL oxidation in the presence of ·NO.
Copper(II) sulfate was purchased from Fisher.
2,2-Azobis-2-amidinopropane hydrochloride (ABAP) was obtained from
Polysciences, Inc.
(Z)-1-{N-Methyl-N-[6-(N-methylammoniohexyl)amino]} diazen-1-ium-1,2-diolate (MNN),
(Z)-1-[N-(3-ammoniopropyl)-N-(n-propyl)amino]}diazen-1-ium-1,2-diolate (PNN),
(Z)-1-{N-[3-aminopropyl]-N-[4-(3-aminopropylammonio)butyl]-amino}diazen-1-ium-1,2-diolate (SNN),
(Z)-1-[N-(2-aminoethyl)-N-(2-ammonioethyl)amino]diazen-1-ium-1,2-diolate (DNN), and 5-amino-3-(4-morpholinyl)-1,2,3-oxadiazolium (SIN-1) were purchased from Cayman Chemical Co. Spermine,
1,1,3,3-tetramethoxypropane, and thiobarbituric acid were purchased
from Sigma. Diethylenetriaminepentaacetic acid (DTPA) was obtained from
Aldrich. LDL was isolated from human plasma as described previously
(19). All concentrations of LDL are given in terms of LDL protein as
measured by the Lowry assay (20).
LDL was incubated in phosphate-buffered saline (PBS) at 37 °C, either alone or in the presence of MNN, PNN, SNN, or DNN. Oxidation was initiated with copper(II) sulfate, ABAP, or SIN-1. Addition of decayed NONOate had no effect upon oxidation at the concentrations used in these experiments. Oxidation was monitored using the following assays.
Thiobarbituric Acid-reactive Substances (TBARS) MeasurementsLDL (10-20 µg) was incubated with thiobarbituric acid (0.5%, w/v, in H2SO4, 50 mM) for 30 min at 100 °C. Samples were centrifuged for 5 min, and the difference in the absorbance was determined between 532 and 580 nm. TBARS concentration was calculated as malondialdehyde (MDA) equivalents using a MDA standard curve. MDA was generated by the acid hydrolysis of 1,1,3,3-tetramethoxypropane. SNN (500 µM) had no effect upon the standard curve.
Conjugated Diene MeasurementsLDL (200 µg/ml), incubated in PBS at 37 °C, was continuously monitored at 234 nm to detect the formation of conjugated dienes as described previously (21).
Apolipoprotein B (ApoB) ModificationOxidative modification of LDL results in an increase in net negative charge that can be detected using agarose electrophoresis (Paragon LIPO lipoprotein electrophoresis kit. Relative electrophoretic mobility (REM) was calculated as the ratio of the electophoretic mobilities of the samples to that of native LDL (22).
-Tocopherol was extracted
into heptane and assayed by high performance liquid chromatography as
described previously (23).
Kinetic simulations were performed using software written by Frank Neese, Facultät für Biologie, Universität Konstanz, Konstanz, Germany. Integration of the equations shown under "Appendix," using the rate constants defined in Table II, was performed using the Stiff Euler algorithm.
|
Each ·NO donor compound was analyzed
under our experimental conditions to determine the kinetics of
·NO release. The decay rate was determined by monitoring the
decrease in absorbance at 250 nm in PBS at 37 °C (24). The decay of
each compound followed first order kinetics with rates of MNN > PNN > SNN DNN in agreement with the known rates of decay.
The presence of both LDL and copper(II) sulfate had no significant upon
the kinetics of decay (data not shown). The calculated half-life and kobs are listed for each donor compound are
listed in Table I. Both ESR spin-trapping
(25) and electrochemical (26) techniques were used to verify that
·NO is released during the breakdown of each compound (data not shown).
|
LDL (200 µg/ml) was incubated with copper(II) sulfate
(20 µM) in PBS at 37 °C, in the presence or absence of
each NONOate compound (10 µM). Oxidation was assayed by
monitoring the formation of TBARS. In the presence of copper(II)
sulfate, TBARS formation increased over time reaching a maximum of
80 ± 1.4 nmol of MDA/mg of LDL protein (Fig.
