Cofactors on the Donor Side of Photosystem II Investigated with EPR Techniques

vorgelegt von

Diplom-Chemiker Michael Kammel aus Laage

von der Fakult¨at II -Mathematik und Naturwissenschaften- der Technischen Universit¨at Berlin zur Erlangung des akademischen Grades

Doktor der Naturwissenschaften - Dr. rer. nat. - genehmigte Dissertation

Promotionsausschuss:

Vorsitzender: Prof. Dr. Martin Lerch Berichter: Prof. Dr. Wolfgang Lubitz Berichter: Prof. Dr. Robert Bittl

Tag der mundlic¨ hen Prufung:¨ 23.06.2003

Berlin 2003 D 83

Zusammenfassung

Michael Kammel: Cofactors on the Donor Side of Photosystem II investigated with EPR techniques Dem Photosystem II, mit dessen Hilfe es höheren Pflanzen und Cyanobakterien gelingt, Wasser zu oxidieren, wird seit Jahrzehnten die Aufmerksamkeit der Forscher vieler Fachgebiete zuteil. Die Kofaktoren im Photosystem II durchlaufen im Rahmen der lichtinduzierten Ladungstren- nung paramagnetische Zustände. Diese können mit den Techniken der Elektronen-paramagnetischen Resonanz erfasst werden. Dies erlaubt Aussagen ub¨ er Kinetik und Mechanismen der Ladungstrennungsreaktionen, sowie ub¨ er Ort und Orientierung der daran beteiligten Kofak- toren. Der erste Teil der Arbeit befasst sich mit dem Beginn der lichtinduzierten Ladungstrennung. Durch eine Blockade des primären Elektronentransfers war es möglich, eine Rekombination am Ort der ursprunglic¨ hen Ladungstrennung zu induzieren. Es entsteht ein Triplettzustand, welcher sich mit zeitaufgelöster EPR detektieren lässt. Durch die Verwendung von Kristallen des PS II und der orientierungsabhängigen Messung des EPR-Signals konnte der Ort der Rekombination ermittelt werden und Aussagen ub¨ er die Ladungstrennung gemacht werden. Der zweite Teil ist der redoxaktiven Aminosäure YD des die Kofaktoren umgebenden Pro- teingerustes¨ gewidmet. Sie liegt im Photosystem als metastabiles Radikal vor. Die magnetischen Hyperfeinwechselwirkungen innerhalb dieses Molekuls¨ wurden mit HYSCORE und 1H-ENDOR untersucht. Durch die Kombination dieser beiden Techniken gelang eine vollständige Bestim- mung und Zuordnung der Hyperfeinkopplungsparameter der Wasserstoffatome des Molekuls.¨ Der dritte Teil untersucht die Umgebung des wasserspaltenden Kofaktors, dessen Herzstuc¨ k ein vierkerniger Mangankomplex ist. An zwei paramagnetischen Oxidationszuständen, dem S2- 1 und dem S 2-Zustand, wurden H-ENDOR-Messungen durchgefuhrt.¨ Dadurch konnten Hy- − perfeinkopplungen von Protonen in der Nähe des Mangankomplexes ermittelt werden. Durch die Anwendung eines erweiterten Punkt-Dipol-Kopplungsmodelles konnten maximale Proton- Mangankomplex-Abstände, sowie die möglichen Aufenthaltssphären der betreffenden Protonen ermittelt werden. Fur¨ die Koordination an den Mangankomplex erwiesen sich protonierte µ- oxo-Bruc¨ ken als wahrscheinlich. Im folgenden Teil war der Mangankomplex selbst anhand von 55Mn-ENDOR-Messungen Gegen- stand der Untersuchung. Im Kontrast zu den Ergebnissen einer anderen Forschungsgruppe wurden hier Signale gefunden, die auf eine andere Zusammensetzung des Mangankomplexes als dort angegeben schließen lassen. Im letzten Teil wurde das primäre Amin Methylamin als Inhibitor der enzymatischen Aktivität des PS II verwendet. Durch begleitende EPR-Messungen des S2-Zustandes am nativen wie auch am inhibierten System konnte festgestellt werden, dass Methylamin wahrscheinlich nicht mit dem Kofaktor Chlorid um eine Bindungsstelle am Enzym konkurriert, sondern möglicherweise mit Wasser.

i ii Publikationen und Konferenzbeitr¨age

Kammel, M., Hofbauer, W., Zouni, A., Fromme, P., Bittl, R., Lendzian, F., Witt, H.T., & Lubitz, W., High ﬁeld EPR studies of the tyrosyl radical YD• in Photosystem II crystals of Syne- chococcus elongatus, In XIth International Congress on Photosynthesis, Budapest, Int. Soc. of Photosynth. Res., 1998.

Kammel, M., Bittl, R., & Lubitz, W., EPR investigations of the g=2 multiline signal of Pho- tosystem II at diﬀerent frequencies of diﬀerent preparations of the S2 state and the S0 MLS, Summer School on: “Advanced methods in electron paramagnetic resonance spectroscopy. Ap- plication to chemistry”, physics and Biology, 12.-19. September 1999, Caorle.

Hofbauer, W., Zouni, A., Bittl, R., Kammel, M., Fromme, P., Lendzian, F., Witt, H.T., Lubitz, W., Krauss, N., & Orth, P., EPR characterization of the tyrosyl radical YD• in active Pho- tosystem II single crystals at 94 GHz, In High Frequency Electron Paramagnetic Resonance, Amsterdam, Royal Netherlands Academy of Arts and Sciences, 2000.

Kammel, M., Lubitz, W., & Messinger, J., Eﬀects of the water analogue methylamine on the S-State EPR multiline signals of the manganese cluster in plant Photosystem II, Joint Meeting of French and German Biophysicists, Mai 2001, Hunfeld.¨

Kammel, M., Kern, J., Lubitz, W., & Bittl, R., Photosystem II single crystals studied by TR EPR at 9.8 GHz: The light-induced triplet state at 10 K, XVIIth readings on Photosynthesis, Juli 2002, Puschchino.

Kammel, M., Kern, J., Lubitz, W., & Bittl, R., Photosystem II single crystals studied by TR EPR at 9.8 GHz: the light-induced triplet at 10 K, Biochim. Biophys. Acta, in press. van Gastel, M., Kammel, M., & Lubitz, W., HYSCORE and DONUT-HYSCORE of the tyrosyl radical YD• in spinach Photosytem II, J. Magn. Res., manuscript in preparation.

iii iv Contents

1 Photosynthesis 5

1.1 Overview ...... 5

1.2 Photosystem II ...... 6

1.2.1 Charge separation, electron pathways and kinetics in PS II ...... 7

1.3 Cofactors of the PS II core complex ...... 9

1.3.1 The primary donor ...... 10

1.3.2 The Oxygen-Evolving Complex (OEC) ...... 11

1.3.3 The tyrosines YZ and YD• ...... 13

2 Principles of EPR 15

2.1 The electron spin and its interaction with a magnetic ﬁeld ...... 15

2.2 The Spin Hamiltonian ...... 16

2.3 The Electron-Zeeman-Interaction and the g Tensor ...... 16

2.4 The Nuclear-Zeeman-Interaction ...... 17

2.5 The Hyperﬁne-Interaction ...... 17

2.6 The Electronic Dipolar Interaction ...... 18

2.6.1 Two Electrons Coupling in an External Magnetic Field ...... 19

2.6.2 Electron Exchange Interaction ...... 20

2.7 Electron Paramagnetic Resonance ...... 22

v 2.7.1 Continuous wave EPR ...... 22

2.7.2 Transient EPR ...... 23

2.7.3 Pulsed EPR ...... 23

2.8 Other techniques of EPR ...... 24

2.8.1 Continuous Wave ENDOR ...... 24

2.8.2 Pulsed ENDOR ...... 25

2.8.3 Electron Spin Echo Envelope Modulation (ESEEM) ...... 27

3 Sample preparation and characterization 33

3.1 PS II membrane fragments ...... 33

3.1.1 Sample treatments ...... 33

4 The Light-induced Triplet State in Photosystem II 35

4.1 Motivation ...... 35

4.2 Experiment ...... 37

4.2.1 Samples ...... 37

4.3 The PS II triplet state powder spectrum ...... 39

4.4 The triplet state spectra in PS II single crystals ...... 39

4.5 Simulation of the Single Crystal Spectra ...... 41

4.6 Discussion ...... 44

1 5 H-ENDOR and HYSCORE of YD• 51

5.1 Motivation ...... 51

5.2 Experiment ...... 53

5.2.1 Samples ...... 53

5.2.2 cw EPR ...... 53

1 5.2.3 Pulsed H-ENDOR of YD• ...... 53

vi 5.2.4 HYSCORE Spectra of YD• ...... 53

5.3 Analysis ...... 55

5.3.1 Pulsed ENDOR ...... 55

5.3.2 HYSCORE ...... 58

5.3.3 Assignment of the observed couplings to the protons of the tyrosine radical 59

5.3.4 The program used for the simulation of the HYSCORE spectra ...... 60

5.3.5 Simulation of the theoretical HYSCORE spectrum of YD• ...... 61

5.4 Discussion ...... 62

5.5 Conclusion ...... 65

1 6 H-ENDOR of the S 2 and the S2 state 67 − 6.1 Motivation ...... 67

6.2 Experiment ...... 70

6.2.1 Samples for the S2 state ...... 70

6.2.2 Samples for the S 2 state ...... 70 − 6.2.3 1H-Davies-ENDOR ...... 71

6.3 Results ...... 71

6.3.1 Field swept ESE of the S2 state ...... 71

6.3.2 Field swept ESE of the S 2 state ...... 72 − 1 6.3.3 H-ENDOR on protonated samples in the S2 state ...... 73

1 6.3.4 H-ENDOR on protonated and deuterated samples in the S 2 state . . . 75 −

6.3.5 Analysis of the ENDOR signals for the S 2 state ...... 78 − 6.4 Discussion ...... 81

6.4.1 The S2 state ...... 81

6.4.2 The S 2 state ...... 83 −

55 7 Mn-ENDOR of the S2-State 87

vii 7.1 Motivation ...... 87

7.2 55Mn-ENDOR spectra of exchange coupled complexes ...... 88

7.3 Experiment ...... 90

7.3.1 Samples and sample treatment ...... 90

7.3.2 Pulsed 55Mn-ENDOR ...... 90

7.4 Results ...... 91

7.5 Discussion ...... 91

8 Methylamine as inhibitor to oxygen evolution 95

8.1 Motivation ...... 95

8.2 Interaction of water analogues with the OEC ...... 96

8.3 Experiment ...... 96

8.3.1 Samples ...... 96

8.3.2 cw-EPR at X-band ...... 97

8.4 Enzyme kinetics and the eﬀect of inhibitors ...... 97

8.4.1 Oxygen measurements in the presence of methylamine ...... 98

8.4.2 cw-EPR ...... 104

8.5 Discussion ...... 104

9 Summary and Outlook 107

10 Appendix 113

10.1 The Program Used for the Simulation of the Triplet Spectra ...... 113

10.2 Calculation of equipotential lines for coupling nuclei ...... 120

viii 1 Introduction

Electron Paramagnetic Resonance (EPR), is the name given to the process of resonant absorption of microwave radiation by paramagnetic ions or molecules, with at least one unpaired electron spin, and in the presence of a static magnetic field. It has a wide range of applications in chemistry, physics, biology, and medicine: it may be used to probe the “static” structure of solid and liquid systems, and is also very useful in investigating dynamic processes. In the last decades the importance of EPR always stood behind the one of NMR. With the rise of the organic chemistry in the second half of the last century there was the need to explore the structure of thousands of newly discovered organic compounds every day. The development of NMR and its applications and techniques was promoted tremendously. Most of these compounds are diamagnetic and therefore not accessible to EPR. At the end of the last century biophysics started to gain more importance. Microbiological processes in cells often including electron transfer processes, intermediate paramagnetic species or enzyme processes in which paramagnetic cofactors play a role, became of great interest. This work will deal with the application of made-to-measure EPR-techniques on cofactors of one particular enzyme system, the Photosystem II (PS II). PS II is one of the enzyme complexes embedded in the thylakoid membrane of chloroplasts and participates in the photosynthetic process. For mankind PS II has two basic meanings, one pointing far back into the history of life on this planet, the other probably influencing the future. Before the advent of PS II several billions of years ago there was already some kind of photosynthesis on earth carried out by simple organisms. These life forms were confronted with a problem that concerns the electron transfer. They needed electrons to start it. At this time these electrons were probably supplied by chemical substances such as H2S, amines and other compounds which could easily be oxidized due to their low redox potential. Life on earth these days must have been slow since the supply with those substances was low. Although there were probably greater amounts of them than today they were all but abundant. There is a lot of water on earth. But it has a very high redox potential. No enzyme at that time was able to extract its electrons. Probably due to a mutation the PS II or a predecessor of it evolved. PS II uses an smart but complicated strategy to extract electrons from the water molecules. This is done stepwise in a cycle driven by the successive absorption of four photons. This meant a revolution to life these days. Not only the 2 supply with electron donors was suddenly plentiful, there was also a side product of water oxidation, which is oxygen. No being on earth was prepared to cope with the sudden existence of this highly aggressive and oxidating molecule. Most of them died out within a short period compared to earth’s evolution. New species had to evolve, and evolution went so far that these species not only got used to oxygen, they indeed need it now. The external energetic situation of the highest developed of these species has always been poor. Exploiting the environment to find new carriers of energy has culminated in the last century. This again led to an apparent improvement of the living situation of this species. Meanwhile it is running out of resources that have been taken hundreds of millions of years to gather. New strategies have to be developed. The only reliable source of energy on earth is sunlight (and perhaps some energy from the interior of the earth). In the last decades much effort has been made to use the energy of the sun. So far production costs of devices using it still exceed their benefits. They are expensive, ineffective and do not “live” long enough. PS II already brings all benefits with it. Nature itself produces it, it is tremendously effective and is repaired/renewed by the cell itself. How PS II eventually can be used to produce more energy than the organism needs to maintain its life functions, is not clear. One neat idea would be to couple enzyme systems that usually have nothing to do with each other. For the case of PS II this could be hydrogenases. The knowledge concerning the structure of PS II and the mechanism of water oxidation are still not sufficient. To bring some more light (sic!) into this problem the work that is documented in the following has been done. It used EPR and its techniques to investigate the structure of cofactors and their surrounding in PS II and the mechanism and the kinetics of the electron transfer process. The first chapter in the experimental section will deal with the start of the electron transfer process, particularly the entity where the light absorbed by surrounding light harvesting pigments is funneled into and the electron is emitted. The location of the start of the charge transfer in PS II is still not known with certainty. If the electron transfer gets blocked beyond a certain stage, the generation of the charge transfer recombination product, which is a triplet state, can be followed by Time-Resolved EPR. These measurements have been done on crystals of PS II. From their evaluation the location of the triplet state can be derived and therefore also information about the start of the charge transfer. The second chapter in this section will focus on a redox-active amino acid within this protein complex whose function is not fully understood yet. The molecule in its paramagnetic form was the first species in PS II investigated by EPR. Although it is one of the best characterized species in the PS II complex its magnetic properties have not completely been resolved yet. The combination of HYperfine Sublevel COrrelation spectroscopy 3

(HYSCORE) and pulsed Electron Nuclear DOuble Resonance (ENDOR) is utilized to clarify the complicated hyperfine coupling structure of the radical. The following chapter takes advantage of 1H-ENDOR, a double resonance technique, to find out more about the unique oxygen evolving complex (OEC), the part of PS II which is solely responsible for the oxidation of water. The OEC has been investigated poised in two different oxidation states, one the well known S2 state and the other the S 2 state, − from which recently an EPR signal has been discovered. With the application of 1H- ENDOR this chapter will focus on the exploration of hydrogen atoms close to the OEC, which could be part of water molecules. With 55Mn-ENDOR another species, the manganese cluster itself will be the matter of observation. To understand the mechanism of water oxidation the structure of the OEC and the oxidation states of its manganese ions are of paramount interest. The complementary use of oxygen-evolution measurements and cw-EPR is the subject of the last chapter in the experimental section. The influence of the PS II enzyme activity inhibiting molecule methylamine will be investigated. It belongs to the group of primary amines which probably compete with essential cofactors of PS II for their binding pockets. This is accompanied by a change of EPR signals of paramagnetic species that belong to PS II. There are two chapters in the beginning of this work conveying some knowledge about structure and function of PS II and EPR and its techniques.

