Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Nitric oxide (NO) and closely related molecules serve a variety of functions in the cardiovascular system and account for most of the endothelium-dependent vasodilation (Ignarro, 1990; Moncada et al., 1991). In addition to controlling vascular tone, NO also exhibits potent antiatherogenic actions. These include the inhibition of platelet function (Radomski et al., 1990) the control of smooth muscle cell proliferation (Garg and Hasid, 1989; Nakaki et al., 1990), and the inhibition of monocyte adhesion and chemoattraction (Bath et al., 1991). The latter has recently been investigated in more detail (De Caterina et al., 1995; Khan et al., 1996; Tsao et al., 1996) and attributed to the inhibition of endothelial activation, i.e., the concerted transcriptional activation of genes involved in leukocyte recruitment to the vascular wall (Gimbrone, 1995). In view of these properties, a possible future use of NO donors as antiatherosclerotic drugs may be envisaged. In a previous study we found that the "spontaneous" NO donors S-nitrosoglutathione (GSNO), 3-morpholinosydnonymine, and sodium nitroprusside inhibited endothelial activation, in particular the surface expression of vascular cell adhesion molecule-1 (VCAM-1) and E-selectin, in human endothelial cells stimulated with inflammatory cytokines. In this model GSNO was the most effective of the three tested compounds. We now examined the capacity of two novel cysteine-containing NO donors, SP/W 3672, a fast spontaneous NO releaser, and its prodrug SP/W 5186, which liberates NO after hydrolysis or bioactivation to SP/W 3672, with regard to their ability to inhibit monocyte adhesion to cultured endothelial cells and leukocyte adhesion molecule expression. The purpose of this study was to gain a better understanding of the structural prerequisites of NO donors with regard to their antiatherogenic potency to facilitate the future design of more potent candidate drugs for possible treatment of atherosclerosis.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Test Compounds and Reagents. N-(3-Nitratopivaloyl)-L-cysteine ethylester (SP/W 3672), 3-nitratopivalic acid (SP/W 4744), N-(3-nitratopivaloyl)-S-(N'-acetylglycyl)-L-cysteinethylester (SP/W 5186), N-(3-hydroxypivaloyl)-S-(N'-acetylglycyl) L-cysteinethylester (SP/W 6373), GSNO, nitroglycerin, and isosorbide dinitrate were obtained from Schwarz Pharma (Monheim, Germany). Several of these compounds have been previously characterized on other types of endpoints (Zanzinger et al., 1994; Wang et al., 1996; Liu et al., 1998). Spermine NONOate was purchased by Calbiochem-Novabiochem (La Jolla, CA) and dissolved in 0.1 N NaOH. Compounds 3672, 4744, 5186, 6373, and GSNO were first dissolved in 99% ethanol up to 0.1 mol/l and subsequently diluted in water. GSNO solutions were stabilized by addition of 0.1% HCl. The structural formulae and molecular weights of the various compounds used are depicted in Fig. 1. IL-1 was obtained from Hoffmann-La Roche (Nutley, NJ). Interferon- (IFN-) was purchased from Genzyme (Cambridge, MA). The monoclonal antibodies (Abs) to VCAM-1 (Ab E1/6), E-selectin (Ab H18/7), and intercellular adhesion molecule-1 (ICAM-1) (Ab HU 5/3) were kindly provided by M. A. Gimbrone, Jr. (Brigham and Women's Hospital, Boston, MA). The monoclonal antibody to major histocompatibility complex class II (MHC II) (HLA-DR, Ab I-2/IA antigen) was kindly provided by Arnold Freedman (Dana Farber Cancer Institute, Boston, MA). Ab E1/1 (from M. A. Gimbrone Jr.), directed against a constitutive and noncytokine-inducible endothelial antigen, was used as a control for the specificity of inhibition by the various compounds.

View larger version (33K): [in this window] [in a new window]   Fig. 1.   Name, structural formula, and molecular weight (MW) of compounds used in the present study.

