Introduction

Top Abstract Introduction Methods Results Discussion References

In 1867, Sir Thomas Lauder Brunton first showed that inhaled ISAN produced favorable effects in patients suffering from angina pectoris. Twelve years later (1879) Murrel showed similar activities with lingual nitroglycerin dosing, and since that time NTG has become a mainstay of pharmacological angina therapy. Because of the similar historical descriptions of ISAN and NTG, and because both agents produce direct vasorelaxation via vascular NO formation (Ignarro et al., 1981; Bauer and Fung, 1995), the actions of these two similar chemical classes (i.e., organic nitrites and organic nitrates) have often been considered to be identical.

The recent interest in NO as a multifacetted physiological messenger has led to an intense interest in the various chemical classes that liberate NO (Moncada and Higgs, 1993; Bauer et al., 1995). It has become increasingly evident, however, that some important dissimilarities exist among these agents. For example, Kowaluk and Fung (1991) showed that the enzymatic requirements for NO generation from organic nitrates was distinctly different from those for organic nitrites. These observations led us to hypothesize that the vascular activities may also be distinct for these two chemical classes. In addition, because reduced vascular NO production has been suggested as a mechanism of nitrate tolerance (Needleman and Johnson, 1973; Forster et al., 1991; Chung and Fung, 1993), we hypothesized that the tolerance properties of organic nitrites may be different as well.

Proposed mechanisms of nitrate tolerance during sustained therapy include biochemical alterations in vascular tissues and/or physiologic counter-regulatory events. We describe in vitro studies using isolated vascular tissues, and intact animal experiments using a relevant animal model of in vivo nitrate tolerance, for the purpose of testing the hypotheses that the pharmacological activities and tolerance properties of organic nitrites and nitrates are distinct.

	Methods

Top Abstract Introduction Methods Results Discussion References

Isolated blood vessel experiments. All animal procedures were approved by the Laboratory Animal Care Committee at the University of Buffalo. Studies using rat thoracic aorta were performed to compare the vascular potencies and in vitro tolerance characteristics of NTG and ISAN, using methods similar to those previously described (Bauer and Fung, 1991a; Bauer et al., 1993). Experiments were also conducted to compare the duration of vasorelaxation in vitro. Briefly, healthy male Sprague-Dawley rats (300-400 g) were killed and the thoracic aorta was rapidly isolated and placed in oxygenated Kreb's buffer. The tissue was carefully cleaned, ring segments were cut (2-3 mm wide), and mounted on hooks connected to force transducers and a physiograph recorder (Grass Instruments, Quincy, MA). Vessel segments were maintained in 10 ml of oxygenated Kreb's buffer, washed twice and equilibrated with a resting tension of 4 g over a 30-min perios. Tissues were then precontracted with PE (0.5 µM). When PE contraction stabilized, cumulative concentration-effect data for NTG or ISAN was obtained. Solutions of NTG were prepared in 5% dextrose (D5W) and ISAN standards were prepared in ethanol and stored in sealed vials on ice. Aliquots of NTG (100 µl) or ISAN (10 µl) solutions were added to the tissue bath in cumulative fashion. All standards were prepared fresh daily, and ISAN was determined to be stable under these conditions using a gas chromatographic assay developed in our laboratory (data not shown).

In vivo model of hemodynamic tolerance. Because vascular tolerance development in vitro may not be identical to hemodynamic (i.e., clinically relevant) tolerance development (Fung and Bauer, 1994), we compared the hemodynamic actions and tolerance properties of NTG, ISAN and ISBN using a relevant animal model previously developed by us (Bauer and Fung, 1990, 1991, a and b). CHF was produced in male Sprague-Dawley rats via coronary artery ligation and myocardial infarction at the left ventricular free wall. After 6 wk recovery, infarcted rats were instrumented for the measurement of left ventricular hemodynamics and intravenous drug infusion. Rats with large ventricular infarcts have elevated venous pressure and decreased cardiac output, similar to hemodynamic abnormalities observed in CHF patients (Pfeffer et al., 1979; Bauer and Fung, 1991b).

