- Email: [email protected]
Mechanism of ubiquinol oxidation by the cytochrome bc1 complex: pre-steady-state kinetics of cytochrome bc1 complexes containing site-directed mutants of the Rieske iron-sulfur protein
BIOCHIMICA ET BIOPHYSICA ACTA
Biochimica et Biophysica Acta, 1365 (1998) 125-134
ELSEVIER
Mechanism of ubiquinol oxidation by the cytochrome bc 1 complex: pre-steady-state kinetics of cytochrome bc 1 complexes containing site-directed mutants of the Rieske iron-sulfur protein Chris Snyder, B e m a r d L. T r u m p o w e r * Department of Biochemistry, Dartmouth Medical School, Hanover, NH 03755, USA
Received 2 February 1998; received in revised form 4 March 1998; accepted 4 March 1998
Abstract To facilitate characterization of mutated cytochrome bCl complexes in S. cerevisiae we have developed a new approach using a rapid scanning monochromator to examine pre-steady-state reduction of the enzyme with menaquinol. The RSM records optical spectra of cytochromes b and Cl at 1-ms intervals after a dead time of 2 ms, and menaquinol fully reduces both cytochromes b H and c~ and a portion of cytochrome b L. The rapid-mixing, rapid-scanning monochromator methodology obviates limitations inherent in previous rapid kinetics methods and permits measurements of rates exceeding 200 s-~. To document the validity of this methodology we have examined the reduction kinetics of the cytochrome bCl complexes from wild-type yeast and yeast that lack ubiquinone. The results establish that menaquinol reacts via the Q cycle pathway both in the presence and absence of ubiquinone. From analyzing bc~ complexes containing Rieske proteins in which the midpoint potential of the iron-sulfur cluster has been altered from +280 to + 105 mV, we propose a mechanism in which the protonated quinol displaces a proton from the imidazole nitrogen of one of the histidines that is a ligand to the iron-sulfur cluster and forms a quinol-imidazolate complex that is the electron donor to the redox active iron. © 1998 Elsevier Science B.V. Keywords: Cytochrome bc~ complex; Q cycle; Rieske iron-sulfur protein; Ubiquinol; Cytochrome c reductase; Saccharomyces cerevisiae
1. Introduction
To facilitate characterization of cytochrome bc~ complexes, in which the kinetic properties of the enzyme have been altered by site-directed mutagenesis of the genes in S. cerevisiae, we have developed a new approach using a rapid-scanning monochromator to examine pre-steady-state reduction of the enzyme on a ms time scale after initiating reduction of the enzyme with menaquinol. The rapidCoiTespondmg author.
scanning monochromator allows simultaneous monitoring of the pre-steady-state reduction of cytochromes b and c 1 with a time resolution of 1 ms after a dead time for mixing of 2 ms, and the potential of menaquinol is low enough to permit partial reduction of cytochrome b L in addition to cytochromes b H and c~. Although rapid scan spectrophotometry has recently been applied to the bovine bc 1 complex [9], the kinetic measurements were limited by a relatively slow 25-ms mixing time and were restricted to examination of the oxidation kinetics due to the lack of a suitably low potential quinol substrate. Photo-
0005-2728/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII: S0005-2728(98)00052-8
126
C. Snyder, B.L. Trumpower / Biochimica et Biophysica Acta 1365 (1998) 125-134
synthetic reaction centers have also been used to drive reduction of the bc~ complex with responses in the ms time range [4,13], but this approach is limited by the fact that a single flash incompletely reduces the cytochromes, and multiple flashes elicit simultaneous reoxidation and reduction. The rapid-mixing, rapid-scanning monochromator (RMRSM) methodology described here obviates all of these problems and permits measurements of rates exceeding 200 S
-1
2. Results and discussion
Menaquinol was chosen as a substrate for reduction of the bc~ complex because it is more strongly reducing (Em7=-74 mV, Ref. [6]) than ubiquinol ( E m T = q - 9 0 mV, Ref. [10]), allowing a greater percentage of cytochrome b to be reduced under equilibrium conditions. Menaquinol has the additional advantage that the quinol, but not the quinone form, is soluble at high concentrations in aqueous solutions, thus obviating problems that might result from the use of organic solvents. From the optical spectra in Fig. 1A one can calculate that menaquinol completely reduced cytochrome c I and approximately 70% of the dithionite reducible cytochrome b. This result is in agreement with the midpoint potentials of menaquinol and the b cytochromes (bH= +90 mV and b L = - 3 0 mV; Ref. [2,10]). At pH 6.0, the predicted extent of reduction of b H is >99% and b L is approximately 20%. Assuming that b H and b e contribute 70 and 30%, respectively, to the total absorbance change at 563 nm, and that the midpoint potential of the b cytochromes is pH independent [12]), the predicted reduction of cytochrome b is 76%, close to the 70% observed experimentally. Eh, and not E m, of the menaquinol/menaquinone couple should be the parameter that determines the extent of b L reduction. However, the observed extent of reduction more closely agrees with what would be predicted from the E m than from the E h. This may indicate that the reported E m of menaquinol [6] is incorrect, with the consequence that the observed extent of reduction would agree more closely with E h. However, we have observed that the extent of b reduction with ubiquinol as substrate also agrees with the E m of the ubiquinol/ubiquinone couple rather
than with the E n (results not shown). An alternative explanation for the less than predicted reduction of cytochrome b is that the extent of reduction of the b hemes through the enzyme-catalyzed pathways does not respond to the potential (Eh) of the quinol/ quinone pool in the absence of mediators. The traces in Fig. 1B illustrate time-resolved optical spectra collected by the rapid-scanning monochromator. In this experiment the bc 1 complex was reduced with a very low concentration (6 IxM) of menaquinol in order to extend the reaction over several seconds. Following a 2-ms dead time, the RSM collects optical spectra every 1 ms, but only 40 spectra, at 50-ms intervals, are displayed in this figure. This low concentration of menaquinol permits a clear display of the triphasic reduction of cytochrome b (see Ref. [2] for a review) when the three-dimensional data from Fig. 1B are analyzed at the wavelength diagnostic of cytochrome b, as shown in Fig. 2. The inset in Fig. 2 shows the first spectra collected in real time during triphasic reduction, and spectra averaged over 10-ms intervals confirm that cytochrome b undergoes partial reoxidation. As a prerequisite to using RMRSM to dissect the kinetics in genetically engineered bc I complexes we performed control experiments to establish that menaquinol reacts through center P in the Q cycle pathway, and alternatively can reduce the b cytochromes through center N if center P is blocked. For this purpose we examined the effects of center N and center P inhibitors on the reduction of cytochromes b and c 1. The time course of reduction of the cytochrome bcl complex when it is rapidly mixed with menaquinol is shown in Fig. 3. In the absence of inhibitors, both cytochromes b and c 1 are reduced by 50 IxM menaquinol within 1 s. When the center P inhibitor stigmatellin was added it blocked the reduction of cytochrome c I (Fig. 3, middle panel), and when both stigmatellin and the center N inhibitor antimycin were added, they blocked the reduction of both cytochromes b and c 1 (Fig. 3, bottom panel). These results establish that menaquinol cannot bypass these center P and center N inhibitors. That menaquinol reacts enzymatically through the Q cycle pathway was confirmed using the Rieske iron-sulfur protein S183C mutant [3]. In this mutant, the Rieske apoprotein is stably incorporated into the cytochrome bc I complex, even though it lacks the
C. Snyder, B.L Trumpower / Biochimica et Biophysica Acta 1365 (1998) 125-134
127
^ 0.05 ~+DT
J!
