Guanylate cyclase from the rat renal medulla. Physical properties and comparison with adenylate cyclase.

Guanylate cyclase from the rat renal medulla is found in both the soluble and particulate fractions of the cell. Sucrose density gradient centrifugation and gel filtration in H2O and D2O indicate that the enzyme from the soluble cell fraction has the following properties: S20w, 6.3 S; Stokes radius, 54 A; partial specific volume, 0.75 ml/g; mass, 154,000 daltons; f/fo, 1.4; axial ratio (prolate ellipsoid), 7. The addition of 0.1% Lubrol PX to this fraction activates the enzyme and changes thartial specific volume, 0.74 ml/g; mass, 148,000 daltons; f/fo, 1.6; axial ratio (prolate ellipsoid), 11. These findings show that detergent activates the enzyme by changing its conformation and not simply by dispersing nonsedimentable membrane fragments. The dimensions of this guanylate cyclase in detergent are very similar to those of detergent-solubilized adenylate cyclase from the same tissue (Neer, E.J. (1974) J. Biol. Chem. 249, 6527-6531). Guanylate cyclase can be solubilized from the particulate cell fraction with 1% Lubrol PX but has properties quite different from those of the guanylate cyclase in the soluble cell fraction. It is a large aggregate with a value of S20,w of about 10 S, Stokes radius of 65 A, and a mass of approximately 300,000 daltons. However, the peaks of guanylate cyclase activity in column effluents and sucrose density gradients are very broad indicating a mixture of different size proteins. The conditions used to solubilize guanylate cyclase from the particulate fraction also solubilize adenylate cyclase, and the two activities can be separated on the same sucrose gradient. Studies of this sort require a rapid, accurate guanylate cyclase assay. We have developed an assay for guanylate cyclase activity which meets these criteria by adapting the competitive protein binding assay for guanosine cyclic 3':5' monophosphate originally described by Murad et al. (Murad, F., Manganiello, V., and Vaughn, M. (1971) Proc. Natl. Acad. Sci. U.S.A. 68, 736-739).

(prolate ellipsoid), 11. These findings show that detergent activates the enzyme by changing its conformation and not simply by dispersing nonsedimentable membrane fragments. The dimensions of this guanylate cyclase in detergent are very similar to those of detergent-solubilized adenylate cyclase from the same tissue (Neer, E. J. (1974) J. Biol. Chem. 249, 652776531).
Guanylate cyclase can be solubilized from the particulate cell fraction with 1% Lubrol PX but has properties quite different from those of the guanylate cyclase in the soluble cell fraction. It is a large aggregate with a value of sZO,,, of about 10 S, Stokes radius of 65 A, and a mass of approximately 300,000 daltons. However, the peaks of guanylate cyclase activity in column effluents and sucrose density gradients are very broad indicating a mixture of different size proteins. The conditions used to solubilize guanylate cyclase from the particulate fraction also solubilize adenylate cyclase, and the two activities can be separated on the same sucrose gradient.
Studies of this sort require a rapid, accurate guanylate cyclase assay. We have developed an assay for guanylate cyclase activity which meets these criteria by adapting the competitive protein binding assay for guanosine cyclic 3':5' monophosphate originally described by Murad  Guanylate cyclase catalyzes the formation of guanosine cyclic 3':5'-monophosphate from GTP. The enzyme exists in two forms, particulate and soluble, but the proportion of each varies from tissue to tissue. In most mammalian tissue, 60 to 90% of guanylate cyclase activity does not sediment on centrifugation at 100,000 x g (l), although there are exceptions such as the rat small intestine in which virtually all of the activity is membrane-bound (1). The activity of both the particulate and the soluble forms of the enzyme can be increased with nonionic detergents (1,2 The procedure used was analogous to that described for adenylate cyclase (6). The activity measured was the same as that obtained with the protein binding assay. Without detergent, the production of cyclic GMP is a linear function of the amount of protein assayed, as is shown in Fig. 2. In the presence of detergent, the production of cyclic GMP by the supernatant is proportional to the amount of protein, but this is not the case for the pellet or the homogenate.
In these studies the amount of protein used was always within the range which gave a linear response.

