| About the journal | | | Subscriptions | | | Authors | | | Users | | | Librarians | | | FAQs |
 |
Key words: cytochrome c´, guanylate cyclase, haem, nitric oxide, X-ray crystallography. Abbreviations used: cyt c´, cytochrome c´; sGC, soluble guanylate cyclase; AXCP, Alcaligenes xylosoxidans cytochrome c´; 5c, 5-coordinate; 6c, 6-coordinate; RR, resonance Raman. 1To whom correspondence should be addressed (e-mail david.lawson@bbsrc.ac.uk). Abstract Cytochrome c´ (cyt c´) is found in the periplasmic space of denitrifying bacteria where it is thought to mediate the transfer of NO between the nitrogen-cycle enzymes dissimilatory nitrite reductase and nitric oxide reductase. It contains a 5-coordinate (5c) His-ligated haem that shares spectroscopic and ligand-binding properties with the haem group in the sensory domain of soluble guanylate cyclase (sGC). The latter is an extremely important enzyme involved in the control of vasodilation and blood clotting. Curiously, the enzyme is activated up to 200-fold by the binding of NO to the haem, whereas the binding of CO gives rise to only a mild stimulation of activity. Through X-ray crystallography we have studied NO and CO binding to cyt c´. CO binds to the distal face to give a 6-coordinate (6c) adduct. By contrast, NO binding gives rise to a 5c adduct through the displacement of the proximal His, to give a novel and unexpected proximal binding mode for NO. These results are also supported by a range of spectroscopies. In the absence of a crystal structure for sGC we propose that cyt c´ provides a structural model for the haem domain of this enzyme and thereby helps to explain the differential effects of NO and CO on its activity. Haem-based sensors Haemoproteins are ubiquitous in nature and display a wide range of different functions including transport, storage and the catalytic activation of substrates. Recently a new class of haemoproteins has emerged that has roles in signalling. These are the haem-based sensors, and relatively few members of this class have been recognized to date [1]. In all cases, these proteins are sensing the presence or absence of diatomic molecules, and signals are transduced via conformational changes when these ligands either bind to, or dissociate from, their haem groups. Several oxygen sensors have been described, including HemAT, an aerotactic transducer [2], the DOS protein of unknown function [3] and FixL. The FixL protein is the most extensively studied. It controls the transcription of the genes that are required for symbiotic bacterial nitrogen fixation, only allowing transcription when O2 levels are sufficiently low to prevent oxidative damage to the nitrogenase enzyme [4,5]. CooA is a transcription factor from the bacterium Rhodospirillum rubrum that controls the transcription of genes involved in the oxidation of CO. The binding of CO to the haem domain activates transcription of the coo operon, enabling the organism to grow on CO as the sole energy source [6,7]. NO is a key signalling molecule in higher organisms [8]. One of the cellular targets for NO is the enzyme soluble guanylate cyclase (sGC), which plays an important role in regulating a diverse range of physiological functions through the elevation of cGMP levels [9]. sGC is composed of an a1 subunit and a haem-containing b1 subunit; the active centre is located at the interface between the two subunits. The single b-type haem is pentaco-ordinate, His-105 being the axial ligand. It produces 6-coordinate (6c) and 5-coordinate (5c) adducts with CO and NO respectively, but is unable to form a stable adduct with O2 [10,11]. Incidently, this is also observed for CooA [6]. This difference in co-ordination, and the formation of the 5c adduct, has been assigned to the trans effect of NO, resulting from the partial transfer of an unpaired electron to the iron. This has a tendency to repel any trans axial ligand from haem [12]. Moreover, NO binding gives rise to a 50–200-fold increase in guanylate cyclase activity, whereas CO binding produces only a 1.4–4-fold stimulation [10,13,14]. To date, there are no crystal structures for sGC, and thus our understanding of its ligand-binding properties relies almost entirely on interpretation of spectroscopic data. The displacement of the proximal His by NO has been implicated as the