Chapter 1

Photosynthesis

1.1 Overview

Photosynthesis is the process in cyano bacteria and higher plants whereby energy from light is harvested to provide carbohydrates for energy production [1]. It is the major path through which carbon as CO2 reenters the biosphere [2]. Photosynthesis is also the major source of oxygen in the earth’s atmosphere. Photosynthetic organisms probably ﬁrst appeared about 3.5 billion years ago. The Earth’s atmosphere before that was probably devoid of oxygen though rich in carbon dioxide [3]. The photosynthesis reaction can be formally written as:

2n H2O + n CO2 + x hν −→ [CH2O]n + n H2O + n O2. (1.1)

The expression, [CH2O], represents a generalized carbohydrate. In this redox reaction molecular oxygen O2 is formed by oxidation of the water oxygen, whereas the carbon in the carbon oxide is reduced. The source of the oxygen released as a result of photosynthesis in plants, algae, and cyanobacteria is H2O. Light energy has no direct eﬀect on this reaction and H2O has no known way to directly reduce CO2. Thus, photosynthesis is a much more intricate process than this equation suggests. There are two main processes occuring during photosynthesis in chloroplasts:

1. Photochemical oxidation of H2O - Energy from light causes electrons from H2O to be + transferred (with protons) to NADP , forming NADPH, and releasing O2. 2. Part of the energy from the sunlight is captured in a process called photophosphory- lation in which ADP is phosphorylated to form ATP. The remaining reactions of photosynthesis are not dependent upon light and are part of what are called the dark reactions of photosynthesis. They do not, however, occur only in

5 6 CHAPTER 1. PHOTOSYNTHESIS

Figure 1.1: The protein complexes occuring within the thylakoid membrane of chloroplasts including the main light induced reactions. Taken from [4]

the dark. They are, in fact, stimulated by light. In the dark reactions of photosynthesis, NADPH and ATP produced by the light reactions are used in the reductive synthesis of carbohydrate from CO2 and water. Figure 1.1 shows a picture of a part of the thylakoid membrane in chloroplasts with the four important protein complexes involved in photosynthesis depicted. The most important reactions and charge transfer pathways are also mentioned. The two photosystems are responsible for the light-induced generation of a proton gradient between the stroma and the lumen. The formation of ATP by ATP-synthase is driven by this gradient. The protein Cyt b6f acts as an electron carrier to supply the primary donor of Photosystem I (PS I) with electrons. The primary donor of Photosystem II gets supplied with electrons by the oxidation of water, a process for which a four-membered manganese cluster is utilized.

1.2 Photosystem II

The protein on which this work will focus is Photosystem II (PS II). PS II is a multisubunit complex embedded in the thylakoid membranes of higher plants, algae and cyanobacteria. It uses light energy to catalyze a series of electron transfer reactions resulting in the splitting of water into molecular oxygen, protons and electrons [5, 6]. These reactions take place on a large scale, being responsible for the production of at- mospheric oxygen and indirectly for almost all the biomass on the planet. Despite its importance, the catalytic properties of PS II have never been reproduced in any artiﬁcial 1.2. PHOTOSYSTEM II 7 system. Understanding its unique chemistry is not only important in its own right, but could have implications for the agricultural industry since PS II is a main site of damage during environmental stress. Structurally PS II is comprised of more than 25 polypeptides [7] and is surrounded by a variety of chlorophyll a (Chl a) and chlorophyll b (Chl b) binding proteins, known as light harvesting complexes, which funnel light energy into the complex. At the enzymatic heart of the complex are two polypeptides known as the D1 and D2 proteins. The primary donor called P680, a chlorophyll a dimer with an optical absorption maximum at 680 nm, is possibly the start of the charge separation process. It is settled between these proteins.

The pheophytin and QA are found on the D2 protein, and the redox-active tyrosine, TyrZ, and the QB plastoquinone are found on the D1 protein. These are further members of the electron transfer chain. In the D2 protein, the QA plastoquinone remains relatively fixed compared to the QB plastoquinone, which is loosely bound, and its reduced and protonated form Q H is free to move in and out of the D1 protein at the Q site“. Inhibitors B 2 ” B of PS II electron transport generally bind at the Q site“, preventing the binding and ” B reduction of the QB plastoquinone. It is possible to extract PS II out of the thylakoid membrane for investigation while re- taining its functionality. For many experiments preparations with PS II still embedded in the lipid membrane are sufficient. Researchers developed preparation methods with high reproducibility to yield so called “membrane fragments” [9]. By further stripping off surrounding proteins one arrives at the PS II reaction centre, the smallest complex of proteins and cofactors still able to evolve oxygen. This complex has been crystallized from the cyanobacterium Synechococcus elongatus and X-ray structure analysis of the resulting crystals has contributed a substantial part of the knowlegde now available on PS II [8]. The resulting protein structure together with the arrangement of cofactors at a resolution of 3.8 A˚ is shown in figure 1.2.

1.2.1 Charge separation, electron pathways and kinetics in PS II

PS II is a protein-cofactor-complex that combines three redox cycles; a four-membered cycle that happens in the oxygen-evolving complex (OEC), which results in the oxidation of water, a two membered leading to the reduction of a plastoquinon (QB) to a plasto- quinol and a single electron redox reaction of a chlorophyll, the primary donor P680. In 1969 Joliot et al. [10] performed experiments measuring the release of oxygen during a series of short (≈ 10µs) ﬂashes applied to previously dark-adapted suspensions of spinach chloroplasts or Chlorella fragments. Maxima of oxygen evolution was detected with a periodicity of four with respect to the number of light ﬂashes. The conclusion by Kok et 8 CHAPTER 1. PHOTOSYNTHESIS

Figure 1.2: Structural model of PS II with an assignment of protein subunits and cofactors, seen perpendicular to the membrane plane (left), and parallel to it (left) [8].

al. [11] was that PS II cycles through four different states. Later this period-four oscilla- tions were assigned to different oxidation states of the manganese cluster which is coupled via a redox-active tyrosine to the primary donor P680. Meanwhile for all electron transfer cycles, which are coupled to each other in PS II kinetic measurements exist. An overview over the transfer times of electrons is given for instance in [12, 13]. Photochemistry in PS II is initiated by light-induced charge separation between the Chl a + species P680 and a pheophytin molecule (Ph), creating the radical pair P680 /Pheo−. Pri- mary charge separation takes about a few picoseconds. Subsequent electron transfer steps have been designed through evolution to prevent the primary charge separation from re- combining. This is accomplished by transferring the electron within 200 picoseconds from pheophytin to a plastoquinone molecule (QA) that is permanently bound to PS II. Al- though plastoquinone normally acts as a two-electron acceptor, it works as a one-electron gate at the QA site. The electron on QA− is then transferred to another plastoquinone molecule that is loosely bound at the QB-site. The plastoquinone at the QB-site differs from QA in that it works as a two-electron acceptor, becoming fully reduced and protonated after two photochemical turnovers of the reaction centre. The full reduction of plastoquinone requires the addition of two electrons and two protons, i.e., the addition of 1.3. COFACTORS OF THE PS II CORE COMPLEX 9

−1.2

−0.8 * P680 3 ps Pheo −0.4 200 ps 100/600 µs QA 0 Q B

E [V] PQ hν 1 ms 0.4

100−800 µ 0.8 H2O s Mn 4 Tyr z P 1.2 200 ps 680 O2

Figure 1.3: Probable electron transfer pathway and kinetics in PS II together with the redox potentials of the participating cofactors.

two hydrogen atoms. The reduced plastoquinone then detaches from the reaction center and diﬀuses into the hydrophobic core of the membrane. After this, another plastoquinone molecule ﬁnds its way to the QB-binding site and the process is repeated. Because the

QB-site is near the outer aqueous phase, the protons added to plastoquinone during its reduction are taken from the outside of the membrane.

To allow a continued light-induced charge separation the primary donor P680 being a cation radical after emitting one electron has to be reduced which is accomplished by the uptake of one electron via a tyrosine of the protein surrounding, called YZ. This redox- active amino acid acts as a mediator between the primary donor and the manganese cluster, which provides the electrons abstracted from water molecules.

1.3 Cofactors of the PS II core complex

A sketch of the reaction centre of PS II showing the cofactors embedded in the protein subunits is depicted in ﬁgure 1.4. The function and possible structure of these cofactors will be described shortly in the following sections which will focus particularly on the cofactors dealt with in this work. 10 CHAPTER 1. PHOTOSYNTHESIS

CP43 CP47

QA QB Fe

Ph Ph P680 Chl Chl

YZ YD Cyt b559

Mn 4 D1 D2 PsbO/CP33

Cyt c550

Figure 1.4: Sketch of a PS II core complex particle containing the cofactors needed to maintain the water splitting ability. Ph: pheophytin, Chl: accessory chlorophylls.

1.3.1 The primary donor

The primary donor in PS II, where the excitation energy is gathered from the surrounding antenna pigments to eventually start the electron transfer process, is believed to consist of a dimer of two Chl a molecules in close distance to each other. Diﬀerent models concernig the nature of the primary donor of PS II exist and are summarized in [16]. Under illumination and absorption of one photon it emits one electron:

+ P680 + hν −→ P680∗ −→ P680• + e− (1.2)

Whether P “ is really the starting point of the charge separation is still under debate. ” 680 P680 has a very high redox potential (≈ +1.1 V) which is the highest found in nature for such a molecule. This oxidizing capacity is needed to oxidize water but is indeed dangerous for the protein surrounding [17]. Although striking similarities exist between the protein organization of PS II and reaction centres from purple bacteria [18] and the arrangement of speciﬁc cofactors in these photosystems, homology considerations concerning the primary donors are not applicable. The primary donor in purple bacteria has already been identiﬁed but there are arguments that its structural homologue in PS II is not the primary donor in PS II [19]. 1.3. COFACTORS OF THE PS II CORE COMPLEX 11

+e− S−2

+e− S−1 hν ν S4 S0 h

2 H2O hν hν

S + − S1 3 O2 + 4 H + 4 e

ν hν h S2

Figure 1.5: States of the manganese cluster during the splitting of two water molecules into molecular oxygen, protons and electrons including the super reduced states which are non-native states of the Kok-cycle.

1.3.2 The Oxygen-Evolving Complex (OEC)

The OEC is unique among the photosystems and enables PS II to use water as an ubiqitous source for electrons. Its structure and detailed function are still unresolved and have remained in a daze for decades now. Indications for the occurence of manganese on the lumenal side of PS II were found by Fowler and Kok [20] - Blankenship and Sauer [21] showed that manganese ions are released to the lumenal side upon Tris-treatment and therefore concluded that it should reside in the fully assembled protein. The binding of manganese ions to specific amino acids was first observed by Coleman and Govindjee [22]. It has been shown that under certain conditions it is possible to reassemble a functional manganese complex in PS II preparations that have been depleted of manganese [23]. This process which also occurs in the native assembly of the manganese cluster is called photoactivation and includes several light-driven steps where manganese is bound and oxidized. The final steps of the assembly consist of the spontaneous binding of two Mn ions which do not require light. In its native state this cluster, which consists of four manganese ions, occurs in 5 different states with four of them more or less stable (S0, S1, S2, and S3) and therefore accessible for observation. One extremely short-lived state (S4) has been postulated. All these states are members of a closed reaction cycle discovered by Kok [11], and which is therefore called Kok-cycle. This cycle is visualized in figure 1.5.

In order to advance from one state to the other a photon is needed. Only the S3 state proceeds via the S4 to the S0 state under uptake of only one photon. At room temperature 12 CHAPTER 1. PHOTOSYNTHESIS

Figure 1.6: Position and alignment of the manganese ions derived from the electron density within the X-ray crystallographic map

the dark stable state is S1 whereas S0 is the state which is most reduced. In the Kok-cycle two water molecules are split into molecular oxygen, protons and electrons. The electrons are withdrawn by the primary donor via the tyrosine YZ . Under strongly reducing conditions it is possible to reduce the manganese cluster further down to the so called super reduced states which are not native [24]. Under illumination these states get oxidized approaching the S0 state through which the native Kok-cycle is reentered.

In its S2 state that is easily accessible and metastable at low temperatures the manganese cluster is paramagnetic and can therefore be detected by EPR. Here two prominent signals one the so called g = 4.1 signal, an unstructured Gaussian line at g = 4.1 [25], and the other the multiline signal (MLS), a highly hyperfine structured signal at g = 2 discovered by Dismukes and Siderer [26], prevail. Several explanations that assign the origin of these signals to magnetic interactions of the manganese ions have been suggested. For an overview see [14]. Together with results from EXAFS and XANES [15], from where manganese-manganese distances and oxidation states of the manganese ions, respectively, are available, many structural models for this cofactor have been proposed. Recently X-ray structure analysis has helped to exclude some of the structural models although the resolution of 3.6 A˚ is still not sufficient to reveal the binding situation of the four manganese ions. One possible arrangement fitted into the 5σ electron density is show in figure 1.6 [8]. 1.3. COFACTORS OF THE PS II CORE COMPLEX 13

1.3.3 The tyrosines YZ and YD•

A redox-active amino acid called YD is situated within the D2 protein (Tyr160) [27] and has a counterpart YZ (Tyr161) in the D1 protein [28]. YZ is directly involved in the light-induced electron transfer and plays a decisive role in the electron transfer between the primary donor P680 and the manganese cluster. It has been identiﬁed as the donor to P680 by side-directed mutagenesis [29]. Gilchrist et al. [30] calculated it to be located about 4.5 A˚ from the manganese cluster. According to X-ray crystallographic results it should rather have a distance of ≈7.5A˚ [8]. Furthermore, YZ can act as a strong proton donor [31] and its rereduction occurs on the same time scale as the release of oxygen from the manganese cluster during the S3-S4-S0 transition [32]. Recently Tommos et al. [33] and Hoganson et al. [34] have concluded that there must be a close association of YZ with the manganese cluster. For a further overview on this cofactor see [16].

