T he rewards of biomedical research can be very exciting and remarkable. The feeling of having done an experiment for the first time and only you know the results that will have an important effect on your research and perhaps broader applications is exhilarating. I suspect that it is not common for scientists to see their fundamental research develop during their research career into such broad clinical application for numerous diseases. While much of our work and experiments were well

more »

... many of our important observations came from luck or serendipitous events-undoubtedly, a common phenomenon in biomedical research. I will recount some of these experiments here. A more detailed review can be found elsewhere. 1,2 I spent my early years as a trainee working with cAMP and adenylyl cyclase. As a result of this experience, after the discovery of cGMP in rat urine in the mid-1960s, I and others thought that cGMP might also be a second messenger to mediate the effects of some hormones and drugs. We transferred much of our work from cAMP to cGMP. Our earliest experiments with guanylyl cyclase in the early 1970s suggested that there were at least two isoforms of the enzyme, soluble and particulate forms, with different kinetic properties. 3, 4 This was of interest since adenylyl cyclase was exclusively particulate. 5 The possibility of soluble and particulate isoforms of guanylyl cyclase suggested that different classes of hormones might regulate these different enzymes to produce separate pools of cGMP that may have different physiological effects. This was indeed the case from our later work with different activators of each isoform as discussed below. We ultimately demonstrated there were different isoforms with purification, cloning, and expression some years later. However, we initially took a shortcut and added various inhibitors of nucleotidases, phosphatases, and phosphodiesterases to soluble and particulate preparations to see if their kinetic behaviors in crude preparations could be altered. Perhaps the kinetic differences of the apparent isoforms could be spurious due to nucleotidases and phosphatases competing for the GTP substrate or phosphodiesterases hydrolyzing our product cGMP. We added fluoride, pyrophosphate, azide, hydroxylamine, nitrite, methylxanthines, and other agents to our incubations. Several hormones that increased cGMP accumulation in intact tissues failed to activate the enzymes in cell-free preparations. One of the goals of the laboratory was to understand the molecular mechanism of hormonal regulation of guanylyl cyclase. Without an effect of hormones in cell-free preparations, this would be a most difficult task. Quite accidentally we found that azide, hydroxylamine, and nitrite activated the soluble guanylyl cyclase. 6,7 What an exciting event this was, and we became committed to understanding their mechanism of activation. We thought that these activators could help us reconstitute activation by some hormones in cell-free systems; after all, fluoride activation of adenylyl cyclase had proven useful in understanding hormonal activation of that enzyme. The activation by azide was oxygen-dependent, increased by thiols, was tissue-specific, and demonstrated a time lag of several minutes before the rate of the activation was maximal. We reasoned that these activators were converted to another activating substance. Furthermore, we found that some hemecontaining proteins, catalase, peroxidase, cytochromes, were required for activation by azide, and hemoglobin or myoglobin caused inhibition. 8 -10 These activators also increased cGMP levels in several tissues. 7 When we added azide, hydroxylamine, or nitrite to tracheal smooth muscle preparations, cGMP levels were increased as expected, and the tissue relaxed. 11 It was quite a coincidence that we had tracheal smooth muscle preparations in the laboratory at the time. We developed this preparation of relatively homogeneous smooth muscle because I thought cGMP might cause smooth muscle contraction. We wanted to correlate the accumulation of cAMP and cGMP in homogenous smooth muscle with contraction and relaxation. We initially avoided working with heterogenous smooth muscle preparations such as blood vessels, since we would be unable to identify which cells accumulated the cyclic nucleotides to affect the smooth muscle directly or indirectly. The opposite was true: cGMP accumulated with azide, hydroxylamine, or nitrite and phosphodiesterase inhibitors to cause relaxation of the preparations. 11 We then tested other smooth muscle relaxants such as nitroglycerin and nitroprusside and found that they also activated soluble guanylyl cyclase and increased cGMP levels in several smooth muscle tissues (trachea and intestine) as well as many other tissues. 11,12 The dose-response and time course experiments convinced us that cGMP accumulation was associated with smooth muscle relaxation. Subsequently, we found that cGMP accumulation caused cGMP-dependent protein kinase activation, altering phosphorylation of a number of proteins and dephosphoryla-