L.
0 es
.(
v 500
?
I
I
F
I
525
550
575
600
625
Wavei en9th (nm)
B
0.04
¢p
J~
2.5
-
0.02 -~
set:)
1 T ¢ l I v ~ ' ! ~-'11~.~!.1! I I I I I i /
U
580 Fig, 1. (A) Optical spectraof 2 I~M purified yeast cytochrome bc~ complex after reduction with 50 I~M menaquinol (MQ) or dithionite (DT). Oxidized cytochrome bc 1 complex in 100 m M sodium phosphate, pH 6.0, containing 2 mM EDTA and 0.1 m g / m l dodecyl maltoside was reduced with menaquinol or dithionite and the resulting difference spectra of reduced versus oxidized enzyme were recorded in the split beam mode at 0.5-nm intervals with an Aminco DW 2A spectrophotometer equipped with an On-Line Instruments Systems, Inc. computer interface. (B) Time-resolved optical spectra obtained by stopped-flow rapid scanning spectrophotometry during reduction of the bc 1 complex by menaquinol. In this experiment 2 ~M bc I complex was reduced by rapid mixing with 6 txM menaquinol. Spectra were recorded on an OLIS rapid-scanning monochromator equipped with a 1200qine/mJn grating blazed at 500 nm, which yields a spectrum of 75 nm width at a resolution of 0.4 nm. Data were collected at 1000 scans/s, and the spectra were centered at 555 nm. The dead time of the stopped-flow mixer is approximately 2 ms, and the end of this period was chosen as time zero.
C. Snyder, B.L. Trumpower I Biochimica et Biophysica Acta 1365 (1998) 125-134
128
cyt b (563.3 nm)
0.06 1
,
/
,,
(-
..Q
"
I,,,
0 Ill
0.04
0.02
8
c B
_
s4o
-
sho Wavel ength (nm)
i
i
0.6
1.2
i
1.8
i
2.4
..... i
3.0
)
3.6
T i m e (s) Fig. 2. Triphasic reduction of cytochrome b observed by time-resolved optical spectra obtained during reduction of cytochromes b and c~ by menaquinol. The traces at 563.3 and 554.6 nm were derived from the three-dimensional data set shown in Fig. lB. The inset shows time-averaged optical spectra obtained across the 10-ms intervals indicated by the solid bars at points A, B, and C at 150 and 555 ms and at 4 s during the reaction.
iron-sulfur cluster. The absence of functional ironsulfur protein will block enzymatic oxidation of menaquinol at center P, but should allow reduction of cytochrome b through center N. In this mutant we no longer observed reduction of cytochrome c~ by menaquinol (data not shown), and in the presence of antimycin reduction of cytochrome b was also blocked (data not shown). These results confirm that menaquinol reduces the cytochrome bc~ complex through its enzymatic pathways, and validate the use of menaquinol as a reductant to study the kinetics of electron transfer within the cytochrome bc~ complex. The concentrations of menaquinol used in these reactions (<100 t~M) are less than saturating, so that the pre-steady-state kinetics are second order. Varying the concentration of menaquinol over a 20-fold range in the presence of either antimycin or stigmatellin, and fitting of the kinetic data, allowed the calculation
of second-order rate constants for reduction of both cytochromes b and c I through center P and for reduction of cytochrome b through center N. The rates of cytochrome b reduction through center P are best fitted by a two-component curve, consisting of a fast and a slow phase with second-order rate constants of 1.3)<106 and 8.8×104 M -1 s -~. As shown in Fig. 4, reduction of cytochrome c l is kinetically matched to the second, slow phase of b reduction, with a rate constant of 7.5X10 ~ M -I s -~ One explanation for these results is that when the first molecule of menaquinol is oxidized one electron equilibrates between iron-sulfur protein and cytochrome c~, and the second electron is transferred to cytochrome b. Under the conditions used in these experiments, the midpoint potential of the iron-sulfur protein is >60 mV more positive than that of cytochrome c~, with the result that the electron from
C. Snyder, B.L. Trumpower I Biochimica et Biophysics Acta 1365 (1998) 125-134
Time
129
(ms)
Fig. 3. Reduction of the cytochrome bc, complex through center P and center N by menaquinol. The top panel shows reduction of 2 FM bc, complex by 50 p,M menaquinol in the absence of inhibitors. The middle panel shows the inhibition of c, reduction by stigmatellin, and the bottom panel shows the inhibition of b and c, reduction by stigmatellin plus antimycin.