In contrast to YZ its counterpart YD is not a member of the electron transfer chain + although it is redox-active. It can donate an electron to P680, which is very fast with t1/2 = 190 ns at high pH (8.5) whereas it is slow at lower pH (6.5) with t1/2 > 1 ms [35]. It was shown that it is involved in the protection of PS II against photoinhibition [36]. The formed radical remains stable for days at room temperature and for months at liquid nitrogen temperature. It is able via P680 to oxidize the S0 state to the S1 state [37] although this reaction is very slow compared to the other electron transfer processe in PS II. The EPR spectra of the two tyrosine radicals in PS II are similar [38]. Experiments indicate that also their binding properties are very similar [16]. But YZ• decays very fast whereas YD• is quite stable. 14 CHAPTER 1. PHOTOSYNTHESIS Chapter 2

Principles of EPR

This chapter deals with the spectroscopic technique Electron Paramagnetic Resonance (EPR), its theoretical background and its applications. Many and valuable books and articles have been published where explanations to phenomena mentioned in the following passages can be found [39–43].

2.1 The electron spin and its interaction with a magnetic ﬁeld

EPR spectroscopy is a technique to perform measurements on the magnetic moment of an electron. Considering the anomalous Zeeman-eﬀect and the Stern-Gerlach-experiment Goudsmith and Uhlenbeck 1925 postulated a new quantum number of the electron, the spin quantum number. It is a quantum mechanical angular momentum with no classical counterpart. A charged particle with an angular momentum implies the existence of a magnetic moment which is parallel to it:

µ~ spin = γ~s (2.1)

1 with γ being the gyromagnetic ratio. The eigenvalues of ~s are given as multiples of 2 of h¯ so one uses the following expression:

γh¯ = −gµB (2.2)

Bohr’s magneton is here symbolized by µB. The so called g-value of the free electron is ge=2.0023193043737(38) [44]. Being exposed to a magnetic ﬁeld the so far degenerate energy levels split according to the quantum numder ms and can now be described by

15 16 CHAPTER 2. PRINCIPLES OF EPR this term:

E = msgeµBB0 (2.3)

1 1 Since ms for one electron can have the values 2 and − 2 this results in two energy levels and one possible transition between them which has this transition energy:

∆E = geµBB0 = hν (2.4)

This phenomenon is used in magnetic resonance techniques here in particular for EPR. An electromagnetic alternating field is applied to a sample containing unpaired electrons. When the energy gap between the two levels corresponds to the energy delivered by this field one observes resonance hence with the appropriate equipment a signal. Since WW II developments in radar technology made microwave sources available for decent prices. EPR nowadays is carried out at X-band frequencies (≈ 9 GHz) which corresponds to a field of ≈ 0.3 T. But meanwhile a field from 0.1 to 6 or more Tesla can be covered with the microwave sources and magnets available.

2.2 The Spin Hamiltonian

A method to describe the energy of a system is by means of the Hamilton-Operator, the Hamiltonian Hˆ . For sytems that are observed in this work the Hamiltonian which is called the Spin-Hamiltonian has the general form:

Hˆ S = Hˆ EZ + Hˆ NZ + Hˆ Hf + Hˆ Q + Hˆ D (2.5) with

Hˆ EZ = Electron-Zeeman-Interaction

Hˆ NZ = Nuclear-Zeeman-Interaction

Hˆ Hf = Hyperﬁne-Interaction

Hˆ Q = Quadrupol-Interaction

Hˆ D = Zero-Field-Interaction

2.3 The Electron-Zeeman-Interaction and the g Tensor

The Electron-Zeeman-Interaction is the magnetic coupling between an electron represented by its spin ~S and a magnetic ﬁeld B~ applied to it: ~ Hˆ EZ = µB~S · ~g · B~ (2.6) 2.4. THE NUCLEAR-ZEEMAN-INTERACTION 17

In this equation ~g is the Zeeman coupling matrix, often referred to as the g-tensor. For a free electron it is isotropic and can be written as a scalar factor ge = 2.0023193043737(38). In chemical compounds where the electrons are subject to restrictions as interactions with the magnetic moments of other spins in their surrounding the g-tensor becomes anisotropic. This anisotropy stems from the coupling of the electron spin with magnetic moments of electrons in higher orbitals than the s-orbital which possess a spin orbit momentum. The resulting effective spin is not a pure spin-state anymore but contains also orbital components. For most chemical compounds the distribution of the effectice magnetic moment, i.e. spatial distribution of orbitals with an orbit quantum number l > 0, is anisotropic. Systems with high symmetry, however, show an isotropic g. The g-tensor itself thereby shows the orientation dependence of the effective g-value. Since it is diagonizable, a relationship between its orientation and the coordinate system and therefore the symmetry of the molecule hosting the spin results.

2.4 The Nuclear-Zeeman-Interaction

Nuclei possess also a spin angular momentum. The Nuclear-Zeeman-Interaction is the magnetic coupling between an nucleus represented by its spin ~I and a magnetic ﬁeld B~ applied to it: ˆ ~ ~ ~ HNZ = µNI · ~gN · B (2.7) ~ The anisotropy of ~gN is small compared to the resolution accessible with EPR. So the ~ anisotropic ~gN can be replaced by the scalar gN.

2.5 The Hyperﬁne-Interaction

This part of the total Hamiltonian, the hyperﬁne coupling hfc, refers to the interaction of the magnetic moment of the electron and the nucleus. It can be split into two parts:

Hˆ hfc = Hˆ contact + Hˆ dipole ~ ~ = Aiso · ~S ·~I + ~S · A~ dip ·~I = ~S · A~ ·~I

2µ0 2 µ0 µ~ S · µ~ I 3(µ~ S ·~r)(µ~ I ·~r) = − µ~ Sµ~ I|ψ(0)| − − . (2.8) 3 4π r3 r5 ! The first term is called the Fermi-Contact-Interaction. It describes a singularity for r→0 by an integration over space for the field. From far away the magnetic field pattern is that of a point dipole. However, if an electron can sample the field close to the nucleus, 18 CHAPTER 2. PRINCIPLES OF EPR then the field distribution differs significantly from that of a point dipole. For example, if an electron can penetrate the nucleus, then the spherical average of the field it experiences is not zero. Electrons residing in orbitals or superposition of orbitals having some s-orbital character contribute to this interaction. This means a probability > 0 for the electron to be found at the place of the nucleus. The isotropic contribution Aiso can be calculated ~ from A~ : 2µ0 1 ~ A = h¯g ν g ν |ψ(0)|2 = trace (A~ ). (2.9) iso 3 e B N N 3 The second term arises from the Dipole-Dipole-Interaction. The two magnetic dipoles

µS and µI couple to each other. The vector ~r is the diﬀerence between their two position ~ vectors. This term has no isotropic contributions. A~ dip is called dipolar hyperﬁne tensor.

2.6 The Electronic Dipolar Interaction

Dipole-Dipole-Interaction concerning two electrons with the magnetic moments µ~ 1 and 1 µ~ 2 is very similar to the coupling between an electron and a nucleus (with I > 2 ). The formula 2.8 can be modiﬁed for this purpose by neglecting the isotropic part:

µ0 µ~ 1 · µ~ 2 3(µ~ 1 · ~r)(µ~ 2 · ~r) Hˆ dip = − . (2.10) 4π r3 r5 !

Using the spin operators Sˆ1 and Sˆ2 this equation becomes:

2 2 ~ˆ ~ˆ ~ˆ ~ˆ g µ µ0 S1 · S2 (S1 · ~r1)(S2 · ~r1) ~ ~ Hˆ = e B − 3 = Sˆ · D · Sˆ (2.11) dip 4π  r3 r5  1 2   The scalar products are:

~ ~ Sˆ1 · Sˆ2 = Sˆ1xSˆ2x + Sˆ1ySˆ2y + Sˆ1zSˆ2z (2.12) ~ Sˆ1 · ~r = Sˆ1xx + Sˆ1yy + Sˆ1zz (2.13) ~ Sˆ2 · ~r = Sˆ2xx + Sˆ2yy + Sˆ2zz (2.14)

It is convenient for tightly coupled electrons to express Hˆ dip in terms of the total spin operator Sˆ: ~ ~ ~ Sˆ = Sˆ1 + Sˆ2 (2.15)

Taking into account the relations between spin operators and their components one eventually arrives at an overall expression for Hˆ dip: 2.6. THE ELECTRONIC DIPOLAR INTERACTION 19

(r2 3x2) 3xy 3xz − ˆ 2 2 r5 −r5 −r5 Sx ge µBµ0 3xy (r2 3y2) 3yz ˆ ˆ ˆ ˆ − ˆ Hdip = Sx, Sy, Sz  −r5 r5 −r5   Sy  , (2.16) 6π 3xy 3yz (r2 3z2)  − − −   Sˆ   r5 r5 r5   z      which can again be written in a short form as:

~ ~ Hˆ dip = Sˆ · D · Sˆ. (2.17)

D is a second rank tensor. It is symmetric and can be diagonalized to dD:

Dx 0 0 d D =  0 Dy 0  (2.18)  0 0 D   z   

The dipolar coupling tensor is traceless. Therefore only two parameters chosen to be D and E are needed to suﬃciently satisfy the condition Dx + Dy + Dz = 0:

1 D = D + E (2.19) x 3 1 D = D − E (2.20) y 3 2 D = − D (2.21) z 3

Since Dx, Dy and Dz are the elements of the diagonal hamiltonian matrix they are the energies of this system in zero magnetic ﬁeld.

2.6.1 Two Electrons Coupling in an External Magnetic Field

To obtain the total Hamiltonian assuming an applied external ﬁeld B~ , equation 2.17 must be added to equation 2.6:

~ ~ ~ Hˆ total = µB~S · ~g · B~ + Sˆ · D · Sˆ. (2.22)

The functions |1>, |0>, and |-1> can be used as a basis set, since they are the eigenfunc- ~ ~ tions for an inﬁnitely large magnetic ﬁeld. But they are not eigenfunctions of Sˆ·D·Sˆ. To acquire the hamiltonian matrix in the form, where an isotropic g factor has been assumed:

ˆ 2 2 2 Htotal = gµBBX Sx + gµBBY Sy + gµBBZ Sz − DxSx − DySy − DzSz, (2.23) 20 CHAPTER 2. PRINCIPLES OF EPR the following spin matrices have to be used:

1 0 √2 0  1 1  Sx = √2 0 √2 (2.24)  1   0 √ 0   2   

i 0 √−2 0  i i  Sy = √2 0 √−2 (2.25)  i   0 √ 0   2    1 0 0

Sz =  0 0 0  (2.26)  0 0 −1      These spin matrices have to be substituted into equation 2.22 and are thereby yielding the complete hamiltonian matrix from which the eigenvalues of the system can be calculated.

2.6.2 Electron Exchange Interaction

The isotropic exchange coupling originates from the electrostatic interaction of the electrons. Pauli’s principle requires the entire wave function to be antisymmetric. The transition between two spin states causes a change in symmetry, i.e. there is a diﬀerence in energy for the two states. The electrostatic interaction energy Ees can be described by the Coulomb interaction: e2 Ees = (2.27) 4π0r Hence the coupling between the participating electrons depends on the overlap between their corresponding wave functions and is described by the coupling constant J:

e2 J = hψA(1)ψB(2)| |ψB(1)ψA(2)i, (2.28) 4π0r for the electrons 1 and 2 and the orbitals described by ψA and ψB. In terms of the hamiltonian using J this leads to: ~ ~ Hˆex = −2J · Sˆ1 · Sˆ2 (2.29)

Knowledge of the sign and the magnitude of J can provide information on the electronic couplings and energetics of the system that hosts the coupling electrons, e.g. the tunneling 2.6. THE ELECTRONIC DIPOLAR INTERACTION 21 probabilities in electron transfer processes. The value of 2J is the energy gap between the singlet and the triplet state neglecting Zeeman and dipolar interaction. Exchange interaction can also be mediated via excited electronic states. Smaller contributions are due to changes in spin-orbit coupling, conformational changes etc. The resulting exchange interaction is therefore not necessarily isotropic. Practically speaking the isotropic part of spin-spin interactions is called “exchange interaction” and all anisotropic components are included in the dipolar coupling term.

The Hamiltonian for exchange coupled complexes

Assuming a system with n unpaired electrons therefore having n magnetic moments and also o nuclear magnetic moments the Hamiltonian consists of terms as follows: o n Hˆ = B~ · gi · Sî + Sî · Ah · Iˆh + Sî · Di · Sî + Iˆh · Pˆh · Iˆh − γhBˆ · Iˆh Xh Xi h i n − JjkSˆj · Sˆk. (2.30) j,k,j=k X6 The terms in this equation are from left to right: electron Zeeman interaction, electron- nuclear coupling (hyperfine interaction), electronic dipolar coupling (zero field splitting interaction), nuclear quadrupole interaction, nuclear Zeeman interaction and electronic exchange coupling between the ions. B~ is the external magnetic field vector, gi are the g-tensors of the single ions, Sî and Iî are the spin angular momentum operator associated with the electron and nuclear magnetic moment of the single ions, respectively. The hyperfine tensor is symbolized by Ai, the quadrupole tensor by Pi, and γi ist the gyrom- gagnetic ratio of the single nuclei. For exchange coupled complexes such as the manganese cluster of PS II and its dinuclear model complexes one in general assumes an effective Spin Seff caused by the antiferromagnetic or ferromagnetic exchange coupling of their electron spins [170]. This effective spin implies the use of the total spin angular momentum operator Sˆtotal in the Hamiltonian for the coupled case: n Hˆcoupled = B~ · g0 · Sˆtotal + Sˆtotal · Ai0 · Iî + Iî · Pi · Iî − γiB~ · Iî Xi h i +Sˆtotal · D0 · Sˆtotal. (2.31)

Here the g0 (electron Zeeman, first term) and D0 (zero field) matrices are related to the effective spin. This spin couples to the nuclei with the effective hyperfine matrices Ai’ (second term). In the equation the third term again symbolizes the quadrupolar interactions of the nuclei, the fourth term the nuclear Zeeman term, and the last the zero field 22 CHAPTER 2. PRINCIPLES OF EPR interaction for effective spins higher than 1/2. The exchange coupling of the electron spin of the metal ions with a coupling constant J leads to effective spin levels. J is negative for antiferromagnetic and positive for ferromagnetic coupling. The magnitude of J depends on the overlap of the electronic wave functions of the participating metal ions and influences the values of the effective hyperfine couplings.