1

1

1

1

1

1

1

	Results

Top Abstract Introduction Methods Results Discussion References

Figure 1 shows the duration of vascular relaxation in vitro for ISAN and NTG. After PE contraction, ISAN or NTG was added to the bath to achieve approximately 50% relaxation, and the duration of this response was continuously recorded for 20 min. Representative tracings from the physiograph are shown in figure 1A, whereas average responses are shown in figure 1B (mean ± S.D., n = 5). Similar initial (peak) reductions in vessel tension were observed at 390 nM NTG and 730 nM ISAN, indicating that ISAN was slightly less potent than NTG. In addition, the duration of relaxation was shorter for ISAN despite similar initial relaxant action. The vasoactivity of ISAN vanished completely 15 min after addition to the vessel bath. A slight loss of activity was also observed with NTG during a 20-min incubation. Statistically significant differences between these two agents was observed at 10, 15 and 20 min after drug addition (P < .05).

View larger version (24K): [in this window] [in a new window]   Fig. 1.   In vitro duration of relaxant effects from NTG or ISAN. A, Representative relaxation data and B, mean data ± S.D. (n = 6). *Statistically different from maximal relaxant effect; 53, significant difference between treatments (P < .05).

Figure 2 shows the cumulative relaxation response curves for NTG or ISAN and the development of self-tolerance for each of these agents. Analyses of the fitted data are presented in table 1. The estimated EC₅₀ values for the vehicle controls showed that NTG was approximately 2-fold more potent than ISAN (table 1), but no change in E_max was observed between these agents. In addition, the estimated Hill coefficient for NTG was statistically smaller than that of ISAN. Incubation of vessel segements with EC₉₅ concentrations for 30 min showed that self-tolerance was induced by NTG but not by ISAN. Preincubation caused a 75-fold increase in the EC₅₀ of NTG, but no significant change for ISAN.

View larger version (30K): [in this window] [in a new window]   Fig. 2.   Vascular concentration-response and tolerance data for NTG or ISAN in vitro (mean ± S.D., n = 6). Solid symbols represent vessel segments incubated with vehicle (control), open symbols represent segments preincubated with either NTG of ISAN at their respective EC95 for 30 min.

Figure 3 shows the effects of NTG, ISAN and ISBN on in vivo left ventricular hemodynamics when administered to CHF rats. Base-line hemodynamic values were not different among the three treatment groups (LVEDP: 19 ± 2, 23 ± 4, 18 ± 6 mmHg; LVPSP: 135 ± 7, 133 ± 8, 135 ± 11 mmHg; HR:381 ± 60, 377 ± 31, 394 ± 27 bpm, NTG, ISAN and ISBN groups, respectively). Constant-rate infusions of these compounds (at 10, 45, and 45 µg min¹, respectively) produced identical initial reductions in LVEDP. However, NTG caused a gradual loss of this initial hemodynamic effect, and complete tolerance development was observed by 8 hr of infusion. In contrast, both organic nitrites maintained efficacy in reducing LVEDP throughout the 10-hr infusions. Thus, the LVEDP effects at 10 hr of nitrite infusion remained statistically different from initial base-line values, and were statistically different from the NTG treatment group at 10 hr. Statistically significant reductions in LVPSP was also observed with organic nitrites, but not with NTG. A subgroup of rats were also continuously infused with organic nitrites for 22 hr (45 µg min¹, n = 7, 3 received ISAN although 4 received ISBN). Statistically significant LVEDP effects were maintained even at 22 hr (35 ± 4.1% reduction from initial baseline, P < .05) and were not statistically different from the 1 or 10 hr effects observed in these animals (1 hr effect 47 ± 5.4%; 10 hr effect 44 ± 4.5% reduction from base-line LVEDP).

View larger version (30K): [in this window] [in a new window]   Fig. 3.   Hemodynamic actions and tolerance properties of continuous infusion of NTG (n = 8) vs. ISAN (n = 7) or ISBN (n = 7), in congestive heart failure rats. *Statistically different from corresponding base-line value; 53, statistically different from the corresponding NTG treatment group.