130
C. Snyder, B.L. Trumpower / Biochimica et Biophysica Acta 1365 (1998) 125-134 •848
cyt b
I
I
I
I
458
68B
758
I
158
~
Ti me I o I~
clt b , , l ° . (-)
(ms) 1
.,oo J~
25
75
50
75
! 00/
H.mi~,~1q i ~ Cyt b, f a s t (El)
25
I
I
I
25
50
75
Menaquinol
1O0
M)
Fig. 4. Kinetics of reduction of the cytochrome bc 1 complex by menaquinol through center P. The traces in A show the time course of reduction of cytochrome b and cytochrome c I when 2 IxM cytochrome bc~ complex is reduced by 50 I~M menaquinol in the presence of antimycin. The plot in (B) shows the effects of increasing menaquinol concentration on the rates of reduction of cytochrome b and cytochrome cl. The inset shows the effects of increasing menaquinol concentration on the slow rates of cytochrome b and cytochrome c~ reduction.
the first molecule of menaquinol oxidized is predominately located in the iron-sulfur protein. Only when the first electron is localized to c 1 can the second molecule of menaquinol be oxidized and transfer
electrons to iron-sulfur protein and cytochrome b. This mechanism is consistent with the results where the slow rate of b reduction matches the rate of cytochrome c I reduction.
C. Snyder, B.L. Trurnpower / Biochimica et Biophysica Acta 1365 (1998) 125-134
From the traces in the top panels of Figs. 4 and 5, which show reduction of cytochrome b through center P and center N with 50 ~M menaquinol, it appears that the former reaction is significantly faster than the latter. This was confirmed by the secondorder rate constant for the reaction through center N,
131
2.5× 105 M -~ s -~ (Fig. 5, bottom panel). This 5-fold difference in rates through center P and center N indicates that these rates are primarily governed by the thermodynamics of the electron transfer reactions. This is consistent with the notion that the reactive head group of ubiquinone can remain resident in the
.868
ma
qD 0 e-
.0 L
L~
•8 4 5
•838
0 .0
~
~/
f
~
~ 1 ~
~
,..
~#lam,.~u.~ ~_. mmk.lm'--Wki~ml-mm--J~m~. ).Jd" e v..?e,,.iF--i/-,'. ~p,~ ~'I ~ r
,
14sec -1
•8 1 5
I 158
I 38@
I
I
458
I
688
758
Ti me (ms) 25 20 I
VI
""
15
¢D 4-1
i
10 5 [] i
I
I
I
I
20
40
60
80
100
Menaquinol
(i,. M)
Fig. 5. Kinetics of reduction of cytochrome b by menaquinol through center N. The trace in A shows the time course of reduction of cytochrorne b when 2 p~M cytochrome bc~ complex is reduced by 50 txM menaquinol in the presence of stigmatellin. The plot in (B) shows the effects of increasing menaquinol concentration on the rate of reduction of cytochrome b through center N.
132
C. Snyder, B.L. Trumpower /Biochimica et Biophysica Acta 1365 (1998) 125-134
bc~ complex for multiple turnovers of the enzyme, and that the accessibility of quinone to the two sites is essentially identical. The preceding kinetic analysis also indicated that there was a secondary, slow reduction of cytochrome c~, which occurred at a rate of <0.01 s ], and this slow reduction was also observed in the presence of stigmatellin. Cytochrome c~ was also slowly reduced in the Rieske iron-sulfur protein S183C mutant, which indicates that menaquinol can directly reduce cytochrome c l in an iron-sulfur protein-independent manner. However, this non-enzymatic rate is several orders of magnitude slower than the iron-sulfur protein-catalyzed reduction of cytochrome c~ through center P, and it does not impact on the kinetics of cytochrome c~ reduction, even when these rates are lowered 10-fold by mutational changes in the ironsulfur protein. We have also examined the reduction kinetics of the cytochrome bc~ complex from yeast mutants that lack endogenous ubiquinone (Fig. 6). The RMRSM analysis of the isolated bc~ complex reveals individual properties of the two quinol/quinone oxidoreductase sites, center P and center N, in their reactivities with menaquinol and menaquinone. Reduction of the cytochromes b and c] through center P and reduction of cytochrome b through center N by menaquinol in the absence of ubiquinone is inhibited by stigmatellin (Fig. 6, middle panel) and antimycin (Fig. 6, bottom panel). The reactivity of menaquinol with bc~ complex lacking ubiquinone may lead to some novel insights into the mechanism of triphasic reduction of cytochrome b. Notably, concentrations of menaquinol that result in triphasic reduction of cytochrome b when ubiquinone is present (Fig. 3, top panel), result in apparently monophasic b reduction when ubiquinone is absent (Fig. 6, top panel), and the rate closely matches the first, rapid phase of the triphasic reduction. The absence of a reoxidation phase suggests that menaquinone cannot reoxidize the b cytochromes at center N, or that reoxidation through center N is slow relative to the rate of reduction through center P, consistent with the low potential of menaquinone relative to cytochrome b~. The crystal structure of the Rieske iron-sulfur protein [5] and the crystal structures of the mitochondrial cytochrome bc~ complexes [1,14,15] pro-
vide an opportunity to test several aspects of ironsulfur protein function in the protonmotive Q cycle mechanism [2,7,8]. We have used RMRSM analysis of pre-steady-state kinetics and steady-state kinetics measurements on bc 1 complexes containing site-directed mutants of the Rieske iron-sulfur protein to test the function of the high redox potential of the Rieske iron-sulfur cluster in the Q cycle mechanism [3]. The midpoint potential of the iron-sulfur cluster was systematically altered from +280 to + 105 mV by eliminating two hydrogen bonds, from a conserved serine and a conserved tyrosine, that otherwise facilitate electron delocalization from the reduced cluster. Eliminating the hydrogen bond from the hydroxyl group of Ser ls3 to S-1 of the cluster lowers the midpoint potential of the cluster by 130 mV, and eliminating the hydrogen bond from the hydroxyl group of Tyr185 to S v of Cys ]59 lowers the midpoint potential by 65 inV. Eliminating both hydrogen bonds has an approximately additive effect, lowering the midpoint potential by 180 inV. Ubiquinol-cytochrome c reductase activity of the bc I complex decreases with the decrease in midpoint potential, confirming that oxidation of ubiquinol by the ironsulfur protein is the rate-limiting partial reaction in the bcj complex. The changes in cytochrome c reductase activity of the bc~ complex resulting from changes in the midpoint potential indicate that the midpoint potential of the iron-sulfur cluster extensively influences the activity of the bc ~ complex. This means that deprotonation of ubiquinol to ubiquinol anion is either not a prerequisite to electron transfer or must occur at a much greater rate than electron transfer, otherwise the rate would be independent of the Rieske midpoint potential. As the midpoint potential of the Rieske protein increased from 220 to 280 mV, the cytochrome c reductase activity increased 2.5-fold [3]. If electron transfer from ubiquinol to cytochrome c~ can be interpreted by Marcus theory, the 2.5-fold change in rate per 60 mV change in potential is most easily explained if electron transfer to the iron-sulfur cluster is from a donor with a redox potential similar to that of the iron-sulfur cluster. Since the potential of the QH /Q couple is approximately - 2 5 mV and that of the QH2/Q couple is approximately +90 mV [10], the observed changes in rate as a function of iron-sulfur protein potentials are more readily ex-
C. Snyder, B.L. Trumpower / Biochimica et Biophysica Acta 1365 (I998) 125-I34
133
8.B4
cyt
b
8,83
cyt c 1
8.1B*Z
8,81
I 158
@
I
I
I
I
3~
I
4~B
~B8
758
988
I
I
388
41~
f~m
• 8"lIB
U
¢ m .Q
.836
L_
0 U~ .Q
,824
• 812
I~
I 158
I'
~
'988I
8.B2
cyt 9.91
c 1 .
