JOURNAL OFInorganicBiochemistry

Journal of Inorganic Biochemistry 98 (2004) 849–855

www.elsevier.com/locate/jinorgbio

Interaction of plastocyanin with oligopeptides: effectof lysine distribution within the peptide

Shun Hirota a,*, Hisano Okumura b, Sachiko Arie c, Kentaro Tanaka c,Mitsuhiko Shionoya c, Teruhiro Takabe d, Noriaki Funasaki a, Yoshihito Watanabe b

a Department of Physical Chemistry, Kyoto Pharmaceutical University, 5 Nakauchi-cho, Misasagi, Yamashina-ku, Kyoto 607-8414, Japanb Department of Chemistry, Graduate School of Science, Nagoya University, Chikusa-ku, Nagoya 464-8602, Japan

c Department of Chemistry, Graduate School of Science, The University of Tokyo, Hongo, Bunkyo-ku, Tokyo 113-0033, Japand Research Institute, Meijo University, Tempaku-ku, Nagoya 468-8502, Japan

Received 5 September 2003; received in revised form 22 October 2003; accepted 31 October 2003

Abstract

We synthesized and purified four oligopeptides containing four lysines (KKKK, GKKGGKK, KKGGGKK, and KGKGKGK)

as models for the plastocyanin (PC) interacting site of cytochrome f. These peptides competitively inhibited electron transfer be-

tween cytochrome c and PC. The inhibitory effect increased as the peptide concentrations were increased. The association constants

between PC and the peptides did not differ significantly (3500–5100 M�1), although the association constant of PC–KGKGKGK

was a little larger than the constants between PC and other peptides. Changes in the absorption spectrum of PC were observed when

the peptides were added to the PC solution: peaks and troughs were detected at about 460 and 630 nm and at about 560 and 700 nm,

respectively, in the difference absorption spectra between the spectra with and without peptides. These changes were attributed to the

structural change at the copper site of PC by interaction with the peptides. The structural change was most significant when tet-

ralysine was used. These results show that binding of the oligopeptide to PC is slightly more efficient when lysines are distributed

uniformly within the peptide, whereas the structural change of PC becomes larger when the lysines are close to each other within the

peptide.

� 2003 Elsevier Inc. All rights reserved.

Keywords: Plastocyanin; Lysine peptide; Protein–peptide interaction; Structural change; Active-site structure

1. Introduction

Structure–function relationship is one of the major

topics in protein researches. Plastocyanin (PC), a mobile

copper protein existing in the thylakoid lumen of pho-tosynthetic organisms, accepts an electron from cyto-

chrome f (cyt f ), a subunit of the cytochrome b6fcomplex, and donates it to the reaction center chloro-

phyll (P700þ) in the photosystem I complex (PSI) [1–3].

PC is classified as a Type 1 copper protein, which ex-

hibits a low energy ligand-to-metal charge transfer

(LMCT) band near 600 nm in the absorption spectra

and a narrow hyperfine coupling constant (jAkj

* Corresponding author. Tel.: +81-75-595-4664; fax: +81-75-595-

4762.

E-mail address: hirota@mb.kyoto-phu.ac.jp (S. Hirota).

0162-0134/$ - see front matter � 2003 Elsevier Inc. All rights reserved.

doi:10.1016/j.jinorgbio.2003.10.022

< 90� 10�4 cm�1) in the electron paramagnetic reso-

nance (EPR) spectra [4,5]. According to the crystal

structures of plant PCs [6–10], PC contains one copper

atom with two histidine nitrogen atoms, one methionine

sulfur atom, and one cysteine sulfur atom coordinatedin a distorted tetrahedral geometry.