	Discussion

Top Abstract Introduction Methods Results Discussion References

NO has recently been shown to be an important endogenous messenger for a wide array of physiologic processes, including cardiovascular regulation, immune function, penile erection and long-term memory. Thus, intense interest in NO donors as potential therapeutic agents has developed. Interestingly, both ISAN and NTG have been used as antianginal agents for more than 100 years and were used long before a detailed understanding of their mechanisms of action had been developed. It is now established that both of these agents cause direct vasorelaxation through vascular generation of NO and relaxation via a cyclic guanosine monophosphate-dependent process (Murad, 1986; Ignarro et al., 1981). Because of their similar actions, the organic nitrites and nitrates have historically been considered to be identical and, as was pointed out recently (Bauer and Fung, 1995), nitrites have commonly (but erroneously) been classified as nitrates.

Despite the obvious general similarities among NO donors (i.e., actions mimicking those of NO) recent evidence suggests that the various chemical classes differ in their rates and extents of NO generation (for reviews see Noack and Feelisch, 1991; Bauer and Fung, 1995), and this may play an important part in their differential pharmacology. For example, studies using bovine coronary artery smooth muscle cells have shown that the predominant enzymatic site for nitrate conversion to NO is found in membrane-bound fractions, is associated with plasma membrane marker enzymes and has an approximate molecular weight of 160 kDa (Chung and Fung, 1990; Chung et al., 1992). In contrast, the primary cellular site of bioactivation of organic nitrites is apparently cytosolic, and the pertinent enzyme has an approximate molecular size of 260 kDa (Kowaluk and Fung, 1991). Thus, both the sites and enzymatic components for NO release are apparently distinct for these two NO donor classes. The bioactivation of both nitrates and nitrites can be enhanced through additional thiols, and can be inhibited by thiol alkylating agents (Chung and Fung, 1990; Kowaluk and Fung, 1991). We asked if the vascular metabolic differences among nitrates and nitrites translate to differences in their vasoactivities and tolerance properties.

When added to in vitro vessel baths, ISAN and NTG produced direct vasorelaxation that was maximal within 2 min, but over time these effects were reduced. When ISAN was added to the bath solution in sufficient concentrations to produce about 50% relaxation, this effect was completely lost 10 min after addition. Similarly, a partial diminution of NTG effect was observed. Feelisch et al. (1995) have recently shown that when NTG is incubated with intact vascular cells, the peak NO release rate is observed within 5 min after which NO production steadily declines to about 50% of maximal within 20 min, a pattern similar to the NTG vascular response we observed (fig. 1). Although the differing duration we observed may be in part due to differing volatilities (both are liquids, highly lipophilic and volatile at room temperature), our preliminary experiments showed that less than 10% of ISAN was lost at 5 min after addition to buffer alone, whereas vasodilating activity was completely lost within 10 min. We therefore postulate that the differences in the time profiles for these two agents may be (at least partly) related to different enzymatic and/or chemical release rates of NO, and differing efficiencies in delivering NO to the target site(s) in vascular tissue.

NTG was found to be slightly more potent than ISAN in vitro (as judged by relaxation EC₅₀ values), but with similar maximal relaxation effects. We also found that the general shape of the concentration response curves were different, and this was exhibited by a statistically significant difference in the fitted Hill coefficients. The reason for this difference is unclear, but other investigators have previously suggested that NTG may have more than one vascular relaxation mechanism, and that the high affinity component may be related to NTG interaction with G-protein rather than NO production (Torfgard et al., 1990; Ahlner et al., 1988). The possible biphasic vascular response is also consistent with our recent observation of two binding sites for radiolabeled NTG when incubated with vascular microsomal fractions (Bauer and Fung, 1996). Thus, it is possible that the vasoactivity of ISAN occurs via a single predominant mechanism (or a single primary enzymatic conversion to NO) whereas the vasoactivity of NTG may be more complex. Interestingly, the apparent biphasic NTG relaxation response observed with naive vessels was lost after vascular tolerance induction (fig. 2; table 1). The significance and relevance of this biphasic response to in vivo conditions remains to be determined.

Incubation of vascular segments with NTG or ISAN with their respective EC₉₅ for 30 min produced differing degrees of tolerance. A 75-fold shift in the EC₅₀ was observed for NTG when vessels were preincubated, indicating a large degree of self-tolerance development. This vascular tolerance phenomenon has been well-documented by others, and may be due to reductions in metabolic conversion of NTG to NO production in vascular tissues (Forster et al., 1991; Chung and Fung 1993), or perhaps by increased scavenging of NO by superoxide anions (Munzel et al., 1995). Previous experiments have shown that nitrate vascular tolerance is not associated with a loss of responsiveness to NO gas itself, suggesting that the vascular tolerance process is primarily related to specific changes in the bioactivation of nitrates (Fung et al., 1989).