The g’, Ai’ and D’ tensors for the coupled case are related to their isolated counterparts in the non-coupled representation by the projection matrices [170]. For weakly exchange coupled cases with a small J also contributions from higher electronic spin levels have to be taken into account, even at low temperatures, which comes along with contributions from the zero field splitting because the effective spin is greater than 1/2. Here the intrinsic zero-field matrix has to be included calculating the projection matrices for the hyperfine coupling matrices. For a Mn(III) ion the axial zero field splitting is in the range 1 of |DIII|=1-4 cm− , whereas for a Mn(IV) ion it is ten times smaller with |DIV |=0.1- 1 0.4 cm− . The connection of the projection matrices for a Mn(III)Mn(IV) dimer with the zero field parameters and the exchange coupling constant is given in Peloquin et al [14]. With the help of these projection matrices ρ(J,D,S) it is possible to calculate the intrinsic hyperfine coupling tensors of the metal ions and hereby for instance deduce the oxidation state. ENDOR measurements can provide the effective hyperfine coupling components.

2.7 Electron Paramagnetic Resonance

The establishment of Electron Paramagnetic Resonance (EPR) originates in the midforties of the last century. Since the power of EPR and its methods concentrates on substances possessing one or more paramagnetic units, its application to the ﬁeld of photosynthetic reactions has been particularly successful since paramagnetic species prevail throughout the processes of these important biological processes.

2.7.1 Continuous wave EPR

Continuous wave (cw) EPR was the first application of EPR at all invented in 1945 by Zavoisky. Though nowadays this very simple technique has been complemented by various other methods using the very principle of it, it is still widely used for the characterization of paramagnetic species. In cw EPR the absorbance of the energy of an electromagnetic field as a function of the magnetic field is detected. All the spin dynamics are more or less 2.7. ELECTRON PARAMAGNETIC RESONANCE 23 incoherent. The interference of the sample with the perturbation imposed by the device is kept low so that the sample can be regarded as close to thermal equilibrium. In contrast to NMR EPR devices are run leaving the frequency constant and changing the magnetic field. The requirement for constant frequency is mainly the use of a resonance structure a sensitivity enhancement device. Cw machines work with effect modulation techniques to enhance the signal with respect to the noise. Spectra are therefore recorded as their first derivatives.

2.7.2 Transient EPR

Of all the techniques stemming from the original idea of electron paramagnetic resonance transient EPR is by far the one most comparable to cw EPR. Here the system is not stationary but gets disturbed and the relaxation kinetics into equilibrium are monitored time-resolved. Bringing the system out of equilibrium can be achieved by light excitation, a magnetic field-jump etc. Hereby it is possible to generate a paramagnetic species whose decay or the generation of a new paragmagnetic species is monitored. Transient EPR spectra are recorded in the direct detection mode. In this mode the signal is observed directly at the output of the pre-amplifier without field-modulation or lock-in technique. The accompanying loss of sensitivity is often compensated by spin-polarization phenomena leading to transient non-Boltzmann populations. The EPR spectra recorded in this way do not appear in the conventional first-derivative form but show directly absorption (A) or emission (E).

2.7.3 Pulsed EPR

Microwave pulses are used in pulsed EPR to bring a spin system out of equilibrium. The resulting transversal magnetization of the precessing spin ensemble is measured. The pulses can be represented as operators acting on the spin system since the microwave pulses interact directly with the spins. This results in a simple model for the description of spin dynamics and is shown in figure 2.1 for a π/2 − π pulse sequence. Using several pulses with specific durations and distances quantum coherences can be created. These are used to probe specific properties of the system. In the rotating frame approximation the effect of applied microwave pulses on the spin system can be described as an additional static magnetic field, giving rise to Larmor precession. This means that through variation of duration and amplitude of the pulses the “tilt” of the spins with respect to their original direction can be tuned. Mostly pulses 24 CHAPTER 2. PRINCIPLES OF EPR

z z

M Z 90° Pulse

x y x y

precessing and spin−dephasing

z z

slow fast

180° Pulse x y τ x y fast slow τ rephasing z

x y

echo

Figure 2.1: The evolution of spins (symbolized by the magnetization) exposed to an external magnetic field B and microwave pulses perpendicular to the magnetic field vector. Exemplified is a “π/2 − π” sequence. are used to switch the polarization of the spin ensemble between longitudinal (along the static magnetic field) and transversal (perpendicular to the static field).

2.8 Other techniques of EPR

2.8.1 Continuous Wave ENDOR

Electron Nuclear DOuble Resonance abbreviated as ENDOR [45] helps to partially overcome one of the drawbacks one has to deal with using conventional EPR. The one or more electrons in singly occupied orbitals of the system under observation couple to the other 1 magnetic moments within this system, such as protons or other nuclei with I≥ 2 . This results in the so called hyperfine coupling apparent in the spectrum as splitting of the 2.8. OTHER TECHNIQUES OF EPR 25 original EPR line. In EPR, due to usually very small relaxtion times, the experimentalist is confronted with quite large line widths of the overlapping hyperfine lines compared to the overall line width of the EPR line. The effect is the blurring of the information provided by hyperfine coupling, and one observes just an inhomogeneously broadened Gaussian line. The application of ENDOR tremendously reduces the number of lines being observed hence these are then better resolved due to the overlap of fewer lines. In EPR the number of lines observed amounts to:

k NEPR = (2NiIi + 1), (2.32) i=1Y whereas using ENDOR it only comes up with:

NENDOR = 2k. (2.33)

Here k ist the number of sets of inequivalent nuclei and Ni the number of equivalent nuclei of type i. The gain in resolution of ENDOR compared to EPR is achieved on cost of the sensitivity since the ENDOR-effect amounts only to some percent of the EPR-effect. 1 1 In figure 2.2 the case of of a system of one electron (S= 2 ) coupling to a proton (I= 2 ) is shown. The allowed EPR- and NMR-transitions are depicted by arrows. On the right of figure 2.2 a practical approach to this technique is explained. All relaxation pathways are drawn with dotted lines. One EPR-transition (solid line MW) is saturated by irradiation of a microwave field of high power. In this case the EPR-signal would disappear. The simualtaneously irradiated radio frequency field induces NMR transitions. This leads to a partial desaturation of the EPR transition and therefore to a rise of microwave absorption. An EPR-signal is observed at the frequency of the NMR-transition when sweeping the RF frequency and is recorded as ENDOR spectrum.

2.8.2 Pulsed ENDOR

The development of powerful amplifiers for microwave and radio frequencies made it possible to also use the opportunities of pulsed ENDOR. An overview on the different pulsed ENDOR methods is given in [46]. The method used in this work was the so called Davies ENDOR [47], together with the so called Mims ENDOR [48] the prevailing technique in the field of pulsed ENDOR. The main principle is similar to the one of cw ENDOR. The sequence of pulses is depicted in figure 2.3. It starts with a selective mw π-pulse to burn a hole into the EPR-line. This can also be described by burning a hole into the EPR line.

The relaxation back into equilibrium is characterized by T1, which is for frozen solutions of the magnitude of several hundred microseconds. In this mixing time a radio frequency 26 CHAPTER 2. PRINCIPLES OF EPR

RF NMR1 NMR1

EPR1 EPR2 MW ~

NMR2 NMR2

1 Figure 2.2: Scheme for the energy levels of a coupled system of an electron spin with S= 2 and a nucleus 1 with I= 2 . The isotropic coupling constant aiso is positive. On the left are depicted the allowed transitions for this system. On the right together with the relaxation pathways (dotted) and two driven absorptive transitions making ENDOR possible are shown.

π π/2 π MW

π RF

Figure 2.3: Pulse sequence for Davies ENDOR. The drawing on top shows the microwave pulses, the bottom shows the radio frequency on the same time scale. Typical pulse lengths for X-band are ≈100 ns for a microwave-π-pulse concerning the electron and ≈ 8 µs for a radiofrequency-π-pulse for protons. 2.8. OTHER TECHNIQUES OF EPR 27

π/2 π/2 π/2 MW

π RF

Figure 2.4: Puls sequence for Mims ENDOR. The drawing on top shows the microwave pulses, the bottom shows the radio frequency on the same time scale.

π-pulse is applied. This pulse is selective and if on-resonant inverts the magnetization of nuclear spins in one of the two ms manifolds. This results in the disappearance of the population difference of the recently excited EPR-transition. Ideally the spin echo vanishes and a π − π/2-sequence is applied for detection. Davies ENDOR is insensitive towards NMR transition frequencies smaller than the width of the spectral hole created by the microwave inversion pulse. The other method of choice as mentioned above is Mims ENDOR. It is based on a stimulated echo sequence experiment with three non-selective microwave π/2 pulses. The first two microwave pulses create a periodic polarization pattern across the inhomogeneously broadened EPR line. During the mixing period a selective radiofrequency π pulse of vari- able frequency is applied. If the radiofrequency is on-resonance with an NMR-transition, the populations in this transition will be changed and consequently also the polarization pattern. This change is then measured as a function of the radiofrequency via the stimulated echo intensity created by the third non-selective π/2 pulse. In a Mims-ENDOR experiment, the ENDOR signal intensity depends also on the hyperfine coupling constant A and on the time τ between the first two microwave pulses. For certain values of τ, there is no ENDOR effect because the timing of the pulse sequence can lead to additional correlations between the subensembles. This leads to so called “blind spots” which are particularly troublesome in ENDOR spectra of disordered systems, where the knowledge of the proper line shape may be essential.

2.8.3 Electron Spin Echo Envelope Modulation (ESEEM)

Two pulse and Three Pulse ESEEM

Electron Spin Echo Envelope Modulation is used to probe other paramagnetic centres with electrons in singly occupied orbitals. Biological interest lets focus on nuclei such as 14N or 1H, which are abundant in all enzymes one should think of. An overview including 28 CHAPTER 2. PRINCIPLES OF EPR also practical approaches is given in [50–53]. Compared to ENDOR the technique ESEEM is rather successful when applied to sytems containing nuclei with a comparably small gyromagnetic ratio. With these it is possible to excite allowed along with forbidden transitions at once. Advancing from the same energy level these transition are related to each other by polarization effects and a coherence transfer can occur due to the mixing of states by application of an ESEEM pulse pattern. In a two-pulse ESEEM experiment the two pulses applied should be non-selective. At the beginning the net magnetization be aligned along the z-axis. The first pulse along x rotates the magnetization by 90◦ into the xy-plane. During the time τ between the first and the second pulse the different spin packets evolve which results in a dephasing of them. The following π pulse inverts the x-component of the individual magnetization vectors. After a time τ the magnetization vectors are in phase along the x-axis and a primary echo is observed. In this experiment, the time τ between the two pulses is increased and the decay of the primary echo is monitored. The echo decays due to spin-spin and spin-lattice relaxation. Superimposed on the decaying echo amplitude, a modulation can occur if nonsecular electron-nuclear spin interactions are present. Fourier transformation of the time-domain modulation patterns yields a spectrum with frequencies directly related to the hyperfine and nuclear quadrupole interactions. The two-pulse modulation formula [54, 55] for a system with S=1/2 and I=1/2 shows that not only the basic nuclear frequencies will be observed, but also peaks at the sum and difference frequencies. The positions of these combination frequencies can be very helpful for the interpretation of two-pulse ESEEM spectra.

V (τ, T ) = Vα(τ, T ) + Vβ(τ, T ) ω τ ω (τ + T ) = 1 − k sin2 α sin2 β 2 2 ! ω τ ω (τ + T ) −ksin2 β sin2 α . (2.34) 2 2 ! The constant k is the modulation factor describing the depth of the modulations in the ESEEM spectrum:

ω0I B 2 2 2 k = with B = Axz + Azy. (2.35) ωαωβ ! The constant B is a measure for the anisotropy of the hyperfine coupling. Complications utilizing ESEEM arise if the anisotropy is small or the direction of the external field is along or close to one of the main axes of the hyperfine coupling tensor since in these cases the modulation depth goes down.

The decay of the echo caused by the spin-spin relaxation time T2, which is in most cases of the order of a few microseconds, is usually a serious drawback, i.e. the echo can not 2.8. OTHER TECHNIQUES OF EPR 29

π/2 π/2 π/2

τ T

Figure 2.5: Pulse sequence for a three-pulse ESEEM. It consists of three non-selective short π/2 pulses. The signal is detected as a function of τ and T. be followed for a sufficient long time. This problem can be overcome using a three-pulse sequence. Here the applied pulse sequence π/2−τ −π/2−T−π/2 creates a stimulated echo at a time τ after the last π/2 pulse. The first π/2 pulse has again the effect of transferring the magnetization from the z-axis into the xy-plane. Dephasing occures within the time τ. Then a second π/2 pulse is applied which rotates the magnetization into the z-axis.

During time T this magnetization decays in the time regime of T1 which is usually much longer than T2. At the time τ + T the third π/2 pulse is applied which rotates the magnetization along the z-axis again into the xy-plane. After the time 2τ + T the echo has built up. Again a modulation is superimposed on this echo in the time domain if the time T between the second and the third pulse is changed (the time τ is kept constant). The diﬀerence compared to the two-pulse experiment is that the spectrum in this case is devoid of sum and diﬀerence frequencies. The resolution here is limited by the spin-lattice relaxation time T1.

HYSCORE

Hyperfine sublevel correlation spectroscopy is a two dimensional EPR experiment using a pulse sequence of π/2-τ-π/2-t1-π-t2-π/2 (see figure 2.6). It was introduced by Hoefer et al. [56, 57] as a four-pulse experiment which correlates the two NMR frequencies associated with a particular hyperfine coupling. The experiment is based on the three-pulse stimulated echo sequence. A mixing π pulse is inserted between the second and the third π/2 pulse to create correlations between the nuclear spin transitions of the two electron spin manifolds ms. To acquire the 2D time-domain modulation signal, the stimulated echo amplitude is observed as a function of pulse separations t1 and t2 assuming a fixed τ. The recorded modulation signal is then 2D-Fourier transformed. The modulation function for this pulse pattern is according to [58–60]:

V (τ, t1, t2) = 1/2 [Vα(τ, t1, t2) + Vβ(τ, t1, t2)] (2.36) The two terms in equation 2.36 can be written as follows:

C0 ωατ ωβτ Vα = 1 − k/2 ∗ [ + Cα cos ωαt1 + + Cβ cos ωβt2 + (2.37) 2 2 2 30 CHAPTER 2. PRINCIPLES OF EPR

π/2 π/2 π π/2

τ t1 t 2

Figure 2.6: Pulse sequence for a HYSCORE experiment. Additional to a three pulse ESEEM a non- selective π pulse is applied which switches the coherences of the nuclear spins between the two ms manifolds.