Plant PC usually possesses two highly conserved sites

which have been considered as molecular recognition

sites for its redox partners, cyt f and PSI: One site is

located at the Cu-coordinated, solvent-accessible histi-

dine (Cu-adjacent hydrophobic patch), and the other

site is located at another solvent-accessible site con-

taining acidic residues (Cu-remote negative patch)(Fig. 1). The negative patch of PC consists of two

clusters: One lower cluster, Asp42/Glu43/Asp44/Glu45,

and another upper cluster, Glu59/Glu60/Asp61 (Fig. 1)

[6–10]. Both of these clusters have been indicated to be

Fig. 1. A schematic view of silene PC (PDB entry 1BYO) interacting

with an oligopeptide which contains lysines.

850 S. Hirota et al. / Journal of Inorganic Biochemistry 98 (2004) 849–855

essential for the binding of PC to cyt f [22] and have

been shown to interact with charged molecules and

proteins [23,24]. It has been shown that the Cu-remote

negative patch of PC and the positively charged site of

cyt f interact through electrostatic interactions to form a

PC–cyt f complex for electron transfer [11–21]. The

hydrophobic patch of PC, however, has also been shown

to be crucial for the PC–cyt f complex [22,25,26].There have been a number of studies on electron

transfer between proteins [27–29], and electron transfer

between PC and cyt f or cytochrome c (cyt c) have been

studied extensively [11–19,22,30–34]. Redox reactions

between PC and small molecules have also been inves-

tigated extensively, and Sykes and co-workers [12,35–38]

have previously shown that small inorganic compounds

can inhibit electron transfer between PC and cyt c or cytf. The association constants between PC and cyt f or cyt

c have been obtained by measuring the increase of the

Soret band intensity of cyt f or cyt c on PC binding

[32,39,40]. An NMR study on the Cd-substituted PC–

cyt c complex indicated that the PC–cyt c complex

consists of a highly dynamic ensemble of structures [41],

whereas, paramagnetic NMR and restrained rigid-body

molecular mechanics study indicated that the electro-static interactions guide PC and cyt f into a position that

is optimal for electron transfer [25]. Photoinduced Zn

cyt c has a high driving force and thus electron transfer

between Zn cyt c and PC occurs very fast [30,31,42].

With the use of Zn cyt c, Kosti�c and co-workers

[30,31,42] proposed that PC and cyt f or cyt c bind and

react with each other in different configurations resulting

from the protein–protein interaction termed as the gat-ing process for electron transfer. Possible configurations

for the diprotein complex of PC and cyt f or cyt c were

shown by computer simulation [21,43].

We have previously shown that charged peptides are

useful for studies on the molecular recognition character

of proteins and their interaction induced structural

changes [24,44–48]. Positively charged lysine peptides

interacted with the consecutive aspartic acid and glu-

tamic acid residues of the negative patch of PC (Fig. 1)

and competitively inhibited electron transfer from re-duced cyt c to oxidized PC [24]. The inhibitory effects of

lysine peptides on electron transfer were explained as

competitive inhibition due to neutralization of the PC

negative patch by formation of PC � peptide complexes.

Changes were also observed in the absorption spectrum

of PC by interaction with lysine peptides [24].

In this study, we investigated the interaction between

PC and oligopeptides with different distribution of ly-sines to obtain detailed information on the association

of PC with a positively charged molecule and the

structural change of PC at its active site induced by the

interaction.

2. Experimental

2.1. Preparation of proteins

Oxidized Silene pratensis (white campion) PC was

purified as described [11], and the purity of the protein

was confirmed by the ratio of the absorbance (Abs) at

280 to that at 597 nm ðAbs280=Abs597 < 1:30Þ. Oxidized

PC was dissolved in 10 mM phosphate buffer, pH 7.0.

Reduced cyt c was purified as described [45].

2.2. Preparation of oligopeptides

Tetralysine (KKKK) was purchased from Sigma. The

peptide resins of GKKGGKK, KKGGGKK, and

KGKGKGK were synthesized by the Fast moc method

(PE Applied Biosystems, Model 433A). Ice cooled tri-

fluoroacetic acid (5 ml) was added to the cooled peptideresin (20 mg), and the solution was stirred for 1.5 h at

room temperature. The obtained solution was filtered to

remove the resin and concentrated to 1 ml. The con-

centrated solution was added with conc. HCl (1 ml) and

again concentrated to 1 ml. The solution was then

dropped slowly into cooled t-butyl methyl ether (50 ml).