In contrast to NTG vascular tolerance, ISAN demonstrated no significant change in vascular potency after preincubation. Zimmerman et al. (1991) recently arrived at a similar conclusion (i.e., organic nitrites produce less vascular tolerance than nitrates) using isolated rabbit aortic strips and isobutyl nitrite, but in their studies the bath concentrations of ISBN were not monitored or maintained, and no in vivo experiments were conducted. In our initial experiments we found a short ISAN duration and complete loss of effect after 15 min (fig. 1). Using spectrophotometric methods and a sensitive gas chromatographic assay with electron capture detection, we also found that ISAN was rapidly lost in the oxygenated organ bath (data not shown), consistent with the observed short duration of vasoactivity. Using these analytical methods we found that less than 10% loss of ISAN occurred within 5 min after bath addition, and vascular relaxation experiments also showed only a slight loss of ISAN effect within this time period (fig. 1). For these reasons, we replaced the preincubation bath with fresh ISAN and NTG every 5 min in our tolerance studies, and by doing so we have removed the confounding factor of NO donor loss during tolerance induction. In addition, to compare self-tolerance properties among NTG and ISAN, we preincubated vessel segments with equally effective concentrations (i.e., EC₉₅ concentrations determined in preliminary experiments), rather than equimolar concentrations, as used by Zimmerman et al. (1991).

The development of in vitro vascular tolerance after isolated blood vessels are incubated with very high NTG concentrations is well established, but the relevance of these experimental conditions and the significance of in vitro nitrate tolerance in an in vivo setting (particularly using pharmacologically relevant doses) is unknown and controversial (Fung and Bauer, 1994). For example, the most widely accepted theory for nitrate tolerance development is that of vascular thiol depletion, which was originally postulated by Needleman and Johnson (1973). More recently, however, two separate studies have shown that tolerance development in vivo is not associated with a depletion of vascular thiols (Boesgaard et al., 1994; Haj-Yehia and Benet, 1995). In addition, no in vivo studies have yet demonstrated a reduction of vascular conversion to NO during hemodynamic tolerance development. In addition to the biochemical hypotheses of in vitro vascular nitrate tolerance, other physiologically based mechanisms may be operative in vivo. Such mechanisms include activation of the renin-angiotensin system, reflex vasoconstriction and vascular fluid shifts (Bauer et al., 1995). Although the exact in vivo mechanism(s) of nitrate tolerance is not defined, it is possible that both biochemical vascular changes and physiologic responses are involved.

Because pharmacologically relevant doses of nitroglycerin produce only slight hemodynamic effects in normal humans and animals, we developed a diseased animal model to allow for the examination of the in vivo mechanisms of nitrate tolerance development. We have previously shown that this CHF rat model mimics humans with respect to NTG hemodynamic effects and tolerance development (Bauer and Fung, 1990, 1991b), and that this model could provide insight into the processes and mechanisms of hemodynamic rebound after abrupt NTG withdrawal (Bauer and Fung, 1993). In addition, this animal preparation predicted a beneficial hemodynamic interaction between NTG and hydralazine (Bauer and Fung, 1991a), an important finding that has recently been confirmed by others in a human CHF trial (Gogia et al., 1995).

NTG, ISAN or ISBN (at doses of 10, 45 or 45 µg min¹, respectively) initially produced similar reductions in ventricular preload (LVEDP) when infused to conscious CHF rats. This preload-reducing effect enhances myocardial blood flow during diastole, helps to relieve pulmonary congestion associated with this disease state and is an important component of NTG therapy in CHF (Greenberg et al., 1975). On a molar basis the in vivo potency difference between NTG and nitrites was approximately 10-fold because 2.6 µmol NTG hr¹, 23 µmol ISAN hr¹, and 25 µmol ISBN hr¹ all produced comparable initial reductions in LVEDP. This in vivo difference is larger than the 2-fold difference in NTG and ISAN vascular EC₅₀ values we observed in vitro, and may be explained by both differing vascular potencies and/or pharmacokinetic characteristics.