Cell Cultures. Human saphenous vein endothelial cells (HSVECs) were harvested enzymatically with type II 0.1% collagenase as described (Libby et al., 1986), and maintained in M199 (Mascia Brunelli, Milan, Italy), containing HEPES (25 mM), heparin (50 U/ml), retinal growth factor from bovine eyes (50 µg/ml), L-glutamine (2 mM), antibiotics, and 5% fetal calf serum (Life Technologies, Milan, Italy). Once grown to confluence, cells were replated on 1.5% gelatin-coated flasks at 20,000 cells/cm². HSVECs isolated by this technique formed a confluent monolayer of polygonal cells and expressed von Willebrand factor as determined by their content of immunoreactive protein. Cell number was assessed after trypsinization in a Burker chamber, and cell viability assessed by Trypan blue exclusion.

Detection of Cell Surface Molecules. The expression of adhesion molecules was determined by cell surface enzyme immunoassay, using mouse anti-human monoclonal antibodies against various surface molecules. The enzyme immunoassay was carried out by first incubating monolayers with saturating concentrations of the specific primary monoclonal antibody against the target molecule for 2 h, followed by biotinylated goat anti-mouse IgG (Delta Biologicals, Roma, Italy) for 1 h, and finally streptavidine-alkaline phosphatase (Sigma, St. Louis, MO). Endothelial monolayers were washed three times between each incubation step, and their integrity was monitored by phase-contrast microscopy after each washing. The surface expression of each protein was quantified spectrophotometrically reading the optical density of the wells at 410 nm 45 to 90 min after the addition of a cromogenic substrate (p-nitrophenylphospate; Sigma), as described (De Caterina et al., 1994).

Assessment of Total Protein Synthesis. The total cell-associated protein content was assessed by the amidoblack assay, as previously described (Libby et al., 1988). Briefly, HSVEC monolayers were fixed with 4% p-formaldehyde in a 0.1 mol/l acetate buffer, pH 3.1, and incubated in amidoblack B (Sigma), to stain cell proteins. Unbound dye was washed away by repeated rinsing with PBS. Dye uptake was determined spectrophotometrically at 620 nm.

Monocyte Adhesion Assay. Monocytoid U937 cells were obtained from American Type Culture Collection (Rockville, MD) and grown in RPMI 1640 (Life Technologies), containing 1 mM sodium-pyruvate, nonessential amino acids, antibiotics (all from Mascia Brunelli), 1 µg/ml gentamycin, and 10% fetal calf serum. For the adhesion assay, HSVECs were grown to confluence on six-well plates, and then treated overnight with 1 ng/ml interleukin-1 (IL-1) for induction of VCAM-1 expression, in the presence or absence of NO donors. For control, some monolayers were treated with the anti-VCAM-1 monoclonal antibody. One milliliter of a suspension of U937 cells, concentrated by centrifugation to 1 × 10⁶ cell/ml, was added to each monolayer under rotating conditions (63 rpm) at room temperature. After 15 min, nonadhering cells were removed by gentle rinsing with M199, and the monolayer fixed with p-formaldehyde. The number of adherent cells was determined by counting six different fields using an ocular grid and an overall magnification of 100× (0.0144 mm²/field). Fields for counting adherent cells were chosen randomly located at half-radius from the center of the well. A minimum of eight readings was performed for each well.

Northern Analysis. HSVEC RNA was isolated with the RNAzol kit (Tel-Test, Friendswood, TX). Total RNA (20 µg) was separated on 1% agarose-formaldehyde gel, transferred to a nylon membrane (Hybond-N; Amersham Life Sciences, Milan, Italy), and immobilized by illumination with short-wave ultraviolet light. Specific cDNA probes were labeled by random hexanucleotide priming. All blots were hybridized at 42°C overnight and washed (0.2× standard saline citrate, 0.1% SDS at 63°C) before autoradiography at 80°C for 24 to 72 h. Densitometric analysis on the exposed films was performed with the aid of the NIH Image software. To assess whether the NO donors tested reduced adhesion molecule expression by decreasing mRNA half-life, Northern analysis was carried out in the presence of the transcription blocker actinomycin D (Sigma). Under these conditions, the rate of mRNA disappearance as a function of time is only due to mRNA instability. Actinomycin D was added at 3 µg/ml 4 h after IL-1, when peak VCAM-1 transcription had occurred; NO donors were added 30 min later. The amount of VCAM-1 mRNA (by densitometry) was compared with the amount of VCAM-1 mRNA at 4 h after IL-1 stimulation (relative intensity) and plotted as a logarithmic function of time (h).

