structure of the manganese complex in photosystem ii ......structure of the mn complex v. k....

Published online 4 October 2002

Structure of the manganese complex in photosystem IIinsights from X-ray spectroscopy

Vittal K YachandraMelvin Calvin Laboratory Physical Biosciences Division Lawrence Berkeley National Laboratory Berkeley

CA 94720 USA

We have used Mn K-edge absorption and K b emission spectroscopy to determine the oxidation states ofthe Mn complex in the various S states We have started exploring the new technique of resonant inelasticX-ray scattering spectroscopy this technique can be characterized as a Raman process that uses K-edgeenergies (1s to 4p ca 6550 eV) to obtain L-edge-like spectra (2p to 3d ca 650 eV) The relevance of thesedata to the oxidation states and structure of the Mn complex is presented We have obtained extended X-ray absorption ne structure data from the S0 and S3 states and observed heterogeneity in the MnndashMndistances leading us to conclude that there may be three rather than two di-m -oxo-bridged units presentper tetranuclear Mn cluster In addition we have obtained data using Ca and Sr X-ray spectroscopy thatprovide evidence for a heteronuclear MnndashCa cluster The possibility of three di-m -oxo-bridged MnndashMnmoieties and the proximity of Ca is incorporated into developing structural models for the Mn clusterThe involvement of bridging and terminal O ligands of Mn in the mechanism of oxygen evolution isdiscussed in the context of our X-ray spectroscopy results

Keywords extended X-ray absorption ne structure oxygen-evolving complex photosystem IIX-ray absorption near-edge structure X-ray absorption spectroscopy

X-ray emission spectroscopy

1 INTRODUCTION

There are two critical questions related to the process ofphotosynthetic water oxidation catalysed by a Mn clusterin the OEC of PSII These questions are as follows

(i) What are the oxidation state(s) and structuralchanges in the Mn complex as the OEC proceedsthrough the S-state cycle

(ii) What is the mechanism by which four electrons areremoved from two water molecules by the Mn com-plex to produce an O2 molecule

We have addressed these questions principally by theuse of XANES Kb XES and EXAFS along with electronparamagnetic resonance spectroscopy (Yachandra et al1996 Robblee et al 2001)

X-ray spectroscopy is element speci c and can monitorMn (or Ca or Sr) directly in the membrane hence it isa good method for studying the structure of the Mn OECwithout interference from pigment molecules the lipidand protein matrix or other metals such as Ca Mg Cuor Fe that are present in active PSII preparations Mn K-edge XANES and Kb XES provide information about oxi-dation states and the site symmetry of the Mn complexand EXAFS at the Mn or Sr or Ca K edge furnishes infor-mation about the number type and distances to neigh-bouring ligand atoms in the S0 S1 S2 and S3 states of the

One contribution of 21 to a Discussion Meeting Issue lsquoPhotosystem IImolecular structure and functionrsquo

Phil Trans R Soc Lond B (2002) 357 1347ndash1358 1347 Oacute 2002 The Royal SocietyDOI 101098rstb20021133

OEC (Yachandra 1995 Yachandra amp Klein 1996) X-rayspectroscopy does not require long-range order the struc-tural studies can be performed on frozen solutions Severalof the intermediate states mentioned above can be stabil-ized as frozen solutions Few other spectroscopic tech-niques provide such speci city for studying the structureof Mn in the OEC

This paper focuses on the application of X-ray spectro-scopic methods to resolve structural questions regardingthe Mn cluster in the OEC and we present our view ofthe structure of the Mn cluster that is re ned based onnew data and a mechanism for water oxidation thatemphasizes the recent results from our laboratory

2 OXIDATION STATES OF THE MANGANESECLUSTER

A key question for the understanding of photosyntheticwater oxidation is whether the four oxidizing equivalentsgenerated by the reaction centre are accumulated on thefour Mn ions of the OEC during S-state turnover orwhether a ligand-centred oxidation takes place especiallybefore the formation and release of molecular oxygen dur-ing the S3 to S4 to S0 transition We have addressed thesequestions by using Mn K-edge XANES (1s to 4pabsorption) K b XES (3p to 1s emission) and the recentlyintroduced RIXS (1s to 3d or 4p absorption followed by2p to 1s Kb emission) to obtain L-edge-like spectra (2pto 3d absorption)

A promising approach to studying the Mn oxidationstates in the native S states is to step samples through theS-state cycle by the application of saturating single-turn-

1348 V K Yachandra Structure of the Mn complex

over ashes and to characterize these samples by X-rayspectroscopy We reported earlier XANES data fromsamples in the S0 through S3 states produced under thesephysiologically relevant conditions (Roelofs et al 1996)The ash-advanced PSII samples prepared from spinachexhibited XANES spectra that showed Mn oxidation fromS0 to S1 and from S1 to S2 but no further oxidation duringthe S2 to S3 transition These results are consistent withour earlier studies by Guiles et al (1990ab) that were pre-pared by illumination at low temperature or by chemicaltreatment

Other spectroscopic techniques such as NMRndashPRE(Sharp 1992) EPR (Dismukes amp Siderer 1981 Hans-son amp Andreasson 1982 Ahrling et al 1997 Messinger etal 1997b) UV absorption (Dekker 1992 Van Leeuwenet al 1993) and Yox

D EPRndashspin echo (Styring amp Rutherford1988) studies show that in PSII the S0 to S1 and S1 toS2 transitions are Mn-centred oxidations and most groupsagree that Mn redox states in the S1 state are (III2 IV2)that are oxidized to (III IV3) in the S2 state There ishowever a long-standing controversy as to whether a Mn-or ligand-centred oxidation occurs in the S2 to S3 tran-sition (Ono et al 1992 Iuzzolino et al 1998)

To resolve the controversy we used high-resolution MnKb X-ray uorescence studies in collaboration with SteveCramer who has constructed a unique high-resolutionemission spectrometer that operates at the Mn K b uor-escence energy and has demonstrated the feasibility ofdetermining the oxidation states of Mn (Bergmann ampCramer 1998) We have completed a comprehensive MnKb XES study in conjunction with XANES on sets ofsamples prepared in a similar manner and characterizedby EPR (Messinger et al 2001)

Due to the fact that the actual S-state composition is ofcritical importance to the interpretation of the results wecharacterized the samples by EPR spectroscopy using theS2 multiline EPR signal as a direct measure for the amountof S2 in our samples ( gure 1) The relative S-state popu-lations in samples given none one two three four veor six ashes were determined from tting the ash-induced EPR multiline signal oscillation pattern to themodel of Kok et al (1970) for each of the samples usedin the X-ray spectroscopy experiments It is essential toobtain high concentrations of PSII in the S states to obtainEPR XANES and Kb spectroscopy data with a goodsignal-to-noise ratio An Nd-YAG laser system (Spectra-Physics PRO 230-10 800 mJ pulse 2 1 at 532 nm 9 nspulse width) was used to illuminate PSII samples fromboth sides simultaneously Due to the fact that the pulsewidths are narrow compared with those of a ash lampdouble hits become negligible The damping of the oscil-lation pattern is minimized thus making it possible toderive an unique set of spectra by deconvolution Whenthe system becomes scrambled by frequent misses anddouble hits the deconvolved spectra are not unique

(a) Manganese K-edge X-ray absorptionnear-edge structure

Our edge data are of suf cient signal-to-noise quality toallow for not only a determination of the absorption edgeposition (using the in ection point energy de ned as thezero-crossing in the second derivative of the edge) but

Phil Trans R Soc Lond B (2002)

20

40

60

80

100

0 1 2 3 4 5 6

flash no

norm

aliz

ed S

2 m

ultil

ine

EP

R s

igna

l (

)

Figure 1 The S2 multiline EPR signal amplitudes are shownas a function of ash number (solid line experimentalaverage) The best t to the S2 multiline oscillation patternis shown as a dashed line (simulation) The S-statecompositions of the samples are derived from the tIndividual S-state spectra can be uniquely determined fromthese data if the oscillation pattern is as pronounced asshown in the gure It is important that a deep oscillationpattern is obtained because all subsequent steps that areused in the deconvolution to generate pure S-state spectraare critically dependent on these data

also an analysis of the structure on the edge as a functionof the number of ashes

The edge spectra of samples given none one two orthree ashes are combined with EPR information to calcu-late the pure S-state edge spectra The shifts in in ectionpoint energy positions and changes in shape are determ-ined by second derivatives of the K-edge spectra and areshown in gure 2 In addition to the shift in edge positionthe S0 to S1 (21 eV) and S1 to S2 (11 eV) transitions areaccompanied by characteristic changes in the shape of theedge also indicative of Mn oxidation The edge positionshifts very little (03 eV) for the S2 to S3 transition andthe edge shape shows only minor changes (Messinger etal 2001)

(b) Kb X-ray emission spectroscopyHigh-resolution Mn XES was performed on S0 S1 S2

and S3 samples prepared using high-power laser ashes asdescribed above In this technique the energy of the MnK b emission (3p to 1s) is measured with a high-resolutiondispersive spectrometer The shape and energy of theK b 1 3 emission re ect the oxidation state(s) of the emittingMn atom(s) The emission occurs from a 3p level that ismainly in uenced by the number of unpaired 3d elec-trons and it is less sensitive to the symmetry and bondingthan the K-edge absorption which involves transitions tothe 4p level (Peng et al 1994) The spin of the unpaired3d valence electrons can be either parallel (Kb 9 ) or anti-parallel (K b 1 3) to the hole in the 3p level The splittingbetween the Kb 9 and K b 1 3 peaks becomes smaller forhigher oxidation states because fewer 3d electrons interactwith the 3p hole Thus in contrast to the in ection points

