Basis of guanylate cyclase activation by carbon monoxide - Europe PMC

10 downloads 0 Views 685KB Size Report
Oct 20, 1994 - and dissociation from carboxy guanylate cyclase have been studied at pH 7.5 ... Soluble guanylate cyclase (GC; EC 4.6.1.2) is a heme protein.
Proc. Nati. Acad. Sci. USA Vol. 92, pp. 2568-2571, March 1995 Biochemistry

Basis of guanylate cyclase activation by carbon monoxide VLADIMIR G. KHARITONOV*, VIJAY S. SHARMA*t, RENATE B. PILZ*, DOUGLAS MAGDEt, AND DORIS KOESLING§ Departments of *Medicine and tChemistry, University of California, San Diego, La Jolla, CA 92093-0652; and §Institute for Pharmacology, Free University of Berlin, Thielallee 67-73, 14195 Berlin, Germany

Communicated by Harvey A. Itano, University of California at San Diego, La Jolla, CA, December 16, 1994 (received for review October 20, 1994)

Knowledge of the three-dimensional structure of the enzyme could help resolve the issue. Unfortunately, this information is not available. In the absence of a crystal structure, information regarding the coordination number of Fe in carboxy derivatives of heme proteins can often be obtained from the rate constants for CO dissociation, as these rate constants for five- and six-coordinate carboxyheme derivatives differ by a factor of 105 (13). We present here results of kinetic studies of CO association with GC and dissociation from GC-CO which support the hypothesis and have important implications regarding the mechanism of GC activity.

Kinetics of CO association with guanylate ABSTRACT cyclase [GTP pyrophosphate-lyase (cyclizing), EC 4.6.1.2] and dissociation from carboxy guanylate cyclase have been studied at pH 7.5 by flash photolysis, yielding rate constants at 23°C of 1.2 ± 0.1 x 105 M-1-sec-1 and 28 ± 2 sec-1, respectively. While the CO combination rate constant is the same as for the T state of hemoglobin, the CO dissociation rate constant is much higher than expected for a six-coordinate carboxyheme protein; yet the absorption spectrum is indicative of a six-coordinate heme. The two observations are reconciled by a reaction mechanism in which CO dissociation proceeds via a five-coordinate intermediate. This intermediate is structurally very similar to the five-coordinate nitrosyl heme derivative of guanylate cyclase and is presumably responsible for the observed 4-fold activation of guanylate cyclase by CO. Thus, we provide a model that explains enzyme activities of the nitrosyl and carboxy forms of the enzyme on the basis of a common mechanism.

METHODS Soluble GC from bovine lung was purified and characterized as described (3). The samples of the native enzyme and its carboxy and nitrosyl derivatives showed Soret absorption bands at 431, 424, and 398 nm, respectively (Fig. 1 and data not shown). The enzyme activity was increased 200-fold by its reaction with NO. These observations are consistent with earlier reports (3-5). Kinetic measurements on GC were made by flash photolysis at 23°C. Details of the laser apparatus and data acquisition and processing have been described (14). Kinetic studies with horse heart myoglobin employed both flash-photolysis and stoppedflow methods. The CO dissociation rate constant was determined by replacing CO with NO (15). Human serum albuminheme (HSA-heme) complex was prepared as described (16). Its carboxy derivative was obtained by equilibrating it with different partial pressures of CO in the presence of small amounts of sodium dithionite. CO combination and dissociation rate constants were determined as in the case of myoglobin. Compositions of reaction mixtures are described in the legends of Figs. 1-3. Absorption spectra were recorded before and after flash-photolysis experiments and confirmed that there was no significant change due to denaturation of the sample.