After waiting for 30 min, the upper ether layer was re-

moved, and the lower layer was added with cooled t-butyl methyl ether (10 ml) to wash the peptide. After

washing the peptide with ether twice, the lower peptide

layer was freeze-dried. The dried peptide was added with

0.1% HCl (1 ml) and purified with a HPLC system

(Shimadzu, LC-10AD) using a C18 column (Vydac).

The purified peptide solution was freeze-dried. Tetraly-

sine was purified by the same method.

Purities of the peptides were checked by mass(MALDI-TOF) spectra (Voyager, DE-Pro). Anal. Calc.

for C24H50N8O5 � 4HCl � 3.5H2O (KKKK � 4HCl � 3.5H2O): C, 38.98; H, 8.31; N, 15.15. Found: 39.07; H,

S. Hirota et al. / Journal of Inorganic Biochemistry 98 (2004) 849–855 851

8.15; N, 14.88%. Anal. Calc. for C30H59N8O5 � 5HCl

� 5H2O (GKKGGKK � 5HCl � 5H2O): C, 36.99; H, 7.66;

N, 15.81. Found: 37.34; H, 7.55; N, 15.43%. Anal. Calc.

for C30H59N8O5 � 5HCl � 5H2O (KKGGGKK � 5HCl � 5H2O): C, 36.99; H, 7.66; N, 15.81. Found: 37.01; H,7.90; N, 15.51%. Anal. Calc. for C30H59N8O5 � 5HCl

� 4.5H2O (KGKGKGK � 5HCl � 4.5H2O): C, 37.33; H,

7.62; N, 15.96. Found: 37.55; H, 7.71; N, 15.71%.

The oligopeptides were first dissolved in 10 mM

phosphate buffer, pH 7.0, with a peptide concentration

of a little larger than 4 mM, and then the pH values and

concentrations of the peptide solutions were readjusted

to pH 7.0 and 4 mM, respectively, by using 10 mMphosphate buffer, pH 7.0, and a little amount of 10 mM

phosphate buffer containing 1 or 0.1 M NaOH. The

same stock solution of each peptide was used for both

kinetic and absorption measurements.

2.3. Kinetic measurements

The electron transfer rate constants from reduced cytc to oxidized PC in the presence of oligopeptides were

obtained by monitoring the absorbance at 420 nm with

an Otsuka Denshi RA601 stopped-flow equipment. A 10

lM solution of PC in 10 mM potassium phosphate

buffer, pH 7.0, was mixed at 15 �C with 1 lM cyt c in the

same buffer. The inhibitory effect of charged peptides on

the electron transfer rate was studied with the PC so-

lution containing peptides (0–500 lM), the peptideconcentration being 0–250 lM after mixing PC and cyt c

solutions.

2.4. Absorption measurements

Absorption spectra of oxidized PC with and without

oligopeptides were measured at 15 �C on a Shimadzu

UV-3100PC spectrophotometer. The difference absorp-tion spectra were calculated between the spectra before

and after mixing PC and peptide solutions with a tan-

dem cell (path length: 4.5 + 4.5 mm).

Fig. 2. Plots of the reciprocal electron transfer rate constants (1=kobs)for the oxidized PC–reduced cyt c system against the concentrations

(0–250 lM) of oligopeptides (tetralysine (d, ——), GKKGGKK (j,

– – –), KKGGGKK (N; � � �), and KGKGKGK (., –�–)), together withleast-squares fitted lines according to Eq. (3). Phosphate buffer (10

mM), pH 7.0, was used. All measurements were performed at 15 �C.