Measurement of In Vitro NO Release from NO Donors. The kinetics of NO release by the various NO donors tested in this study was assessed by measurement of the chemiluminescence obtained in the reaction of NO with ozone, using a CLD 780 TR Apparatus (Eco Physics, Heidelberg, Germany). The setup was calibrated each day with a standard gaseous mixture of NO and nitrogen. NO donors were incubated at 37°C in a reaction chamber (total volume 2 ml) under anaerobic conditions, using nitrogen as a vehicle. The reaction chamber and gas lines were designed such that the NO generated in the reaction chamber reached the chemiluminescence detector within 5 s. Stock solutions of the NO donors (1 mM) in formamide (in citric acid, pH 2, for GSNO) were added to a stirred 100 mM PBS buffer solution, pH 7.4, by injection through a septum. NO release was expressed in picomoles of NO per time unit (30 min) from 2 ml of a 0.1 mM solution of each donor. Measurements were performed in the absence and presence of thiols.

Statistics. Multiple comparisons were performed by one-way ANOVA and individual differences tested by the Fisher's protected least-significant difference test after the demonstration of significant intergroup differences by ANOVA.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Inhibition of IL-1-Stimulated Cell Adhesion by NO Donors. Treatment of HSVEC monolayers with IL-1 induced a dramatic increase in U937 monocytoid cell adhesion (Fig. 2). Nitroglycerin and isosorbide dinitrate (25-100 µM) did not induce any significant change in this phenomenon (data not shown). All other compounds induced a significant decrease in cell adhesion, with a potency ranking between 65% inhibition for SP/W 5186 and 26% inhibition for SP/W 4744 (Fig. 2). Intermediate results were obtained for SP/W 6373, GSNO (Fig. 2), and spermine NONOate (data not shown). SP/W 3672 produced 62% inhibition of monocytoid cell adhesion. None of the test compounds at the concentrations used in this study resulted in any inhibition of cell number, viability total protein synthesis, or E1/1 antigen (data not shown), indicating the specificity of action of the NO donors on endothelial activation.

View larger version (83K): [in this window] [in a new window]   Fig. 2.   Inhibition of U937 cell adhesion by NO donors. HSVEC monolayers were grown to confluence and stimulated overnight with 1 ng/ml IL-1 in the presence of 150 µM the indicated NO donors or the NO-lacking control compound SP/W 6373, respectively. SP/W 5186 and GSNO inhibited IL-1-stimulated U937 cell adhesion significantly (P < .05 versus stimulated control), whereas the degree of inhibition seen with SP/W 4744 and 6373 was not significant and considered unspecific. The depicted results are representative of three individual experiments.

Comparative Potency of NO Donors in Inhibiting IL-1-Stimulated Adhesion Molecule Expression in HSVECs. The effect of the NO donors on the expression of VCAM-1, ICAM-1, and E-selectin was investigated at three different concentrations (25, 50, and 100 µM). The various compounds tested were found to exhibit the same rank order of potency as for the inhibition of monocytoid cell adhesion, with nitroglycerin and isosorbide dinitrate being ineffective (data not shown); SP/W 5186 > spermine NONOate > GSNO = SP/W 6373 = SP/W 4744 (Fig. 3). SP/W 3672 appeared equieffective to SP/W 5186 [at 100 µM, inhibition of VCAM-1 was 81 ± 5% for SP/W 5186 and 71 ± 6% for SP/W 3672 (nonsignificant difference)] (Fig. 3). At 200 µM GSNO and spermine NONOate were roughly equipotent, and inhibited VCAM-1 expression significantly more than SP/W 4744 and SP/W 6373 (data not shown). Estimated IC₅₀ values for VCAM-1 inhibition were >400 µM for SP/W 4744, 200 µM for GSNO, 80 µM for SP/W 3672, and 50 µM for SP/W 5186. 

