Introduction

Top Abstract Introduction Methods Results Discussion References

NO is a chemical messenger and effector molecule involved in the regulation of neural communication, blood vessel tone, platelet activation and immune responses (Moncada et al., 1991; Moilanen and Vapaatalo, 1995). NO activates guanylate cyclase and increases the concentration of cGMP, which mediates several physiological actions of NO (Ignarro, 1991; Moncada et al., 1991).

In the immune system, NO has been recognized as an important effector molecule for macrophages in their cytotoxic and cytostatic activity toward pathogens and tumor cell targets (see MacMicking et al., 1997). In addition, NO regulates host immunity as a suppressor of T lymphocyte responses. The production of large amounts of NO by activated macrophages was found to suppress lectin- and alloantigen-induced lymphocyte proliferation in vitro (Albina et al., 1991; Mills, 1991). The principal modulatory effect of NO was down-regulation of T cell proliferation but not cytokine production (Merryman et al., 1993; Marcinkiewicz et al., 1996). In addition, NO produced by macrophages inhibits T cell proliferation in experimentally induced infections such as murine trypanosomiasis and mycobacterial infections (Schleifer and Mansfield, 1993; Maw et al., 1997). There also are data supporting the mediator role of NO in the tumor-induced suppression of lymphocyte proliferation (Lejeune et al., 1994).

Although the mechanism of action of nitrovasodilators became evident only a decade ago (see Feelisch, 1993), they have been in clinical use for more than a century for indications such as angina pectoris, myocardial infarction and congestive heart failure. Because reduced generation and/or accelerated inactivation of NO has been implicated in a number of clinical conditions, NO donors are believed to have therapeutic potential in a range of cardiovascular and other diseases (see Moncada and Higgs, 1995). Classic nitrovasodilator compounds like nitroglycerin undergo enzymatic activation to generate NO. These reactions take place in the vessel wall, and therefore the effects of organic nitrates are restricted mainly to the cardiovascular system (Feelisch, 1993). Recently, pharmaceutical companies have been active in developing several novel drug candidates that release NO in enzymatic and nonenzymatic reactions. The present study was designed to investigate the immunological effects of a group of novel NO donors. We measured the effects of two NO-releasing oxatriazole derivatives, GEA 3162 and GEA 3175 (Kankaanranta et al., 1996a), and an earlier known NO donor, SNAP (Feelisch, 1993), on human lymphocyte proliferation. The results suggest that NO donors have an immunosuppressive action that involves inhibition of lymphocyte proliferation by a cGMP-independent manner.

	Methods

Top Abstract Introduction Methods Results Discussion References

Cell isolation and proliferation assay. PBMCs were isolated by Ficoll-Paque gradient centrifugation from human venous blood obtained from healthy volunteers who had abstained from any drugs for 1 week before blood samples were taken. PBMCs were suspended in RPMI 1640 Glutamax-1 supplemented with 10% heat-inactivated fetal bovine serum, penicillin (100 units/ml), streptomycin (100 µg/ml) and amphotericin B (250 ng/ml). PBMCs were cultured in 96-well plates (2 × 10 cells 200 µl). PDE inhibitors, guanylate cyclase inhibitors and NO donors were added into the culture at the beginning of the incubations. Lymphocyte proliferation was induced by lectin mitogen ConA (1 µg/ml). Cellular proliferation was determined by the incorporation of ³H-thymidine into cellular DNA (Kankaanranta et al., 1996b). The cells were incubated for 2 days at 37°C (in 5% CO₂) and then pulsed for 20 hr with 0.1 µCi of ³H-thymidine (5.0 Ci/mmol). The cells were harvested, and the incorporated radioactivity was measured with a -counter. ³H-Thymidine incorporation in ConA (1 µg/ml)-stimulated cells amounted to 1106 ± 168 cpm/well (n = 19), and that in unstimulated cells was 158 ± 34 cpm/well (n = 18). None of the treatments decreased cell viability as measured by Trypan blue staining and lactate dehydrogenase release.