Structure of the Mn complex V K Yachandra 1349

6535 6545 6555 6565 6575

d2 (FI

0)dE

2

6484 6488 6492 6496

X-ray energy (eV)

S1ndashS0

S2ndashS1

S3ndashS2

S0ndashS3

(a) (b)

Figure 2 (a) Second derivatives of the normalized pure S-state Mn K-edge spectra of the Mn cluster of PSII For clarity avertical line has been drawn at the in ection-point energy of the S1 state The states are as follows topmdashsolid line S0 dashedline S1 middlemdashsolid line S2 dashed line S1 bottommdashsolid line S2 dashed line S3 The change in IPE ( n in eV) for eachS-state advance is as follows top 21 eV middle 11 eV bottom 03 eV For reference the K-edge spectrum of the S1 stateis also plotted (dotted line) (b) The S12 S0 S22 S1 S32 S2 and S02 S3 difference high-resolution Mn X-ray K b -emissionspectra The derivative shapes indicate shifts in the energy of the spectra between the S0 and S1 and S1 and S2 spectraindicating oxidations of Mn The S32 S2 difference spectra demonstrates the similarity of the spectra indicating that thisadvance does not involve an Mn-centred oxidation

of the XANES edges the Kb 1 3 peaks shift to a lowerenergy with higher oxidation states

Using a rst-moment analysis the position of the mainKb 1 3 peak has been calculated for each S state Based onthe shifts or lack thereof of the rst moments we con-cluded that Mn is not oxidized during the S2 S3 tran-sition Our results (shown as difference spectra in gure2) show a shift in energy between the S0- and S1-statespectra and between the S1- and S2-state spectra and seenin the derivative-shaped difference spectra Howeverthere is very little change between the S2 and S3 states asseen in the difference spectrum (Messinger et al 2001)

We have used two independent spectroscopic tech-niques K b and XANES spectroscopy along with EPRto examine the Mn redox states in PSII Figure 3 summar-izes the oscillation patterns of the rst moments from theKb 1 3 spectra and of the XANES spectra in ection pointenergies Both patterns show large shifts between the 0Fand 1F samples which re ect the Mn-centred oxidationduring the S1 S2 transition A much smaller shiftoccurs between the 1F and 2F samples which is inconsist-ent with a Mn-centred oxidation during the S2 S3 tran-sition The second derivatives of the S-state XANESspectra and the Kb difference spectra ( gure 2) con rmthis conclusion It is unlikely that both techniques wouldfail to detect a Mn-centred oxidation for the S2 S3

transition if it did in fact occur Support for this con-clusion has also come from XANES and Kb studies oftwo sets of structurally homologous Mn compounds indifferent oxidation states (Visser et al 2001) We conclude


that probably no direct Mn oxidation is involved in thistransition The proposed Mn-oxidation state assignmentsare as follows S0 (II III IV IV) or (III III III IV) S1

(III III IV IV) S2 (III IV IV IV) S3 (III IV IV IV)(Messinger et al 2001)

(c) Resonant inelastic X-ray scatteringspectroscopy

Mn L-edge spectra have several advantages over Mn K-edge spectra because multiplet calculations can beapplied to L-edge spectra that are sensitive to the numberof holes in the 3d level the oxidation state of Mn andthe spin state of Mn This is a rich amount of informationabout oxidation states relative to what is available from K-edge XANES spectra L-edge spectra have already beenobtained for a number of Fe Cu and Ni metalloproteinsand have been simulated using multiplet calculations toprovide quantitative conclusions about the oxidationstates of the metals involved (Wang et al 1996 1997)However the main disadvantage of using soft X-rayswhere the Mn L-edge occurs (ca 650 eV) is that thesamples are subjected to ultra-high vacuum and are moreprone to X-ray damage

We have recently started exploring the RIXS techniquethat circumvents the problems mentioned above in a veryelegant manner It is possible to combine the two types ofK- and L-edge experiments by mapping the 1s2p RIXSlandscape as shown in gure 4 in a simpli ed two-stepprocess (Caliebe et al 1998) RIXS involves exciting apre-edge transition (1s to 3d) of the K shell using hard X-


015

017

019

021

023

flash no

ltE

gtndash 6490 eV

0 1 2 3

25

30

35

40

0 1 2 3 4 5 6

IPE

ndash 6

550

eV

(a) (b)

Figure 3 (a) IPE (in eV) of the Mn K-edge of PSII membranes as a function of the number of applied ashes (b) The rstmoments of the Mn K b uorescence spectra ( E ) as a function of ash number The Mn K-edge shifts to higher energy asMn is oxidized The rst moment of the Mn K b uorescence shifts to lower energy as Mn is oxidized The complementaryand mutually reinforcing results provide a strong case for Mn oxidation during the S0 to S1 and S1 and S2 advances but noMn-centred oxidation at the S2 to S3 advance

bind

ing

ener

gy

K edge

1s22p6hellip3d4

1s12p6hellip3d44p1

1s12p6hellip3d4p 1

1s12p6hellip3d5

J =32

J =12

L edge

1s22p6hellip3d4

1s22p5hellip 3d5

1s22p5hellip3d5

RIXS

EB = ndash f

f2p 1s

1s22p6hellip3d4

1s12p6hellip3d44p1

1s12p6hellip3d4p 1

1s12p6hellip3d5

1s22p5hellip3d51s22p5hellip3d4p 1

1s22p5hellip3d4p 1

1s22p5hellip3d5

u

u u

u

Figure 4 The electronic energy diagram of the transitions that occur in Mn K-edge L-edge and RIXS spectroscopy L-edge-like information is extracted from RIXS by subtracting the emission energy f from the excitation energy y (energy transfer( y ndash f ) versus the incident energy y ) However more nal states are present in the RIXS spectra than in the L-edge spectraTherefore RIXS gives a wealth of information about the Mn-oxidation state ligand environment and local symmetry

rays Then the high-resolution X-ray emission analysercan be tuned to detect the K a uorescence of Mn whichis a 2p to 1s transition The nal state from this transitionis a 2p hole the same nal state as the L-edge uorescence(3d to 2p) An advantage of RIXS over both conventionaltechniques is that the resonances are separated accordingto the nal states they decay into As a result the reson-ances appear as islands in the resonant landscape and theirspectral shapes can be analysed A 2D RIXS spectrum isobtained by plotting the scanned incident X-ray energy y(1s to 3d) versus EB the binding energy which is equalto y 2 f ) f is the energy of the scanned high-resolutionX-ray emission analyser Furthermore this techniqueoffers an unprecedented opportunity to selectively excite


into speci c molecular orbitals with 3d character by tun-ing the incoming X-rays to look at a speci c pre-edge tran-sition at least two to three resolvable pre-edge transitionsare seen in the Mn K-edge spectrum of PSII This meansthat it would be possible to obtain lsquonestedrsquo RIXS spectrathat correspond to uorescence from different molecularorbitals with a 3d character

We have collected preliminary data from a PSII samplein the S1 and S2 states The shape and ratio of the LIII toLII-like peaks of the PSII RIXS spectrum are in betweenthose for the MnII I and MnIV complex and very differentfrom that for the MnII complex This indicates a mixed(III) and (IV) Mn oxidation state in the S1 state ie(III2IV2) a conclusion in accord with that derived from


other studies The cross-sections of the 2D RIXS spectrareveal the L-edge-like character showing two well-resolved peaks due to spin-orbit coupled states J = 32(LII I-like edge) and J = 12 (LII-like edge) The ratio of theLII I to LII peaks increases from (II) to (III) to (IV) theratio for PSII changes from S1 to S2 in a manner consistentwith an oxidation from (III)2(IV)2 to (III)(IV)3

3 STRUCTURE OF THE MANGANESE CLUSTER

We have used Mn EXAFS to study the local structureof the Mn complex and have proposed topologically con-sistent models that can be reconciled with several struc-tural arrangements of the four Mn atoms (DeRose et al1994) The proposed model consists of two or three di-m -oxo-bridged binuclear Mn units with MnndashMn distancesof ca 27 Ac that are linked to each other by a mono-m -oxo bridge with a MnndashMn separation of ca 33 Ac andwith Ca at a distance of 34 Ac to Mn (Yachandra et al1993 DeRose et al 1994) The MnndashMn distances are lar-gely invariant in the native S1 and S2 states (DeRose et al1994) EXAFS experiments on the S3 and S0 states havebeen dif cult to perform on PSII samples using single ashes The requirement of optically dilute samples toensure saturation by single actinic ashes generally resultsin a low signal-to-noise ratio We have recently been ableto extend these studies to the S3 and S0 states preparedby ash illumination

(a) S3 stateEarlier EXAFS data from the chemically treated S3-

state samples produced by a double turnover method indi-cated that the two di-m -oxo-bridged MnndashMn dimer unitsmay become non-equivalent (Guiles et al 1990b) TheEXAFS spectra from Ca-depleted S9

3 samples (S2YZ) pre-

pared by low pH treatment in a citrate buffer do not exhi-bit similar heterogeneity in 27 Ac MnndashMn distances(Latimer et al 1998)