Soluble guanylate cyclase (GC; EC 4.6.1.2) is a heme protein that catalyzes the conversion of guanosine 5'-triphosphate

(GTP) to guanosine 3',5'-cyclic monophosphate (cGMP) (1). Enzymatic activity of GC is increased about 100- to 200-fold by its reaction with nitric oxide (NO). This change in activity is accompanied by a shift in the Soret absorption band of the enzyme from 431 nm to 398 nm (2-5). Spectroscopic studies with both heme model compounds and hemoglobin have attributed the 398-nm Soret band to a five-coordinate nitrosyl heme (6-10). These and several other observations suggest that the active form of GC is obtained by breaking the bond between the heme iron and the proximal base and that NO is able to achieve this result by virtue of its strong negative trans effect on the bond between Fe and the proximal base (11, 12).

+

NO

-

-

[1]

RESULTS For three wavelengths, Fig. 2 shows reaction time courses of CO rebinding to GC following photolysis of GC-CO by a laser pulse of 5-ns duration. The isosbestic point of the reaction is around 429 nm. Values of k.b, were obtained by fitting the data to a single exponential equation; they varied with CO concentration. In Fig. 3, k.b, is plotted against CO concentration. The intercept and slope yielded rate constants for CO dissociation (1 = 28 + 2 sec-1) and CO combination (1 = 1.2 ± 0.1 X 105 M- 1-sec- 1), respectively. Under identical solution conditions, the CO association rate constant for myoglobin was 1.7 ± 0.1 X 106 M-1'sec-1 and the CO dissociation rate constant was 0.025 ± 0.001 sec-1. In the absence of 50% glycerol the rate constants were as follows: CO combination, 6 + 0.1 x 105 M-1 sec-1; CO dissociation, 0.016

On the basis of data obtained in the present study, we propose as a working hypothesis that the very same mechanism involving a free, nonbonded proximal base (presumably histidine) and a five-coordinate Fe(II)-heme is responsible for the much smaller activation of GC by CO and, perhaps, even for the level of basal activity. A strong objection can be raised immediately: in all known cases CO, unlike NO, shows a strong positive trans effect; binding CO actually increases the strength of the bond to a proximal base (11, 12). We will return to this objection after considering evidence in favor of the hypothesis. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Abbreviations: GC, guanylate cyclase; HSA, human CTAB, cetyltrimethylammonium bromide. tTo whom reprint requests should be addressed.

2568

serum

albumin;

Proc. Natt Acad. Scd USA 92 (1995)

Biochemistry: Kharitonov et al.

2569

NI

a)

0 C



co 0

A2

0.0

0.2

0.1

0.3

0.4

0.5

CO, mM 0.010 350

425

FIG. 3. Characteristic rates for recovery of transient absorbance changes, kobs, plotted against CO concentration. Solution parameters, except CO concentration, are the same as in Fig. 2. Each point is the mean of two bleaches at 424 nm and two absorptions at 435 nm; each of those four traces was recorded as the mean of 100 laser shots. The straight line is a linear least-squares fit to the four points, weighted by their uncertainties, which gave the results listed in the text.

500

Wavelength, nm FIG. 1. Absorption spectra of the native (A) and carboxy (B) [at 1 atmosphere (101.3 kPa) of CO] GC in 25 mM triethanolamine buffer in 50% (vol/vol) glycerol containing 0.5 mM EDTA and 1 mM glutathione at pH 7.5. + 0.001 sec-t. For HSA-heme-CO, the CO combination rate constant was 8.5 0.5 x 106 M-1 sec-1 and the CO dissociation rate constant, 100 + 10 sec-1; both were determined in aqueous buffer (without glycerol) at pH 7.0.

x

10

0.6

DISCUSSION The most important finding of this investigation is the unusual CO dissociation rate constant of GC-CO (28 sec-1), which is

-3

5rF "Lam.aw

WO

0

I

ALliuL.

ply-WMA",

iLA

a

li

lif wrv

wo"-1-T -Ioif f -

.-

r-W rw

.5

20

40

80

60

100

120

-3

4

K

10

2 AUl

2

l

0

20

40

20

40

__

l

_ _. __

l

60

80

100

120

60

80

100

120

:10

lor

Time, msec Transient

0.29

i0-3

x

M

10-7 M; [CO]

1.2 x absorption for GC-CO in aqueous 50% glycerol. All buffer conditions were the same as in Fig. 1. [GC] (equilibrated with 50 kPa of CO and 50 kPa of argon). (Top) Bleaching at 424 nm. (Bottom) Transient absorption

FIG. 2.

at 435 nm.