3. Results and discussion

3.1. Effects of peptides on electron transfer between cyt c

and PC

The positively charged lysines within the oligopep-

tides could interact with the negative patch of PC

through electrostatic interaction to form a PC–peptide

complex. To investigate the effect of the lysine distri-

bution within the peptide, we performed stopped-flow

measurements on electron transfer between reduced cytc and oxidized PC in the presence of oligopeptides

containing lysines. The inhibitory effect was seen for all

the peptides we studied. This effect can be explained as

competitive inhibition by formation of PC–peptide

complexes, which was consistent with our previous re-

sult obtained using lysine peptides [24].

To investigate the inhibitory effect in detail, we

studied the effect of the peptide concentration on theelectron transfer rate constant (kobs). The electron

transfer rate constants decreased as the concentrations

of the peptides were increased. We can consider for-

mation of two complexes: A PC � cyt c complex, where

electron transfer occurs subsequently and a PC � peptidecomplex, which competitively inhibits formation of the

PC � cyt c complex and thus inhibits electron transfer.

The complex formations can be expressed by the fol-lowing equations:

PCox þ peptide ¢Ki ðPCox � peptideÞ ð1Þ

PCox þ cytcred ¢KOS ðPCox � cyt credÞ!

ke ðPCred � cyt coxÞð2Þ

where Ki and KOS are the association constants forPCox � peptide and PCox � cyt cred complexes, respec-

tively, and ke represents the electron transfer rate con-

stant. The suffixes ox and red refer to the oxidized and

reduced states, respectively. If we write the observed rate

constant as kobs, KOS � ke as k, and the concentrations of

PC and peptide as [PC] and [peptide], respectively, we

obtain the relationship

1

kobs¼ Ki

k � ½PC� � ½peptide� þ 1

k � ½PC� ð3Þ

Plots of 1=kobs against [peptide] gave lines shown in

Fig. 2, substantiating the validity of the assumptions

leading to Eq. (3). k (ð13:7� 0:4Þ � 106 M�1 s�1) and Ki

852 S. Hirota et al. / Journal of Inorganic Biochemistry 98 (2004) 849–855

values (Table 1) were obtained by least-squares fitting

the data in Fig. 2 with Eq. (3). Since the fitting was

successful, it is also reasonable to assume that only one

peptide is bound to the negative patch of PC.

The association constants between PC and the pep-tides were all in the same range for all the peptides

studied (3500–5100 M�1) (Table 1). The association

constant of PC–KGKGKGK, however, was slightly

larger than the constants between PC and other pep-

tides. Lysines are more uniformly distributed within the

KGKGKGK peptide compared with other peptides.

The larger association constant for KGKGKGK may

indicate that binding of the peptide to PC is a little moreefficient when lysines are distributed a little widely

within the peptide. The negative patch of PC consists of

two negatively charged clusters, a lower Asp42/Glu43/

Asp44/Glu45 cluster and an upper Glu59/Glu60/Asp61

cluster, both of which interact with positively charged

molecules (Fig. 1) [6–10]. The stronger association of PC

with a peptide having lysines with a relatively wide

distribution might be due to the relatively broad rangeof the molecular interaction site for PC. The obtained

result is consistent with the report that both of the lower

and upper clusters at the negative patch are essential for

the binding of PC to cyt f and cyt c [22,23]. The lysines

and arginine at the positive patch of cyt f are not next to

each other but are distributed close to each other with

certain distances [49]. The distances between the side

chain N atoms of the nearest-neighbor lysines and ar-ginine at the positive patch are 6–12 �A. The distance

between the Ca atoms of the neighboring amino acids in

a peptide is about 3.8 �A, although the pepitde structure

is flexible. Comparing these distances, it would be rea-

sonable to have one extra residue between the lysines in

the oligopeptide to increase its affinity with PC.