2 ω+τ 2 ω τ +Cc{c cos ωαt1 + ωβt2 + − s cos ωαt1 + ωβt2 + − }] 2 2 C0 ωατ ωβτ Vβ = 1 − k/2 ∗ [ + Cα cos ωαt2 + + Cβ cos ωβt1 + (2.38) 2 2 2 2 ω+τ 2 ω τ +Cc{c cos ωαt2 + ωβt1 + − s cos ωβt1 + ωαt2 + − }] 2 2

The constants C0, Cα, Cβ and Cc are the expressions:

2 2 C0 = 3 − cos(ωβτ) − cos(ωατ) − s cos(ω+τ) − c cos(ω τ) (2.39) − 2 ωατ 2 ωατ ωατ Cα = c cos ωβτ − + s cos ωβτ + − cos (2.40) 2 2 2 2 ωβτ 2 ωβτ ωβτ Cβ = c cos ωατ − + s cos ωατ + − cos (2.41) 2 2 2 ωατ ωβτ Cc = −2 sin sin (2.42) 2 2 2 2 with ω+ = ωα + ωβ, ω = ωα − ωβ and k = s c (2.43) −

2 2 2 2 The parameters s = If = sin (δ/2) and c = Ia = cos (δ/2) give values for the transition probabilities for the forbidden and allowed NMR-transitions. Assuming an isotropic g- factor and isotropic hyperﬁne coupling they amount to:

2 2 ω0 − 1/4 ∗ (ωα − ωβ) Ia = , (2.44) ωαωβ 2 2 ω0 − 1/4 ∗ (ωα + ωβ) If = . (2.45) ωαωβ

The angle between the two eﬀective ﬁelds at the nucleus of the two mS states is δ, the actual correlation of the nuclear frequencies ωα and ωβ is given by Cc. Using more than 2 pulses one always has to cope with unwanted echoes originating from the combinations of the applied pulses. A pulse sequence with N pulses generates NEcho = N 1 N 1 3 − −2 − echoes [60]. To minimize the interference of these unwanted echoes a so called phase cycle is used. Table 2.1 shows a typical phase cycling scheme used for a HYSCORE experiment. 2.8. OTHER TECHNIQUES OF EPR 31

Number and phase position of the pulses 1 2 3 4 Recording of the signal +x +x +x +x + +x +x +x +x − +x +x −x +x + +x +x −x −x −

Table 2.1: Phase cycle used for the recording of HYSCORE spectra.

Evaluation of HYSCORE spectra

The HYSCORE spectrum of an S=1/2 spin coupled to a nuclear I=1/2 spin shows correlation cross-peaks according to the nuclear frequencies (να, νβ, where α and β stand for the ms electron spin manifolds) [56, 119]. In general the main values of a rhombic tensor can be written as:

A = (Axx, Ayy, Azz) (2.46)

= (Aiso − T − ∆, Aiso − T + ∆, Aiso − 2T). (2.47)

Aiso is the isotropic, T the axial and ∆ the rhombic component of the hyperﬁne tensor. According to Dikanov et al [120] for an I=1/2 nucleus the nuclear frequencies are as follows: 2 2 2 2 2 2 2 2 2 να(β) = [νzα(β) cos θ + νyα(β) sin θ sin φ + νxα(β) sin θ cos φ] (2.48) with

νzα(β) = −ν1 (Aiso + 2T )/2 (2.49)

νxα(β) = −ν1 (Aiso − T + ∆)/2 (2.50)

νyα(β) = −ν1 (Aiso − T − ∆)/2, (2.51) θ and φ are the rotation angles of the magnetic field vector with respect to the principal axis system of the hyperfine tensor. Regarding the axial case for the hyperfine tensor, ∆ becomes zero, and the equations above are now: 2 2 2 2 2 να(β) = [ν α(β) − ν α(β)] cos θ0 + ν α(β) (2.52) || ⊥ ⊥

2 ν α(β) = −ν1 (Aiso + 2T )/2 (2.53) || 2 ν α(β) = −ν1 (Aiso − T )/2. (2.54) ⊥

Plotting να versus νβ for an axial A tensor keeping φ constant while varying θ this leads to arc-shaped cross-peaks in the HYSCORE spectrum. For a rhombic A tensor diﬀerent ν ν I I

ν ν I I

A ∆ 3T/2 B

Figure 2.7: Two examples for axial (left) and rhombic (right) hyperﬁne coupling tensors in HYSCORE spectra, respectively. νI is the frequency of the free proton. The other variables are described in the text.

φ values lead to diﬀerent arcs, all of them having a common point for θ = 90◦. This is visualized in [118]. In an ideal spectrum the values Aiso, T and ∆ can roughly be ascertained from the spectrum itself. This is shown in [118] and is described here shortly.

The axial and the rhombic case

In the axial case the projected distance between two corresponding arcs along the antidiagonal to one of the frequency axes is equal to 3T/2. The ends of two corresponding arcs pointing towards the diagonal have a projected distance of Aiso − 2T when Aiso/T < 0, and Aiso−T when Aiso/T > 0. Rhombic hyperﬁne tensors in the HYSCORE spectrum looking like spherical triangles allow the determination of T and ∆ by evaluation of the extreme points in their structure. According to Deligiannakis et al [118] these two equations are valid:

3T = A + B (2.55) ∆ = A − B. (2.56)

B is the projected distance between the extreme point of the triangle being the nearest to the diagonal and the one farthest from it, whereas value A is the projected distance between the nearest point and the extreme point between the farthest and the nearest point. Two examples are axial and rhombic hyperﬁne tensors are depicted in ﬁgure 2.7. Chapter 3

Sample preparation and characterization

3.1 PS II membrane fragments

The samples used in this work were mostly prepared from fresh market spinach according to the method described by Berthold/Babcock/Yocum (“BBY”) [9] sometimes with the modifications described in [115]. Sucrose at a concentration of 0.4 M, 1 M betain or 30% glycerol depending on the experiment were used as cryoprotectants. Chl concentrations were determined as described by Porra et al[176]. The samples were in general brought to a concentration of ≈4 mg(Chl)/ml and stored as beads at -70◦C. Depending on the experiment the beads were thawed and the required additions made (specified in the according chapters). Then the suspension of BBY particles was filled into EPR sample tubes. To increase the concentration of reaction centres and therefore the anticipated EPR signal these sample tubes were centrifuged in special sample adaptors to yield a final concentration of 15-20 mg(Chl)/ml. The procedure is visualized in figure 3.1. Specific samples and sample treatments are described in the respective chapters.

3.1.1 Sample treatments

To induce the formation of paramagnetic species the samples in some cases had to be illuminated at low temperatures. An illumination temperature of 200 K, which was used in most of the cases was achieved by a dry ice/ethanol bath in a glas dewar. The sample

33 34 CHAPTER 3. SAMPLE PREPARATION AND CHARACTERIZATION

Sample

Figure 3.1: To achieve a high concentration of PS II reactions centres in the sample tube the procedure mentioned in the text is used. After the sample with an approximate concentration of 4 mg(Chl)/ml is filled into to sample tube the latter is put into the adaptor appropriate for an SS 34 rotor. The ensemble is then centrifuged at 4000-5000 rpm for 30-60 min. tube was transferred into this dewar. The dewar was placed in between two optical banks consisting of a arrangement of optical lenses which were adjusted to focus the illumination light. The illumination was carried out using two halogen lamps (P=250 W each) whose light was directed through a water filter and an infrared cutoff filter (750 nm) to remove most of the infrared part of the light. The illumination time depended on the sample concentration and the light-induced species to generate.

T 7 1 2 3 4 5 6

Figure 3.2: Illumination device, 1: halogen lamp, 2,3,5: lenses, 4: water filter, 6: infrared cutoff filter, 7: dewar with dry ice/ethanol bath, 8: sample tube with sample; to double the light intensity the same illumination arrangement is present on the right. A thermo element is used to control the temperature of the dry ice/ethanol bath. Chapter 4

The Light-induced Triplet State in Photosystem II

4.1 Motivation

In Photosystem II the actual charge separation starts on one of the chlorophyll species in the reaction centre. Consensus exists that this is either the so called P680 or one of the “accessory chlorophylls” [61, 62]. For bacterial reactions centres which resemble the reaction centre of PS II [18, 63] but lack the oxygen evolving complex the charge separation starts on P865, the counterpart to P680 in PS II. Many structural [8, 64, 65] aspects of PS II have been revealed in the last years [8, 64–66]. EPR has proven to be particularly helpful for this system where paramagnetic species play a main role [67]. In all photosynthetic systems triplet states of the pigment molecule are usually not formed since the singlet energy transfer and subsequent charge separation process in the RC are very fast. When forward electron transfer to the ﬁrst stable electron acceptor (quinone) in RCs is blocked, the primary singlet radical can, however, transform into the triplet radical pair that recombines to form a pigment triplet 3P with a lifetime of ≈100µs (in bRC) to 1 ms (in PS II) [68]. It is not a priori clear on which pigment 3P is located in PS II. 3 Traditionally the triplet state in PS II is denoted P680. The triplet state is potentially harmful since it reacts with O2 to form singlet oxygen which leads to cell damage. Nature has provided protection against this reaction by incorporating carotenoid molecules in close contact to chlorophylls, therby enabling eﬃcient triplet energy transfer. 3Car can- not react with oxygen [69]. Although the triplet is not a functional state in the electron transfer process. Its investigation is important to understand the primary donor excited state electronic properties and triplet energy transfer. Furthermore, the triplet can be

35 36 CHAPTER 4. THE LIGHT-INDUCED TRIPLET STATE IN PHOTOSYSTEM II used as a paramagnetic probe for EPR studies of the photosystems [70, 71]. The generation of 3P in a magnetic field from the strongly spin-polarized radical pair state leads to very intense signals and a characteristic polarization pattern in the time resolved transient EPR powder spectrum [72–74]. The interaction between the two unpaired electron spins of 3P leads to zero field splitting (ZFS) in the spectra, described by the parameters D and E, which reflect the size and symmetry of the two singly occupied orbitals. From a comparison with the respective ZFS parameters of monomeric (bacte- rio)chlorophyll in organic glasses it has been concluded that at cryogenic temperatures (≈10 K) the triplet exciton is delocalized over a pair of such molecules in the bRC [75], whereas in PS II it is located on a single chlorophyll [76]. At elevated temperatures, a delocalization onto a second chlorophyll was, however, postulated from the observed spectral changes [77]. Information on the ZFS tensor axes has been obtained from studies of 3P in single crystals of bRCs which supported the model of triplet exciton delocalization over a dimeric bacteriochlorophyll (BChl), the so called special pair [75]. Additional information on the electronic structure of the triplet state in the bRC and PS II has been gained by application of transient [78] and pulse ENDOR techniques [67, 79, 80] by which several hyperfine coupling constants could be resolved. In the RC of purple bacteria the sequence of charge transfer events after photo excitation of the primary donor is well established. The initial charge separation occurs between the photo-excited primary donor P∗, a BChl dimer, with ring planes perpendicular to the photosynthetic membrane and the “accessory” monomeric BChl followed by reduction of the BPh molecule [81, 82]. The initially photo-excited “special pair” remains as the oxi- + 3 dized species P · and also the triplet state ( P), formed in prereduced bRC, is localized on the same species. In PS II the corresponding reaction sequence is not fully established + yet. A central question is whether the oxidized species P · resides on the same molecule as the initially photo-excited species P∗ that initiates the primary charge transfer. Due to the strong overlap of the absorption spectra of the individual chlorins in PS II resolving this question by optical spectroscopy alone is difficult although some information has recently been obtained using specific PS II mutants [83] and analyzing FTIR spectroscopic data [84]. Information regarding the energetics of the involved molecules and, therefore, on the mechanism of the charge separation can be gained by analysis of the location of the ex- 3 + cited triplet state P and of the location of the oxidized species P ·. It has been shown + by analysis of pulse EPR techniques on the radical pair P680· QA−· [85, 86] and a comparison [67] with the PS II crystal structure that the cation radical is located on one of the chlorophylls with planes parallel to the membrane normal (PD1 or PD2) at least on the time scale of the EPR experiment (ns to µs). Rutherford and coworkers [76] have analyzed the orientation dependence of the PS II 4.2. EXPERIMENT 37 triplet EPR signal in one-dimensionally ordered PS II enriched membrane fragments at cryogenic temperatures. Initially the authors have concluded that the triplet resides on a chlorophyll with the ring plane parallel to the photosynthetic membrane [87, 88]. Later this result has been refined using improved sample preparation and a 30◦ tilt of the chlorophyll plane with respect to the membrane has been given [76]. According to the recent X ray crystal structure of PS II cc this orientation would correspond to the “accessory” chlorophylls (ChlD1 or ChlD2) in the PS II RC. This result required further support by investigation of 3P in single crystals of PS II. I will report here on such experiments and 3 analyze the ZFS tensor magnitude and orientation of the PS II triplet state P680 at low temperatures using time-resolved transient EPR.

4.2 Experiment

4.2.1 Samples

PS II cc from Synechococcus elongatus were purified according to Dekker et al [90] using a weak anion exchange chromatography in the final stage of purification of PS II cc [91]. All chemicals were of analytical grade, and triply destilled water (Millipore-Q) was used as solvent. According to SDS/PAGE and matrix-assisted laser desorption ionization-time-of-flight mass spectrometry, the PS II cc of S. elongatus are composed of at least 17 subunits [92], of which 14 are located within the photosynthetic membrane: the RC proteins D1 and D2, the heterodimeric cytochrome b559, the two chlorophyll a-binding inner light harvesting antenna proteins CP 43 and CP 47 and some smaller subunits. The membrane-extrinsic cytochrome c550, the 12-kDa, and the manganese stabilizing 33-kDa protein are located on the lumenal side of the PS II cc [93]. From this preparation, three-dimensional PS II single crystals were grown in the group of H.T. Witt and P. Fromme [8] that are active in light-induced electron transfer and water oxidation [94] and diffract to a resolution of at least 3.8 A˚ in X-ray structure analysis. The typical size of the PS II crystals obtained is

0.5mm×0.3mm×0.2mm. The crystals are of the orthorhombic space group P212121 and contain four PS II dimers per unit cell. In the crystals PS II occurs as a homodimer with non-crystallographic C2 symmetry. This non-crystallographic C2 axis is itself parallel to the pseudo C2 axis relating the cofactors of the D1 and D2 protein to each other [8]. 38 CHAPTER 4. THE LIGHT-INDUCED TRIPLET STATE IN PHOTOSYSTEM II

Sample preparation

The original PS II crystals contain a stripped but nevertheless native and functioning Photosystem II. In order to be able to generate the triplet state electron transfer had to be blocked by reducing the first electron acceptor, the quinone QA. This can be accomplished in different ways. In this work the reduction was carried out using sodium dithionite (Na2S2O4) in an inert atmosphere (Argon) to prevent dithionite from being oxidized by athmospheric oxygen. The crystals are very vulnerable towards any chemicals added to the mother liquor they are stored in. If their optimized environment is perturbed they dissolve easily. So any additions of the reactant had to be done with care. First of all the dithionite was dissolved in the mother liquor (without any crystals) prior to addition to the crystals. All solutions were flushed with argon to get rid of the dissolved oxygen. After this they were stored in a vessel under argon. The prereduction itself was done in the dark (to prevent electron transfer and keep the quinone unreduced prior to chemical reduction) and again under argon. Several concentrations of the reductant were used to find the optimum conditions under which the formed triplet state signal was at its maximum. After 20 minutes incubation time excess of the solution was removed and the crystals were placed under argon in sample tubes (Suprasil) and quickly frozen in liquid nitrogen.