Determination of cGMP production. PBMCs (5 × 10⁶ cells in 500 µl of RPMI) were incubated with ConA (1 µg/ml) and the NO donors in either the presence or absence of a PDE inhibitor (zaprinast 25 µM or IBMX 100 µM) or guanylate cyclase inhibitor (ODQ 10 µM or LY 83583 1 µM) for indicated time at 37°C. The incubations were terminated by the addition of ice-cold trichloroacetic acid (final concentration, 6%). The samples were sonicated and centrifuged (10 000 × g for 5 min). The supernatants were washed four times with water-saturated ether, diluted with an equal volume of 100 mM sodium acetate buffer, pH 6.2, and stored at 20°C until assayed for cGMP. The cGMP samples were acetylated and measured by radioimmunoassay as described previously (Moilanen et al., 1993).

Drugs and chemicals. The two mesoionic 3-aryl-substituted oxatriazole-5-imine derivatives GEA 3162 and GEA 3175, as well as SNAP and 8-p-chlorophenylthio-cGMP, were kindly provided by GEA (Copenhagen, Denmark). Culture media and supplements were from GIBCO (Paisley, Scotland). ConA and Ficoll-Paque were from Pharmacia (Uppsala, Sweden). IBMX was from EGA-Chemie (Steinheim, Germany). ¹²⁵I-Labeled cGMP was from DuPont (Boston, MA). [methyl-³H]Thymidine (5.0 Ci/mmol) was from Amersham International (Buckinghamshire, UK). ODQ was from Tocris Cookson (Bristol, UK). LY 83583 was from Calbiochem (La Jolla, CA). Zaprinast, 8-bromo-cGMP and 8-bromo-cAMP were from Sigma Chemical (St. Louis, MO).

Statistics. The results are expressed as mean ± SEM. Statistical significance was calculated by analysis of variance for repeated measures supported by Bonferroni's multiple-comparisons test and by Friedman nonparametric repeated-measures test followed by Dunn's multiple-comparisons test.

	Results

Top Abstract Introduction Methods Results Discussion References

Effects of NO donors on human lymphocyte proliferation and cGMP synthesis. The three NO donors, GEA 3162, GEA 3175 and SNAP, inhibited in a dose-dependent manner the proliferative responses of human PBMCs stimulated with ConA (1 µg/ml) (fig. 1, solid bars). At corresponding concentrations, the NO donors also induced an increase in cGMP production in ConA-stimulated human PBMCs as measured after 30-min or 2-hr exposure to the NO donor (table 1). Incubation of PBMCs with the NO donors for 24 hr resulted in cGMP levels comparable with the pretreatment levels. When red blood cells (red cell-to-PBMC ratio, 10-100:1) were added into the cultures, the antiproliferative action of NO donors was inhibited by 43% to 87% (n = 4), and the cGMP response was attenuated by 42% to 47% (n = 3), suggesting that NO is involved in both of these responses.

View larger version (33K): [in this window] [in a new window]   Fig. 1.   The effect of a guanylate cyclase inhibitor ODQ on the inhibitory action of NO donors GEA 3162, GEA 3175 and SNAP on human PBMC proliferation. The cells were cultured with ConA (1 µg/ml) and NO donor for 68 hr in either the presence or absence of ODQ. PBMC proliferation was determined by uptake of 3H-thymidine (0.1 µCi/well), which was added 20 hr before the end of culture. Results are expressed as percentage of inhibition of proliferation. The actual thymidine uptake for control conditions was 1731 ± 264 cpm/well (n = 7). ODQ (10 µM) alone (i.e., in the absence of NO donors) did not alter ConA-induced PBMC proliferation (4 ± 9% inhibition; n = 7). The values are the mean ± SEM of seven triplicate experiments.