3.2. Absorption spectral changes of PC by oligopeptide

binding

We have previously shown that peaks and troughs are

observed in the difference absorption spectrum between

the spectra of PC with and without positively charged

lysine peptides [24,48]. These peaks revealed that the Cu

active-site structure of PC is perturbed by interaction

with a lysine peptide. To investigate the structuralchange induced by the interaction with a positively

Table 1

Association constants (Ki) for PC–oligopeptide complexes obtained

from Eq. (3)a

Oligopeptide Ki (M�1)

KKKK 4100� 300

GKKGGKK 4000� 300

KKGGGKK 3500� 300

KGKGKGK 5100� 300

a In 10 mM phosphate buffer, pH 7.0, at 15 �C.

charged peptide in more detail, we measured the changes

in the absorption spectra of PC by interaction with the

peptides with different lysine distribution.

The absorption spectra of PC with and without tet-

ralysine and their difference spectrum are shown inFig. 3. The intense absorption band around 600 nm

observed in the absorption spectra of oxidized PC has

been assigned to the cysteine thiolate (S(Cys))-to-Cu(II)

LMCT band [5]. One might be afraid that the concen-

tration of PC is not the same between the spectra with

and without the peptides when we use a tandem cell: PC

concentration is reduced to half after mixing the PC

solution with the peptide solution, while the pass lengthof the cell is doubled. However, no significant difference

was detected in the difference spectrum just by diluting

(pass length doubled) the PC solution with a tandem cell.

The difference absorption spectra of oxidized PC in

the presence and absence of various peptides are shown

in Fig. 4. A similar pattern was observed for each pep-

tide: Peaks (at about 460 and 630 nm) and troughs (at

about 560 and 700 nm) were detected in the differencespectra of PC. The peaks and troughs in the difference

spectra indicate a shift of the 600 nm band to a slightly

longer wavelength and an increase in the intensity of the

460 nm band. Since the 600 nm band is assigned to

the S(Cys)-to-Cu(II) LMCT band [4,5], the structure of

the Cu site (at least the Cu–S(Cys) bond) should be al-

tered by addition of the peptide. These spectral changes

were previously attributed to binding of peptide to thenegative patch of PC by the use of the negative patch

mutants of PC [24]. Resonance Raman measurements

showed that the Cu–Scys bond of PC becomes longer by

interaction with the peptide and electrochemical studies

Fig. 3. The absorption spectra of oxidized PC (200 and 400 lM for A

and B, respectively) with (A) and without (B) tetralysine (1.2 mM) and

their difference spectrum multiplied by ten (C). Cell pass lengths were

0.9 and 0.45 mm for A and B, respectively. The baseline of the dif-

ference spectrum is shown as a dotted line. Phosphate buffer (10 mM),

pH 7.0, was used. All measurements were performed at 15 �C.

8

Fig. 5. Plots of DDAbs vs [peptide], together with least-squares

fitted lines according to Eq. (4): tetralysine (d, ——), GKKGGKK

(j, – – –), KKGGGKK (N, � � �), and KGKGKGK (., –�–). Each data

contains an error of �0.0003. Experimental conditions were the same

as those for Fig. 3.

Table 2

Difference in the absorption coefficient changes (DDe) at 630 and 560

nm by peptide binding obtained from Eq. (4)a

Oligopeptide DDe (M�1 cm�1)

KKKK 49� 3

GKKGGKK 35� 3

KKGGGKK 39� 3

KGKGKGK 34� 3

a In 10 mM phosphate buffer, pH 7.0, at 15 �C.

Fig. 4. Difference absorption spectra of oxidized PC obtained by

subtracting the spectrum without oligopeptides from the spectra with

oligopeptides (1.2 mM); tetralysine (A), GKKGGKK (B),

KKGGGKK (C), and KGKGKGK (D). Experimental conditions

were the same as those for Fig. 3.