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Inhibitory profile of the tested NO donors on VCAM-1, ICAM-1, and E-selectin expression. HSVECs were grown to confluence in 96-well plates and adhesion molecule surface expression was measured by enzyme immunoassay, as detailed under Materials and Methods. Inhibition of VCAM-1 expression was evaluated after cell stimulation with IL-1 (1 ng/ml) in the absence and presence of various NO donors (25, 50, and 100 µM). Stimulation was carried on for 16 h with VCAM-1 and ICAM-1 and for 6 h with E-selectin. Each point represents the mean of eight independent replicates. Note that SP/W 3672 and SP/W 5186 are considerably more potent in inhibiting adhesion molecule expression than the other compounds. Inhibition compared with stimulated control at 100 µM is significantly different in the following rank order: SP/W 5186 = SP/W 3672 > spermine NONOate > GSNO = SP/W 6373 = SP/W 4744.

Different Potency of Various NO Donors in Inhibiting IFN--Stimulated MHC-II Expression in HSVECs. The concentration-response relationship for the various NO donors was also investigated with regard to inhibition of MHC II molecule expression. Again, the compounds tested ranked in the same order of potency as in the adhesion molecule assays (Fig. 4). Inhibition by SP/W 5186 and SP/W 3672 was greatest, spermine NONOate intermediate, GSNO slightly less, and SP/W 4744 and 6373 the weakest. These differences in potency were statistically significant (P < .05).

View larger version (18K): [in this window] [in a new window]   Fig. 4.   MHC II inhibitory activity profile of the tested NO donors. HSVECs were grown to confluence in 96-well plates and MHC II surface expression was measured by enzyme immunoassay, as detailed under Materials and Methods. Inhibition of MHC II expression was evaluated 48 h after stimulation with 100 U/ml IFN- in the presence of three different NO donor concentrations. Each point represents the mean of eight independent replicates. Note that SP/W 3672 and SP/W 5186 are by far more potent in terms of MHC II inhibitory activity than any of the other compounds.

SP/W 5186 Is More Potent Than GSNO in Inhibiting IL-1-Stimulated VCAM-1 mRNA. SP/W 5186 strongly reduced the steady-state level of VCAM-1 mRNA at Northern analysis after IL-1-stimulation and this inhibition was quantitatively greater than that observed with GSNO (Fig. 5), indicating less VCAM-1 mRNA accumulated after exposure of cells to IL-1 in the presence of SP/W 5186 compared with GSNO.

View larger version (49K): [in this window] [in a new window]   Fig. 5.   Northern analysis of IL-1-stimulated VCAM-1 mRNA levels in the absence and presence of 100 µM SP/W 5186 and GSNO. Top, ethidium bromide staining of the membrane for 28S and 18S rRNA. The blot is representative of three identical experiments. Note that the inhibitory activity of SP/W 5186 is greater than that of GSNO (85 versus 60% inhibition by densitometric analysis) at comparable concentration. Densitometry analysis accounted for minor differences in total RNA loading.

S/PW 5186 Inhibits VCAM-1 Transcription. To determine whether the strong inhibition of VCAM-1 mRNA levels was due to inhibition of transcription or acceleration of mRNA degradation, a time course study at Northern analysis was carried out in the presence of the transcription inhibitor actinomycin D, added at the peak of VCAM-1 mRNA transcription as determined in previous experiments (De Caterina et al., 1994, 1995). No evidence for a post-transcriptional activity of the compound was obtained because plots of the densitometric readings of the Northern autoradiography after actinomycin D in the presence and absence of SP/W 5186 were practically superimposable (Fig. 6). Hence, the effect of SP/W 5186 on mRNA levels was attributed to an inhibition of transcription. Identical results were obtained with SP/W 3672 in two other separate experiments (data not shown).