(Middle) Near the isosbestic point at 429 nm. Each trace is the average of 40 laser shots and was measured with a spectral bandpass of 8 nm. For the calculation of rate constants, traces which were an average of 100 or more laser shots and of considerably better signal/noise ratio were used.

2570

Proc. Natl. Acad ScL USA 92

Biochemistry: Kharitonov et al.

much higher than expected on the basis of values reported for other heme proteins. This, we suggest, is strong evidence for five-coordinate heme-CO. The CO combination rate constant, on the other hand, is similar to that in deoxyhemoglobin and within a factor of 10 of that in myoglobin, in both of which the iron is pentacoordinated and the binding site is free. Rate constants for CO dissociation from five- and sixcoordinate carboxyheme derivatives have been determined for model compounds in dry benzene and aqueous cetyltrimethylammonium bromide (CTAB) at 25°C (13). For

Im-heme-CO

12 -> Im-heme

+

CO,

[2]

where 12 = 0.028 sec-1 in benzene and 0.008 sec-1 in CTAB (Im is imidazole), and for

heme-CO

13

-> heme

+

CO,

[3]

where 13 = 1.3 x 104 sec-I in benzene and 400 sec-1 in CTAB. Based on the ratio of 12 to 13 in benzene, one might expect 13 3000 sec-1 in aqueous CTAB. The fact that it is only 400 has been attributed to the presence of the weak-field ligand H20 in the trans position. Regardless of such fine points, it is clear that in the absence of a trans base, CO dissociates much faster. Equivalent data for oxygen-binding heme proteins is not available. In order to generate five-coordinate CO-hemeprotein, the pH of the solution would have to be lowered to values at which most proteins would denature. However, Beavan et al. (16) have shown that HSA-heme is stable at pH 7.0. The heme binds to the protein mainly by hydrophobic contacts; i.e., heme does not seem to have a coordinate bond between the protein and the heme iron. We extended the investigation to show that the nitrosyl derivative shows a Soret band at 400 nm, close to that observed in GC-NO (398 nm), whereas the CO derivative of HSA-heme shows a Soret band at 414 nm and loses CO with a rate constant (14) of 100 sec-1: 14 HSA-heme-CO

--

HSA-heme + CO.

[4]

This value of 14 iS much higher than the rate constant for CO dissociation from six-coordinate carboxyheme proteins (carboxymyoglobin, carboxy horseradish peroxidase, and carboxyhemoglobin) and is indicative of the presence of a five-coordinate species (HSA-heme-CO). For six-coordinate species, the CO dissociation rates are 0.02 sec-1 for carboxymyoglobin, 7 x 10-5 sec-1 for carboxy horseradish peroxidase, and 0.01 sec-1 and 0.1 sec-1, respectively, for hemoglobin in the R and T states (17-19). The UV-visible spectrum of GC seems, at first, to contradict this interpretation of the kinetic data. The prosthetic group in soluble GC is protoheme. Its Soret absorption band is at 431 nm (5) in the deliganded form (i.e., the native enzyme), suggesting coordination to the protein via an imidazole side chain, as is the case in myoglobin and hemoglobin. On reaction with CO the Soret band shifts to 424 nm, which is the same as in carboxymyoglobin or a CN-heme-CO model compound (20), suggesting that most of the carboxy derivative of GC is six-coordinate. This combination of high CO dissociation rates with a Soret band characteristic of six-coordinate heme iron can be reconciled by the following reaction mechanism for CO dissociation:

GC6-CO

K

k

GC5-CO -> GC + CO,

[5]

where subscripts to GC represent the coordination number of the heme iron in the liganded state. For GC5 the possibility of the presence of a weak-field ligand such as H20 at the free coordination site is not ruled out. Although the actual reaction mechanism required to fit both on- and off-rate constants may well be more complex than assumed in Eq. 5, and in obtaining

(1995)

on- and off-rate constants from plots such as Fig. 3, the only way such a high off-rate constant could be found is if there were a state, fractionally represented, with an even larger rate constant than the observed 28 sec-1, and it is plausible to assume that it is indeed five-coordinate. In Eq. 5 the fivecoordinate species (GC5) exists as a reaction intermediate in preequilibrium with the six-coordinate species. It follows that

kobs = k-K.

