|
ENABLING SCIENCE Nitric oxide-releasing compounds: From basic research to promising drugsBy attaching the anionic [N(O)NO]- functional group to various carrier molecules, new biomedical research tools that spontaneously generate bioactive NO at physiological pH have been made possible.Larry K. KeeferW hen the first stunning revelations about the many roles of nitric oxide (NO) in normal physiology began to appear in the literature a little more than a decade ago, those of us who had been studying NO's toxic effects were skeptical. It seemed unlikely that such a simple diatomic radical, whose status as a noxious air pollutant and toxic cigarette smoke constituent had long been known, could possibly be a natural anticoagulant, vasodilator, neurotransmitter, and mediator of immune system function--to name only a few of the bioregulatory phenomena that have been attributed to NO (1). However, the evidence was overwhelming. By 1988, it was clear that a major revolution in the biomedical sciences had begun, and with it, a new opportunity to contribute. Having spent the previous 20 years studying NO indirectly as a precursor and metabolite of the carcinogenic nitrosamines, I decided to try my hand as a NO chemist.
"Pure" research turns practical
Our findings were not very interesting as a nitrosamine formation pathway, because the ability of aerobic NO to nitrosate amines had long since been established, and salts such as (C2H5)2NH2+ (C2H5)2N[N(O)NO]- were regarded as mere laboratory curiosities--so esoteric that we were able to find only one prior review of their chemistry (4). The fact that such ions could release NO spontaneously on dissolution in aqueous media became extremely interesting, however, when the first announcements of NO's multifaceted bioregulatory roles appeared a few years later. Fundamental diazeniumdiolate chemistry. With the hypothesis that practical pharmacological benefits might be built on this finding, in 1988 we began a detailed reappraisal of our earlier study. First, we synthesized a variety of nucleophile/NO adducts and characterized their physicochemical properties. Most adducts could be prepared conveniently and simply by placing anaerobic solutions of a nucleophile (X-) under several atmospheres of NO and filtering off the solid product containing the X[N(O)NO]- moiety as it formed (6). Even though many of these compounds are stable to prolonged storage as the dry powders, they spontaneously release up to 2 mol of NO per mole of dissociating ion, as anticipated from Figure 1. Half-lives at physiological pH and 37 °C reported in the literature thus far range from 2 s to 20 h. Decomposition slows markedly with increasing pH, permitting preparation of stable alkaline stock solutions that generate significant NO only when the pH is lowered to that of the bloodstream or culture medium. Dissociation of the anions to NO proceeds with simple first-order kinetics at constant hydrogen ion activity. All have a strong chromophore around 250 nm in their UV spectra that can be used for convenient detection, quantification, and verification of purity. Diazeniumdiolates as biomedical research tools. This physicochemical profile (7) suggested that the diazeniumdiolates offered important advantages as probes for elucidating the biological effects of NO. Unlike sodium nitroprusside, molsidomine, and the nitrate esters (NO donor drugs used clinically in 1988 that are now known to require redox activation), the anionic diazeniumdiolates spontaneously release NO when dissolved in aqueous media. Moreover, unlike the more recently introduced agents of the S-nitrosothiol family, the diazeniumdiolates' rates of spontaneous NO generation are not dramatically affected by the presence of thiols or trace metal contaminants in the medium. This reliability of the NO generation rate has allowed us to reach some important conclusions about NO biology. For example, by comparing the effects of diazeniumdiolates that have widely differing half-lives, we confirmed that a given amount of NO released slowly over a long period of time has a less potent but more prolonged duration of vasodilatory action than the same amount of NO released all at once (8, 9). In vitro and in vivo anticoagulant activity similarly correlates with the rate of NO release (10). In contrast, NO's ability to inhibit the proliferation of vascular smooth muscle cells in culture proved more effective when a given total NO exposure was spread evenly over the observation period than when it was released abruptly into the culture medium (11). Anionic diazeniumdiolates also have been used to show that exposure to NO can
Clinical potential One proposed mechanism for this occurrence hypothesizes that the hemoglobin in the clot surrounding an artery induces these spasms by robbing the muscle controlling vessel diameter of the endogenous NO (called endothelium-derived relaxing factor) that it needs to remain properly relaxed, thereby permitting contracting factors to run amok. Using this reasoning, Pluta et al. have studied the ability of several diazeniumdiolates to reverse vasospasms in monkeys (33). After infusing DEA/NO (Figure 2,1) into the carotid artery immediately upstream from the contracting segment of the middle cerebral artery, the researchers observed a complete reversal of the spastic condition. The effects of administration of this drug on an artery are shown in Figure 3. However, the diazeniumdiolate used, which has a 2-min half-life for NO release, significantly lowered blood pressure. Such a drop in blood pressure is a dangerous complication in a patient whose hemodynamic system is already pathologically out of balance.