trigger for activation through a conformational change [15]; a direct role in catalysis has also been proposed for the liberated proximal base [11]. Cytochromes c´ (cyts c´) are a unique class of c-type cytochromes that are found in the periplasm of certain nitrogen-fixing, photosynthetic, denitrifying, methanotrophic and sulphur-oxidizing bacteria [16]. Although the physiological function of cyt c´ has yet to be proven, several studies have indicated that the binding of NO to the haem may be functionally important [17,18]. Indeed, in denitrifying bacteria it has been implicated in the transfer of NO between dissimilatory nitrite reductase and NO reductase during the denitrification process, effectively sequestering potentially toxic levels of NO [19]. Cyt c´ is of relevance to the haem sensors, as it shares similar atypical spectroscopic and ligand-binding properties with sGC and to some extent with CooA [10,16,20,21]. Crystal structures of Alcaligenes xylosoxidans cyt c´ (AXCP) The structures of several cyts c´ have been determined by X-ray crystallography [22,23], although this review will focus on only one, AXCP, which was originally determined at ambient temperature by Dobbs et al. [24]. AXCP exists as a homodimer, each subunit being composed of an anti-parallel four-helix bundle containing a partially exposed 5c c-type haem (Figure 1). A conserved Cys-Xaa-Xaa-Cys-His motif provides the proximal His ligand (residue 120 in AXCP) and two Cys–thioether bonds to the haem. Access to the vacant sixth co-ordination site on the distal face of the haem is mediated by the side chain of a hydrophobic residue (Leu-16 in AXCP). Indeed, the distal pocket is generally hydrophobic, sterically crowded and less solvent-accessible than the proximal pocket, which is positively polarized, less crowded and readily accessible to the surrounding solvent [24]. In order to understand the molecular details of the unusual ligand-binding properties of cyts c´, the crystal structures of AXCP were subsequently determined under cryogenic conditions in the oxidized, reduced, CO-bound and NO-bound forms [25]. The oxidized and reduced structures were determined at resolutions of 2.05 and 1.90 Å respectively and are very similar (Figures 2a and 2b). In both, the iron lies out of the haem plane, which is distinctly puckered, by approx. 0.3 Å towards the proximal His to which it is ligated; Leu-16 in the distal pocket occludes the sixth co-ordination site. Additionally, the proximal His is hydrogen-bonded to an ordered water molecule through its Nd1. However, the structures differ from one another in one important respect: in the oxidized structure, the planar guanidinium moiety of Arg-124, which lies adjacent to His-120, is parallel to the haem; in the reduced structure it stacks against His-120. The CO-bound form of AXCP was produced by the incubation of reduced crystals in solutions saturated with CO. This structure, determined at 1.95 Å resolution, differs markedly with respect to the reduced structure: CO is bound in a near-linear conformation (Fe–C–O angle of 167°) to the distal face of the haem, thereby displacing the side chain of Leu-16 to one side; the iron is now within the plane of the haem, which is essentially flat (Figure 2c) [25]. In addition, a second CO is hydrogen bonded to Nd1 of His-120, mimicking the interaction seen in other haemoproteins between the proximal His and either a main-chain carbonyl or a side-chain carboxyl group [26,27]. On superimposition of the CO-bound structure upon the reduced structure, it is apparent that the positions of the proximal His and the iron are essentially unchanged: it is the movement and flattening of the haem that restores the iron into the plane of the haem. This is in contrast to the situation in haemoglobin where the binding of O2 pulls the iron into the plane of the haem. This in turn pulls the proximal His towards the haem, thereby triggering allosteric changes [28].The NO-bound form of AXCP was produced by the incubation of reduced crystals in solutions saturated with NO. As expected from spectroscopy [20], the crystal structure, resolved at 1.35 Å, shows that the NO adduct is 5c with the proximal His displaced [25]. Unexpectedly, however, the NO is ligated to the proximal face of the haem and is present as two equal conformers with