.IJ~Jla~.~[ i l...a .~..~,~,..~,. a~a..,...., ~a.,, t,~.,. ~j~ ~i...t,, ~,"~m rw ,'1r - ~ T~-~Ir-I,-rrr,, r.... l,'e'lmw~ n'r"'l~l~ ~'~' l'~n~' 3 ~,r' cyt b
I lSA
I ~
1
I
458
688
[ ~
I 988
Ti m e (ms) Fig. 6. Reduction of the cytochrome bc~ complex through center P and center N by menaquinot, in the absence of endogenous ubiquinone. Figure shows the time course of reduction of the cytochrome bc I complex isolated from a mutant yeast strain (ACoQ2) that lacks endogenous ubiquinone. The top panel shows reduction of 2 p~M bca complex by 50 >M menaquinol in the absence of inhibitors. The middle panel shows the inhibition of c~ reduction by stigmatellin, and the bottom panel shows the inhibition of b and c~ reduction by stigmatellin plus antimycin.
plained if ubiquinol, and not ubiquinol anion, is the redox donor for the Rieske protein. Furthermore, deprotonation of ubiquinol, which has an estimated p K a = l l . 2 5 [10], would require an extremely alkaline local environment on the iron-
sulfur protein, to provide a binding site proximal to the iron-sulfur cluster, and would result in tight binding of ubiquinol anion in the absence of electron transfer. No such amino acid sequence conservation is observed in the iron-sulfur proteins sequenced to
134
C. Snyder, B.L. Trumpower / Biochimica et Biophysica Acta 1365 (1998) 125-134
date [2], and no high-affinity binding site for ubiquinol has ever been demonstrated. We suggest an alternative mechanism in which the protonated ubiquinol competitively displaces a proton from the imidazole nitrogen of one of the histidines that is a ligand to the iron-sulfur cluster and forms a quinolimidazolate complex, and this complex is the electron donor to the redox active iron. This mechanism circumvents prerequisite deprotonation of ubiquinol. We are also testing whether conserved aromatic amino acids form a conduit or otherwise facilitate electron transfer from the iron-sulfur cluster to the surface of the iron-sulfur protein en route to the heme of cytochrome c~ [11]. Sequence alignments of all Rieske proteins characterized to date identify a maximum of 10 aromatic amino acids that are conserved, if conservative exchanges and a liberal allowance of sequence alignments are allowed. These correspond to W l l l , F l l 7 , W152, F173, W176, F177, H184, Y185, Y205, and F207 in the S. cerevisiae Rieske iron-sulfur protein. We have constructed yeast mutants in which each of these aromatic residues in the Rieske iron-sulfur protein has been replaced with a non-aromatic residue, and in some instances yeast mutants with combined replacements have been constructed. Preliminary analysis of the bc r complexes from these mutants suggests that there is no obligatory electron transfer conduit through aromatic residues in the S. cerevisiae Rieske iron-sulfur protein. Small changes in the ubiquinolcytochrome c reductase activities of the bc 1 complexes containing the W176L, F177L, and the W176L, F177L forms of the Rieske protein indicate that any facilitation of electron transfer from 7r orbitals in aromatic amino acids results in only modest rate enhancements.
Acknowledgements This research has been supported by an NRSA Fellowship GM 18811 to CS and by NIH Grant GM 20379.
References [1] E.A. Berry, L.S. Huang, T.N. Earnest, B.K. Jap, J. Mol. Biol. 224 (1992) 1161-1166. [2] U. Brandt, B.L. Trumpower, Crit. Rev. Biochem. Mol. Biol. 29 (1994) 165-197. [3] E. Denke, T. Merbitz-Zahradnik, O.M. Hatzfeld, C. Snyder, T.A. Link, B.L. Trumpower, J. Biol. Chem. 273 (1998) 9085 -9093. [4] H. Ding, D.E. Robertson, F. Daldal, P.L. Dutton, Biochemistry 31 (1992) 3144-3158. [5] S. Iwata, M. Saynowits, T.A. Link, H. Michel, Structure 4 (1996) 567-579. [6l A. Kr6ger, G. Unden, in: Q. Coenzyme, G. Lenaz (Ed.), John Wiley & Sons, Ltd., New York, 1985, pp. 285-300. 17] A. Matsuno Yagi, Y. Hatefi, J. Biol. Chem. 271 (1996) 6164-6171. [8] A. Matsuno Yagi, Y. Hatefi, J. Biol. Chem. 272 (1997) 16928-16933. [9] Y. Orii, T. Miki, J. Biol. Chem. 272 (1997) 17594-17604. [10] P.R. Rich, Biochim. Biophys. Acta 768 (1984) 53-79. [11] C. Snyder, Trumpower, B.L. (submitted). [12] A.L. Tsai, G. Palmer, Biochim. Biophys. Acta 722 (1983) 349-363. [13] S.R. Van Doren, R.B. Gennis, B. Barquera, A.R. Crofts, Biochemistry 32 (1993) 8083-8091. [14] D. Xia, C.A. Yu, H. Kim, J.Z. Xia, A.M. Kachurin, L. Zhang, L. Yu, J. Deisenhofer, Science 277 (1997) 60-66. [15] Z. Zhang, L. Huang, V.M. Shulmeister, Y.I. Chi, K.K. Kim, L.W. Hung, R. Antony, A.R. Crofts, E.A. Berry, S.H. Kim, Nature (1998) in press.