Effects of guanylate cyclase inhibitors on the action of NO donors. The guanylate cyclase inhibitor ODQ was used to test whether inhibition of the cGMP formation would reverse the inhibitory action of NO donors. ODQ (10 µM) alone did not alter ConA-induced PBMC proliferation (96 ± 9% of control; n = 7). ODQ did not reverse the inhibitory effects of NO donors on PBMC proliferation (fig. 1), although NO donor-induced cGMP production was decreased by 26% to 84% (table 1). To further confirm this finding, we tested the effects of an other guanylate cyclase inhibitor, LY 83583. LY 83583 (10-50 µM) had its own inhibitory effect on PBMC proliferation (i.e., in the absence of NO donors), and therefore only a low concentration (1 µM) could be used in combination with NO donors. LY 83583 at that concentration decreased NO donor-induced cGMP production by 12% to 37% (n = 3) but did not alter ConA-induced cell proliferation (n = 3).

Effects of PDE inhibitors on the action of NO donors. Two PDE inhibitors were used to test whether a further increase in the concentration of cGMP would enhance NO donor-induced suppression of lymphocyte proliferation. A selective inhibitor of cGMP-specific type V PDE (Beavo, 1995), zaprinast (25 µM), increased NO donor-induced cGMP levels substantially during 30-min or 2-hr incubations (table 2). The NO donor-induced inhibition of the proliferative response in PBMCs was not, however, enhanced in the presence of zaprinast 25 µM (fig. 2). Zaprinast (25 µM) alone did not alter ConA-induced PBMC proliferation (96 ± 17% of control; n = 6). Also, the effects of IBMX, a nonspecific inhibitor of PDE, were tested on PBMC proliferation. IBMX alone (i.e., in the absence of NO donors) had an inhibitory action on PBMC proliferation. IBMX (100 µM) suppressed the proliferative response down to 51 ± 13% of control (n = 5). This may be due to simultaneous inhibition of cAMP degradation because IBMX is a nonselective inhibitor of PDEs. IBMX (100 µM) did not potentiate the inhibitory effects of NO donors on PBMC proliferation, although a clear increase in cGMP levels during 30-min incubations was seen (table 3). Taken together, PDE inhibitors enhanced the NO donor-induced increase in cGMP concentration but did not alter their inhibitory effect on proliferation.

View larger version (30K): [in this window] [in a new window]   Fig. 2.   The effect of a PDE inhibitor zaprinast (zap; 25 µM) on the inhibitory action of NO donors GEA 3175 and SNAP on human PBMC proliferation. The cells were cultured with ConA (1 µg/ml) and NO donor for 68 hr in either the presence or absence of zaprinast (25 µM). PBMC proliferation was determined by uptake of 3H-thymidine (0.1 µCi/well), which was added 20 hr before the end of culture. Results are expressed as percent of inhibition of proliferation. The actual thymidine uptake for control conditions was 691 ± 165 cpm/well (n = 6). Zaprinast alone (i.e. in the absence of NO donors) did not alter ConA-induced PBMC proliferation (4 ± 17% inhibition; n = 6). The values are the mean ± SEM of six triplicate experiments.

Effects of analogs of cGMP on human lymphocyte proliferation. The effects of two analogs of cGMP, 8-bromo-cGMP and a more cell-permeable compound, 8-p-chlorophenylthio-cGMP, were tested on lymphocyte proliferation. Submillimolar concentrations of cGMP analogs (8-bromo-cGMP, 100-300 µM; 8-p-chlorophenylthio-cGMP, 10-300 µM) did not inhibit lymphocyte proliferation, whereas 8-bromo-cGMP at millimolar concentrations (1 and 3 mM) had a moderate inhibitory action (fig. 3). The addition of the PDE inhibitor zaprinast did not further augment the inhibitory action of 8-bromo-cGMP (data not shown). The effects of the higher concentrations of 8-bromo-cGMP may be complicated by the unspecific effects such as increased cAMP due to inhibition of type III PDE (Beavo, 1995) or activation of cAMP-dependent protein kinase (Lincoln et al., 1995). Therefore, the effects of an analog of cAMP, 8-bromo-cAMP, were tested. 8-Bromo-cAMP inhibited ConA-stimulated lymphocyte proliferation in a concentration-dependent manner, being more potent than 8-bromo-cGMP (fig. 3).