Time Resolved EPR

EPR measurements were carried out at X-band (9.83 GHz) using a Bruker (Rheinstetten, Germany) ESP 380E spectrometer. An Oxford CF 935 helium cryostat was used to achieve a temperature of 10 K. Transient EPR signals were generated by excitation with laser pulses (8 ns duration, ≈10 mJ light energy per pulse) at 532 nm from a frequency doubled NdYAG laser (Spectra Physics GCR 130). From the complete time/ﬁeld data sets transient spectra were extracted by integrating the signal intensity in a chosen time window following the laser pulse. All spectra presented are plotted with absorption (A) positive and emission (E) negative.

Spectral Analysis

EPR spectra were simulated using a self-written computer program based on a spin Hamil- tonian including the electron Zeeman and the electron-electron dipolar term. The hfc 4.3. THE PS II TRIPLET STATE POWDER SPECTRUM 39 interactions were subsummed in the linewidth. Crystal spectra of the inequivalent sites, related by crystallographic (P212121) and non-crystallographic (C2 dimer axis) symme- tries. For these PS II crystals exist two C2 axes,one being the local pseudo-C2 axis relating the D1-protein and its cofactors to the D2-protein. The other one is the C2 dimer axis relating the two dimer halves to each other. These two axes are parallel to each other.

Therefore both are referred to as C2 axis. The simulation parameters include the D-tensor principal values DX, DY and DZ and the orientation of D for one site in the unit cell, the orientation of the crystal in the magnetic ﬁeld, and a linewidth parameter. The D tensor principal values DX, DY and DZ can also be represented by the values D and E since the 1 tensor is traceless. These are related to the principal values as follows: DX = 3 D + E , 1 2 DY = 3 D − E and DZ = 3 D.

4.3 The powder spectrum for the triplet state in PS II particles

In order to understand the spectra of the triplet state in PS II single crystals it is important to know how the triplet powder spectrum from PS II particles derived. Such a spectrum is depicted in figure 4.1 (recorded by Lendzian et al) [80]. This spectrum can be viewed as a superposition of two powder spectra, one from the absorptive transitions (between the ms = 0 and ms = +1 triplet levels) with only positive contributions extending from Z to X (≈320 mT to ≈373 mT) and one from the emissive transitions (between the | | ms = 0 and ms = −1 triplet levels) extending from X to Z (≈328 mT to ≈381 mT) || || with only negative contributions. Contributions with different polarization overlap in the field range X to X . Due to the fact that the two powder spectra are shifted with respect || | to each other, the resulting spectrum shows the AEEAAE pattern which is characteristic for the triplet state spectrum of PS II.

4.4 The triplet state spectra in PS II single crystals

The PS II single crystals available have a volume of only ≈ 35 nl. Taking into account the size of the unit cell of 0.9 nm3 (308A˚×227×130A)˚ with eight PS II cc per unit cell one arrives at a maximum number of ≈ 3.2 × 1013 spins in one crystal. EPR experiments on a stationary paramagnetic species could so far only be carried out with these small crystals using W-band (94 GHz) EPR that makes use of the increased sensitivity of high- ﬁeld/high frequency EPR [89]. However, for a light-induced spin-polarized species as the triplet state of PS II part of the sensitivity enhancement in high-ﬁeld EPR is lost since 40 CHAPTER 4. THE LIGHT-INDUCED TRIPLET STATE IN PHOTOSYSTEM II

Z| X|| Y|| Amplitude

Y| X| Z||

310 320 330 340 350 360 370 380 Magnetic field [mT]

Figure 4.1: Powder spectrum of the triplet state in PS II particles (D1/D1/cytb559-complexes) [80]. The spectrum itself is a superposition of two single spectra (absorptive and emissive) which are depicted dashed and dotted, respectively. Labeled are the ﬁeld positions of the single spectra, from which the principal values of the D tensor can be derived.

in this case the population difference between the spin levels is not increased at higher fields according to Boltzmann’s law but stays constant due to an exclusive population of the T0 state from the spin-polarized radical pair. Therefore, a high filling factor/high-Q dielectric ring resonator at X-band (≈10 GHz) was used to perform the experiments. By using TR EPR the strong polarization leads to a significant signal enhancement of the triplet state EPR signals. Since at low temperatures the triplet yield is high it was possible to acquire spectra. The first recorded spectra (not shown) for the triplet state in PS II single crystals had a very low signal-to-noise ration. The concentration of sodium dithionite was comparably low (the final concentration was ≈ 10µM). Even though the signal/noise ratio (S/R) was low it was clearly visible that a light-induced species had formed which shows to be dependent on the rotation angle with respect to the external magnetic field. Absorptive and emissive signals are visible which have a pairwise relation to each other. Compared to the powder spectrum shown on top of this figure the field range fits with the stretch of the signals for the single crystal spectra. Although this clearly proves that it is possible to generate the triplet state in single crystals with the method used here the low S/R prevents a quantitative evaluation. Nevertheless one can conclude that the dithionite treatment described led to reduction of QA in the single crystal and that the light-induced 4.5. SIMULATION OF THE SINGLE CRYSTAL SPECTRA 41 spectra observed can be assigned to the triplet state of PS II. Shown in figure 4.2 together with the powder spectrum for comparison are the light- induced spectra recorded after treatment of the PS II crystals with a sodium dithionite solution to a final concentration of 100µM. The S/R has improved significantly. The orientation dependent spectra show a spectral width and polarization patterns of the individual single crystal spectra consitent with the frozen solution spectrum. In the field ranges between Z and X , and X and Z , respectively, the single crystal | || | || spectra exclusively show the same polarization as the powder spectrum. Only contributions from either the absorptive or the emissive spectrum occur in these spectral ranges since there is no overlap between the transitions. In the range from X to X , where lines || | of different polarization overlap in the powder spectrum. lines with the same and with opposite polarization as in the powder spectrum are visible in the single crystal spectra.

The single crystal spectra show the required 180◦ periodicity. Therefore only one 180◦ interval in steps of 5◦ is depicted in figure 4.2. The crystal unit cell contains 4 inequivalent dimers of PS II with two monomers related by a non-crystallographic C2 axis in the unit cell. Therefore, eight magnetically inequivalent PS II complexes are present in the unit cell resulting in a maximum of sixteen EPR lines (eight absorptive and eight emissive) for an arbitrary orientation of the crystal with respect to the magnetic field. In the pattern shown in figure 4.2 at most eight lines (four absorptive and four emissive) are resolved. This implies that the rotation axis is almost parallel to one of the crystallographic axes. Under this condition two of the four PS II dimers in the unit cell are pairwise magnetically equivalent. Close to the angles 0◦ and 180◦ the magnetic field is aligned roughly parallel to one of the crystallographic axes and all four dimers are magnetically equivalent. The remaining two absorptive and two emissive transitions are due to the magnetically inequivalent monomers of one PS II dimer. Close to 90◦ another crystallographic axis is roughly parallel to the magnetic field.

4.5 Simulation of the Single Crystal Spectra

To obtain the orientation of the zero-ﬁeld splitting tensor D principal axes within the crystal axis frame, a simulation of the orientation dependent spectra had to be performed. The simulation of this rotation pattern requires 9 parameters. These are the D and E values derived from the traceless D tensor, one isotropic line width parameter, 3 angles for the orientation of the crystal axes with respect to the laboratory frame, and 3 angles for the orientation of the D tensor principal axes of one of the PS II monomers with respect to the crystal axes. The orientation of the remaining seven inequivalent PS II monomers 42 CHAPTER 4. THE LIGHT-INDUCED TRIPLET STATE IN PHOTOSYSTEM II

180

150

120 Angle 90

310 320 330 340 350 360 370 380 Magnetic Field [mT]

Figure 4.2: Light-induced triplet state spectrum in PS II crystals. The concentration of the reductant sodium dithionite had been adjusted to 100 mM. For comparison the frozen solution spectrum is also given on top. Experimental conditions: T = 10K, νMW = 9.86GHz, accumulation time = 30 min per trace. 4.5. SIMULATION OF THE SINGLE CRYSTAL SPECTRA 43

Figure 4.3: Symmetry operations within the unit cell of a PS II crystal with P212121 symmetry leading to the generation of the sites depicted in the respective pictures. For each picture rotation around one of the crystal axes leads to the generation of a diﬀerent site.

is generated by the crystallographic symmetry operations and the non-crystallographic C2 symmetry between the monomers within one PS II dimer. These symmetry constraints have been taken from the X-ray structure [8] and were used before in the analysis of ori- ox entation dependent W-band spectra of TyrD • in the PS II single crystals [89]. To reduce the number of simulation parameters the D and E values were taken from the frozen solution spectrum by Lendzian et al [80]. First overall simulations were performed with all three angles describing the orientation of the D tensor principal axes stepped through the entire sphere with a step width of

10◦. The sum of least squares (SLS) for all combinations of these angles was calculated. This procedure was done for the three possible orientations with either one of the crystallographic axes parallel to the experimental rotation axis. For the b and the c axes no clear minima for the SLS nor any resemblance to the measured spectra or polarization pattern resulted. For the third orientation, which turned out to coincide with a rotation along the crystallographic a axis, the SLS showed a minimum for a particular set of Euler angles and the resemblance to the experimental rotation pattern was striking. In general the symmetry properties of the P212121 space group do not allow to distinguish the crystallographic axes by means of EPR. The existence of a non-crystallographic C2 axis relating the dimer halves in the PS II crystal unit cell to each other allows resolution of this ambiguity [89].

A step width of 5◦ was then applied for the rotation along the a-axis over the whole range of angles to exclude the existence of another minimum for the SLS. Even though the rotation axis was close to one of the crystallographic axes, no second minimum was found. This again is a consequence of the non-crystallographic symmetry axis relating the two PS II monomers (dimer halves) to each other for each of the non-equivalent sites in the 44 CHAPTER 4. THE LIGHT-INDUCED TRIPLET STATE IN PHOTOSYSTEM II

cos 6 X Y Z C2 a -0.572 +0.525 +0.630 -0.282 b -0.122 -0.814 +0.567 -0.558 c +0.811 +0.248 +0.530 +0.781

Table 4.1: Orientation of the chloropyll triplet D tensor axes X,Y and Z in the crystal site deﬁned by the given C2 axis orientation. Note that the non-crystallographic C2 axis is parallel to the local pseudo

C2 axis of the individual PS II cc (cf. Fig 4.5). The directional cosines are given with respect to the crystallographic abc axis system. The orientations for the other sites can be obtained using the symmetry operations of the P212121 space group. crystallographic unit cell. For any given orientation of the magnetic field the resonance positions originating from the triplet states in the different halves of the dimer have to be related by the non-crystallographic symmetry axis. This imposes severe constraints on the possible orientations of the D tensor principal axes for each individual spectrum of the whole rotation pattern. This results in only one orientation of the D tensor principal axes being consistent with the entire pattern. In the vicinity of the found angles a step width of 1◦ was applied to determine the precise values. In a further refinement concerning the angle of the crystallographic a-axis with respect to the experimental rotation axis a tilt angle of 4◦ gave the best result. In the simulations this results initially in eight resolved emissive and absorptive transitions, respectively. Due to the inhomogeneous broadening, however, only four of these lines are finally seen in the experimental spectra. For the simulations shown in figure 4.4 the line width parameter was finally adjusted to properly reflect the overlap of lines originating from different inequivalent PS II monomers. The direction cosines with respect to the crystallographic abc axis system calculated from the Euler angles are listed in table 4.1.

4.6 Discussion

3 4 1 The D and E values obtained for P680 from PS II (|D| = 287 9 × 10− cm− and 4 1 3 |E| = 432×10− cm− ) [80, 95] were found to be very close to those obtained for Chl a in 4 1 4 1 frozen organic solvents (glasses) (D = 275×10− cm− and E = 36×10− cm− ) [96]. This 3 is in contrast to the results for the recombination triplet state P865 in bRCs [79], where the D and E values are reduced in magnitude as compared to monomeric 3BChl [72, 75]. This was interpreted as resulting from delocalization of the triplet exciton over a BChl a 3 3 dimer in case of P865 [75]. For P680 in PS II the triplet state is clearly conﬁned to a Chl a monomer of the electron transfer chain -at least at low temperatures- whose 4.6. DISCUSSION 45

powder spectrum X Y X Z Z Y 180

150

120 Angle 90

310 320 330 340 350 360 370 380 Magnetic Field [mT]

Figure 4.4: Top: Transient spin-polarized powder spectrum of 3P in PS II. The arrows denote prominent features in the spectrum from which the principal values of the D tensor are obtained. Bottom: TR-EPR spectra (experiment: black, simulation: dashed blue) of the light-induced triplet state in single crystals of PS II. The crystal was rotated about an axis close to the crystallographic a axis. The coloured lines represent simulations of the 8 diﬀerently oriented triplet molecules in the single crystal unit cell. Note that for each site a pair of lines is obtained (absorptive and emissive). For the simulations constant D and E values obtained from the powder spectrum and an optimized linewidth of 0.4 mT were used. 46 CHAPTER 4. THE LIGHT-INDUCED TRIPLET STATE IN PHOTOSYSTEM II

Fe Q Q A * B C2 − Axis

PhD1 PhD2

PD1 PD2 Z

ChlD1 ChlD2

YZ YD

Figure 4.5: Orientation of the principal axis of the D tensor component Z in the PS II core complex. The panel shows a view parallel to the membrane plane. The angle of DZ determined from EPR with ◦ ◦ respect to the intrinsic pseudo C2 axis is 30 2 that ﬁts to one of the accessory chlorophylls (ChlD1 or