Biochimica et Biophysica Acta, 1365 (1998) 125-134
ELSEVIER
Mechanism of ubiquinol oxidation by the cytochrome bc 1 complex: pre-steady-state kinetics of cytochrome bc 1 complexes containing site-directed mutants of the Rieske iron-sulfur protein Chris Snyder, B e m a r d L. T r u m p o w e r * Department of Biochemistry, Dartmouth Medical School, Hanover, NH 03755, USA
Received 2 February 1998; received in revised form 4 March 1998; accepted 4 March 1998
Abstract To facilitate characterization of mutated cytochrome bCl complexes in S. cerevisiae we have developed a new approach using a rapid scanning monochromator to examine pre-steady-state reduction of the enzyme with menaquinol. The RSM records optical spectra of cytochromes b and Cl at 1-ms intervals after a dead time of 2 ms, and menaquinol fully reduces both cytochromes b H and c~ and a portion of cytochrome b L. The rapid-mixing, rapid-scanning monochromator methodology obviates limitations inherent in previous rapid kinetics methods and permits measurements of rates exceeding 200 s-~. To document the validity of this methodology we have examined the reduction kinetics of the cytochrome bCl complexes from wild-type yeast and yeast that lack ubiquinone. The results establish that menaquinol reacts via the Q cycle pathway both in the presence and absence of ubiquinone. From analyzing bc~ complexes containing Rieske proteins in which the midpoint potential of the iron-sulfur cluster has been altered from +280 to + 105 mV, we propose a mechanism in which the protonated quinol displaces a proton from the imidazole nitrogen of one of the histidines that is a ligand to the iron-sulfur cluster and forms a quinol-imidazolate complex that is the electron donor to the redox active iron. © 1998 Elsevier Science B.V. Keywords: Cytochrome bc~ complex; Q cycle; Rieske iron-sulfur protein; Ubiquinol; Cytochrome c reductase; Saccharomyces cerevisiae
1. Introduction
To facilitate characterization of cytochrome bc~ complexes, in which the kinetic properties of the enzyme have been altered by site-directed mutagenesis of the genes in S. cerevisiae, we have developed a new approach using a rapid-scanning monochromator to examine pre-steady-state reduction of the enzyme on a ms time scale after initiating reduction of the enzyme with menaquinol. The rapidCoiTespondmg author.
scanning monochromator allows simultaneous monitoring of the pre-steady-state reduction of cytochromes b and c 1 with a time resolution of 1 ms after a dead time for mixing of 2 ms, and the potential of menaquinol is low enough to permit partial reduction of cytochrome b L in addition to cytochromes b H and c~. Although rapid scan spectrophotometry has recently been applied to the bovine bc 1 complex [9], the kinetic measurements were limited by a relatively slow 25-ms mixing time and were restricted to examination of the oxidation kinetics due to the lack of a suitably low potential quinol substrate. Photo-
0005-2728/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII: S0005-2728(98)00052-8
126
C. Snyder, B.L. Trumpower / Biochimica et Biophysica Acta 1365 (1998) 125-134
synthetic reaction centers have also been used to drive reduction of the bc~ complex with responses in the ms time range [4,13], but this approach is limited by the fact that a single flash incompletely reduces the cytochromes, and multiple flashes elicit simultaneous reoxidation and reduction. The rapid-mixing, rapid-scanning monochromator (RMRSM) methodology described here obviates all of these problems and permits measurements of rates exceeding 200 S
-1
2. Results and discussion
Menaquinol was chosen as a substrate for reduction of the bc~ complex because it is more strongly reducing (Em7=-74 mV, Ref. [6]) than ubiquinol ( E m T = q - 9 0 mV, Ref. [10]), allowing a greater percentage of cytochrome b to be reduced under equilibrium conditions. Menaquinol has the additional advantage that the quinol, but not the quinone form, is soluble at high concentrations in aqueous solutions, thus obviating problems that might result from the use of organic solvents. From the optical spectra in Fig. 1A one can calculate that menaquinol completely reduced cytochrome c I and approximately 70% of the dithionite reducible cytochrome b. This result is in agreement with the midpoint potentials of menaquinol and the b cytochromes (bH= +90 mV and b L = - 3 0 mV; Ref. [2,10]). At pH 6.0, the predicted extent of reduction of b H is >99% and b L is approximately 20%. Assuming that b H and b e contribute 70 and 30%, respectively, to the total absorbance change at 563 nm, and that the midpoint potential of the b cytochromes is pH independent [12]), the predicted reduction of cytochrome b is 76%, close to the 70% observed experimentally. Eh, and not E m, of the menaquinol/menaquinone couple should be the parameter that determines the extent of b L reduction. However, the observed extent of reduction more closely agrees with what would be predicted from the E m than from the E h. This may indicate that the reported E m of menaquinol [6] is incorrect, with the consequence that the observed extent of reduction would agree more closely with E h. However, we have observed that the extent of b reduction with ubiquinol as substrate also agrees with the E m of the ubiquinol/ubiquinone couple rather
than with the E n (results not shown). An alternative explanation for the less than predicted reduction of cytochrome b is that the extent of reduction of the b hemes through the enzyme-catalyzed pathways does not respond to the potential (Eh) of the quinol/ quinone pool in the absence of mediators. The traces in Fig. 1B illustrate time-resolved optical spectra collected by the rapid-scanning monochromator. In this experiment the bc 1 complex was reduced with a very low concentration (6 IxM) of menaquinol in order to extend the reaction over several seconds. Following a 2-ms dead time, the RSM collects optical spectra every 1 ms, but only 40 spectra, at 50-ms intervals, are displayed in this figure. This low concentration of menaquinol permits a clear display of the triphasic reduction of cytochrome b (see Ref. [2] for a review) when the three-dimensional data from Fig. 1B are analyzed at the wavelength diagnostic of cytochrome b, as shown in Fig. 2. The inset in Fig. 2 shows the first spectra collected in real time during triphasic reduction, and spectra averaged over 10-ms intervals confirm that cytochrome b undergoes partial reoxidation. As a prerequisite to using RMRSM to dissect the kinetics in genetically engineered bc I complexes we performed control experiments to establish that menaquinol reacts through center P in the Q cycle pathway, and alternatively can reduce the b cytochromes through center N if center P is blocked. For this purpose we examined the effects of center N and center P inhibitors on the reduction of cytochromes b and c 1. The time course of reduction of the cytochrome bcl complex when it is rapidly mixed with menaquinol is shown in Fig. 3. In the absence of inhibitors, both cytochromes b and c 1 are reduced by 50 IxM menaquinol within 1 s. When the center P inhibitor stigmatellin was added it blocked the reduction of cytochrome c I (Fig. 3, middle panel), and when both stigmatellin and the center N inhibitor antimycin were added, they blocked the reduction of both cytochromes b and c 1 (Fig. 3, bottom panel). These results establish that menaquinol cannot bypass these center P and center N inhibitors. That menaquinol reacts enzymatically through the Q cycle pathway was confirmed using the Rieske iron-sulfur protein S183C mutant [3]. In this mutant, the Rieske apoprotein is stably incorporated into the cytochrome bc I complex, even though it lacks the
C. Snyder, B.L Trumpower / Biochimica et Biophysica Acta 1365 (1998) 125-134
127
^ 0.05 ~+DT
J!