EXAFS data of S3 samples created under physiologicalconditions with saturating ash illumination show signi -cant changes in the MnndashMn distances in the S3 state com-pared with the S1 and the S2 states The two 27 Ac MnndashMn distances that characterize the di-m -oxo centres in theS1 and S2 states are lengthened to ca 28 and 30 Ac in theS3 state respectively ( gure 5) The 33 Ac MnndashMn andMnndashCa distances also increase by 004ndash02 Ac Thesechanges in the MnndashMn distances are interpreted as conse-quences of the onset of substrate water oxidation in theS3 state Mn-centred oxidation is evident during the S0 toS1 and S1 to S2 transitions During the S2 to S3 transitionwe propose that the changes in the MnndashMn distances arethe result of ligand or water oxidation leading to the for-mation of an oxyl radical intermediate formed at a bridg-ing or terminal position We propose that substratendashwateroxidation occurring at this transition provides the triggerfor the formation of the OndashO bond the critical step in thewater oxidation reaction The reaction of the oxyl radicalwith OH 2 H2O or an oxo group during the subsequentS-state conversion is proposed to lead to the formation ofthe OndashO bond (Liang et al 2000)


35 55 75 95 115

_50

50

00

photo-electron wavevector k (Aringndash1)

(k

) times

k3

I II

III

FT a

mpl

itud

e

00

02

apparent distance Rrsquo (Aring)

00 40302010

c

(a)

(b)

Figure 5 (a) Fourier- ltered k-space EXAFS data from S2

(dashed line) and S3 (solid line) state of PSII samples Thedifferences in phase frequency and amplitudes between theraw S2 and S3 state EXAFS spectra are very obvious in thesespectra (b) FT power spectra of S2 (dashed line) and S3

(solid line) states of PSII The major Fourier peaks arelabelled I II and III The spectra are clearly differentbetween the S2 (dashed line) and S3 states (solid line)There is a reduction in amplitude in all three peaks in the S3

state compared with the S2 state More importantly peaks IIand III are at a greater apparent distance R9 for the S3 statecompared with the S2 state as shown

(b) S0 stateAlthough many topological structures have been pro-

posed on the basis of EXAFS (Penner-Hahn 1998 Dauet al 2001 Robblee et al 2001) and EPR (Hasegawa etal 1999a Peloquin amp Britt 2001) studies for the Mn clus-ter in PSII the issue of whether there are two or three di-m -oxo-bridged moieties in the Mn cluster has been anopen question However the determination of the exactnumber of such interactions is important for narrowingthe number of options from the many that have been pro-posed We have obtained data from the S0 state that showthat there may be three di-m -oxo-bridged motifs that arecharacterized by the approximate 27 Ac MnndashMn distance

The S0 samples were generated by three ash illumi-nation followed by incubation with carbonyl cyanide 4-



FT

mag

nitu

de

0 2 6 8 104

01

02

I II

III

Figure 6 FTs of the S0 (dashed line) and S1 state (solidline) of the k3-weighted Mn-EXAFS spectra (35ndash115 Ac 2 1)The main Fourier peaks are labelled I II and III There areclear differences in the amplitude and position of these peaksbetween the S0 and S1 states The peak labelled II is fromMnndashMn interactions at ca 27 Ac

(tri uoromethoxy) phenyl hydrazone (Messinger et al1997a) The multiline EPR spectra of the S0 and S2 stateswere used to characterize all the samples

Figure 6 shows the FTs of the k 3-weighted EXAFS datafrom the S0 and S1 states The Fourier peak II was isolatedand t to the EXAFS equation with one or two distancesfor the S0 and S1 states Since there has to be an integralnumber of MnndashMn interactions the ts were constrainedto a ratio of 1 1 and 2 1 for the two MnndashMn interac-tions The ts for the S1 state showed that the two dis-tances are separated by signi cantly less than what can beresolved using the dataset at present However the situ-ation is very different for the ts for the S0 state The qual-ity of t parameter the laquo 2 value improved considerablyon going from a one-shell to a two-shell t especially inthe case of the 2 1 ratio of MnndashMn interactions Theresult of the tting provides evidence that there may bethree short (ca 27ndash28 Ac ) MnndashMn interactions in theMn cluster

Figure 7 shows the best ts to two MnndashMn distancesfor the S0 state as a function of the number of each of theMnndashMn interactions The contour plot graphically illus-trates that the global minimum at two MnndashMn interac-tions at 27 Ac and one MnndashMn interaction at 285 Ac isvery well de ned and hence gives us con dence in itsreliability We have re-examined the data from the S curren

0 theg = 41 S2 the NH3- or F 2 -inhibited S2 and the S3 stateswhere we have shown that there is distance heterogeneityin the MnndashMn vectors In each case the ratio of the num-ber of the two MnndashMn interactions was ca 2 1 Thisleads us to think that there may be three Mn interactionsat ca 27 Ac in the Mn cluster in its native state one ofwhich is perturbed in the S0 F 2 - and NH3-inhibited S2

and the g = 41 S2 states and all three are perturbed inthe S3 state

We showed earlier that there are several topologicalmodels that are compatible with our EXAFS data Thesemodels have been described in detail (DeRose et al 1994Robblee et al 2001) However only one of those options(A) is widely used as a working model although options


0 321

2

1

number of 27 Aring MnndashMn vectors

num

ber

of 2

85

Aring M

nndashM

n ve

ctor

s

Figure 7 The contour plot shows the quality of tparameter laquo 2 for the S0 state plotted versus the number ofthe two MnndashMn interactions The dashed lines show thedistinct minimum at one MnndashMn vector at 285 Ac and twoMnndashMn vectors at 27 Ac

E F and G (the nomenclature is from DeRose et al(1994)) have been shown to be preferred on the basis ofEPR and electron nuclear double resonance simulations(Hasegawa et al 1999b Peloquin amp Britt 2001) On thebasis of the present results from the S0 state we think itis important to consider the options that include three di-m -oxo-bridged moieties in the Mn cluster Figure 8 showsseveral such options G I and J that were proposed earl-ier and two newer options L and M among severalothers that can be conceived Options J and I are lessprobable structures because J lacks a MnndashMn interactionat 33 Ac and I has two such interactions The EXAFSdata from an inorganic compound (Auger et al 1990) withthe motif in option J are very different from those obtainedfrom a PSII sample (R M Cinco and V K Yachandraunpublished data) Options G L and M all have three di-m -oxo bridges and one mono-m -oxo bridge that is three27 Ac MnndashMn interactions and one 33 Ac MnndashMn inter-action and are qualitatively in agreement with the asym-metry seen in the electron density of the Mn cluster(Zouni et al 2001) Options L and M are also similar tothe structure proposed on the basis of density functionaltheory calculations (Siegbahn 2000) We are in the pro-cess of testing the compatibility of various Mn models thatsatisfy the criteria set by the distance data and orientationof the various MnndashMn MnndashCa vectors obtained fromEXAFS with the electron-density data obtained from X-ray crystallography

4 THE ROLE OF THE CALCIUM COFACTOR

Calcium and chloride are the necessary cofactors for theproper function of the OEC of PSII (Debus 1992) Whilethe Mn complex has been extensively studied by X-rayabsorption techniques (XANES and EXAFS) compara-tively less is known about the Ca2 1 cofactor The fewernumber of studies on the Ca2 1 cofactor have sometimesrelied on substituting the native cofactor with Sr or othermetals (Ghanotakis et al 1985 Boussac amp Rutherford1988) and have stirred some debate about the structure


Mn

O

OMn

O

OMn

O

Mn

O

O

Mn

O

Mn

MnO

O Mn

Mn

O

OMn

O

Mn

Mn O Mn

O

OMn

Mn

O

OMn

OCa

O

Mn

O O

Mn

Mn

O

OMn

O

MnO

O Mn

Mn O

O Mn

Mn O

O Mn

O

Mn

O

O

Mn

O

Mn O

OMn

O

Mn

Mn

O

Mn

O

O

O

OMn

Mn

O

OMnMn

O

OMn

Mn

O

Mn

O

Mn Mn

O

Mn

OOO

O

Mn

Mn

O

OMn

O

Mn

OMn

O

O

Mn

O

MnMnMn

O

OO

(a) (b) (c) (d )

(e) ( f ) (g)

(h) (i) ( j )

(k) (l) (m )

Figure 8 Structural models for the Mn cluster with three di-m -oxo-bridged Mn moieties Options G I and J are fromDeRose et al (1994) and L and M are two other viable options One or two mono-m -oxo-bridged motifs are also present inoptions G I L and M No such motif exists in option J

of the binding site Past efforts using Mn EXAFS on Sr-substituted PSII have indicated a close link between theMn cluster and Sr within 35 Ac (Latimer et al 1995) Themost recently published study using Sr EXAFS on similarsamples con rms this nding of a 35 Ac distance betweenMn and Sr (Cinco et al 1998) This nding was based ona second Fourier peak in the Sr EXAFS from functionalsamples but is absent from inactive hydroxylamine-treated PSII This Fourier peak II was found to t bestto two Mn at 35 Ac rather than lighter atoms (carbon)Nevertheless other experiments have given contrastingresults (RiggsGelasco et al 1996)

(a) Orientated strontium extended X-rayabsorption ne structure

We have extended the technique using polarized SrEXAFS on layered Sr-substituted samples to provideimportant angle information Polarized EXAFS involvescollecting spectra for different incident angles ( u ) betweenthe membrane normal of the layered sample and the X-rayelectric eld vector Dichroism in the EXAFS can occurdepending on how the particular absorberndashbackscatterer(AndashB) vector is aligned with the electric eld Throughanalysis of the dichroism we extract the average orien-tation (f ) of this AndashB vector relative to the membranenormal and the average number of scatters per absorbing


atom (Niso) Constraints on the structural model are thenimposed by these parameters