ChlD2). Here ChlD1 has been chosen to visualize this. 4.6. DISCUSSION 47

Ph Chl D2 D1 PD2 Y

PD1 ChlD2 PhD1

Figure 4.6: Orientation of the principal axes of the D tensor components X and Y in the PS II core complex. The panel shows a view perpendicular to the membrane plane. The position of X and Y reveals the position of the fifth ring of the chlorophyll. 48 CHAPTER 4. THE LIGHT-INDUCED TRIPLET STATE IN PHOTOSYSTEM II structure seems to be slightly disturbed probably due to interactions with the protein environment. Recently this assumption has been confirmed by a pulse ENDOR analysis 3 3 of the hyperfine structure of “ P680” and Chl in organic solvents [95]. The question is to which of the chlorophylls in the crystal structure of PS II this triplet state must be assigned. This information is obtained from evaluation of the triplet axes. The analysis of the orientation-dependent EPR spectra in the single crystals of PS II yielded precise data for the orientation of the D tensor in the unit cell. Because the D tensor is closely related to the geometrical structure of the porphyrin ring of the chlorophyll, the assignment and full orientation in space of this molecule can be derived. Due to the existence of the non-crystallographic symmetry axis of the PS II dimer the orientation of the D tensor in the unit cell can be assigned to one of the four dimers in the unit cell [89]. Furthermore, since the non-crystallographic symmetry axis of the dimer is perpendicular to the membrane plane, the orientation of the chlorophyll hosting the triplet relative to the membrane normal can be derived by calculating the angles between the D tensor axes and this C2 axis (see table 4.1). This allows a comparison of the data on single crystals of PS II cc with those obtained earlier on oriented PS II membrane fragments [76]. However, the parallel alignment of the non-crystallographic dimer symmetry axis and the pseudo C2 axis relating the D1 and D2 subunits in each monomer prevents a distinction of the cofactors on the D1 and the D2 side of PS II, respectively. The Z principal axis of the D tensor is perpendicular to the plane of the chlorophyll molecule [97]. The directional cosine between the triplet Z principal axis and the pseudo

C2 axis of the PS II cc as given in table 4.1 corresponds to an angle of 30◦ 2◦. This value is in good agreement with the 29◦ angle calculated from the crystal structure between the pseudo C2 axis and the normal of the “accessory” chlorophylls ChlD1 or ChlD2 (see ﬁgure 4.5). This result corroborates the earlier ﬁnding of van Mieghem et al [76] who found an angle of 30◦ 3◦ between the triplet Z axis and the membrane normal and concluded 3 that the the triplet state P680 is located in PS II on a Chl similarly arranged as the “accessory” BChls in the bRC. The directional cosines between the X and Y axes of the

D tensor and the pseudo C2 axis given in table 4.1 correspond to angles of 62◦ 2◦ and

89◦ 2◦, respectively. The angle of 89◦ between the triplet Y axis and the pseudo C2 axis shows that the triplet Y axis is aligned parallel to the photosynthetic membrane. In order to assign the in-plane axes X and Y to the molecular structure we have to re- 3 sort to some recent ENDOR data [95]. In the pulse ENDOR study on P680 a geometric relation between the hyperﬁne tensor axes of the methine protons of the porphyrin ring and the D tensor principal axes was obtained. Since the principal axes of these hyperﬁne tensors are directly related to the molecular structure it is possible to unambiguously identify the D tensor principal axes X and Y with the molecular axes [95]. The in-plane X axis is parallel to a line connecting the carbon atoms 10 and 20 (according to IUPAC 4.6. DISCUSSION 49

5 A B N N X 20 Mg 10

N N H3 C D C H 15 H (CH 2 )2 E H COO COOCH3 O Phytyl

Figure 4.7: Picture of a Chl a molecule. The triplet X and Y axes are drawn with respect to the molecule symmetry[95]. The carbon atoms of the molecule are numbered according to the IUPAC numbering scheme. nomenclature), and the Y axis is perpendicular to it. This is visualized in figure 4.7. The assignment of the triplet X and Y axes corresponds well with the in-plane orientation of the “accessory” chlorophylls ChlD1 and ChlD2 in the present X-ray structural model [98] as shown in figure 4.6 for the case of ChlD1. Based on this assignment it is also possible to localize the position of the fifth isocyclic ring in the chlorophyll and thus the axes of the optical transition moments QX and QY [95]. This has not been possible so far with confi- dence based on the crystallographic data alone. Interestingly, the obtained arrangement of the chlorophyll macrocycle ChlD1 is very similar to that of the “accessory” BChl in bRC, and is also in agreement with the PS II RC computer model of Svensson et al [63]. The result suggests that the orientation of the other chlorophyll species in PS II might also be similar to those in the bRC. An assignment resulting from this assumption is depicted in the figure 4.6. 3 The data on the orientation of the D tensor of the P680 principal axes clearly show that this triplet state is located at low temperatures on one of the accessory Chls in the PS II reaction centre. Due to the dimeric nature of PS II in the single crystals a distinction of the two symmetrically arranged chlorophylls ChlD1 and ChlD2 is not possible. The location of the triplet state on an “accessory” Chl in PS II is remarkably different from bRC. In the bRC the triplet state is localized on the “special pair” primary donor P with ring plane perpendicular to the photosynthetic membrane [75, 99] that also carries the positive charge after photoinduced electron transfer [100]. However, in PS II it has been shown by distance measurements using pulse EPR techniques on radical pair states that the cation 50 CHAPTER 4. THE LIGHT-INDUCED TRIPLET STATE IN PHOTOSYSTEM II

+ radical P680• is localized on PD1 or PD2, the chlorophyll molecules which correspond to the “special pair” in the bRC [85, 86]. The results from EPR on wild-type PS II preparations on the location of the triplet and the cation radical state in PS II are in agreement with a recent optical study by Diner et al [83] on genetically modified PS II complexes. In the latter work replacement of the D1-His 198 and D2-His 197, which are believed to ligate PD1 and PD2 respectively, + 3 showed clear effects on the P •-absorbance difference spectrum but did not affect the P- 1 + P-difference spectrum. These results were interpreted as the cation P680• being localized 3 on PD1 or PD2 while the P680 triplet state does not reside on these Chl species. Based on the result of van Mieghem et al localization on an “accessory” Chl was proposed [76]. Assuming similar singlet-triplet splittings for all Chl molecules within the PS II reaction 3 centre the localization of the triplet state “ P680” at cryogenic temperatures on an “accessory” Chl implies that the same “accessory” Chl could also be the trap for the singlet excitation and initiate the charge separation process. In such a scenario the excited ChlD1 + would reduce PhD1, and the created primary cation radical ChlD1• subsequently oxidize + + PD1. In this model PD1 acts as an intermediary electron donor to ChlD1•. PD1• removes an electron from the active tyrosine YZ that in turn oxidizes the manganese cluster, finally leading to water oxidation. While EPR spectroscopy can not directly address the different possible models for charge separation in PS II a detailed EPR spectroscopic analysis of 3 3 the slow rise of the “ P680” state in PS II compared to P in bRC could provide additional arguments to decide between the models. Chapter 5

1 Pulsed H-ENDOR and HYSCORE • of the Stable Tyrosine Radical YD in Photosystem II

5.1 Motivation

Amino acid radicals often play a crucial role in enzyme-mediated reactions [101]. Par- ticularly the amino acid tyrosine and its derivatives which have unique redox properties and play an important role as centre for bond-building, bond-breaking or as electron- transfer cofactor [102–104]. It is possible to oxidize tyrosine in aqueous solution at neutral pH. During oxidation the phenoxyl proton is released and a neutral tyrosine radical is formed [105, 106]. The redox potential Em = 0.93 V vs. NHE of the neutral radical is one of the lowest among those of the 20 common amino acids [105, 106]. This is the reason why neutral tyrosine radicals in proteins have relatively long lifetimes without them oxidizing their protein environment. Therefore these radicals are often found in high potential biological enzyme systems.

Photosystem II contains two symmetry-related redox-active tyrosine residues YD (D2-

Tyr160) and YZ (D1-Tyr161). The oxidized radicals are generated by the action of the + oxidizing species P680, the photogenerated cation of the primary donor of PS II. Of these two amino acids the function of YZ in the electron transfer process is straightforward. It mediates the electron transfer between the tetranuclear manganese cluster (oxygen evolv- + ing complex, OEC) and the photooxidized chlorophyll moiety P680[5]. There are also speculations that YZ is directly involved in water oxidation as a photogenerated base and

51 1 52 CHAPTER 5. H-ENDOR AND HYSCORE OF YD• H H

Ha 2 3

¡ ¡ ¡ ¡ ¡ ¡ ¡

¢¡¢¡¢¡¢¡¢¡¢¡¢¡¢ ¡ ¡ ¡ ¡ ¡ ¡ ¡

¢¡¢¡¢¡¢¡¢¡¢¡¢¡¢ 1 . 4 ¡ ¡ ¡ ¡ ¡ ¡ ¡

¢¡¢¡¢¡¢¡¢¡¢¡¢¡¢ C O

¡ ¡ ¡ ¡ ¡ ¡ ¡

¢¡¢¡¢¡¢¡¢¡¢¡¢¡¢

¡ ¡ ¡ ¡ ¡ ¡ ¡ ¢¡¢¡¢¡¢¡¢¡¢¡¢¡¢ Protein 6 5 Hb H H

• Figure 5.1: Schematic representation of YD in PS II of spinach including numbering scheme within and outside of the phenolic ring. The delocalization of the electron over the ring system and the oxygen atom is symbolized by dotted lines.

abstracts protons from water bound to the manganese cluster [30, 107].

In contrast, YD is bypassed in the fast electron transfer between the manganese cluster + and P680. It is typically present as a dark stable neutral radical (YD• ) and slowly under- goes redox reactions with the lower oxidation states of the OEC in the dark by which it is reduced back to YD [37].

Since YZ and YD reside at homologous positions of the D1 and the D2 protein, respectively, which themselves form a 1:1 complex with C2 symmetry (similar to the L and M subunits in bRC), they are expected to be symmetrically placed about the primary donor and the differences in the observed oxidation rates must arise from factors such as their orientations, environments and the proximity of the OEC. Tyrosine radicals are paramagnetic. They can be studied with EPR and ENDOR spectroscopy. The spectrum of YD• in PS II is broadened by proton hyperfine interaction which at X-band is only partially resolved. The signal can be observed in the dark since after the radical is generated by illumination it is stable at liquid nitrogen temperature for months. The EPR spectra of YD• in several organisms are similar, and for this reason data from different organisms have been used interchangeably. Several attempts have been made in the past to resolve to complex hyperfine structure implied by EPR spectra of YD• . Nevertheless hyperfine data for YD• presented in the literature show remarkable variations [30, 108–111, 114].

In order to obtain hyperﬁne coupling constants for YD• , ENDOR and ESEEM are the methods of choice. In this case a proton with a strong anisotropic hyperﬁne coupling is present. This leads to broad and overlapping signals in the ENDOR spectrum. This problem can be overcome using 2D-ESEEM, particularly by HYSCORE, as is shown in the following sections. 5.2. EXPERIMENT 53

5.2 Experiment

5.2.1 Samples

The samples used in this work were prepared from fresh market spinach according to the method described by Berthold/Babcock/Yocum (“BBY”) [9] with the modiﬁcations described in [115]. Sucrose at a concentration of 0.4 M was used as cryoprotectant. The ﬁnal concentration achieved by centrifugation was ≈20 mg(Chl)/ml. To get the maximum signal of the tyrosyl radical the sample was illuminated at 200 K for 10 min. At these conditions the Photosystem II turns into the S2 state. To keep a maximum of the YD• but with the PS II in the S1 state the frozen sample was stored in the dark at 30◦C for 15 h.

This resulted in the S2 state relaxing back to the S1 state.

5.2.2 cw EPR

The samples were checked at a temperature of 8.5 K with cw EPR at X-band to evaluate the signal-to-noise ratio of the YD• EPR signal and the redox state of the manganese cluster. A Bruker ESP 300 spectrometer equipped with an Oxford cryostat ESR 9 was used.

1 5.2.3 Pulsed H-ENDOR of YD•

Pulsed ENDOR experiments were performed on the concentrated and illuminated samples described above at a temperature of 5 K at X-band (9.75 GHz) frequency. A Bruker ESP 380E spectrometer equipped with an ESP 360D-P pulsed ENDOR accessory and an ENI A-500 RF ampliﬁer was used. A Davies pulse sequence with an RF pulse of 8µs duration was applied. The spectra were recorded at 347.5 mT, which corresponds to the echo maximum of the ﬁeld swept electron spin echo.

5.2.4 HYSCORE Spectra of YD•

The same sample was measured at 5 K with the PS II in the S1 and in the S2 state, respectively. A Bruker ESP 380 spectrometer (resonator: dielectric saphire ring, ESP 380- 1052 DLQ-H) attached to an Oxford CF 935 helium cryostat was used.

HYSCORE spectra were recorded using the sequence “π/2 − τ − π/2 − t1 − π − t2 − 1 54 CHAPTER 5. H-ENDOR AND HYSCORE OF YD•

3' 3 2 Amplitude

1 4' 5' 5 4

-12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 ν _ ν RFRadio frequencyH [MHz] [MHz]

1 • Figure 5.2: H-ENDOR spectrum for the YD radical in PS II from spinach. The frequency of the free proton is 14.8 MHz. Corresponding hyperﬁne coupling features are grouped symmetrically around

0 MHz. Experimental conditions: T=4.5 K, B = 347.5 mT, νMW = 9.75 GHz.

feature hyperﬁne coupling value [MHz] assignment to proton(s) 1 31.6 7a (A ) || 2 25.6 7a (A ) ⊥ 3,3’ 20 3 or 5 (AZ )

4,4’ 18.6 3 or 5 (AZ )

5,5’ 4 2 or 6 (AX )

Table 5.1: Hyperﬁne values and assignment to the protons of the tyrosine molecule according to the spectrum in ﬁgure 5.2. The assignment is further corroborated by the works from Rigby et al [111] and Hofbauer [112]. 5.3. ANALYSIS 55

π/2 − echo”, where the echo was measured as a function of t1 and t2. The duration of the π/2 and π pulses was 16 and 32 ns, respectively. The resolution in the time domain was 256 × 256 for the spectrum of the PS II in the S1 state where a τ-value of 120 ns was used and 512 × 512 for the system in the S2 state with a τ-value of 112 ns. The variables t1 and t2 were incremented in steps of 16 ns from their initial value. The spectra were recorded at B0=3475 G for νMW=9.72 GHz at echo maximum with respect to ﬁeld swept electron spin echo.

5.3 Analysis

5.3.1 Pulsed ENDOR

Figure 5.2 shows the frequency corrected Davies ENDOR spectrum measured on a concentrated (c ≈ 20mg(Chl)/ml) sample of PS II membrane fragments. Due to the very small g anisotropy of YD• (gx = 2.00767, gy = 2.00438, gz = 2.00219) almost no orientation selection eﬀects are to be expected.