2). Addition of each NONOate compound led to an inhibition of TBARS formation, the extent of which was dependent on the compound used. MNN, with a t1/2 of 35.6 ± 1.8 s, had little effect on the oxidation of LDL, whereas PNN,
with a t1/2 of 14.3 ± 0.2 min, inhibited LDL
oxidation for approximately 3 h. A slower rate of ·NO
release (i.e. SNN, t1/2 = 72.4 ± 2.7 min), inhibited LDL oxidation for approximately 2 h, while
DNN, with a t1/2 of 21.9 ± 0.7 h,
inhibited oxidation for 1 h. This indicates that there is an
optimal rate of ·NO production for maximal suppression of
Cu2+-dependent LDL oxidation.
The oxidation of LDL has previously been shown to cause an increase in
the net negative charge of the LDL particle (22, 27). This is thought
to result from the modification of lysine residues of the ApoB protein
by aldehydes formed during lipid peroxidation. When LDL (200 µg/ml)
was incubated with copper(II) sulfate (20 µM) for 4 h in PBS at 37 °C, an increase in the electrophoretic mobility of
the LDL particle was observed (Fig.
3A). The presence of each
NONOate compound (10 µM) had a time-dependent
inhibitory effect upon this change in mobility (Fig. 3B).
Both the fastest releasing (MNN) and slowest releasing (DNN) ·NO
donors (see Table I) were much less effective inhibitors than PNN and
SNN.
The formation of conjugated dienes is a sensitive and direct
measurement of lipid peroxidation. LDL (200 µg/ml) was incubated in
PBS at 37 °C in the presence of copper(II) sulfate (20 µM). After an initial lag period, there was a rapid
increase in the rate of conjugated diene formation. As shown in Fig.
4A, the length of the lag time
was calculated by determining the time at which the absorbance reached
half of the maximum absorbance. As previously demonstrated (12),
addition of SNN (4 µM) led to an increase in the length
of the lag time by approximately 2.5 h (Fig. 4A). The
concentration dependence of the inhibition of conjugated diene formation for each of the four NONOates is shown in Fig. 4B.
The presence of all NONOate compounds led to an increase in the lag time of LDL oxidation. The extent of inhibition exhibited a biphasic relationship to the NONOate decay rate (see Table I).
The Effect of ·NO Release Rate on ABAP-mediated LDL Oxidation
ABAP, a peroxyl radical generator, has been shown to
oxidize LDL by initiating lipid peroxidation (28). Addition of ABAP (1 mM) to LDL (200 µg/ml protein) led to a
time-dependent increase in TBARS formation over a period of
10 h after a 2-h lag period (Fig.
5). MNN had no effect on
ABAP-dependent LDL oxidation, PNN increased the lag period
by 2 h, and both SNN and DNN inhibited LDL oxidation over the full
10-h period (Fig. 5). The ·NO donor compounds also inhibited the
change in REM of LDL during ABAP-dependent oxidation as
shown in Fig. 6. The inhibitory effect was again of the order SNN > DNN > PNN > MNN.
The rate of ABAP-dependent LDL oxidation can be sensitively
measured by following the decrease in -tocopherol levels. In the
absence of ·NO donors,
-tocopherol was consumed over a 2-h
period at an approximately linear rate (Fig.
7). MNN had no effect on
-tocopherol
decomposition, in agreement with the observation that this compound did
not inhibit lipid peroxidation (Fig. 5) and protein modification (Fig.
6). PNN inhibited
-tocopherol depletion for approximately 1 h
and SNN for approximately 2 h (Fig. 7). Surprisingly, although DNN inhibited LDL oxidation (Figs. 5 and 6), it had little inhibitory effect on
-tocopherol depletion (Fig. 7), indicating that the sparing of
-tocopherol is not an absolute requirement for the inhibitory effect of ·NO.