S. Hirota et al. / Journal of Inorganic Biochemistry 98 (2004) 849–855 853

showed that this structural change raises the redox po-

tential of Cu, making PC adapted for receiving an

electron from its redox partner [24].Since similar changes were observed in the absorption

spectra by interaction with the peptides (Fig. 4), a sim-

ilar structural change was induced at the active site of

PC by interaction with the peptides. The result of a

similar structural change with different peptides dem-

onstrates that electrostatic interactions between PC and

oligopeptides are non-specific without specific salt

bridges. This character is in agreement with that of thePC–cyt c complex with a highly dynamic ensemble of

structures [41]. The character of non-specific interaction

might be important for the gating process and formation

of a transient complex suitable for electron transfer.

The structural change, however, was most significant

when tetralysine was used (Fig. 4). To investigate the

effect of the lysine distribution within the peptide on the

structural change of PC, we titrated the absorptionchanges with the peptide concentrations. The structural

change due to peptide binding should be based on

equilibrium between oxidized PC and the peptide, be-

cause the intensities of the peaks and troughs in the

difference spectra increased with the concentration of

peptide. The observed intensity changes in the absorp-

tion spectra are interpreted by considering formation of

a 1:1 PC � peptide complex by Eq. (1). If we write thedifference between the observed absorption changes at

630 and 560 nm upon peptide binding as DDAbs, the

difference between the absorption coefficient changes at

630 and 560 nm upon peptide binding as DDe, the

concentrations of PC and peptide as [PC] and [peptide],

respectively, and the cell length used for the measure-

ments as l, we obtain the relationship [46,47]:

DDAbs ¼ DDe� 1

2½PC�

�<: þ ½peptide� þ 1

Ki

�

� ½PC��"

þ ½peptide� þ 1

Ki

�2

� 4½PC�½peptide�#1=2

9=; ð4Þ

Plots ofDDAbs vs [peptide] for all the peptides studied are

shown in Fig. 5. The Ki values obtained from Fig. 2 are

more accurate than those obtained by simply fitting the

DDAbs data with Eq. (4) to obtain both Ki and DDe atonce, since the DDAbs values are small and contain someerrors. To obtain DDe, the data in Fig. 5 thus were per-

formed least-squares fitting with Eq. (4) using the Ki

values obtained from Eq. (3) (Table 1). The obtained DDevalues are listed in Table 2. Since the fitting was relatively

successful, it is reasonable to assume that only one pep-

tide is bound to the negative patch of PC, which is in

agreement with the results obtained by the kinetic studies.

The DDe and thus the structural change at the coppersite of PC were largest when tetralysine was used. When

lysines are distributed close to each other within the

peptide, lysines may be less flexible among each other

within the peptide and thus the protein might feel more

854 S. Hirota et al. / Journal of Inorganic Biochemistry 98 (2004) 849–855

strain. As a result, the structural change of PC by the

interaction may become larger for the peptide with ly-

sines close to each other within the peptide. On the other

hand, when the lysines are distributed widely within the

peptide, the peptide could changes its conformationmore easily to fit with the protein and could cause less

strain to the protein on binding.

4. Summary

Tetralysine, GKKGGKK, KKGGGKK, and

KGKGKGK peptides competitively inhibited electron

transfer between cyt c and PC. The inhibitory effect in-creased as the peptide concentrations were increased,

and the association constants between PC and the pep-

tides did not differ significantly (3500–5100 M�1). The

association constant of PC–KGKGKGK, however, was

slightly larger than the constants between PC and other

peptides, which showed that binding to PC is slightly

more efficient when lysines are distributed a little widely

within the peptide.Peaks and troughs were detected at about 460 and

630 nm and at about 560 and 700 nm, respectively, in

the difference absorption spectra of PC between the

spectra with and without the peptides. These changes

were attributed to the structural change at the copper

site of PC by the interaction with the peptides. The

structural change at the copper site was most significant

when tetralysine was used, which showed that thestructural change of PC by interaction with the peptide

becomes larger when the charges are close to each other

within the peptide.