Sr-substituted PSII samples were made by a process ofCa2 1 depletion Sr2 1 reactivation and Chelex treatment toremove any excess Sr2 1 (Cinco et al 1998) Orientatedsamples were made by layering onto at Mylar lms(Mukerji et al 1994 Dau et al 1995) The FTs from thepolarized Sr EXAFS showed extreme dichroism in Fourierpeak II ( gure 9) Preliminary nonlinear least-squaresregression analysis produced the solid curve shown in g-ure 9 as the best t of the 15 data points (angles from sixseparate samples) and the result translates to 1ndash2 SrndashMnvectors with an average angle of 23 plusmn 4deg from the mem-brane normal

The orientation data from the Sr EXAFS experimentscan be combined with the dichroism data from MnEXAFS data to calculate the orientation of the 33 Ac MnndashMn vector Fourier peak III in the Mn EXAFS whichcontains MnndashMn (33 Ac ) and MnndashCa (34 Ac ) contri-butions is dichroic with an average angle of 43 plusmn 10deg withrespect to the membrane normal (Mukerji et al 1994) Byincluding the MnndashCa vector at 23deg we calculate an angleof ca 62deg for the 33 Ac MnndashMn vector Previous polarizedMn EXAFS experiments on PSII have shown angles of55deg and 67deg for the 27 Ac MnndashMn vectors (Mukerji et al1994 Dau et al 1995) All MnndashMn vectors then lie at


0 2 4 6 8 10

FT

mag

nitu

de


I

II

01

02

03

= 10 Ecirc

= 80 Ecirc

X-raye-vector

Mn

0

30

6090

120

210

240270

300

330

experimental angle (deg)

005101520253035

Nap

p

(a) (b)

q

q

q

Figure 9 (a) FTs of k3-weighted Sr EXAFS from orientated SrndashPSII samples at two angles ( u ) The dichroism is mostreadily apparent in Fourier peak II (R 9 = 30 Ac ) which is assigned to backscattering from Mn (b) The polar plot shows Nexp

(black circles) plotted versus the angle of detection u The solid line is the best t from which we obtain Niso (number of Mnneighbours to Sr) and f the angle the MnndashSr vector makes with the membrane normal

roughly the same angle (ca 60deg) with respect to the mem-brane plane but are not restricted to being collinearbecause the PSII membranes are ordered in one dimen-sion only

(b) Calcium extended X-ray absorption nestructure

In a complementary and de nitive experiment we haveused Ca K-edge EXAFS studies to probe the binding siteof the native cofactor for any nearby Mn within ca 4 Ac This is analogous to the Sr EXAFS studies already pub-lished (Cinco et al 1998) but it focuses on the nativecofactor and avoids the treatments involving Ca depletionand Sr substitution This technique is a more sensitive anddirect probe of the Ca binding site in PSII than SrEXAFS

Chelex-treated PSII (2 Ca per 4 Mn per PSII) wasaccumulated on the at surface of solid Plexiglas (Ca free)while not affecting oxygen-evolving activity or the S2 EPRmultiline signal To prepare the parallel control (inactive)sample 40 m l of hydroxylamine (NH2OH 100 mM) wasadded to the surface of the layered PSII and was allowedto soak and dry without loss of PSII material

The FT of the Ca EXAFS is presented in gure 10 Thespectra are remarkably similar to the Fourier transforms ofthe earlier Sr EXAFS study with Sr substituted for CaThe rst (largest) Fourier peak corresponds to the coordi-nating oxygen atoms closest to Ca In contrast to the con-trol (NH2OH-treated) sample the Chelex-treated PSIIshows a second Fourier peak When this peak II is isolatedand simulated with possible scattering atoms it corre-sponded best to Mn at 34 Ac rather than to light atom(C O or Cl) neighbours These results were consistentwith the earlier Sr EXAFS studies (Cinco et al 1998)

The Ca EXAFS protocol has directly probed the Cacofactor in its native state avoiding potentially disruptivetreatments such as the low pH exposure involved in metalsubstitution This result provides proof of the MnndashCa hetero-nuclear cluster as the catalytic site of oxygen evolutionin PSII


0

04

08

12

2 4 6 8 10

FT

mag

nitu

de

apparent distance R rsquo (Aring)

I II

Figure 10 FT of Ca EXAFS from Chelex-treated layeredsamples with 2 Ca per PSII (k3 weighted k = 25ndash105 Ac 2 1The FTs show the presence of a second Fourier peak in the2 Ca per PSII sample that ts to CandashMn that is absent inthe control sample Solid line intact 2 Ca per PSII (Chelextreated) dashed line inactive 2 Ca per PSII (NH2OHtreated)

Due to the fact that signi cant angle information aboutMnndashMn and MnndashCa vectors is now available as describedabove other topological models previously discussed(DeRose et al 1994) can be re ned to include the pres-ence of Ca and account for the dichroism data Threepossible re ned models are presented in gure 11 RecentEXAFS data from the S0 state has indicated that theremay be three di-m -oxo-bridged MnndashMn units (see above)However such motifs were not included in gure 11 butconsidering them is a subject for future work

5 MECHANISM OF WATER OXIDATION

The many mechanisms of water oxidation that havebeen proposed can be broadly divided into four groups


Ca

O

O

Mn

MnMn

MnO

O

O

Ca

OO

Mn

Mn Mn

O O

O

Mn

Ca

O

Mn

O

O Mn Mn

Mn O O

membranenormal

(a) (b) (c)

Figure 11 Re ned models for the active site of the OEC in PSII These models combine the nding from orientated Sr-substituted PSII samples with previous results from Mn EXAFS on orientated PSII samples These are derived from corestructures that have been described in earlier studies ((a) and (b) from option A and (c) from option F in DeRose et al(1994)) Mn and Ca atoms are as labelled and the white circles with those with circles in represent oxygen atoms

in which the kind of oxygen atom (Mn-terminal oxygenligand Mn-bridging oxygen ligand or exogenous oxygenfrom water or hydroxide) involved in OndashO bond formationis different In these four groups the oxygen atoms can dederived from two terminal oxygen atoms bound to Mnatoms (Haumann amp Junge 1999 Hoganson amp Babcock2000 Limburg et al 2000 Messinger et al 2001 Renger2001 Robblee et al 2001 Hillier amp Wydrzynski 2001)one bridging or terminal oxygen and one exogenous oxy-gen (Siegbahn amp Crabtree 1999 Dau et al 2001 Hillier ampWydrzynski 2001 Kuzek amp Pace 2001 Messinger et al2001 Robblee et al 2001 Vrettos et al 2001) one ter-minal and one bridging oxygen (Limburg et al 2000 Nug-ent et al 2001) or two bridging oxygens (Brudvig ampCrabtree 1986 Christou amp Vincent 1987 Messinger et al2001 Robblee et al 2001) The results that have a directbearing on this subject are the H2

1 6OndashH21 8O exchange

studies by Messinger and co-workers that support thepresence of two non-equivalent exchangeable sites in theS3 state (Messinger et al 1995 Hillier et al 1998)

We have proposed earlier a model for the S3 state inwhich an oxyl radical is generated on one of the m -oxobridges that results in an increase of one of the MnndashMndistances to ca 295 Ac ( gure 5) (Yachandra et al 1996Liang et al 2000) Consistent with our XANES results isthe implication in this structure that the oxidative equival-ent is not stored on the Mn atoms per se during theS2 S3 transition but is delocalized with signi cantcharge and spin density on the bridging oxo ligand Thesedata are reinforced by our Mn K b emission studies of thevarious S states (Messinger et al 2001) Second the MnndashMn distance in both of the di-m -oxo-bridged unitsincreases from 272 to 282 Ac and 295 Ac upon the forma-tion of the S3 state These changes imply a signi cantstructural change in the Mn cluster as it proceeds to theS3 state It is dif cult to rationalize such changes in MnndashMn distance as arising purely from Mn oxidation orinvolving Mn terminal ligands It is dif cult to understandhow changes in terminal ligation can generate such a pro-found change on the MnndashMn distances in the S3 stateReplacement of terminal ligands in di-m -oxo-bridgedmodel compounds has a minimal effect on the MnndashMndistance of 27 Ac that is characteristic of such di-m -oxo-bridged Mn compounds (Wieghardt 1989 Pecoraro


1992) However it is easier to rationalize increases in MnndashMn distances as being due to changes in the bridgingstructure We propose that substratendashwater oxidationchemistry is occurring at the S2 to S3 transition as evi-denced by the lack of Mn oxidation and leading to thesigni cant structural changes as seen in the increase in theMnndashMn distances Our proposed mechanism where theOndashO bond is formed between one critical bridging oxygenand another oxygen atom derived from a bridging or ter-minal oxygen ligand or an exogenous oxygen also avoidsthe formation of the OndashO bond until the most oxidizedstate (S4) is reached This precludes the formation andrelease of peroxide or other oxidation products of waterin the earlier S states thus preventing the system fromlsquoshort circuitingrsquo and avoiding the risk of damaging thepolypeptides of PSII