The Effect of ·NO Release Rate on SIN-1-mediated LDL Oxidation
SIN-1, which releases ·NO and O2 to
generate peroxynitrite (29), has been shown to oxidize LDL (30) and
cause a stoichiometric conversion of
-tocopherol to
-tocopherol
quinone (24, 31, 32). To determine the effect of ·NO release on
SIN-1-dependent LDL oxidation, LDL was incubated with SIN-1
(1 mM) in the presence and absence of the four ·NO
donor compounds. As shown in Fig.
8A, addition of SIN-1 resulted in an approximately linear depletion of
-tocopherol over a 10-min period, with a concomitant formation of
-tocopherol quinone (Fig. 8B). MNN inhibited
-tocopherol depletion for 7 min. PNN
was the most effective inhibitor of
-tocopherol oxidation,
inhibiting depletion for 15-20 min. SNN inhibited
-tocopherol
depletion for 5 min, and DNN was not inhibitory. After the inhibitory
period,
-tocopherol was predominantly oxidized to
-tocopherol
quinone in the presence of all NONOate compounds (Fig.
8B).
Several effects of ·NO have been shown to be dependent on the rate of ·NO delivery. For example, ·NO is more toxic to Chinese hamster ovary cells when slowly infused through a sylastic membrane than when added as a bolus (18). Previous work with NONOate compounds has shown that slower rates of ·NO release are much more effective at inhibiting DNA synthesis and the proliferation of smooth muscle cells (33). ·NO has also been shown to have both oxidant and antioxidant effects, depending on the rate of formation (15). If ·NO is either added to LDL in bolus amounts (12) or continuously bubbled through a solution of LDL (34), oxidation may occur. ·NO may also cause LDL oxidation in the presence of superoxide (30). If ·NO is slowly released from donor compounds, it is an efficient antioxidant, preventing LDL oxidation.
The mechanism by which ·NO inhibits lipid peroxidation is thought to be the scavenging of lipid radicals that propagate the peroxidative chain reaction. The reaction between the lipid peroxyl radical and ·NO creates LOONO, an organic peroxynitrite. It has been suggested that such compounds decay via homolytic cleavage to yield an alkoxyl radical (LO·) and ·NO2 (35). If this were true of LOONO in LDL, the action of ·NO may be expected to be pro-oxidant, as both LO· and ·NO2 are able to oxidize LDL lipid. It is likely, therefore, that LOONO decays by internal re-arrangement to LONO2, a stable organic nitrite. Alternatively, homolysis of LOONO could be followed by rapid re-association of LO· and ·NO2, within the solvent cage, to form LONO2. The mechanistic details of how ·NO inhibits lipid peroxidation have yet to be clearly established.
Here we have investigated the effect of different rates of ·NO
release on copper(II) sulfate, ABAP and
peroxynitrite-dependent LDL oxidation. Oxidation of LDL by
copper(II) sulfate was inhibited most effectively by PNN and SNN, the
order of effectiveness being SNN PNN
DNN > MNN
(Figs. 2, 3, 4). This indicates a biphasic relationship between the rate
of ·NO production and the efficacy of ·NO as an
antioxidant. This relationship was observed at all concentrations of
the NONOate compounds tested (Fig. 4). It is interesting to note that
when concentrations as high as 12 µM of MNN and DNN were
used, they were still less effective than 4 µM SNN.
When oxidation was initiated with ABAP, the order of antioxidant
efficiency was SNN>DNNPNN>MNN (Figs. 5 and 6), again showing a
biphasic relationship between the rate of ·NO release and its
antioxidant efficacy. A comparison between Figs. 2 and 5 indicates that
copper(II) sulfate-dependent oxidation is a more rapid
process than ABAP-dependent oxidation. When oxidation occurred at a slow rate (i.e. with ABAP), a slower rate of
·NO release was most effective at inhibiting LDL oxidation. In contrast, when oxidation occurred more rapidly (i.e. with
copper(II) sulfate), a faster ·NO release rate was most
effective. This suggests that there is an optimal rate of ·NO
production for maximal suppression of LDL oxidation, and that this
optimal rate is dependent upon the rate of oxidation.