L.
0 es
.(
v 500
?
I
I
F
I
525
550
575
600
625
Wavei en9th (nm)
B
0.04
¢p
J~
2.5
-
0.02 -~
set:)
1 T ¢ l I v ~ ' ! ~-'11~.~!.1! I I I I I i /
U
580 Fig, 1. (A) Optical spectraof 2 I~M purified yeast cytochrome bc~ complex after reduction with 50 I~M menaquinol (MQ) or dithionite (DT). Oxidized cytochrome bc 1 complex in 100 m M sodium phosphate, pH 6.0, containing 2 mM EDTA and 0.1 m g / m l dodecyl maltoside was reduced with menaquinol or dithionite and the resulting difference spectra of reduced versus oxidized enzyme were recorded in the split beam mode at 0.5-nm intervals with an Aminco DW 2A spectrophotometer equipped with an On-Line Instruments Systems, Inc. computer interface. (B) Time-resolved optical spectra obtained by stopped-flow rapid scanning spectrophotometry during reduction of the bc 1 complex by menaquinol. In this experiment 2 ~M bc I complex was reduced by rapid mixing with 6 txM menaquinol. Spectra were recorded on an OLIS rapid-scanning monochromator equipped with a 1200qine/mJn grating blazed at 500 nm, which yields a spectrum of 75 nm width at a resolution of 0.4 nm. Data were collected at 1000 scans/s, and the spectra were centered at 555 nm. The dead time of the stopped-flow mixer is approximately 2 ms, and the end of this period was chosen as time zero.
C. Snyder, B.L. Trumpower I Biochimica et Biophysica Acta 1365 (1998) 125-134
128
cyt b (563.3 nm)
0.06 1
,
/
,,
(-
..Q
"
I,,,
0 Ill
0.04
0.02
8
c B
_
s4o
-
sho Wavel ength (nm)
i
i
0.6
1.2
i
1.8
i
2.4
..... i
3.0
)
3.6
T i m e (s) Fig. 2. Triphasic reduction of cytochrome b observed by time-resolved optical spectra obtained during reduction of cytochromes b and c~ by menaquinol. The traces at 563.3 and 554.6 nm were derived from the three-dimensional data set shown in Fig. lB. The inset shows time-averaged optical spectra obtained across the 10-ms intervals indicated by the solid bars at points A, B, and C at 150 and 555 ms and at 4 s during the reaction.
iron-sulfur cluster. The absence of functional ironsulfur protein will block enzymatic oxidation of menaquinol at center P, but should allow reduction of cytochrome b through center N. In this mutant we no longer observed reduction of cytochrome c~ by menaquinol (data not shown), and in the presence of antimycin reduction of cytochrome b was also blocked (data not shown). These results confirm that menaquinol reduces the cytochrome bc~ complex through its enzymatic pathways, and validate the use of menaquinol as a reductant to study the kinetics of electron transfer within the cytochrome bc~ complex. The concentrations of menaquinol used in these reactions (<100 t~M) are less than saturating, so that the pre-steady-state kinetics are second order. Varying the concentration of menaquinol over a 20-fold range in the presence of either antimycin or stigmatellin, and fitting of the kinetic data, allowed the calculation
of second-order rate constants for reduction of both cytochromes b and c I through center P and for reduction of cytochrome b through center N. The rates of cytochrome b reduction through center P are best fitted by a two-component curve, consisting of a fast and a slow phase with second-order rate constants of 1.3)<106 and 8.8×104 M -1 s -~. As shown in Fig. 4, reduction of cytochrome c l is kinetically matched to the second, slow phase of b reduction, with a rate constant of 7.5X10 ~ M -I s -~ One explanation for these results is that when the first molecule of menaquinol is oxidized one electron equilibrates between iron-sulfur protein and cytochrome c~, and the second electron is transferred to cytochrome b. Under the conditions used in these experiments, the midpoint potential of the iron-sulfur protein is >60 mV more positive than that of cytochrome c~, with the result that the electron from
C. Snyder, B.L. Trumpower I Biochimica et Biophysics Acta 1365 (1998) 125-134
Time
129
(ms)
Fig. 3. Reduction of the cytochrome bc, complex through center P and center N by menaquinol. The top panel shows reduction of 2 FM bc, complex by 50 p,M menaquinol in the absence of inhibitors. The middle panel shows the inhibition of c, reduction by stigmatellin, and the bottom panel shows the inhibition of b and c, reduction by stigmatellin plus antimycin.
130
C. Snyder, B.L. Trumpower / Biochimica et Biophysica Acta 1365 (1998) 125-134 •848
cyt b
I
I
I
I
458
68B
758
I
158
~
Ti me I o I~
clt b , , l ° . (-)
(ms) 1
.,oo J~
25
75
50
75
! 00/
H.mi~,~1q i ~ Cyt b, f a s t (El)
25
I
I
I
25
50
75
Menaquinol
1O0
M)
Fig. 4. Kinetics of reduction of the cytochrome bc 1 complex by menaquinol through center P. The traces in A show the time course of reduction of cytochrome b and cytochrome c I when 2 IxM cytochrome bc~ complex is reduced by 50 I~M menaquinol in the presence of antimycin. The plot in (B) shows the effects of increasing menaquinol concentration on the rates of reduction of cytochrome b and cytochrome cl. The inset shows the effects of increasing menaquinol concentration on the slow rates of cytochrome b and cytochrome c~ reduction.
the first molecule of menaquinol oxidized is predominately located in the iron-sulfur protein. Only when the first electron is localized to c 1 can the second molecule of menaquinol be oxidized and transfer
electrons to iron-sulfur protein and cytochrome b. This mechanism is consistent with the results where the slow rate of b reduction matches the rate of cytochrome c I reduction.