View larger version (18K): [in this window] [in a new window]   Fig. 3.   The effects of analogs of cGMP and cAMP on ConA-induced proliferation of human PBMCs. Proliferation was determined by uptake of 3H-thymidine. Results are expressed as percent of control (i.e., the cells cultured without analogs of cGMP or cAMP). The actual thymidine uptake for control conditions was 850 ± 277 cpm/well (n = 6). Values are mean ± SEM of three to six triplicate experiments. **P < .01 and ***P < .001 vs. control.

	Discussion

Top Abstract Introduction Methods Results Discussion References

Although the regulatory role of endogenously produced NO in the human immune system has been difficult to demonstrate (see Moilanen et al., 1997) the present data suggest that exogenous NO in the form of NO-releasing compounds may have immunosuppressive effects in humans. The present results demonstrate that chemically different NO-releasing compounds inhibit proliferative responses in human lymphocytes stimulated with T helper mitogen ConA and increase cGMP production in the same cells. However, the data from experiments with PDE and guanylate cyclase inhibitors suggest that the antiproliferative effects of NO donors are mediated by a cGMP-independent mechanism.

In the present study, we used two novel oxatriazole-derived NO-releasing compounds, GEA 3162 and GEA 3175, and an earlier-known NO donor, SNAP. NO-releasing properties of GEA 3162 and GEA 3175 have been recently characterized, documenting their ability to inhibit platelet aggregation, induce cGMP synthesis in platelets, convert oxyhemoglobin to methemoglobin, generate nitrite and nitrate in aqueous solutions and form nitrosyl-hemoglobin complex (Karup et al., 1994; Kankaanranta et al., 1996a). GEA 3162 and GEA 3175 have been shown to have vasodilator, antiplatelet, fibrinolytic (Corell et al., 1994) and antibacterial (Virta et al., 1994) activities, as well as to inhibit neutrophil functions (Moilanen et al., 1993, 1994), suppress tumor cell growth (Vilpo et al. 1994), regulate glycosaminoglycan synthesis in articular cartilage (Järvinen et al., 1995) and inhibit oxidation of low-density lipoprotein (Malo-Ranta et al., 1994). In our preliminary study, we found that NO-releasing oxatriazole derivatives inhibit lymphocyte proliferation in NO-dependent and reversible manner (Kosonen et al., 1997).

At physiological concentrations of NO, the enzyme guanylate cyclase is the principal target of NO (Ignarro, 1991). Increased synthesis of cGMP mediates the vasodilatory and antiaggregatory actions of NO, as well as its effects on neurotransmission. Some studies have shown that NO inhibits cell proliferation by a cGMP-mediated mechanism in vascular smooth muscle cells (Etienne et al., 1996) and endothelial cells (Yang et al., 1994). In lymphocytes, the role of cGMP is controversial: cGMP correlates positively with T cell proliferation in some studies (Hersey et al., 1989), whereas in some others, cGMP seems to inhibit cell proliferation (Fecho et al., 1995). The possible mediator role of cGMP in the antiproliferative action of NO donors in human lymphocytes was tested in the present study.

First, NO donors caused a rapid and transient increase in cGMP production in PBMCs. After 30-min or 2-hr incubation with the NO donors, cGMP levels in PBMCs were substantially increased. Because the early events after addition of the stimuli are critical in the regulation of mitogen-induced lymphocyte proliferation, the time course of the increase in cGMP concentration is consistent with the hypothesis that antiproliferative actions of NO donors are mediated by cGMP. However, the capacity of different NO donors to generate cGMP did not always correlate with their antiproliferative action. For example, GEA 3175 and GEA 3162 inhibited cell proliferation about equally, whereas GEA 3175 induced a lower generation of cGMP.