5. Abbreviations

PC plastocyanincyt f cytochrome f

cyt c cytochrome c

PSI photosystem I

LMCT ligand-to-metal charge transfer

KKKK lysine–lysine–lysine–lysine

GKKGGKK glycine–lysine–lysine–glycine–

glycine–lysine–lysine

KKGGGKK lysine–lysine–glycine–glycine–glycine–lysine–lysine

KGKGKGK lysine–glycine–lysine–glycine–

lysine–glycine–lysine

References

[1] A.G. Sykes, Struct. Bond. (Berlin) 75 (1991) 175–224.

[2] M.R. Redinbo, T.O. Yeates, S. Merchant, J. Bioenerg. Biomembr.

26 (1994) 49–66.

[3] E.L. Gross, Photosynth. Res. 37 (1993) 103–116.

[4] A.A. Gewirth, E.I. Solomon, J. Am. Chem. Soc. 110 (1988) 3811–

3819.

[5] E.I. Solomon, M.J. Baldwin, M.D. Lowery, Chem. Rev. 92 (1992)

521–542.

[6] P.M. Colman, H.C. Freeman, J.M. Guss, M. Murata, V.A.

Norris, J.A.M. Ramshaw, M.P. Venkatappa, Nature 272 (1978)

319–324.

[7] J.M. Guss, H.C. Freeman, J. Mol. Biol. 169 (1983) 521–563.

[8] J.M. Guss, P.R. Harrowell, M. Murata, V.A. Norris, H.C.

Freeman, J. Mol. Biol. 192 (1986) 361–387.

[9] H.C. Freeman, in: J.L. Laurent (Ed.), Coordination Chemistry,

vol. 21, Pergamon Press, Oxford, UK, 1981, pp. 29–51.

[10] H. Sugawara, T. Inoue, C. Li, M. Gotowda, T. Hibino, T.

Takabe, Y. Kai, J. Biochem. 125 (1999) 899–903.

[11] B.H. Lee, T. Hibino, T. Takabe, P.J. Weisbeek, T. Takabe, J.

Biochem. 117 (1995) 1209–1217.

[12] S.K. Chapman, C.V. Knox, A.G. Sykes, J. Chem. Soc., Dalton

Trans. (1984) 2775–2780.

[13] G.P. Anderson, D.G. Sanderson, C.H. Lee, S. Durell, L.B.

Anderson, E.L.Gross, Biochim. Biophys. Acta 894 (1987) 386–398.

[14] Z. Adam, R. Malkin, Biochim. Biophys. Acta 975 (1989) 158–163.

[15] E.L. Gross, A. Curtiss, Biochim. Biophys. Acta 1056 (1991) 166–

172.

[16] L.Z. Morand, M.K. Frame, K.K. Colvert, D.A. Johnson, D.W.

Krogmann, D.J. Davis, Biochemistry 28 (1989) 8039–8047.

[17] L. Qin, N.M. Kosti�c, Biochemistry 31 (1992) 5145–5150.

[18] T.E. Meyer, Z.G. Zhao, M.A. Cusanovich, G. Tollin, Biochem-

istry 32 (1993) 4552–4559.

[19] X.-S. Gong, J.Q. Wen, N.E. Fisher, S. Young, C.J. Howe, D.S.

Bendall, J.C. Gray, Eur. J. Biohem. 267 (2000) 3461–3468.

[20] V.A. Roberts, H.C. Freeman, A.J. Olson, J.A. Tainer, E.D.

Getzoff, J. Biol. Chem. 266 (1991) 13431–13441.

[21] G.M. Ullmann, E.-W. Knapp, N.M. Kosti�c, J. Am. Chem. Soc.

119 (1997) 42–52.

[22] J. Illerhaus, L. Altschmied, J. Reichert, E. Zak, R.G. Herrmann,

W. Haehnel, J. Biol. Chem. 275 (2000) 17495–17590.

[23] M.M. Crnogorac, M. Ullmann, N.M. Kosti�c, J. Am. Chem. Soc.

123 (2001) 10789–10798.

[24] S. Hirota, K. Hayamizu, M. Endo, T. Hibino, T. Takabe, T.