an average Fe–N–O angle of 128° (Figure 2d). In fact, overall this structure overlaps more closely with the reduced structure than does the CO-bound structure, as the positions of the haem, the iron and Leu-16 are closely similar. The major differences are the ligation of the NO, the displacement of His-120 and the flipping of Arg-124 to an orientation similar to that seen in the oxidized structure, where it stacks against the haem plane. In this conformation, Arg-124 can hydrogen-bond to one of the NO conformers. Spectroscopic studies of AXCP AXCP has been studied extensively by spectroscopy and the results of these studies not only support the crystallographic analysis, but also provide additional kinetic and mechanistic insights. Excitation of the haem at 413.1 nm yields characteristic resonance Raman (RR) frequencies that identify the co-ordination number and spin state of the iron. These measurements show that the ferrous haem is 5c in both the absence of ligands and the presence of NO. This agrees with the crystal structures where, in both cases, the iron is out of the haem plane and there is only one axial ligand. By contrast, the CO-bound form gives RR frequencies characteristic of 6c low-spin ferrous iron. Again this is consistent with the crystallographic analysis where the iron is co-ordinately saturated and in the haem plane [29]. Using RR spectroscopy it is possible to predict the environments of NO and CO through the analysis of backbonding correlations. In the case of AXCP these indicate that NO is bound in a positively polarized environment, in line with it being located in the proximal pocket adjacent to Arg-124, whereas the results for CO indicate that it resides in a neutral environment, consistent with it lying in the distal pocket [29]. Furthermore, this analysis shows that the proximal His is protonated in the CO-bound form, in contrast with the situation in the absence of CO when it has imidazolate character. This agrees with the CO-bound crystal structure, because the hydrogen bond seen between His-120 and the second CO molecule requires that Nd1 of the His is protonated. Finally, RR spectroscopy on the NO adduct gives a broad stretching frequency for the N–O bond that could be interpreted as multiple NO conformers [29], again in accordance with the crystallographic data where two alternative conformations of NO were observed [25]. Pre-steady-state studies of changes in the visible absorption spectra on the binding of NO to reduced AXCP unambiguously demonstrated the initial formation of an intermediate 6c-NO species (Emax 415 nm), preceding the production of a 5c-NO adduct (Emax 395 nm) [30]. The 6c-NO intermediate corresponds to a distally bound NO with the proximal His still ligated to the haem. This species presumably resembles the 6c CO adduct in structure. Second-order kinetics were observed for both the formation of the initial 6c-NO adduct and the subsequent formation of the 5c-NO adduct, indicating that both steps are dependent on the concentration of NO. The rate constants were (4.6±0.2)×104 M-1·s-1 and (8.1±0.7)×103 M-1·s-1 respectively [30]. The appearance and disappearance of the 6c-NO species and the appearance of the 5c-NO product were also monitored using stopped-flow Fourier transform IR spectroscopy, showing similar kinetics and a dependence on the NO concentration for both steps [31]. Discussion When combined together, the X-ray analysis and the spectrosopic results for AXCP can be rationalized according to the following model for NO binding: (i) NO binds to the distal face of the ferrous haem to give the 6c-NO adduct, weakening but not breaking the His–Fe bond due to the repulsive trans effect of NO [12]; (ii) a second NO displaces the proximal His and binds to the proximal face of the haem, giving a putative transient 6c dinitrosyl species and (iii) the second NO repels the first, distally bound, NO to leave the 5c-NO adduct (Figure 3). An alternative mechanism whereby the His must spontaneously dissociate from the haem to give the 5c-NO adduct is ruled out by the second-order kinetics [30] and, moreover, would leave the NO on the distal side of the haem, which is inconsistent with the crystallographic analysis [25]. What are the factors that govern the mechanism and kinetics of NO binding to AXCP? Firstly, the