[6]

NO activates GC by a factor between 100 and 200 (3, 5), and the Soret absorption spectrum of GC-NO suggests almost complete conversion to five-coordinate species. If we assume that the five-coordinate species, irrespective of the nature of the ligand, is 100-200 times more active than the basal form of the enzyme, then from the reported 4-fold activation of GC by CO (5) we can make an estimate of the percentage of the five-coordinate intermediate induced by binding CO-namely, 3% to 1.5%. Using the mean value (-2 ± 1%) and kobs = 28 sec-1, Eq. 5 requires that k 1 x 103 sec-1. This value of k is intermediate between rate constants for CO dissociation from a five-coordinate heme-CO model compound in benzene (1.3 x 104 sec-1) and in aqueous CTAB (400 sec-1) (13). The discussion above implies that adding CO to heme in GC weakens the proximal bond, although not as much as is the case of NO. It seems that any bond-strengthening trans effect of CO ligation is more than cancelled by steric factors around the prosthetic group. Such a negative trans effect for CO is unexpected. Perhaps the orientation of the proximal base and other nearby side chains is such that the binding of heme to the native enzyme, which is already weak, becomes even weaker as iron moves towards the plane of the porphyrin ring. This effect may be due to increased steric interaction between the proximal base and its vicinal side chains and the porphyrin ring. Steric interactions of this type, particularly due to the orientation of the proximal base, are well characterized in hemoglobin, myoglobin, and model compounds and are considered responsible for the low ligand affinities of the T state of hemoglobin (21-25). The reported weak binding of heme to the protein in the native GC enzyme, which results in heme depletion during enzyme purification unless special precautions are taken (1, 5), may also arise from the steric factor discussed above. We suggest further that it is possible that even in native GC, there could be at all times a small population of a species in which heme is held only by hydrophobic contacts and the proximal base is free. Basal enzyme activity of the native GC can be attributed to such a species. Finally, we note that steric hindrance on the proximal side should lead to unusually low affinities for ligands in GC, even to the point that a ligand like 02, which usually binds less well than NO and CO, may bind very little to GC. The 02 affinity of GC can be estimated from its CO equilibrium constant (1'/l 3 x 103 M-1) with the use of a scaling factor of 40 as is the case with myoglobin. This estimate of equilibrium constant indicates that at ambient 02 tension about 2% of GC will be liganded with 02, which accounts for the fact that experimental measurements did not detect 02 binding (4, 5). At physiological concentrations of 02 and CO, ligation or activation of GC by these two ligands may be insignificant. It seems that both of the enzymatically active derivatives (GCs-CO and GC5-NO) are characterized by two structural features. (i) Heme is held in the protein matrix by hydrophobic contacts only. This packing of heme into the protein provides an overall structure. (ii) The proximal base is free, presumably to participate, independently or in concert with other side chains, in the base-catalyzed conversion of GTP to cGMP as suggested by Senter et al. (26) (Fig. 4). To summarize, the unexpectedly high CO dissociation rate from GC-CO combined with a Soret absorption band indicative of six-coordinate heme (protein-heme-CO) suggests a

Biochemistry: Kharitonov et al.