Despite similar initial reductions in LVEDP, complete hemodynamic tolerance to NTG was observed within 8 hr of continuous infusion but nitrite effects were maintained throughout the 10-hr infusion regimen. Additionally, we were able to continue the nitrite infusions in a subgroup of CHF rats and found that the LVEDP actions were still maintained at 22 hr of continuous infusion. Unfortunately, difficulties in maintaining the patency of implanted ventricular catheters did not allow us to study the hemodynamic effects beyond this timepoint, and further studies are warranted to determine if hemodynamic tolerance to nitrites might eventually develop. However, these in vivo studies illustrate that the LVEDP tolerance property of NTG is apparently not shared by ISAN or ISBN, and that a lack of tolerance development (or at least lesser tolerance relative to NTG) may be a common characteristic of all members of the organic nitrite class.

What makes the tolerance properties of organic nitrates and nitrites distinct? One possible explanation may be that the enzymes involved for conversion of these two classes to NO are distinct and are regulated differently. Thus, nitrate tolerance may be due to reduced enzymatic activity (due to changes in enzyme behavior or cofactor availability), whereas no such modification occurs for organic nitrites. Such a hypothesis is consistent with previous findings from our laboratory (Kowaluk and Fung, 1991), and would support the view that nitrate tolerance in vivo occurs via reduced vascular bioconversion to NO. Alternatively, the apparent tolerance differences may be related to potential differences in hemodynamic activities in vivo. In addition to the hemodynamically favorable reductions in LVEDP observed with organic nitrites, these agents caused slight but statistically significant reductions in LVPSP at 1 and 2 hr of infusion. Although this afterload effect was not maintained throughout the continuous nitrite infusions (suggesting tolerance to this effect), it is possible that beneficial afterload actions (at least initially) with no significant tachycardia may be an important determinant of the sustained LVEDP effects of nitrites. We have previously shown that coadministration of the arterial vasodilator hydralazine prevented NTG tolerance in this animal preparation, and suggested that the predominant preload reducing effects of nitrates may be related to activation of physiological counterregulatory mechanisms and be a determinant of their hemodynamic tolerance development (Bauer and Fung, 1991a). In addition, we and others have previously reported that an S-nitrosothiol produced "balanced" preload and afterload effects and produced less hemodynamic tolerance than NTG (Bauer and Fung, 1991b; Shaffer et al., 1992). The apparent afterload actions of nitrites may be particularly favorable in CHF therapy because nitrate administration is typically combined with other arterially acting agents, such as hydralazine or an angiotensin converting enzyme inhibitor, to achieve therapeutic efficacy (Massie et al., 1977; Leier et al., 1981).

ISAN is currently approved in the United States for the treatment of angina, but it is seldom used clinically. Owing to its volatility, it is currently administered via inhalation. Although inhaled ISAN has a rapid onset of action, its hemodynamic effects last only a few minutes because of its short biological half-life and the current mode of administration is useful only for acute administration. In addition to its occasional use as an antianginal agent, ISAN is also used for the diagnostic evaluation of cardiac murmurs (Lembo et al., 1988) and the treatment of cyanide toxicity associated with sodium nitroprusside overdose. Despite the long history of ISAN as an acute vasodilating agent, no previous studies have described its hemodynamic action during long-term continuous administration. In preliminary experiments, we developed the ability to accurately and reliably infuse nitrites as neat oils to small animals at extremely low flow rates (3 µl hr¹), thus allowing us to conduct these studies. We used this approach because the aqueous solubility of these nitrites is poor (approximately 1 mg ml¹), and it allowed us to deliver appropriate doses without the use of cosolvent systems or infusion of large volumes in these small animals. Throughout the nitrite infusion studies the animals behaved normally and no untoward effects were observed. Gross autopsies also revealed no abnormalities after nitrite infusions.

Our findings indicate that the classical view that nitrites and nitrates are pharmacologically equivalent is incorrect. In keeping with vascular enzyme differences demonstrated previously (Kowaluk and Fung 1991), we have found that organic nitrites have different vascular actions and hemodynamic effects and cause less tolerance when compared to NTG in vitro or in vivo. Further evaluation of organic nitrites as therapeutic agents for the treatment of cardiovascular diseases appears warranted.