The depletion of -tocopherol is a sensitive
measurement of oxidation. ABAP-dependent depletion of
-tocopherol was inhibited by SNN and PNN but not by MNN and DNN
(Fig. 7). Although DNN did not inhibit
-tocopherol depletion, it was
a potent inhibitor of LDL oxidation (Figs. 5 and 6). This suggests that
the level of ·NO generated by DNN is unable to compete with
-tocopherol as a peroxyl radical scavenger but is still sufficiently
high to inhibit LDL oxidation after
-tocopherol has been
consumed.
The effect of
·NO on ABAP-dependent -tocopherol depletion can
be modeled by the following set of reactions (see Table
II for k values).
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
The set of differential equations generated from Reactions 1-7 (see
"Appendix") can be solved using numerical methods to give the time
course for each species in the reaction scheme. Fig. 9 shows a plot of the time required for
total -tocopherol depletion as a function of
k2 (the rate constant for NONOate decay). This model agrees with the experimental data, if k7
is set to 2.5 s
1, and shows a biphasic dependence on the
rate of ·NO formation. The rate of ·NO formation at which
maximal inhibition of
-tocopherol depletion is observed
(k2max) depends strongly on the
rate of oxidation (k1). The model predicts a
linear relationship between k1 and the optimal
value for k2max (Fig.
10). As the rate of oxidation
increases, the rate at which ·NO is most effective at inhibiting
oxidation also increases. Therefore, one can predict the extent of the
antioxidant effect of any ·NO donor compound by determining the
rate of oxidation and the rate of ·NO release from the donor.
Many spontaneous ·NO donor compounds, such as NONOate and NOR
(39) compounds, are commercially available with easily determined rates
of release.
An interesting prediction of this model is that the maximal inhibition
of lipid hydroperoxide formation will occur at a
k2 lower than
k2max (Fig. 9). This provides a
kinetic explanation for the inhibition of lipid peroxidation by DNN
(Fig. 5) when DNN was unable to inhibit -tocopherol depletion (Fig.
7) during ABAP-mediated LDL oxidation.
SIN-1 decomposes to form stoichiometric amounts of
·NO and superoxide in the presence of oxygen (29), effectively
making it a peroxynitrite donor. Fig. 8 shows that SIN-1 oxidized
-tocopherol to
-tocopherol quinone, the two-electron oxidation
product (31-32). ·NO inhibited SIN-1-dependent
-tocopherol oxidation, and this inhibition had a biphasic dependence
on the rate of ·NO release. Our data are consistent with a
direct reaction between ·NO and peroxynitrite. In support of
this, ·NO was shown to react with peroxynitrite, with a rate
constant of 9.1 × 104 M
1
s
1 (40).
This investigation has shown that the release of ·NO from all NONOate compounds conferred an inhibitory effect upon the oxidation of LDL. The extent of the antioxidant effect, however, exhibited a biphasic dependence upon the rate of ·NO release. The rate of ·NO formation that was most effective at inhibiting LDL oxidation varied with the rate of oxidation. This indicates that there is an optimal rate of ·NO production for maximal suppression of LDL oxidation, and that this optimal rate is dependent upon the rate of oxidation. We conclude that the antioxidant capability of ·NO is critically dependent not only on the rate of its release from ·NO donors, but also upon the relative rate of lipid oxidation. The slow generation of ·NO by the vascular endothelium may represent a continuous source of antioxidant, playing an integral role in suppressing oxidative reactions within the vasculature. Impairment of ·NO generation or acceleration of the rate of oxidation may be a critical component in both the early stages and the development of atherosclerosis.
The following differential equations were constructed from
Reactions 1-7. This set of differential equations is not analytically soluble. Therefore we have used numerical methods to calculate the
kinetics of each species. Initial concentrations were: NONOate, 10 µM; TH, 0.9 µM.
![]() |
(Eq. 1) |
![]() |
(Eq. 2) |
![]() |
(Eq. 3) |
![]() |
(Eq. 4) |
![]() |
(Eq. 5) |
![]() |
(Eq. 6) |
![]() |
(Eq. 7) |
![]() |
(Eq. 8) |
![]() |
(Eq. 9) |