C. Snyder, B.L. Trurnpower / Biochimica et Biophysica Acta 1365 (1998) 125-134
From the traces in the top panels of Figs. 4 and 5, which show reduction of cytochrome b through center P and center N with 50 ~M menaquinol, it appears that the former reaction is significantly faster than the latter. This was confirmed by the secondorder rate constant for the reaction through center N,
131
2.5× 105 M -~ s -~ (Fig. 5, bottom panel). This 5-fold difference in rates through center P and center N indicates that these rates are primarily governed by the thermodynamics of the electron transfer reactions. This is consistent with the notion that the reactive head group of ubiquinone can remain resident in the
.868
ma
qD 0 e-
.0 L
L~
•8 4 5
•838
0 .0
~
~/
f
~
~ 1 ~
~
,..
~#lam,.~u.~ ~_. mmk.lm'--Wki~ml-mm--J~m~. ).Jd" e v..?e,,.iF--i/-,'. ~p,~ ~'I ~ r
,
14sec -1
•8 1 5
I 158
I 38@
I
I
458
I
688
758
Ti me (ms) 25 20 I
VI
""
15
¢D 4-1
i
10 5 [] i
I
I
I
I
20
40
60
80
100
Menaquinol
(i,. M)
Fig. 5. Kinetics of reduction of cytochrome b by menaquinol through center N. The trace in A shows the time course of reduction of cytochrorne b when 2 p~M cytochrome bc~ complex is reduced by 50 txM menaquinol in the presence of stigmatellin. The plot in (B) shows the effects of increasing menaquinol concentration on the rate of reduction of cytochrome b through center N.
132
C. Snyder, B.L. Trumpower /Biochimica et Biophysica Acta 1365 (1998) 125-134
bc~ complex for multiple turnovers of the enzyme, and that the accessibility of quinone to the two sites is essentially identical. The preceding kinetic analysis also indicated that there was a secondary, slow reduction of cytochrome c~, which occurred at a rate of <0.01 s ], and this slow reduction was also observed in the presence of stigmatellin. Cytochrome c~ was also slowly reduced in the Rieske iron-sulfur protein S183C mutant, which indicates that menaquinol can directly reduce cytochrome c l in an iron-sulfur protein-independent manner. However, this non-enzymatic rate is several orders of magnitude slower than the iron-sulfur protein-catalyzed reduction of cytochrome c~ through center P, and it does not impact on the kinetics of cytochrome c~ reduction, even when these rates are lowered 10-fold by mutational changes in the ironsulfur protein. We have also examined the reduction kinetics of the cytochrome bc~ complex from yeast mutants that lack endogenous ubiquinone (Fig. 6). The RMRSM analysis of the isolated bc~ complex reveals individual properties of the two quinol/quinone oxidoreductase sites, center P and center N, in their reactivities with menaquinol and menaquinone. Reduction of the cytochromes b and c] through center P and reduction of cytochrome b through center N by menaquinol in the absence of ubiquinone is inhibited by stigmatellin (Fig. 6, middle panel) and antimycin (Fig. 6, bottom panel). The reactivity of menaquinol with bc~ complex lacking ubiquinone may lead to some novel insights into the mechanism of triphasic reduction of cytochrome b. Notably, concentrations of menaquinol that result in triphasic reduction of cytochrome b when ubiquinone is present (Fig. 3, top panel), result in apparently monophasic b reduction when ubiquinone is absent (Fig. 6, top panel), and the rate closely matches the first, rapid phase of the triphasic reduction. The absence of a reoxidation phase suggests that menaquinone cannot reoxidize the b cytochromes at center N, or that reoxidation through center N is slow relative to the rate of reduction through center P, consistent with the low potential of menaquinone relative to cytochrome b~. The crystal structure of the Rieske iron-sulfur protein [5] and the crystal structures of the mitochondrial cytochrome bc~ complexes [1,14,15] pro-
vide an opportunity to test several aspects of ironsulfur protein function in the protonmotive Q cycle mechanism [2,7,8]. We have used RMRSM analysis of pre-steady-state kinetics and steady-state kinetics measurements on bc 1 complexes containing site-directed mutants of the Rieske iron-sulfur protein to test the function of the high redox potential of the Rieske iron-sulfur cluster in the Q cycle mechanism [3]. The midpoint potential of the iron-sulfur cluster was systematically altered from +280 to + 105 mV by eliminating two hydrogen bonds, from a conserved serine and a conserved tyrosine, that otherwise facilitate electron delocalization from the reduced cluster. Eliminating the hydrogen bond from the hydroxyl group of Ser ls3 to S-1 of the cluster lowers the midpoint potential of the cluster by 130 mV, and eliminating the hydrogen bond from the hydroxyl group of Tyr185 to S v of Cys ]59 lowers the midpoint potential by 65 inV. Eliminating both hydrogen bonds has an approximately additive effect, lowering the midpoint potential by 180 inV. Ubiquinol-cytochrome c reductase activity of the bc I complex decreases with the decrease in midpoint potential, confirming that oxidation of ubiquinol by the ironsulfur protein is the rate-limiting partial reaction in the bcj complex. The changes in cytochrome c reductase activity of the bc~ complex resulting from changes in the midpoint potential indicate that the midpoint potential of the iron-sulfur cluster extensively influences the activity of the bc ~ complex. This means that deprotonation of ubiquinol to ubiquinol anion is either not a prerequisite to electron transfer or must occur at a much greater rate than electron transfer, otherwise the rate would be independent of the Rieske midpoint potential. As the midpoint potential of the Rieske protein increased from 220 to 280 mV, the cytochrome c reductase activity increased 2.5-fold [3]. If electron transfer from ubiquinol to cytochrome c~ can be interpreted by Marcus theory, the 2.5-fold change in rate per 60 mV change in potential is most easily explained if electron transfer to the iron-sulfur cluster is from a donor with a redox potential similar to that of the iron-sulfur cluster. Since the potential of the QH /Q couple is approximately - 2 5 mV and that of the QH2/Q couple is approximately +90 mV [10], the observed changes in rate as a function of iron-sulfur protein potentials are more readily ex-
C. Snyder, B.L. Trumpower / Biochimica et Biophysica Acta 1365 (I998) 125-I34
133
8.B4
cyt
b
8,83
cyt c 1
8.1B*Z
8,81
I 158
@
I
I
I
I
3~
I
4~B
~B8
758
988
I
I
388
41~
f~m
• 8"lIB
U
¢ m .Q
.836
L_
0 U~ .Q
,824
• 812
I~
I 158
I'
~
'988I
8.B2
cyt 9.91
c 1 .