This work was supported by a National Institutes of Healthgrant (GM 55302) and by the Director Of ce of ScienceOf ce of Basic Energy Sciences Division of Energy Biosci-ences US Department of Energy under contract no DE-AC03-76SF00098 The important contributions to theresearch presented in this article by Dr R Cinco Dr JRobblee Dr J Messinger Dr H Visser Dr U Bergmann DrP Glatzel Dr C Fernandez Dr E Anxolabehere-Mallart DrS Pizarro Dr K McFarlane Dr J Yano Professor M KleinProfessor K Sauer and Professor S Cramer are gratefullyacknowledged The author thanks his collaborators ProfessorW H Armstrong Professor G Christou Professor J-JGirerd Professor R N Mukherjee and Professor K Wiegh-ardt for providing all the inorganic Mn compounds Synchro-tron radiation facilities were provided by the StanfordSynchrotron Radiation Laboratory (SSRL) and the AdvancedPhoton Source (APS) both supported by the US Departmentof Energy The Biotechnology Laboratory at SSRL andBioCAT at the APS are supported by the National Center forResearch Resources of the National Institutes of Health

REFERENCES

Ahrling K A Peterson S amp Styring S 1997 An oscillatingmanganese electron paramagnetic resonance signal from theS0 state of the oxygen evolving complex in photosystem IIBiochemistry 36 13 148ndash13 152

Auger N Girerd J-J Corbella M Gleizes A amp Zimmer-mann J-L 1990 Synthesis structure and magnetic proper-ties of the stable triangular [Mn(IV)3O4]41 core J AmChem Soc 112 448ndash450


Bergmann U amp Cramer S P 1998 A high-resolution large-acceptance analyzer for x-ray uorescence and Raman spec-troscopy In Proc SPIE Crystal and Multilayer Optics vol3448 pp 198ndash210 San Diego CA SPIE

Boussac A amp Rutherford A W 1988 Nature of the inhibitionof the oxygen-evolving enzyme of photosystem-II inducedby NaCl washing and reversed by the addition of Ca21 orSr21 Biochemistry 27 3476ndash3483

Brudvig G W amp Crabtree R H 1986 Mechanism for photo-synthetic oxygen evolution Proc Natl Acad Sci USA 834586ndash4588

Caliebe W A Kao C-C Hastings J B Taguchi M Kot-ani A Uozumi T amp DeGroot F M F 1998 1s2p res-onant inelastic x-ray scattering in a -FeO3 Phys Rev B 5813 542ndash13 548

Christou G amp Vincent J B 1987 The molecular double-pivot mechanism for water oxidation Biochim Biophys Acta895 259ndash274

Cinco R M Robblee J H Rompel A Fernandez CYachandra V K Sauer K amp Klein M P 1998 StrontiumEXAFS reveals the proximity of calcium to the manganesecluster of oxygen-evolving photosystem II J Phys ChemB 102 8248ndash8256

Dau H Andrews J C Roelofs T A Latimer M J LiangW C Yachandra V K Sauer K amp Klein M P 1995Structural consequences of ammonia binding to the manga-nese center of the photosynthetic oxygen-evolving com-plexmdashan X-ray-absorption spectroscopy study of isotropicand oriented photosystem-II particles Biochemistry 345274ndash5287

Dau H Iuzzolino L amp Dittmer J 2001 The tetra-manga-nese complex of photosystem II during its redox cyclemdashX-ray absorption results and mechanistic implicationsBiochim Biophys Acta 1503 24ndash39

Debus R J 1992 The manganese and calcium-ions of photo-synthetic oxygen evolution Biochim Biophys Acta 1102269ndash352

Dekker J P 1992 Optical studies on the oxygen-evolvingcomplex of photosystem II In Manganese redox enzymes (edV L Pecoraro) pp 85ndash103 New York VCH

DeRose V J Mukerji I Latimer M J Yachandra V KSauer K amp Klein M P 1994 Comparison of the manga-nese oxygen-evolving complex in photosystem II of spinachand Synechococcus sp with multinuclear manganese modelcompounds by X-ray absorption spectroscopy J Am ChemSoc 116 5239ndash5249

Dismukes G C amp Siderer Y 1981 Intermediates of a poly-nuclear manganese center involved in photosynthetic oxi-dation of water Proc Natl Acad Sci USA 78 274ndash278

Ghanotakis D F Babcock G T amp Yocum C F 1985Structure of the oxygen-evolving complex of photosystem IIcalcium and lanthanum compete for sites on the oxidizingside of photosystem II which control the binding of water-soluble polypeptides and regulate the activity of the manga-nese complex Biochim Biophys Acta 809 173ndash180

Guiles R D Yachandra V K McDermott A E ColeJ L Dexheimer S L Britt R D Sauer K amp KleinM P 1990a The S0 state of photosystem-II induced byhydroxylaminemdashdifferences between the structure of themanganese complex in the S0 and S1 states determined byX-ray absorption-spectroscopy Biochemistry 29 486ndash496

Guiles R D (and 10 others) 1990b The S3 state of photosys-tem II differences between the structure of the manganesecomplex in the S2 and S3 states determined by X-ray absorp-tion spectroscopy Biochemistry 29 471ndash485

Hansson O amp Andreasson L-E 1982 EPR-detectable mag-netically interacting manganese ions in the photosyntheticoxygen-evolving system after continuous illuminationBiochim Biophys Acta 679 261ndash268


Hasegawa K Ono T-A Inoue Y amp Kusunoki M 1999aHow to evaluate the structure of a tetranuclear Mn clusterfrom magnetic and EXAFS data case of the S2-state Mn-cluster in photosystem II Bull Chem Soc Jpn 72 1013ndash1023

Hasegawa K Ono T-A Inoue Y amp Kusunoki M 1999bSpin-exchange interactions in the S2-state manganesetetramer in photosynthetic oxygen-evolving complexdeduced from g=2 multiline EPR signal Chem Phys Lett300 9ndash19

Haumann M amp Junge W 1999 Photosynthetic water oxi-dation a simplex-scheme of its partial reactions BiochimBiophys Acta 1411 86ndash91

Hillier W amp Wydrzynski T 2001 Oxygen ligand exchange atmetal sitesmdashimplications for the O2 evolving mechanism ofphotosystem II Biochim Biophys Acta 1503 197ndash209

Hillier W Messinger J amp Wydrzynski T 1998 Kineticdetermination of the fast exchanging substrate water mol-ecule in the S3 state of photosystem II Biochemistry 3716 908ndash16 914

Hoganson C W amp Babcock G T 2000 Mechanistic aspectsof the tyrosyl radical-manganese complex in photosyntheticwater oxidation In Manganese and its role in biological pro-cesses vol 37 (ed A Sigel amp H Sigel) pp 613ndash656 NewYork Marcel Dekker

Iuzzolino L Dittmer J Dorner W Meyer-Klaucke W ampDau H 1998 X-ray absorption spectroscopy on layeredphotosystem II membrane particles suggests manganese-centered oxidation of the oxygen-evolving complex for theS0ndashS1 S1ndashS2 and S2ndashS3 transitions of the water oxidationcycle Biochemistry 37 17 112ndash17 119

Kok B Forbush B amp McGloin M 1970 Cooperation ofcharges in photosynthetic oxygen evolution I A linear fourstep mechanism Photochem Photobiol 11 457ndash475

Kuzek D amp Pace R J 2001 Probing the Mn oxidation statesin the OEC Insights from spectroscopic computational andkinetic data Biochim Biophys Acta 1503 123ndash137

Latimer M J DeRose V J Mukerji I Yachandra V KSauer K amp Klein M P 1995 Evidence for the proximityof calcium to the manganese cluster of photosystem IImdashdetermination by X-ray-absorption spectroscopy Biochemis-try 34 10 898ndash10 909

Latimer M J DeRose V J Yachandra V K Sauer K ampKlein M P 1998 Structural effects of calcium depletion onthe manganese cluster of photosystem II determination byX-ray absorption spectroscopy J Phys Chem B 1028257ndash8265

Liang W Roelofs T A Cinco R M Rompel A LatimerM J Yu W O Sauer K Klein M P amp YachandraV K 2000 Structural change of the Mn cluster during theS2 S3 state transition of the oxygen-evolving complex ofphotosystem II Does it re ect the onset of watersubstrateoxidation Determination by Mn X-ray absorption spec-troscopy J Am Chem Soc 122 3399ndash3412

Limburg J Brudvig G W amp Crabtree R H 2000 Modelingthe oxygen-evolving complex in photosystem II In Biomi-metic oxidations catalyzed by transition metal complexes (ed BMeunier) pp 509ndash541 London Imperial College

Messinger J Badger M amp Wydrzynski T 1995 Detectionof one slowly exchanging substrate water molecule in theS3 state of photosystem-II Proc Natl Acad Sci USA 923209ndash3213

Messinger J Nugent J H A amp Evans M C W 1997aDetection of an EPR multiline signal for the S0 state in pho-tosystem II Biochemistry 36 11 055ndash11 060

Messinger J Robblee J H Yu W O Sauer K Yachan-dra V K amp Klein M P 1997b The S0 state of the oxygen-evolving complex in photosystem II is paramagnetic detec-


tion of an EPR multiline signal J Am Chem Soc 11911 349ndash11 350

Messinger J (and 13 others) 2001 Absence of Mn-centeredoxidation in the S2 to S3 transition implications for themechanism of photosynthetic water oxidation J Am ChemSoc 123 7804ndash7820