Second, the effects of an inhibitor of guanylate cyclase, ODQ, was evaluated. ODQ is a potent inhibitor of NO-stimulated soluble guanylate cyclase without actions on particulate guanylate cyclase or adenylate cyclase (Garthwaite et al., 1995). ODQ inhibited NO donor-induced cGMP production, but the antiproliferatory action of NO donors was not altered. Also, LY 83583, which is widely used as an inhibitor of guanylate cyclase (Mulsch et al., 1988), tended to attenuate the stimulation of cGMP formation but did not alter the antiproliferatory action of NO-releasing compounds.

Third, the effects of the PDE inhibitors zaprinast and IBMX were tested on the antiproliferative and the cGMP-increasing actions of NO donors. Drug concentrations exceeding the earlier documented IC₅₀ values for inhibition of PDEs were used (zaprinast, IC₅₀ = 0.8 µM; IBMX, IC₅₀ = 2-50 µM; Beavo, 1995). Zaprinast, an inhibitor of cGMP-specific type V PDE (Beavo, 1995), did not potentiate the antiproliferative action of NO donors, although a clear increase in cGMP production was seen. A nonselective PDE inhibitor, IBMX, failed to enhance the antiproliferative effects of NO donors, whereas cGMP levels were further increased. These data indicate that the NO-induced antiproliferative action is not augmented when the breakdown of cGMP is inhibited by PDE inhibitors, suggesting that the process is not mediated by cGMP. Another explanation for these results might be that metabolism through zaprinast- or IBMX-inhibitable PDEs is not the principal inactivation mechanism of cGMP in lymphocytes. The importance of efflux rather than enzymatic metabolism in the inactivation of cGMP has been reported in some other cell types (Tjörnhammar et al., 1983; Radziszewski et al., 1995).

Fourth, the effects of two analogs of cGMP, 8-bromo-cGMP and a more cell-permeable compound, 8-p-chlorophenylthio-cGMP, were measured on lymphocyte proliferation. Up to concentrations of 300 µM, 8-bromo-cGMP or 8-p-chlorophenylthio-cGMP did not inhibit ConA-stimulated human PBMC proliferation. At millimolar concentrations, 8-bromo-cGMP had a moderate inhibitory action. At these higher concentrations, unspecific effects of cGMP, like increased cAMP due to inhibition of type III PDE (Beavo, 1995) or activation of cAMP-dependent protein kinase (Lincoln et al., 1995), cannot be ruled out. Consistently, an analog of cAMP, 8-bromo-cAMP, inhibited lymphocyte proliferation in a concentration-dependent manner, being more potent than 8-bromo-cGMP.

Taken together, when the mediator role of cGMP in the antiproliferative action of NO donors was tested, the characteristics originally presented by Sutherland et al. (1968) for a second messenger could not be demonstrated. These data suggest that NO donors inhibit lymphocyte proliferation in a cGMP-independent manner and thus adds to the list of cGMP-independent cellular actions of NO. These could be explained by the direct effects of NO on various other enzymes in addition to guanylate cyclase. NO inhibits iron-containing enzymes including aconitase, NADH-ubiquinone oxidoreductase and succinate-ubiquinone oxidoreductase of the mitochondrial respiratory chain (Stadler et al., 1991) and ribonucleotide reductase, the rate-limiting enzyme in DNA synthesis (Lepoivre et al., 1990). Therefore, NO can directly affect mitochondrial respiration and DNA synthesis, leading to suppressed proliferative responses. Another possible explanation arises from the finding that NO inhibits Ia antigen expression on antigen presenting cells (Sicher et al., 1994) because ConA-stimulation is dependent on MHC II expression in antigen presenting cells (Ahmann et al., 1978).

In conclusion, we demonstrated that NO-releasing oxatriazole derivatives at pharmacologically relevant drug concentrations (Karup et al., 1994; Corell et al., 1994; Kankaanranta et al., 1996a) inhibit human lymphocyte proliferation by a cGMP-independent manner. These data add our knowledge of the effects of NO on human immune system and suggest novel indications for NO donors. In addition, the immunosuppressive action should be kept in mind as a potential adverse effect when these drugs are used in other indications.