formation of the 6c-NO intermediate and its conversion to the 5c-NO are both slow steps [30] because conformational changes are involved (see above and Figure 2) [25]. Steric hindrance in the distal pocket may be more conducive to ligands that prefer a linear conformation, such as CO, but discriminate against those adducts that favour a bent geometry, such as O2 and NO [32]. Furthermore, O2 requires a strong electron donor, such as His, as the trans ligand and thus cannot bind in the proximal pocket like NO. When NO is the ligand, it binds first to the vacant distal site, where it may be forced into an unfavourable configuration, consistent with the RR Fe–NO stretching frequency [30]. When the second NO binds to the proximal face, it can adopt the preferred bent geometry and, moreover, can hydrogen-bond to Arg-124. This is clearly the most stereochemically and electrostatically favoured position, and thus it is the first, distally bound, NO that is displaced through trans repulsion.Several attempts have been made to model the NO-mediated activation process in sGC, some of which invoke a 6c-nitrosyl intermediate [11,15,33–35]. The Marletta group has also postulated alternative NO binding sites [15] and a requirement for a second NO, possibly attacking on the proximal side [35]. However, prior to the crystallographic analysis of AXCP [25], all of these mechanisms assumed that the product is a 5c distally bound NO adduct – the so-called ‘distal dogma’. Nevertheless, a transient dinitrosyl ferric haem intermediate has been proposed for haemoglobin [36], and the binding of NO and CO to opposite faces of the haem in CooA has also been postulated [6]. Because of the obvious similarities between the two systems [10,16,20,21], cyt c´ has been put forward as a structural model for the haem domain of sGC [25,29–31], in the absence of any crystal structures for the latter. Recently, two conflicting reports have been published in this area. The first describes the use of classical receptor theory to reinterpret the data from the Marletta group [35], and proposes that these results are consistent with a single NO binding event to the distal face of the haem [37]. By contrast, the second report, which comes from the Marletta group, discounts this theory and states that the initial interpretation of these data was essentially correct [38]. Furthermore, they have revised their ideas to embrace the activation model based on the AXCP data [25,29–31]. The fact that CO results in only mild stimulation of sGC activity suggests that the main activation event(s) occurs during the conversion of the 6c-NO adduct to the final 5c-NO product. It still remains a distinct possibility that conformational changes induced by the cleavage of the bond to the proximal His are important [15], and it has recently been speculated that the putative dinitrosyl intermediate could have substantial guanylate cyclase activity [38]. NO release from sGC in vitro may take several minutes [39], whereas much higher rates are observed in vivo [40], suggesting that some additional factor is required to deactivate the enzyme [38,41]. It is possible that forcing the proximal His to rebind may promote NO dissociation. It has also been postulated that the loss of NO-bound haem may be a mechanism whereby sGC is down-regulated in vivo, on account of the tendency for preparations of sGC to lose the haem group [12]. An alternative theory is that the haem becomes oxidized, which would significantly lower its affinity for NO [32]. Indeed, it has been proposed that cyt c´ loses NO through the coupled transfer of the ligand and an electron to NO reductase during denitrification (R.R. Eady, unpublished work). In conclusion, the combined crystallographic and spectroscopic analysis of AXCP provides a more or less complete picture of the complex ligand binding events that occur upon the addition of NO. We believe that the comparison with sGC is a valid one, but AXCP is unlikely to provide us with information concerning how the NO signal is transduced from the haem domain to the active centre of sGC. Clearly there remains a pressing need for a crystal structure of sGC. We thank the Biotechnology and Biological Sciences Research Council and National Institutes of Health grant GM 34468 (to C.R.A.) for financial support. References 