Mukerji I Andrews J C Derose V J Latimer M JYachandra V K Sauer K amp Klein M P 1994 Orien-tation of the oxygen-evolving manganese complex in a pho-tosystem-II membrane preparationmdashan X-ray-absorptionspectroscopy study Biochemistry 33 9712ndash9721

Nugent J H A Rich P amp Evans M C W 2001 Photosyn-thetic water oxidation towards a mechanism BiochimBiophys Acta 1503 138ndash146

Ono T Noguchi T Inoue Y Kusunoki M MatsushitaT amp Oyanagi H 1992 X-ray-detection of the period-4 cyc-ling of the manganese cluster in photosynthetic water oxidiz-ing enzyme Science 258 1335ndash1337

Pecoraro V L 1992 Manganese redox enzymes New YorkVCH Publishers

Peloquin J M amp Britt R D 2001 EPRENDOR characteriz-ation of the physical and electronic structure of the OECMn cluster Biochim Biophys Acta 1503 96ndash111

Peng G (and 10 others) 1994 High-resolution manganese X-ray- uorescence spectroscopymdashoxidation-state and spin-state sensitivity J Am Chem Soc 116 2914ndash2920

Penner-Hahn J E 1998 Structural characterization of the Mnsite in the photosynthetic oxygen-evolving complex StructBond (Berlin) 90 1ndash36

Renger G 2001 Photosynthetic water oxidation to molecularoxygen apparatus and mechanism Biochim Biophys Acta1503 210ndash228

RiggsGelasco P J Mei R Ghanotakis D F YocumC F amp PennerHahn J E 1996 X-ray absorption spec-troscopy of calcium-substituted derivatives of the oxygen-evolving complex of phostosystem II J Am Chem Soc 1182400ndash2410

Robblee J H Cinco R M amp Yachandra V K 2001 X-rayspectroscopy-based structure of the Mn cluster and mech-anism of photosynthetic oxygen evolution Biochim BiophysActa 1503 7ndash23

Roelofs T A Liang M C Latimer M J Cinco R MRompel A Andrews J C Sauer K Yachandra V K ampKlein M P 1996 Oxidation states of the manganese clusterduring the ash-induced S-state cycle of the photosyntheticoxygen-evolving complex Proc Natl Acad Sci USA 933335ndash3340

Sharp R R 1992 Proton NMR relaxation due to the photo-synthetic oxygen-evolving center In Manganese redoxenzymes (ed V L Pecoraro) pp 177ndash196 New YorkVCH

Siegbahn P E M 2000 Theoretical models for the oxygenradical mechanism of water oxidation and of the water oxid-izing complex of photosystem II Inorganic Chem 392923ndash2935

Siegbahn P E M amp Crabtree R H 1999 Manganese oxylradical intermediates and OndashO bond formation in photosyn-thetic oxygen evolution and a proposed role for the calciumcofactor in photosystem II J Am Chem Soc 121 117ndash127

Styring S A amp Rutherford A W 1988 The microwave-power saturation of SIIslow varies with the redox state of theoxygen-evolving complex in photosystem-II Biochemistry 274915ndash4923

Van Leeuwen P J Heimann C amp Vangorkom H J 1993Absorbency difference spectra of the S-state transitions inphotosystem-II core particles Photosynth Res 38 323ndash330

Visser H Anxolabehere-Mallart E Bergman U GlatzelP Robblee J H Cramer S P Girerd J-J Sauer KKlein M P amp Yachandra V K 2001 Mn K-edge XANES


and K b XES studies of two Mn-oxo binuclear complexesInvestigation of three different oxidation states relevant tothe oxygen-evolving complex of photosystem II J AmChem Soc 123 7031ndash7039

Vrettos J S Limburg J amp Brudvig G W 2001 Mechanismof photosynthetic water oxidation combining biophysicalstudies of photosystem II with inorganic model chemistryBiochim Biophys Acta 1503 229ndash245

Wang H Peng G Miller L M Scheuring E M GeorgeS J Chance M R amp Cramer S P 1997 Iron L-edge X-ray absorption spectroscopy of myoglobin complexes andphotolysis products J Am Chem Soc 119 4921ndash4928

Wang X de Groot F amp Cramer S P 1996 X-ray magneticcircular dichroism spectra and distortions at Fe21 L23 edgesJ Electron Spectrosc Relat Phenom 78 337ndash340

Wieghardt K 1989 The active-sites in manganese-containingmetalloproteins and inorganic model complexes AngewChem Int Ed English 28 1153ndash1172

Yachandra V K 1995 X-ray absorption spectroscopy andapplications in structural biology In Methods in enzymologyvol 246 (ed K Sauer) pp 638ndash675 Orlando FL Aca-demic

Yachandra V K amp Klein M P 1996 X-ray absorption spec-troscopy determination of transition metal site structures inphotosynthesis Adv Photosynth 3 337ndash354

Yachandra V K Derose V J Latimer M J Mukerji ISauer K amp Klein M P 1993 Where plants make oxygenmdasha structural model for the photosynthetic oxygen-evolvingmanganese cluster Science 260 675ndash679

Yachandra V K Sauer K amp Klein M P 1996 Manganesecluster in photosynthesis where plants oxidize water todioxygen Chem Rev 96 2927ndash2950

Zouni A Witt H T Kern J Fromme P Krauss NSaenger W amp Orth P 2001 Crystal structure of photosys-tem II from Synechococcus elongatus at 38 angstrom resol-ution Nature 409 739ndash743

DiscussionW Junge (Abteilung Biophysik Universitat Osnabruck

Osnabruck Germany) Is there any evidence for theinvolvement of MnV in the S-state cycle

V K Yachandra We have not detected any MnV in theS0 through to the S3 state MnV has a distinctive pre-edgefeature in the Mn XANES and if it was present it wouldbe quite easy to detect

C Dismukes (Department of Chemistry Princeton Univer-sity Princeton NJ USA) Since there is a large structuralchange in the S2 to S3 transitions (as evident in yourEXAFS data) would that not also result in a substantialchange in the XANES K-edge data that would obscurethe sharp edge seen in the other S-state changes In otherwords how can you conclude that there is no Mn oxi-dation on the S2 to S3 transition based on the absence ofa sharp edge shift

V K Yachandra There is a lengthening of the MnndashMn distance from 27 Ac in the S2 state to ca 282 and ca295 Ac in the S3 state In three Mn model compounds insimilar oxidation states investigated by Vince Pecoraroand Jim Penner-Hahn where the MnndashMn distancechanges from 27 to 28 or to 29 Ac the Mn K-edge is atthe same energy position in all three cases However ifyou are looking at complexes with completely differentstructures then it is possible that the oxidation state corre-lation with K-edge in ection point energy could be com-plicated However to address such questions precisely wecarried out Mn K b emission studies As I showed in my


talk using model compounds the Kb emission spectra arein uenced very little by structural change and are pre-dominantly dependent on the oxidation state of Mn Sothe combination of Mn K-edge XANES and K b emissionstudies of PSII in all the S-states leads us to conclude thatthere is no Mn-centred oxidation in the S2 to S3 transition

L Hammarstrom (Department of Physical ChemistryUppsala University Uppsala Sweden) Longer distances inthe S0 state are consistent with either MnII or MnIII Butyou have indicated that two MnndashMn distances becomelonger in the S2 to S3 transition If you put a radical onone oxygen then would you expect MnndashMn distancechanges in more than one of the binuclear Mn units

V K Yachandra Your question raises an importantpointmdashhow can we account for the increase in two MnndashMn distances if the radical is formed on only one di-m -oxo-bridged unit The oxidation of the bridging oxo groupprovides an explanation for an increase in distance of onedi-m -oxo-bridged MnndashMn moiety from 27 Ac in the S2

state to ca 30 Ac in the S3 state But it is dif cult to under-stand why in model A the other di-m -oxo MnndashMn moietyalso increased in distance even though it was somewhatisolated from the proposed oxyl radical Thus it would bemore logical if the structure of the OEC was in fact morelsquotied togetherrsquo than that shown in A which would moreeasily explain the lengthening of all di-m -oxo MnndashMn


motifs in the S3 state From the topological models modelG seems best suited to understand the structural changesduring the S2 to S3 transition If G is used formation ofan oxyl radical at one of the di-m -oxo bridges for examplewould give rise to the longer ca 30 Ac MnndashMn distancein the S3 state The lengthening of the other di-m -oxo MnndashMn moieties can be explained if some of the spin densityof the oxyl radical in G is present on the m 3-oxo bridge orthe other m 2-oxo bridges Similarly for structures L andM one of the di-m -oxo-bridged oxygen becomes the oxylradical and gives rise to the 30 Ac MnndashMn distance in theS3 state In a manner similar to what was proposed abovefor G the increase in the other MnndashMn distances can berationalized by some of this spin density being present onthe m 3-oxo-bridged oxygen or the other m 2-oxo-bridgedoxygens

GLOSSARY

EPR electron paramagnetic resonanceEXAFS extended X-ray absorption ne structureFT Fourier transformOEC oxygen-evolving complexPSII photosystem IIRIXS resonant inelastic X-ray scattering spectroscopyXANES X-ray absorption near-edge structureXES X-ray emission spectroscopy


over ashes and to characterize these samples by X-rayspectroscopy We reported earlier XANES data fromsamples in the S0 through S3 states produced under thesephysiologically relevant conditions (Roelofs et al 1996)The ash-advanced PSII samples prepared from spinachexhibited XANES spectra that showed Mn oxidation fromS0 to S1 and from S1 to S2 but no further oxidation duringthe S2 to S3 transition These results are consistent withour earlier studies by Guiles et al (1990ab) that were pre-pared by illumination at low temperature or by chemicaltreatment