View larger version (26K): [in this window] [in a new window]   Fig. 6.   Determination of IL-1 induced-VCAM-1 mRNA half-life at Northern analysis in the presence and absence of SP/W 5186. Actinomycin D (3 µg/ml) and SP/W 5186 (100 µM) were added 4 and 4.5 h, respectively, after IL-1 (1 ng/ml) stimulation (left). The amount of VCAM-1 mRNA recovered at different time points in the presence or absence of SP/W 5186 was compared by densitometry and plotted on a logarithmic scale as a function of time (right). The blot is representative of two identical experiments. No difference in the rate of mRNA degradation was observed, indicating that SP/W 5186 does not act by accelerating VCAM-1 mRNA degradation.

Comparison of Thiol-Dependent and Spontaneous NO Release of Different NO Donors. To gain further insight into the possible reasons for the superiority of SP/W 5186 and 3672 over other compounds to inhibit endothelial activation, we assessed the capability of the most relevant NO donors used in this study to generate NO, in the absence and presence of thiols (Table 1). Although SP/W 3672 and GSNO released comparable amounts of NO in the presence of thiols, SP/W 3672 was at least 4 times more potent than GSNO in their absence. Conversely, SP/W 5186, which was equipotent to SP/W 3672 and thus more potent than GSNO as an inhibitor of endothelial activation, was a relatively poor NO donor, both in the absence and presence of thiols (Table 1). However, the potency of SP/W 5186 can be explained by the fact that this compound is a prodrug for SP/W 3672, i.e., it hydrolyses to form a spontaneous NO donor with free sulfhydryl group at physiological pH (Knüttel et al., 1996).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The results of the present study demonstrate potent inhibitory effects of two novel NO donors, SP/W 5186 and 3672, on monocytoid cell adhesion, the expression of endothelial leukocyte adhesion molecules, their messenger RNA, and on MHC II. Their mechanism of inhibition appeared to be consistent with that previously reported for GSNO (De Caterina et al., 1995), i.e., decreased gene transcription, but their potency was found to be considerably higher. The difference in inhibitory potency may be explained by the ability of both SP/W compounds to generate NO over prolonged periods of time, enabling NO to interfere with gene transcription at times critical for their expression of specific markers of endothelial cell activation.

Antiatherogenic effects are a new and interesting feature of the overall biological action profile of NO donors. Previous work from our group (De Caterina et al., 1995), as well as that of others (Khan et al., 1996; Tsao et al., 1996) has unmasked the importance of NO as an inhibitor of endothelial cell activation. This property differs from other previously known cardiovascular effects of NO, including vasodilatation and inhibition of platelet aggregation, in that it does not require soluble guanylate cyclase stimulation. Consequently, effects of NO on adhesion molecule expression are not mimicked by 8-bromo- or dibutyryl-cGMP (De Caterina et al., 1995) but may involve the scavenging of superoxide anions by NO, leading to a reduced generation of hydrogen peroxide and reduced activation of the transcription factor nuclear factor-B (Schreck et al., 1992; De Caterina et al., 1995). Alternatively, a modification of gene transcription by S-nitrosylation of nuclear transcription factors (Tabuchi et al., 1994; DelaTorre et al., 1997) cannot be excluded.