Fortunately, Pluta et al. were able to sidestep the problem of hypotension by taking advantage of diazeniumdiolates' wide array of half-lives. When the ultrafast NO donor PROLI/NO (2-s half-life at physiological pH; Figure 2, 2) was infused into the spastic vessel instead of DEA-NO, the available NO was completely released at the site where it was needed, before it could escape downstream. The spasm was reversed, and there was no detectable lowering of systemic blood pressure. If clinical trials now in the planning stages are successful, a considerably improved prognosis could become available for the several thousand individuals who are affected by cerebral vasospasm each year in this country. Treatment of impotence. It has been estimated that 20-30 million American men suffer from an inability to achieve and maintain an erection sufficient for the completion of sexual intercourse. A few years ago, it was discovered that NO is a critical physiological effector of penile erection; by inducing relaxation of the corpus cavernosum, NO allows blood to engorge the penis and maintain its tumescent state. Capitalizing on these basic research findings, Hellstrom and colleagues have studied the ability of PIPERAZI/NO (Figure 2, 3) and PROLI/NO to induce erections in male cats (34). Although the researchers had previously shown that NO was effective when donor drugs were injected intracavernosally (i.e., injected into the corpus cavernosum through the wall of the penile shaft) in cats (35), they wanted to determine whether comparable results could be achieved using transurethral administration (i.e., into the genitourinary tract via the tip of the penis, without piercing the skin). The transurethral delivery method is far more comfortable for the patient than the intracavernosal route. Hellstrom and colleagues found that the diazeniumdiolates both increased the length of the penis and raised the pressure within the corpus cavernosum when administered transurethrally under the experimental conditions (34). It is important to note that whereas the control drug combination of papaverine, phentolamine, and prostaglandin E1 injected intracavernosally induced systemic hypotension in these cats, the transurethral administration of diazeniumdiolates did not. Nonthrombogenic blood-contact surfaces. Unwanted clot formation is a major problem for the medical practitioner. Many devices that have surfaces that come into contact with blood (e.g., biosensors, extracorporeal membrane oxygenation equipment, heart-lung bypass machines, and renal dialysis systems) can be fouled or, in the case of vascular grafts and shunts, sealed by adhering blood platelets. To minimize this thrombogenic foreign-body response, blood thinners such as heparin, coumadin, and aspirin are often used. Unfortunately, systemic administration of such antiplatelet agents concomitantly increases the risk of uncontrolled bleeding elsewhere in the body. One way to prevent this potentially life-threatening side effect is to coat the foreign body with an immobilized drug that limits antiplatelet action to the contact surface between the blood and the device. Using this strategy, the interface is the only place in the circulatory system with other than a normal clotting time. NO technology offers an attractive way to accomplish this because
Another excellent illustration of the potential improvements in clinical practice that such coatings might allow has been provided by Espadas-Torre et al. (37). In seeking to prevent attachment of platelets to biosensors that they were developing for continuous measurement of analytes in patients' blood, the investigators coated the sensor surface with blends of the zwitterionic diazeniumdiolate MAHMA/NO (Figure 2, 4) with either polyurethane or poly(vinyl chloride). By judiciously changing the plasticizer content of these blends, materials with varying NO release rates were devised. The profound effect of these coatings on platelet function is shown in Figure 4. Control coatings allowed considerable adhesion when exposed to platelet-rich sheep plasma, but platelets were almost absent when the diazeniumdiolate-containing coating was used under otherwise identical conditions (37).
Many other applications for such diazeniumdiolated polymers can be conceived. These include wound dressings (24); and materials that can be extruded, cast, or spun into forms suitable for fabricating containers, sutures, plugs, valves, implants, blood conduits, or other medical devices for which an NO-releasing surface might be deemed beneficial. Work along these lines is under way in several laboratories.
Future directions However, because many organs, tissues, and cell types are difficult or impossible to access by local administration, our major current research effort is aimed at prodrug development--that is, at converting spontaneously releasing diazeniumdiolate ions to stable derivatives by attaching protecting groups that will allow them to circulate freely after intravenous or oral administration. By selecting a protecting group that can be metabolically removed by enzymes unique to the target organ, tissue, or cell type, NO generation should be concentrated at the target site. This approach has been used successfully to design a liver-selective NO
donor of the diazeniumdiolate class (39). By affixing a
vinyl group to the O-terminus of an anion that generates NO with a 3-s
half-life at physiological pH, a prodrug candidate dubbed V-PYRRO/NO
was synthesized (39). Consistent with our hypothesis that
drug-metabolizing enzymes in the liver should be able to remove the
vinyl group, hepatocytes (but not the other cell types examined in
culture) metabolized V-PYRRO/NO to NO. Furthermore, V-PYRRO/NO
administered intravenously stimulated cyclic guanosine 5`-monophosphate
(cGMP) synthesis--a biochemical consequence of NO release--in the liver
while minimally affecting systemic blood pressure in rats. The compound
also protected the animals from the liver-toxic effects of tumor
necrosis factor- Other O-substituted diazeniumdiolates are being examined as candidate prodrugs for
Acknowledgments
References |
Figure 1.
Figure 2.
Figure 3.
Figure 4.
, suggesting a possible therapeutic strategy for
fulminant liver failure.
 |
 |
|