Other spectroscopic techniques such as NMRndashPRE(Sharp 1992) EPR (Dismukes amp Siderer 1981 Hans-son amp Andreasson 1982 Ahrling et al 1997 Messinger etal 1997b) UV absorption (Dekker 1992 Van Leeuwenet al 1993) and Yox

D EPRndashspin echo (Styring amp Rutherford1988) studies show that in PSII the S0 to S1 and S1 toS2 transitions are Mn-centred oxidations and most groupsagree that Mn redox states in the S1 state are (III2 IV2)that are oxidized to (III IV3) in the S2 state There ishowever a long-standing controversy as to whether a Mn-or ligand-centred oxidation occurs in the S2 to S3 tran-sition (Ono et al 1992 Iuzzolino et al 1998)

To resolve the controversy we used high-resolution MnKb X-ray uorescence studies in collaboration with SteveCramer who has constructed a unique high-resolutionemission spectrometer that operates at the Mn K b uor-escence energy and has demonstrated the feasibility ofdetermining the oxidation states of Mn (Bergmann ampCramer 1998) We have completed a comprehensive MnKb XES study in conjunction with XANES on sets ofsamples prepared in a similar manner and characterizedby EPR (Messinger et al 2001)

Due to the fact that the actual S-state composition is ofcritical importance to the interpretation of the results wecharacterized the samples by EPR spectroscopy using theS2 multiline EPR signal as a direct measure for the amountof S2 in our samples ( gure 1) The relative S-state popu-lations in samples given none one two three four veor six ashes were determined from tting the ash-induced EPR multiline signal oscillation pattern to themodel of Kok et al (1970) for each of the samples usedin the X-ray spectroscopy experiments It is essential toobtain high concentrations of PSII in the S states to obtainEPR XANES and Kb spectroscopy data with a goodsignal-to-noise ratio An Nd-YAG laser system (Spectra-Physics PRO 230-10 800 mJ pulse 2 1 at 532 nm 9 nspulse width) was used to illuminate PSII samples fromboth sides simultaneously Due to the fact that the pulsewidths are narrow compared with those of a ash lampdouble hits become negligible The damping of the oscil-lation pattern is minimized thus making it possible toderive an unique set of spectra by deconvolution Whenthe system becomes scrambled by frequent misses anddouble hits the deconvolved spectra are not unique

(a) Manganese K-edge X-ray absorptionnear-edge structure

Our edge data are of suf cient signal-to-noise quality toallow for not only a determination of the absorption edgeposition (using the in ection point energy de ned as thezero-crossing in the second derivative of the edge) but


20

40

60

80

100

0 1 2 3 4 5 6

flash no

norm

aliz

ed S

2 m

ultil

ine

EP

R s

igna

l (

)

Figure 1 The S2 multiline EPR signal amplitudes are shownas a function of ash number (solid line experimentalaverage) The best t to the S2 multiline oscillation patternis shown as a dashed line (simulation) The S-statecompositions of the samples are derived from the tIndividual S-state spectra can be uniquely determined fromthese data if the oscillation pattern is as pronounced asshown in the gure It is important that a deep oscillationpattern is obtained because all subsequent steps that areused in the deconvolution to generate pure S-state spectraare critically dependent on these data

also an analysis of the structure on the edge as a functionof the number of ashes

The edge spectra of samples given none one two orthree ashes are combined with EPR information to calcu-late the pure S-state edge spectra The shifts in in ectionpoint energy positions and changes in shape are determ-ined by second derivatives of the K-edge spectra and areshown in gure 2 In addition to the shift in edge positionthe S0 to S1 (21 eV) and S1 to S2 (11 eV) transitions areaccompanied by characteristic changes in the shape of theedge also indicative of Mn oxidation The edge positionshifts very little (03 eV) for the S2 to S3 transition andthe edge shape shows only minor changes (Messinger etal 2001)

(b) Kb X-ray emission spectroscopyHigh-resolution Mn XES was performed on S0 S1 S2

and S3 samples prepared using high-power laser ashes asdescribed above In this technique the energy of the MnK b emission (3p to 1s) is measured with a high-resolutiondispersive spectrometer The shape and energy of theK b 1 3 emission re ect the oxidation state(s) of the emittingMn atom(s) The emission occurs from a 3p level that ismainly in uenced by the number of unpaired 3d elec-trons and it is less sensitive to the symmetry and bondingthan the K-edge absorption which involves transitions tothe 4p level (Peng et al 1994) The spin of the unpaired3d valence electrons can be either parallel (Kb 9 ) or anti-parallel (K b 1 3) to the hole in the 3p level The splittingbetween the Kb 9 and K b 1 3 peaks becomes smaller forhigher oxidation states because fewer 3d electrons interactwith the 3p hole Thus in contrast to the in ection points


6535 6545 6555 6565 6575

d2 (FI

0)dE

2

6484 6488 6492 6496

X-ray energy (eV)

S1ndashS0

S2ndashS1

S3ndashS2

S0ndashS3

(a) (b)













015

017

019

021

023

flash no

ltE

gtndash 6490 eV

0 1 2 3

25

30

35

40

0 1 2 3 4 5 6

IPE

ndash 6

550

eV

(a) (b)


bind

ing

ener

gy

K edge

1s22p6hellip3d4

1s12p6hellip3d44p1

1s12p6hellip3d4p 1

1s12p6hellip3d5

J =32

J =12

L edge

1s22p6hellip3d4

1s22p5hellip 3d5

1s22p5hellip3d5

RIXS

EB = ndash f

f2p 1s

1s22p6hellip3d4

1s12p6hellip3d44p1

1s12p6hellip3d4p 1

1s12p6hellip3d5


1s22p5hellip3d4p 1

1s22p5hellip3d5

u

u u

u
















35 55 75 95 115

_50

50

00


(k

) times

k3

I II

III

FT a

mpl

itud

e

00

02


00 40302010

c

(a)

(b)










FT

mag

nitu

de

0 2 6 8 104

01

02

I II

III









0 321

2

1


num

ber

of 2

85

Aring M

nndashM

n ve

ctor

s






Mn

O

OMn

O

OMn

O

Mn

O

O

Mn

O

Mn

MnO

O Mn

Mn

O

OMn

O

Mn

Mn O Mn

O

OMn

Mn

O

OMn

OCa

O

Mn

O O

Mn

Mn

O

OMn

O

MnO

O Mn

Mn O

O Mn

Mn O

O Mn

O

Mn

O

O

Mn

O

Mn O

OMn

O

Mn

Mn

O

Mn

O

O

O

OMn

Mn

O

OMnMn

O

OMn

Mn

O

Mn

O

Mn Mn

O

Mn

OOO

O

Mn

Mn

O

OMn

O

Mn

OMn

O

O

Mn

O

MnMnMn

O

OO

(a) (b) (c) (d )

(e) ( f ) (g)

(h) (i) ( j )

(k) (l) (m )










0 2 4 6 8 10

FT

mag

nitu

de


I

II

01

02

03

= 10 Ecirc

= 80 Ecirc

X-raye-vector

Mn

0

30

6090

120

210

240270

300

330


005101520253035

Nap

p

(a) (b)

q

q

q










0

04

08

12

2 4 6 8 10

FT

mag

nitu

de


I II






Ca

O

O

Mn

MnMn

MnO

O

O

Ca

OO

Mn

Mn Mn

O O

O

Mn

Ca

O

Mn

O

O Mn Mn

Mn O O

membranenormal

(a) (b) (c)









REFERENCES
















































































GLOSSARY



6535 6545 6555 6565 6575

d2 (FI

0)dE

2

6484 6488 6492 6496

X-ray energy (eV)

S1ndashS0

S2ndashS1

S3ndashS2

S0ndashS3

(a) (b)













015

017

019

021

023

flash no

ltE

gtndash 6490 eV

0 1 2 3

25

30

35

40

0 1 2 3 4 5 6

IPE

ndash 6

550

eV

(a) (b)


bind

ing

ener

gy

K edge

1s22p6hellip3d4

1s12p6hellip3d44p1

1s12p6hellip3d4p 1

1s12p6hellip3d5

J =32

J =12

L edge

1s22p6hellip3d4

1s22p5hellip 3d5

1s22p5hellip3d5

RIXS

EB = ndash f

f2p 1s

1s22p6hellip3d4

1s12p6hellip3d44p1

1s12p6hellip3d4p 1

1s12p6hellip3d5


1s22p5hellip3d4p 1

1s22p5hellip3d5

u

u u

u
















35 55 75 95 115

_50

50

00


(k

) times

k3

I II

III

FT a

mpl

itud

e

00

02


00 40302010

c

(a)

(b)










FT

mag

nitu

de

0 2 6 8 104

01

02

I II

III









0 321

2

1


num

ber

of 2

85

Aring M

nndashM

n ve

ctor

s






Mn

O

OMn

O

OMn

O

Mn

O

O

Mn

O

Mn

MnO

O Mn

Mn

O

OMn

O

Mn

Mn O Mn

O

OMn

Mn

O

OMn

OCa

O

Mn

O O

Mn

Mn

O

OMn

O

MnO

O Mn

Mn O

O Mn

Mn O

O Mn

O

Mn

O

O

Mn

O

Mn O

OMn

O

Mn

Mn

O

Mn

O

O

O

OMn

Mn

O

OMnMn

O

OMn

Mn

O

Mn

O

Mn Mn

O

Mn

OOO

O

Mn

Mn

O

OMn

O

Mn

OMn

O

O

Mn

O

MnMnMn

O

OO

(a) (b) (c) (d )