AND DISCUSSION
Distribution of Guanylate Cyclase in Cell Homogenutes-When the renal medulla was homogenized in 0.05 M Tris. HCl, pH 7.6, 1 mM EDTA, 1 mM dithiothreitol and centrifuged at 100,000 x g for 45 to 60 min, 70 to 80% of the guanylate cyclase activity of the homogenate was found in the supernatant.
Lubrol PX (0.1%) increases the activity of both this fraction and the pelleted enzyme by a factor of 2 to 4.
The enzyme in the pellet could not be solubilized with 0.1% Lubrol PX but was solubilized when the detergent concentration was raised to 1%. Some guanylate cyclase remains particulate even after two washes with 1% Lubrol PX and 0.2 M Tris.HCl.
A typical example is shown in Table I. The distribution of adenylate cyclase in the same preparation is shown for comparison. These results are qualitative at present since the recovery of the two enzymes is different.
The total activity of adenylate cyclase and guanylate cyclase is similar, the main difference being in the partition between the soluble and particulate forms. The intracellular concentration of cyclic AMP in the kidney is about 1 x lo-' mol/kg of tissue. Cyclic GMP is about 3 to 5 x lo-" mol/kg of tissue (2, [14][15][16]. This 20-to 30-fold difference in cyclic nucleotide concentration does not seem to be reflected in a difference in activity of the respective cyclase.
Physical Properties of Guanylate Cyclase-Figs. 3 and 4 show the pattern of activity obtained by gel filtration and sucrose density gradient centrifugation of supernatant guanylate cyclase. The sedimentation coefficient and Stokes radius experimentally determined for this form of the enzyme, both in the presence and absence of 0.1% Lubrol PX, are given in Table  II.
In the presence of detergent, the sedimentation coefficient becomes smaller by about 1 S compared to that determined without detergent, while the Stokes radius of the enzyme increases from 54 A to 62 A. Such a change could be due to either of two events. First, the enzyme might bind a large amount of detergent. This binding would increase the Stokes radius and would increase the partial specific volume (6), thus slowing the rate of sedimentation.s The mass of the particle would increase. Second, the detergent might cause a conformational change such that the enzyme would become more asymmetric.
It would then have a larger Stokes radius and a smaller sedimentation coefficient but there would be no change in its mass or in its partial specific volume. In this case, the enzyme might bind some detergent but in an amount too small to cause a measurable change in these parameters.
To decide between these alternatives, we measured the partial specific volume of soluble guanylate cyclase by comparing the sedimentation of the enzyme in sucrose density gradients made up in H,O or D,O with and without detergent. Both with and without Lubrol PX the value for 0 is that of a typical soluble protein (see Table II).
To exclude the possibility that detergent is released in D1O, we measured the Stokes radius of guanylate cyclase by gel filtration on columns made up in DzO, with and without 0.1% Lubrol PX. The results were the same as those obtained in H,O. This is in agreement with the findings of Clarke (11)  Measurement of 0 allows the explicit calculation of the molecular weight of the enzyme. This, too, is unchanged in the presence of detergent. Therefore, we conclude that the changes in Stokes radius and sedimentation coefficient produced by the detergent must be primarily due to a conformational change in the enzyme with unfolding of the polypeptide chain. The fact that the molecular weight of soluble guanylate cyclase is low and is not changed by detergent shows that detergent activation is not a matter of exposing cryptic enzyme from nonsedimentable membrane vesicles. There has been some controversy over whether nonionic detergents change the conformation of proteins. Kirkpatrick and Sandberg (17) studied the effect of a number of detergents 'The partial specific volume of Lubrol PX is 0.98 ml/g (information from the manufacturer, ICI, America, Inc.). PX does lead to a measurable conformational change, presumably because of a small amount of specific binding.

7908
The values for the sedimentation coefficient and Stokes radius of guanylate cyclase solubilized from the particulate fraction by 1% Lubrol PX are also given in Table II. These values must be viewed as approximations, for the peak of solubilized guanylate cyclase activity is much broader than that of the marker enzymes used in calibration.
Because it is so broad, the comparison of sedimentation in H,O and D,O gradients is not meaningful, and the partial specific volume of solubilized guanylate cyclase could not be measured. The molecular weight given is calculated using a "typical" U of 0.74 ml/g and is probably underestimated.
The procedure used to solubilize guanylate cyclase from the particulate fraction of the cell also solubilized adenylate cyclase. We were, therefore, able to compare the sedimentation behavior of the solubilized cyclases in the same sucrose gradient. As can be seen in Fig. 5 the two activities are clearly separable. The sedimentation coefficient for adenylate cyclase in the experiment shown was 5.7 S.
The tact that the adenylate cyclase activity sediments as expected from previous experiments (10) shows that the 10 S guanylate cyclase is not the result of nonspecific protein aggregation in the sample.
For purposes of comparison, in Table II we have included the  physical parameters for the predominant form of detergent solubilized adenylate cyclase previously reported by one of us (10). There is a striking similarity between adenylate cyclase solubilized with detergent and guanylate cyclase from the cell supernatant analyzed in Lubrol PX. Illiano et al. (4) have raised the interesting hypothesis that adenylate and guanylate cyclase are the same enzyme which changes its substrate specificity depending on whether or not it is membrane-bound.
This hypothesis cannot be rigorously tested until the two activities have been purified to homogeneity, and the proteins have been characterized. However, the fact that adenylate cyclase and guanylate cyclase can both exist in soluble form and that both activities can be found in membrane fractions argues against a change in substrate specificity which depends only on the state of aggregation or on incorporation into a membrane.
There is general agreement that adenylate cyclase is a plasma membrane enzyme. Guanylate cyclase activity has been variously reported to be 90% soluble in some cells and to be 90% particulate in others (1 is a membrane-bound enzyme in uiuo but differs from adenylate cyclase in the ease with which it is dislodged from the membrane.
An alternative explanation for the two classes of guanylate cyclase activity is that there are in fact two different enzymes presumably with different functions. Which of these alternatives is correct is important to establish because different control mechanisms and different physiological roles would be predicted depending on the conclusions.