1 Rodgers, K.R. (1999) Curr. Opin. Chem. Biol. 3, 158–167 2 Hou, S., Larsen, R.W., Boudko, D., Riley, C.W., Karatan, E., Zimmer, M., Ordal, G.W. and Alam, M. (2000) Nature (London) 403, 540–544 3 Delgado-Nixon, V.M., Gonzalez, G. and Gilles-Gonzalez, M.A. (2000) Biochemistry 39, 2685–2691 4 Gilles-Gonzalez, M.A., Gonzalez, G. and Perutz, M.F. (1995) Biochemistry 34, 232–236 5 Pellequer, J.L., Brudler, R. and Getzoff, E.D. (1999) Curr. Biol. 9, R416–R418 6 Reynolds, M.F., Parks, R.B., Burstyn, J.N., Shelver, D., Thorsteinsson, M.V., Kerby, R.L., Roberts, G.P., Vogel, K.M. and Spiro, T.G. (2000) Biochemistry 39, 388–396 7 Thorsteinsson, M.V., Kerby, R.L., Youn, H., Conrad, M., Serate, J., Staples, C.R. and Roberts, G.P. (2001) J. Biol. Chem. 276, 26807–26813 8 Cooper, C.E. (1999) Biochim. Biophys. Acta 1411, 290–309 9 Denninger, J.W. and Marletta, M.A. (1999) Biochim. Biophys. Acta 1411, 334–350 10 Vogel, K.M., Hu, S.Z., Spiro, T.G., Dierks, E.A., Yu, A.E. and Burstyn, J.N. (1999) J. Biol. Inorg. Chem. 4, 804–813 11 Sharma, V.S. and Magde, D. (1999) Methods 19, 494–505 12 Traylor, T.G. and Sharma, V.S. (1992) Biochemistry 31, 2847–2849 13 Tomita, T., Ogura, T., Tsuyama, S., Imai, Y. and Kitagawa, T. (1997) Biochemistry 36, 10155–10160 14 Stone, J.R. and Marletta, M.A. (1995) Biochemistry 34, 16397–16403 15 Stone, J.R. and Marletta, M.A. (1996) Biochemistry 35, 1093–1099 16 Meyer, T.E. and Kamen, M.D. (1982) Adv. Protein Chem. 35, 105–212 17 Yoshimura, T., Iwasaki, H., Shidara, S., Suzuki, S., Nakahara, A. and Matsubara, T. (1988) J. Biochem. (Tokyo) 103, 1016–1019 18 Ferguson, S.J. (1991) Biochim. Biophys. Acta 1058, 17–20 19 Moir, J.W.B. (1999) Biochim. Biophys. Acta 1430, 65–72 20 Iwasaki, H., Yoshimura, T., Suzuki, S. and Shidara, S. (1991) Biochim. Biophys. Acta 1058, 79–82 21 Stone, J.R. and Marletta, M.A. (1994) Biochemistry 33, 5636–5640 22 Shibata, N., Iba, S., Misaki, S., Meyer, T.E., Bartsch, R.G., Cusanovich, M.A., Morimoto, Y., Higuchi, Y. and Yasuoka, N. (1998) J. Mol. Biol. 284, 751–760 23 Tahirov, T. H., Misaki, S., Meyer, T.E., Cusanovich, M.A., Higuchi, Y. and Yasuoka, N. (1996) J. Mol. Biol. 259, 467–479 24 Dobbs, A.J., Anderson, B.F., Faber, H.R. and Baker, E.N. (1996) Acta Crystallogr. D 52, 356–368 25 Lawson, D.M., Stevenson, C.E.M., Andrew, C.R. and Eady, R.R. (2000) EMBO J. 19, 5661–5671 26 Poulos, T.L. and Kraut, J. (1980) J. Biol. Chem. 255, 8199–8205 27 Harutyunyan, E.H., Safonova, T.N., Kuranova, I.P., Popov, A.N., Teplyakov, A.V., Obmolova, G.V., Valnshtein, B.K., Dodson, G.G. and Wilson, J.C. (1996) J. Mol. Biol. 264, 152–161 28 Perutz, M.F. (1970) Nature (London) 228, 726–739 29 Andrew, C.R., Green, E.L., Lawson, D.M. and Eady, R.R. (2001) Biochemistry 40, 4115–4122 30 Andrew, C.R., George, S.J., Lawson, D.M. and Eady, R.R. (2002) Biochemistry 41, 2353–2360 31 George, S.J., Andrew, C.R., Lawson, D.M., Thorneley, R.N.F. and Eady, R.R. (2001) J. Am. Chem. Soc. 123, 9683–9684 32 Tsai, A. (1994) FEBS Lett. 341, 141–145 33 Makino, R., Matsuda, H., Obayashi, E., Shiro, Y., Iizuka, T. and Hori, H. (1999) J. Biol. Chem. 274, 7714–7723 34 Zhao, Y.D., Hoganson, C., Babcock, G.T. and Marletta, M.A. (1998) Biochemistry 37, 12458–12464 35 Zhao, Y., Brandish, P.E., Ballou, D.P. and Marletta, M.A. (1999) Proc. Natl. Acad. Sci. U.S.A. 96, 14753–14758 36 Addison, A.W. and Stephanos, J.J. (1986) Biochemistry 25, 4104–4113 37 Bellamy, T.C., Wood, J. and Garthwaite, J. (2002) Proc. Natl. Acad. Sci. U.S.A. 99, 507–510 38 Ballou, D.P., Zhao, Y., Brandish, P.E. and Marletta, M.A. (2002) Proc. Natl. Acad. Sci. U.S.A. 99, 12097–12101 39 Brandish, P.E., Buechler, W. and Marletta, M.A. (1998) Biochemistry 37, 16898–16907 40 Bellamy, T.C. and Garthwaite, J. (2001) J. Biol. Chem. 276, 4287–4292 41 Negrerie, M., Bouzhir, L., Martin, J.L. and Liebl, U. (2001) J. Biol. Chem. 276, 46815–46821 42 Kraulis, P.J. (1991) J. Appl. Crystallogr. 24, 946–950 43 Merritt, E.A. and Bacon, D.J. (1997) Methods Enzymol. 277, 505–524 Received 16 December 2002 © 2003 Biochemical Society |
| ||||||||||||||||
 |
Journal home page | |||||||||||||||||
| Current issue | |||||||||||||||||
|
Browse archive by issue | |||||||||||||||||
|
Browse archive by meeting | |||||||||||||||||
|
Search archive | |||||||||||||||||
|
Medal lectures | |||||||||||||||||
|
Biochemical Society Symposia |