(e) ( f ) (g)

(h) (i) ( j )

(k) (l) (m )










0 2 4 6 8 10

FT

mag

nitu

de


I

II

01

02

03

= 10 Ecirc

= 80 Ecirc

X-raye-vector

Mn

0

30

6090

120

210

240270

300

330


005101520253035

Nap

p

(a) (b)

q

q

q










0

04

08

12

2 4 6 8 10

FT

mag

nitu

de


I II






Ca

O

O

Mn

MnMn

MnO

O

O

Ca

OO

Mn

Mn Mn

O O

O

Mn

Ca

O

Mn

O

O Mn Mn

Mn O O

membranenormal

(a) (b) (c)









REFERENCES
















































































GLOSSARY



015

017

019

021

023

flash no

ltE

gtndash 6490 eV

0 1 2 3

25

30

35

40

0 1 2 3 4 5 6

IPE

ndash 6

550

eV

(a) (b)


bind

ing

ener

gy

K edge

1s22p6hellip3d4

1s12p6hellip3d44p1

1s12p6hellip3d4p 1

1s12p6hellip3d5

J =32

J =12

L edge

1s22p6hellip3d4

1s22p5hellip 3d5

1s22p5hellip3d5

RIXS

EB = ndash f

f2p 1s

1s22p6hellip3d4

1s12p6hellip3d44p1

1s12p6hellip3d4p 1

1s12p6hellip3d5


1s22p5hellip3d4p 1

1s22p5hellip3d5

u

u u

u
















35 55 75 95 115

_50

50

00


(k

) times

k3

I II

III

FT a

mpl

itud

e

00

02


00 40302010

c

(a)

(b)










FT

mag

nitu

de

0 2 6 8 104

01

02

I II

III









0 321

2

1


num

ber

of 2

85

Aring M

nndashM

n ve

ctor

s






Mn

O

OMn

O

OMn

O

Mn

O

O

Mn

O

Mn

MnO

O Mn

Mn

O

OMn

O

Mn

Mn O Mn

O

OMn

Mn

O

OMn

OCa

O

Mn

O O

Mn

Mn

O

OMn

O

MnO

O Mn

Mn O

O Mn

Mn O

O Mn

O

Mn

O

O

Mn

O

Mn O

OMn

O

Mn

Mn

O

Mn

O

O

O

OMn

Mn

O

OMnMn

O

OMn

Mn

O

Mn

O

Mn Mn

O

Mn

OOO

O

Mn

Mn

O

OMn

O

Mn

OMn

O

O

Mn

O

MnMnMn

O

OO

(a) (b) (c) (d )

(e) ( f ) (g)

(h) (i) ( j )

(k) (l) (m )










0 2 4 6 8 10

FT

mag

nitu

de


I

II

01

02

03

= 10 Ecirc

= 80 Ecirc

X-raye-vector

Mn

0

30

6090

120

210

240270

300

330


005101520253035

Nap

p

(a) (b)

q

q

q










0

04

08

12

2 4 6 8 10

FT

mag

nitu

de


I II






Ca

O

O

Mn

MnMn

MnO

O

O

Ca

OO

Mn

Mn Mn

O O

O

Mn

Ca

O

Mn

O

O Mn Mn

Mn O O

membranenormal

(a) (b) (c)









REFERENCES
















































































GLOSSARY












35 55 75 95 115

_50

50

00


(k

) times

k3

I II

III

FT a

mpl

itud

e

00

02


00 40302010

c

(a)

(b)










FT

mag

nitu

de

0 2 6 8 104

01

02

I II

III









0 321

2

1


num

ber

of 2

85

Aring M

nndashM

n ve

ctor

s






Mn

O

OMn

O

OMn

O

Mn

O

O

Mn

O

Mn

MnO

O Mn

Mn

O

OMn

O

Mn

Mn O Mn

O

OMn

Mn

O

OMn

OCa

O

Mn

O O

Mn

Mn

O

OMn

O

MnO

O Mn

Mn O

O Mn

Mn O

O Mn

O

Mn

O

O

Mn

O

Mn O

OMn

O

Mn

Mn

O

Mn

O

O

O

OMn

Mn

O

OMnMn

O

OMn

Mn

O

Mn

O

Mn Mn

O

Mn

OOO

O

Mn

Mn

O

OMn

O

Mn

OMn

O

O

Mn

O

MnMnMn

O

OO

(a) (b) (c) (d )

(e) ( f ) (g)

(h) (i) ( j )

(k) (l) (m )










0 2 4 6 8 10

FT

mag

nitu

de


I

II

01

02

03

= 10 Ecirc

= 80 Ecirc

X-raye-vector

Mn

0

30

6090

120

210

240270

300

330


005101520253035

Nap

p

(a) (b)

q

q

q










0

04

08

12

2 4 6 8 10

FT

mag

nitu

de


I II






Ca

O

O

Mn

MnMn

MnO

O

O

Ca

OO

Mn

Mn Mn

O O

O

Mn

Ca

O

Mn

O

O Mn Mn

Mn O O

membranenormal

(a) (b) (c)









REFERENCES
















































































GLOSSARY




FT

mag

nitu

de

0 2 6 8 104

01

02

I II

III









0 321

2

1


num

ber

of 2

85

Aring M

nndashM

n ve

ctor

s






Mn

O

OMn

O

OMn

O

Mn

O

O

Mn

O

Mn

MnO

O Mn

Mn

O

OMn

O

Mn

Mn O Mn

O

OMn

Mn

O

OMn

OCa

O

Mn

O O

Mn

Mn

O

OMn

O

MnO

O Mn

Mn O

O Mn

Mn O

O Mn

O

Mn

O

O

Mn

O

Mn O

OMn

O

Mn

Mn

O

Mn

O

O

O

OMn

Mn

O

OMnMn

O

OMn

Mn

O

Mn

O

Mn Mn

O

Mn

OOO

O

Mn

Mn

O

OMn

O

Mn

OMn

O

O

Mn

O

MnMnMn

O

OO

(a) (b) (c) (d )

(e) ( f ) (g)

(h) (i) ( j )

(k) (l) (m )










0 2 4 6 8 10

FT

mag

nitu

de


I

II

01

02

03

= 10 Ecirc

= 80 Ecirc

X-raye-vector

Mn

0

30

6090

120

210

240270

300

330


005101520253035

Nap

p

(a) (b)

q

q

q










0

04

08

12

2 4 6 8 10

FT

mag

nitu

de


I II






Ca

O

O

Mn

MnMn

MnO

O

O

Ca

OO

Mn

Mn Mn

O O

O

Mn

Ca

O

Mn

O

O Mn Mn

Mn O O

membranenormal

(a) (b) (c)









REFERENCES
















































































GLOSSARY



Mn

O

OMn

O

OMn

O

Mn

O

O

Mn

O

Mn

MnO

O Mn

Mn

O

OMn

O

Mn

Mn O Mn

O

OMn

Mn

O

OMn

OCa

O

Mn

O O

Mn

Mn

O

OMn

O

MnO

O Mn

Mn O

O Mn

Mn O

O Mn

O

Mn

O

O

Mn

O

Mn O

OMn

O

Mn

Mn

O

Mn

O

O

O

OMn

Mn

O

OMnMn

O

OMn

Mn

O

Mn

O

Mn Mn

O

Mn

OOO

O

Mn

Mn

O

OMn

O

Mn

OMn

O

O

Mn

O

MnMnMn

O

OO

(a) (b) (c) (d )

(e) ( f ) (g)

(h) (i) ( j )

(k) (l) (m )










0 2 4 6 8 10

FT

mag

nitu

de


I

II

01

02

03

= 10 Ecirc

= 80 Ecirc

X-raye-vector

Mn

0

30

6090

120

210

240270

300

330


005101520253035

Nap

p

(a) (b)

q

q

q










0

04

08

12

2 4 6 8 10

FT

mag

nitu

de


I II






Ca

O

O

Mn

MnMn

MnO

O

O

Ca

OO

Mn

Mn Mn

O O

O

Mn

Ca

O

Mn

O

O Mn Mn

Mn O O

membranenormal

(a) (b) (c)









REFERENCES
















































































GLOSSARY



0 2 4 6 8 10

FT

mag

nitu

de


I

II

01

02

03

= 10 Ecirc

= 80 Ecirc

X-raye-vector

Mn

0

30

6090

120

210

240270

300

330


005101520253035

Nap

p

(a) (b)

q

q

q










0

04

08

12

2 4 6 8 10

FT

mag

nitu

de


I II






Ca

O

O

Mn

MnMn

MnO

O

O

Ca

OO

Mn

Mn Mn

O O

O

Mn

Ca

O

Mn

O

O Mn Mn

Mn O O

membranenormal

(a) (b) (c)









REFERENCES
















































































GLOSSARY



Ca

O

O

Mn

MnMn

MnO

O

O

Ca

OO

Mn

Mn Mn

O O

O

Mn

Ca

O

Mn

O

O Mn Mn

Mn O O

membranenormal

(a) (b) (c)









REFERENCES
















































































GLOSSARY















































































GLOSSARY


structure of the manganese complex in photosystem ii ......structure of the mn complex v. k....

Documents