Proc. NatL Acad ScL USA 92 (1995)

2571

0

pp1-l"';----o PPIO 0

HO'

Gua

+

HO +

NO

IF PP

+

i

of~~~~~~

I

!~~~~~~~~~~~~~~~~~~~_-

.0 F;-

L

L

FIG. 4. Reaction model postulated for base-catalyzed conversion of GTP to cGMP by soluble GC. PPi, pyrophosphate; Gua, guanine; Im, imidazole.

small population of an intermediate in which heme is fivecoordinate (heme-CO). This intermediate is similar to the dominant species in the nitrosyl derivative of GC and is responsible for the observed 4-fold activation of the enzyme by CO. Thus, both CO and NO activate GC by a common mechanism. This work was supported in part by grants from the National Institutes of Health (HL 13581, 48014, 48018). 1. Ignarro, L. J. (1989) Semin. HematoL 26, 63-76. 2. Ignarro, L. J., Adams, J. B., Horwitz, P. M. & Wood, K S. (1986) J. Biol. Chem. 261, 4997-5002. 3. Humbert, P., Niroomand, F., Fischer, G., Mayer, B., Koesling, D., Hinsch, K-D., Gausepohl, H., Frank, R., Schultz, G. & Bohme, E. (1990) Eur. J. Biochem. 190, 273-278. 4. Gerzer, R., Bohme, E., Hofmann, F. & Schultz, G. (1981) FEBS Let. 132, 71-74. 5. Stone, J. R. & Marletta, M. A. (1994) Biochemistry 33, 56365640. 6. Rose, E. J. & Hoffinan, B. M. (1983) J. Am. Chem. Soc. 105, 2866-2873. 7. Suzuki, S., Yoshimura, T., Nakahara, A., Iwasaki, H., Shidara, S. & Matsubara, T. (1987) Inorg. Chem. 26, 1006-1008. 8. Yoshimura, T. & Ozaki, T. (1984) Arch. Biochem. Biophys. 229, 126-135. 9. Maxwell, J. C. & Caughey, W. S. (1976) Biochemistry 15, 388396. 10. Szabo, A. & Perutz, M. F. (1976) Biochemistry 15, 4427-4428. 11. Traylor, T. G. & Sharma, V. S. (1992) Biochemistry 31, 28472849.

12. Yu, A. E., Hu, S., Spiro, T. G. & Burstyn, J. N. (1994) J. Am. Chem. Soc. 116, 4117-4118. 13. White, D. K., Cannon, J. B. & Traylor, T. R. (1979)J. Am. Chem. Soc. 101, 2443-2454. 14. Walda, K N., Liu, X. Y., Sharma, V. S. & Magde, D. (1994) Biochemistry 33, 2198-2209. 15. Moore, E. G. & Gibson, Q. H. (1976) J. Biol. Chem. 251, 27882794. 16. Beaven, G. H., Chen, S.-H., D'Albis, A. & Gratzer, W. B. (1974) Eur. J. Biochem. 41, 539-546. 17. Coletta, M., Ascoli, F., Brunori, M. & Traylor, T. (1986) J. Biol. Chem. 261, 9811-9814. 18. Sharma, V. S., Schmidt, M. R. & Ranney, H. M. (1976) J. Biol. Chem. 251, 4267-4272. 19. Sharma, V. S., Ranney, H. R., Geibel, J. F. & Traylor, T. G. (1975) Biochem. Biophys. Res. Commun. 66, 1301-1306. 20. Antonini, E. & Brunori, M. (1971) Hemoglobin and Myoglobin in TheirReactions with Ligands (Elsevier, New York), pp. 19 and 60. 21. Perutz, M. F. (1979) Annu. Rev. Biochem. 48, 327-286. 22. Sharma, V. S., Geibel, J. F. & Ranney, H. M. (1978) Proc. Natl. Acad. Sci. USA 75, 3747-3750. 23. Makinen, M. W., Houtchens, R. A. & Caughey, W. S. (1979) Proc. Natl. Acad. Sci. USA 76, 6042-6046. 24. Friedman, J. M., Rousseau, D. L., Ondrias, M. R. & Stepnoski, R. A. (1982) Science 218, 1244-1246. 25. Rifkind, J. M. (1987) inAdvances in Inorganic Biochemistry, eds. Eichhorn, G. L. & Marzilli, L. G. (Elsevier, New York), Vol. 7, pp. 155-244. 26. Senter, P. D., Eckstein, F., Mulsch, A. & Bohme, E. (1983)J. Biol. Chem. 258, 6741-6745.