Kohzuma, O. Yamauchi, J. Am. Chem. Soc. 120 (1998) 8177–

8183.

[25] M. Ubbink, M. Ejdeb€ack, B.G. Karsson, D.S. Bendall, Structure

6 (1998) 323–335.

[26] P.B. Crowley, N. Vintonenko, G.S. Bullerjahn, M. Ubbink,

Biochemistry 41 (2002) 15698–16705.

[27] V.L. Davidson, Biochemistry 35 (1996) 14035–14039.

[28] G. McLendon, R. Hake, Chem. Rev. 92 (1992) 481–490.

[29] C.C. Moser, J.M. Keske, K. Warncke, R.S. Farid, P.L. Dutton,

Nature 355 (1992) 796–802.

[30] M.M. Ivkovi�c-Jensen, N.M. Kosti�c, Biochemistry 35 (1996)

15095–15106.

[31] M.M. Ivkovi�c-Jensen, N.M. Kosti�c, Biochemistry 36 (1997) 8135–

8144.

[32] S. Modi, M. Nordling, L.G. Lundberg, €O. Hansson, D.S. Bendall,

Biochim. Biophys. Acta 1102 (1992) 85–90.

[33] G.M. Soriano, M.V. Ponamarev, G.-S. Tae, W.A. Cramer,

Biochemistry 35 (1996) 14590–14598.

[34] B.G. Schlarb-Ridley, D.S. Bendall, C.J. Howe, Biochemistry 41

(2002) 3279–3285.

[35] J. McGinnis, J.D. Sinclair-Day, A.G. Sykes, J. Chem. Soc.,

Dalton Trans. (1986) 2007–2009.

[36] S.K. Chapman, I. Sanemasa, A.G. Sykes, J. Chem. Soc., Dalton

Trans. (1983) 2549–2553.

[37] S.K. Chapman, A.D. Watson, A.G. Sykes, J. Chem. Soc., Dalton

Trans. (1983) 2543–2548.

[38] D. Beoku-Betts, S.K. Chapman, C.V. Knox, A.G. Sykes, Inorg.

Chem. 24 (1985) 1677–1681.

S. Hirota et al. / Journal of Inorganic Biochemistry 98 (2004) 849–855 855

[39] S. He, S. Modi, D.S. Bendall, J.C. Gray, EMBO J. 10 (1991)

4011–4016.

[40] S. Modi, S. He, J.C. Gray, D.S. Bendall, Biochim. Biophys. Acta

1101 (1992) 64–68.

[41] M. Ubbink, D.S. Bendall, Biochemistry 36 (1997) 6326–6335.

[42] M.M. Crnogorac, C. Shen, S. Young, €O. Hansson, N.M. Kostic,

Biochemistry 35 (1996) 16465–16474.

[43] G.M. Ullmann, N.M. Kosti�c, J. Am. Chem. Soc. 117 (1995) 4766–

4774.

[44] S. Hirota, M. Endo, T. Tsukazaki, T. Takabe, O. Yamauchi, J.

Biol. Inorg. Chem. 3 (1998) 563–569.

[45] S. Hirota, M. Endo, K. Hayamizu, T. Tsukazaki, T. Takabe, T.

Kohzuma, O. Yamauchi, J. Am. Chem. Soc. 121 (1999) 849–

855.

[46] S. Hirota, K. Hayamizu, T. Okuno, M. Kishi, H. Iwasaki, T.

Kondo, T. Hibino, T. Takabe, T. Kohzuma, O. Yamauchi,

Biochemistry 39 (2000) 6357–6364.

[47] S. Hirota, K. Hayamizu, T. Takabe, O. Yamauchi, Inorg. React.

Mech. 2 (2000) 113–118.

[48] S. Hirota, O. Yamauchi, Eur. J. Inorg. Chem. (2002) 17–25.

[49] S.E. Martinez, D. Huang, A. Szczepaniak, W.A. Cramer, J.L.

Smith, Structure 2 (1994) 95–105.