The present study attempts to address systematically the interference of several NO-generating compounds with the expression of adhesion molecules on endothelial cells. Classical nitrates, such as nitroglycerin and isosorbide dinitrate were totally devoid of any inhibitory activity at any concentration tested. Likewise, SP/W 4744, a simple nitrated fatty acid that represents the backbone of the two cysteine-containing NO donors without sharing their spontaneous NO-liberating properties, also lacked inhibitory activity. The minor inhibitory activity of monocytoid cell adhesion that occurred with both SP/W 4744 and SP/W 6373 at high concentrations appeared to be unrelated to the effect on adhesion molecule expression. Inhibition by NO donors was maximal for VCAM-1 and MHC II antigen, less for E-selectin, and minimal for ICAM-1. This parallels earlier observations also with different types of inhibitors, such as n-3 fatty acids (De Caterina et al., 1994). The relative resistance of ICAM-1 to inhibition by NO donors or n-3 fatty acids may, at least in part, reflect the relatively high degree of constitutive expression of this molecule in the absence of cytokines (Marlin and Springer, 1987). The inhibition of VCAM-1 expression was attributed to inhibition of gene transcription because the slopes of VCAM-1 mRNA degradation after blockade of transcription with actinomycin D was identical in the absence and presence of either compound. The noncytokine inducible endothelial antigen E 1/1 was not influenced at all by the various NO donors, indicating that the inhibitory action of NO was highly specific. Moreover, none of the compounds tested showed any significant toxicity at the concentration range used. A remarkable finding was the consistently superior potency of SP/W 5186 and 3672 on all endpoints measured, i.e., monocytoid cell adhesion, adhesion molecule expression, and MHC II antigen expression. The reason for this is not clear at present but may be due to either the kinetics of NO release from these compounds or an additional antioxidative component inherent to their cysteine moiety. Although an investigation into the potential antioxidative properties of these NO donors was beyond the scope of the present study such an activity may act synergistically with NO in inhibiting endothelial cell adhesion molecule expression via an independent pathway of inactivation of nuclear factor-B (Spiecker et al., 1998).

One essential feature present in all NO donors able to inhibit endothelial activation is the ability to release NO spontaneously. This may explain, at least in part, the lack of efficacy of the organic nitrates, nitroglycerin, isosorbide dinitrate, and SP/W 4744, which do not release NO in the absence of sulfhydryl groups (Noack and Feelisch, 1991; Feelisch, 1993). Accordingly, the greater activity of SP/W 3672 compared with GSNO cannot be explained by the ability of SP/W 3672 to release NO in the presence of thiols because the total amount of NO released under these conditions, as detected by gas phase chemiluminescence, was similar for these two compounds, at least within the first 30 min of incubation. On the other hand, SP/W 3672 released much more NO in the absence of thiols, possibly accounting for its greater potency on adhesion molecule and MHC II antigen expression. Besides the thiol dependence the kinetics of NO release may account for some of the differences among the different compounds tested. Because the expression of various adhesion molecules after cytokine stimulation requires hours for the induction of gene transcription [peaking earliest (2-4 h) for E-selectin (Bevilacqua et al., 1989), later for VCAM-1 (Cybulsky and Gimbrone, 1991) and ICAM-1 (Marlin and Springer, 1987), and latest (>12 h) for MHC II antigens (Pober et al., 1983; Lapierre et al., 1988)], one might expect that a prolonged NO release best meets the requirements for an effective inhibitor of gene transcription for such molecules. This assumption was not specifically addressed in our study, but may be particularly relevant for SP/W 5186, which is a relatively poor spontaneous NO releaser per se, but which spontaneously hydrolyzes, at physiological pH, to SP/W 3672. In fact, SP/W 5186 was designed as a prodrug for SP/W 3672 (Knüttel et al., 1995). Hydrolysis of SP/W 5186 to SP/W 3672 could therefore account for the similar potencies of both compounds. In agreement with the lack of an NO-generating nitrate group, SP/W 6373 failed to show significant inhibitory activity in the test systems investigated.

As the main result of this investigation, two compounds were identified that exhibited a considerably higher potency to interfere with endothelial activation than the hitherto best described NO donor compound GSNO. The availability of these interesting compounds may foster new research into the therapeutic potential of NO donors beyond their classical cardiovascular indications. One of the next steps along this line of research will be to assess whether the in vitro properties here described will translate into inhibition of early atherogenesis, fatty streak, and raised lesion formation in experimental animal models of atherosclerosis, where adhesion molecule expression plays a major role (Gimbrone, 1995).