.IJ~Jla~.~[ i l...a .~..~,~,..~,. a~a..,...., ~a.,, t,~.,. ~j~ ~i...t,, ~,"~m rw ,'1r - ~ T~-~Ir-I,-rrr,, r.... l,'e'lmw~ n'r"'l~l~ ~'~' l'~n~' 3 ~,r' cyt b
I lSA
I ~
1
I
458
688
[ ~
I 988
Ti m e (ms) Fig. 6. Reduction of the cytochrome bc~ complex through center P and center N by menaquinot, in the absence of endogenous ubiquinone. Figure shows the time course of reduction of the cytochrome bc I complex isolated from a mutant yeast strain (ACoQ2) that lacks endogenous ubiquinone. The top panel shows reduction of 2 p~M bca complex by 50 >M menaquinol in the absence of inhibitors. The middle panel shows the inhibition of c~ reduction by stigmatellin, and the bottom panel shows the inhibition of b and c~ reduction by stigmatellin plus antimycin.
plained if ubiquinol, and not ubiquinol anion, is the redox donor for the Rieske protein. Furthermore, deprotonation of ubiquinol, which has an estimated p K a = l l . 2 5 [10], would require an extremely alkaline local environment on the iron-
sulfur protein, to provide a binding site proximal to the iron-sulfur cluster, and would result in tight binding of ubiquinol anion in the absence of electron transfer. No such amino acid sequence conservation is observed in the iron-sulfur proteins sequenced to
134
C. Snyder, B.L. Trumpower / Biochimica et Biophysica Acta 1365 (1998) 125-134
date [2], and no high-affinity binding site for ubiquinol has ever been demonstrated. We suggest an alternative mechanism in which the protonated ubiquinol competitively displaces a proton from the imidazole nitrogen of one of the histidines that is a ligand to the iron-sulfur cluster and forms a quinolimidazolate complex, and this complex is the electron donor to the redox active iron. This mechanism circumvents prerequisite deprotonation of ubiquinol. We are also testing whether conserved aromatic amino acids form a conduit or otherwise facilitate electron transfer from the iron-sulfur cluster to the surface of the iron-sulfur protein en route to the heme of cytochrome c~ [11]. Sequence alignments of all Rieske proteins characterized to date identify a maximum of 10 aromatic amino acids that are conserved, if conservative exchanges and a liberal allowance of sequence alignments are allowed. These correspond to W l l l , F l l 7 , W152, F173, W176, F177, H184, Y185, Y205, and F207 in the S. cerevisiae Rieske iron-sulfur protein. We have constructed yeast mutants in which each of these aromatic residues in the Rieske iron-sulfur protein has been replaced with a non-aromatic residue, and in some instances yeast mutants with combined replacements have been constructed. Preliminary analysis of the bc r complexes from these mutants suggests that there is no obligatory electron transfer conduit through aromatic residues in the S. cerevisiae Rieske iron-sulfur protein. Small changes in the ubiquinolcytochrome c reductase activities of the bc 1 complexes containing the W176L, F177L, and the W176L, F177L forms of the Rieske protein indicate that any facilitation of electron transfer from 7r orbitals in aromatic amino acids results in only modest rate enhancements.
Acknowledgements This research has been supported by an NRSA Fellowship GM 18811 to CS and by NIH Grant GM 20379.
References [1] E.A. Berry, L.S. Huang, T.N. Earnest, B.K. Jap, J. Mol. Biol. 224 (1992) 1161-1166. [2] U. Brandt, B.L. Trumpower, Crit. Rev. Biochem. Mol. Biol. 29 (1994) 165-197. [3] E. Denke, T. Merbitz-Zahradnik, O.M. Hatzfeld, C. Snyder, T.A. Link, B.L. Trumpower, J. Biol. Chem. 273 (1998) 9085 -9093. [4] H. Ding, D.E. Robertson, F. Daldal, P.L. Dutton, Biochemistry 31 (1992) 3144-3158. [5] S. Iwata, M. Saynowits, T.A. Link, H. Michel, Structure 4 (1996) 567-579. [6l A. Kr6ger, G. Unden, in: Q. Coenzyme, G. Lenaz (Ed.), John Wiley & Sons, Ltd., New York, 1985, pp. 285-300. 17] A. Matsuno Yagi, Y. Hatefi, J. Biol. Chem. 271 (1996) 6164-6171. [8] A. Matsuno Yagi, Y. Hatefi, J. Biol. Chem. 272 (1997) 16928-16933. [9] Y. Orii, T. Miki, J. Biol. Chem. 272 (1997) 17594-17604. [10] P.R. Rich, Biochim. Biophys. Acta 768 (1984) 53-79. [11] C. Snyder, Trumpower, B.L. (submitted). [12] A.L. Tsai, G. Palmer, Biochim. Biophys. Acta 722 (1983) 349-363. [13] S.R. Van Doren, R.B. Gennis, B. Barquera, A.R. Crofts, Biochemistry 32 (1993) 8083-8091. [14] D. Xia, C.A. Yu, H. Kim, J.Z. Xia, A.M. Kachurin, L. Zhang, L. Yu, J. Deisenhofer, Science 277 (1997) 60-66. [15] Z. Zhang, L. Huang, V.M. Shulmeister, Y.I. Chi, K.K. Kim, L.W. Hung, R. Antony, A.R. Crofts, E.A. Berry, S.H. Kim, Nature (1998) in press.