For a given effective hyperfine coupling constant a and a nuclear Zeeman frequency νn, the ENDOR resonance condition is νENDOR = |νn a/2|. In this case, the proton Zeeman frequency is νn = 14.8 MHz. This frequency has been subtracted from the experimental radio frequency values in figure 5.2 to adjust the center of the spectrum to 0 MHz to acquire the hyperfine coupling values. The spectral features in figure 5.2 correspond to

|a| < 2νn are therefore aligned symmetrically about this frequency. There are two β-protons bound to the carbon atom in para-position to the hydroxylic group of the tyrosine. So one could expect two strong dipolar coupling tensors in the high-field region of the spectrum. According to the coupling scheme observed by Hof- bauer et al [89, 112] apparently only one of the β protons at the phenolic ring contributes to the spectrum. Obviously the other one lies in the plane of the ring and has no overlap with the π electron density of the phenolic ring. Contributions from this proton are assumed in the proton matrix region [111, 113]. Its relatively small anisotropy shows, that the dipolar part contributing to this hyperfine interaction is not the dominant part. The by far largest contribution stems from the isotropic Fermi-contact interaction with the spin density of the phenoxyl ring. This spin density is supposed to be conveyed by hyperconjugation [40, 89]. As in earlier works its value is in agreement with an assignment to one of the β-protons (7a) at the carbon atom in para-position to the phenolic oxygen in figure 5.1. The counterpart on the low-frequency part of the ENDOR-spectrum is not observable, since it would lie in a region of the spectrum (below 1.5 MHz) which is not 1 56 CHAPTER 5. H-ENDOR AND HYSCORE OF YD•

30 Kammel

300000

20 280000

10 260000

240000 0 Frequency [MHz]

220000 -10

200000

-20

180000

-30 0 5 10 15 20 25 30 Frequency [MHz]

• Figure 5.3: HYSCORE spectrum presented as frequency-domain spectrum for the YD radical in PS II from spinach with the manganese cluster in the S1 state. Shown are the [+,+] (top) and the [+,-] (bottom) quadrants. Experimental conditions: B0=3475 G, νMW=9.72 GHz, τ=120 ns, T=5K. 5.3. ANALYSIS 57

Kammel 300000 30

20 250000

200000

0 Frequency [MHz] 150000

-10

100000

-20

50000 -30

0 5 10 15 20 25 30 Frequency [MHz]

• Figure 5.4: HYSCORE spectrum presented as frequency-domain spectrum for the YD radical in PS II from spinach with the manganese cluster in the S2 state. Shown are the [+,+] (top) and the [+,-] (bottom) quadrants. Experimental conditions: B0=3475 G, νMW=9.72 GHz, τ=112 ns, T=5K. 1 58 CHAPTER 5. H-ENDOR AND HYSCORE OF YD• easily accessible by this method. Two further features outside the matrix region are visible in this spectrum. At ν =≈9.8 MHz and ν =≈10 MHz there are two peaks which lie very close to each other. They have their counterparts at ν =≈ −9.8 MHz and ν =≈ −10 MHz in the low frequency region. The “split peak” they form stretches vastly into the matrix region on one side and merges with the signal for the aforementioned axial β-proton tensor on the other. This leads to the assumption that these two features are the middle components of two very anisotropic, rhombic hyperfine tensors. Since they lie very close to each other they should belong to two very similar protons within the phenolic ring of the tyrosine, which are either the 3,5-protons or the 2,6-protons. The spin density for the tyrosine radical is particularly high in ortho-position to the oxygen. This aspect leads to the conclusion that only the protons at these positions are responsible for these hyperfine couplings. The result of the high spin density is also corroborated by theoretical studies [116, 117]. The other components of these two tensors are not visible in the ENDOR-spectrum. According to an earlier work [89] related to a different organism the larger components should show up as shoulders in the region of ≈ 11 MHz with a similar splitting. The smaller components are congested in the matrix region of the spectrum. It was concluded that the small splitting of the observed middle components of the hyperfine tensor of the two equivalent protons (3 and 5) is mostly caused by a slightly different spin density [89]. Close to the proton matrix region at 2 MHz one coupling component, that has been assigned to AX [111] of the hyperfine tensor of the equivalent proton in ortho-position (2 and 6) is visible. The missing components of this tensor are low in intensity and overlap with contributions from other proton hyperfine couplings close to the proton matrix region. The hyperfine couplings that could umambiguously be assigned to tyrosine protons are listed in table 5.1. Although ENDOR is usually a feasible technique to find out 1H couplings in organic radicals it is not really successful in this case. Here the strong anisotropic α−1H couplings cause a large spectral extent. Signals overlap or stretch far into to matrix region.

5.3.2 HYSCORE

Hyperfine sublevel correlation spectroscopy, called HYSCORE, offers the opportunity to separate different and overlapping hyperfine tensors in ENDOR or 1D-ESEEM spectra from each other. The 2D-character of HYSCORE-spectra with the generation of off- diagonal cross-peaks belonging to opposite electron spin manifolds allows to disentangle complicated hyperfine pattern. The qualitative and quantitative assignment of features in the spectra is a first step in 5.3. ANALYSIS 59

understanding the hyperﬁne structure of the YD• radical. For the case of an isotropic g-tensor Dikanov, Bowman and Tyryshkin derived analytical expressions for the shape and the intensities of the features of the related HYSCORE spectrum [119, 120]. These works are applicable for the case of YD• with its g-tensor (gx = 2.00767, gy = 2.00438 and gz = 2.00219 [89, 118]) which is suﬃciently isotropic at X-band.

5.3.3 Assignment of the observed couplings to the protons of the tyrosine radical

For the assignment of the proton couplings the HYSCORE spectrum of YD• with the PS II residing in the S2 state depicted in figure 5.4 has been chosen since the signal-to-noise ratio is by far better than for the S1 state in figure 5.3. This is probably due to different τ value used to record both spectra. The method of assigning the hyperfine values is described in the theory chapter and is also used in the work by Deligiannakis et al [118]. The values extracted from the spectrum shall later be used together with the assigned hyperfine coupling from the ENDOR spectrum to simulate the theoretical HYSCORE spectrum. Figure 5.4 shows the [+,+]- and the [+,-]-quadrants of the experimental HYSCORE spectrum. In the [+,+]-quadrant corresponding signals are grouped along the antidiagonal and passes through the Larmor frequency. In the [+,+]-quadrant this must be the cou- 1 plings of protons having an isotropic coupling part with Aiso < 2νI ( H) = 29.6 MHz. At the outer edges of the antidiagonal there are two very strong peaks at the positions ≈ [2 MHz, 28.5 MHz] and ≈ [28.5 MHz, 2 MHz]. These signals belong to a proton with a hyperfine coupling near the matching condition. In the ENDOR spectrum these couplings lie within the proton matrix region and can not be identified. The hyperfine coupling tensor has a small anisotropy, since there is almost no deviation from the antidiagonal. The projected length of the arcs is ≈ 2.5 MHz which gives T≈ 1.7 MHz. The centre-to-centre-distance of the two peaks ranges from 2 MHz to 29 MHZ which means a difference of 27 MHz. This proton with the approximate hyperfine coupling tensor of [25.3 MHz, 25.3 MHz, 30.4 MHz] has been assigned in the literature [118, 121] to one of the β-protons, in figure 5.1 it would be the Ha protruding out of the ring plane. There are striking features deviating approximately ∆ =2 MHz in projection from the antidiagonal. They belong to a rhombic tensor with strong anisotropy. The projected edge-to-edge distance is 7 MHz, which corresponds to T = 4.7 MHz. The centre-to-centre distance is ≈ 17 MHz, leading to a hyperfine tensor of ≈[10.3 MHz, 14.3 MHz, 24.4 MHz]. It is assigned to the α protons in position 3 and 5 of the tyrosine. 1 60 CHAPTER 5. H-ENDOR AND HYSCORE OF YD•

Hyperﬁne values Protons x y z 3,5 9 14 27 7a 26 26 32 2,6 4 4 8.5

Table 5.2: Hyperﬁne values yielded by a graphical evaluation of the HYSCORE spectrum in ﬁgure 5.4 according to the method by Deligiannakis et al [118].

Close to the inner edge of the aforemetioned rhombic tensor there is a third pair of peaks.

Their centre-to-centre distance is about 8 MHz which results in an Aiso of ≈ 5 MHz. Its anisotropy is very low. This weak coupling corresponds to the α-protons in positions 2 and 6. At these positions the electron density is small. It could also belong to the other β-proton which lies within the plane of the tyrosyl ring where the electron density is also small. The values extracted from the HYSCORE spectrum with the method described by Deligiannakis et al [118] corroborate the ﬁndings of this group with only small deviations.

5.3.4 The program used for the simulation of the HYSCORE spectra

The simulations of the HYSCORE spectra have been performed with a program for simulating ESEEM spectra, described in reference [122]. The program has been adapted to incorporate the modulation formula for HYSCORE [58–60] and to store the modulation amplitudes in a 2-dimensional instead of a 1-dimensional array. The program calculates the HYSCORE spectrum at a fixed magnetic field setting and microwave frequency in the following way. From all possible orientations of the molecules in the frozen solution, only those are selected for which the effective g value is such that the resonance condition is fulfilled. The resonance condition is fulfilled if the effective g value (geff ) determined by the fixed magnetic field setting and microwave frequency are within the intrinsic linewidth of the spectrum of the molecule. Parameters such as the linewidth can be taken from the spectrum and used in the simulation procedure. The molecule is assigned a weight, i.e., how much it contributes to the HYSCORE spectrum, determined by geff − gres and the effective linewidth. Subsequently, the matrix which represents the nuclear spin Hamilto- nian is computed and diagonalized. The eigenvalues and eigenvectors are used directly in the modulation formula for the HYSCORE experiment to compute the modulation frequencies and amplitudes. These are stored in a 2-dimensional array in the frequency domain and assigned a linewidth of 70 kHz. When all possible orientations have been 5.3. ANALYSIS 61

30 25 a b 20 15 10 5 0

[MHz] 30 2 ν 25 c d 20 15 10 5 0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 ν1 [MHz]

• Figure 5.5: Simulated single contributions of protons to the HYSCORE spectrum of YD in PS II introducing the protons in the positions 3 and 5 (b), the β proton at position 7a (c) and the protons in the positions 2 and 6 (d). The experimental spectrum (a) is shown for comparison. Simulation parameters: τ=112 ns, B0=347.8 mT, νmw=9.72 GHz. considered (the orientational averaging includes typically 10000 orientations), the array can be compared directly with the experimental spectrum.

5.3.5 Simulation of the theoretical HYSCORE spectrum of YD•

In ﬁgure 5.5 a simulation of single contributions of the protons of YD• to the HYSCORE spectrum is depicted. The simulation includes the stronlgy anisotropic α-protons at the positions 3 and 5, the β-proton at position 7a, and a set of parameters that can be assigned to the α-protons at the positions 2 and 6. For the protons at the positions 3 and

5 as well as for the protons at positions 2 and 6 an angle of 20◦ between AY and gY has been assumed which is in accordance with the findings of other researchers [89, 108]. For comparison the experimental spectrum is given also. Shown in figure 5.6 is the complete simulated spectrum with the hfc’s of all contributing protons. The simulation parameters are given in table 5.3 and compared to the hyperfine 1 62 CHAPTER 5. H-ENDOR AND HYSCORE OF YD•

300

250

20 200

15 150 Freq [MHz]

10 100

50 5

0 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 Freq [MHz]

• Figure 5.6: Simulated resulting HYSCORE spectrum of YD in PS II showing all contributions from all protons and their corresponding correlations. Simulation parameters: τ=112 ns, B0=347.8 mT,

νmw=9.72 GHz. coupling values found by other groups.

5.4 Discussion

In the simulation of the HYSCORE spectrum of YD• all prominent features of the experimental spectrum could be reproduced. In the evaluation of the quality of the simulation parameters the data extracted from the 1H-ENDOR spectra are important benchmarks. All spectra have been measured on the same sample, and therefore the hyperfine data should be the same. With the collection of two datasets acquired with two different methods the result is much more reliable. This is important since there are some deviations from the data obtained by other groups. The β proton at position 7a yields an axial hyperfine tensor. The values for A and A derived from the ENDOR spectrum as || ⊥ well as from the graphical evaluation of the HYSCORE spectrum are almost identical. In the simulated HYSCORE spectrum the small contribution of this proton could be reproduced with these values within 1 MHz. For the value ofA there are only deviations ⊥ 5.4. DISCUSSION 63

direction cosines lij Hyperﬁne Deligi- Hofbauer Rigby et Farrar et Protons values Ai x y z annakis et et al. al. [111] al. [110] [MHz] al. [118] [89, 112] (spin.) (spin.) (spin.) (S. elon.) −25.6, 3,5 Ax =26.1 0.940 0.342 0.000 26.0 23 -26.1 −27.5 18.8 Ay =8.1 0.342 0.940 0.000 15.0 -8.0 -8 20.3 19.1, 12.3 Az =19.5 0.000 0.000 1.000 10.0 -19.5 20.5 14.8

7a Ax =32.8 1.000 0.000 0.000 31.7 31.5 30.5 32.8

Ay =25.8 0.000 1.000 0.000 26.6 27.2 27.0 27.2

Az =25.8 0.000 0.000 1.000 26.6 27.2 27.0 27.2 4.4, 2,6 Ax =9.3 0.940 0.342 0.000 8.8 4.8 7.15, Ay =5.5 0.342 0.940 0.000 3.5 7.4

Az =5.5 0.000 0.000 1.000 3.5

• Table 5.3: Principal values of the hfc tensor and direction cosines of the three proton species of YD with respect to the g principal axis system used in the simulation shown in figure 5.5 and to give the best result compared to the HYSCORE spectrum in figure 5.4. Also given for comparison are the hyperfine values found by other groups. Written in parentheses is the source of PS II; spin.: spinach, S. elon.: Synechococcus elongatus. The PS II samples measured by Deligiannakis et al. [118] had been depleted of manganese. 1 64 CHAPTER 5. H-ENDOR AND HYSCORE OF YD• of at most 1.4 MHz compared to the results by other groups listed in table 5.3. They all got very similar values for A , which yields a very prominent signal in the ENDOR ⊥ spectrum. Rigby et al.[111] showed that the β-proton coupling is species dependent while the ring proton coupling constants are not. As shown in table 5.3 the hfc’s found for spinach PS II do not differ from the results of Hofbauer et al. [89, 112] who measured a PS II sample from Synechococcus elongatus. The value for A deviates in all works || since in the ENDOR spectra it results only in a low intensity shoulder. The HYSCORE simulation in this work reproduced the corresponding feature assuming a value for A , || which is approximately 1 MHz smaller. The ring protons at the positions 3 and 5 have highly anisotropic hyperfine coupling tensors. These protons are magnetically almost equivalent. Due to a small difference in the spin density distribution they show a small splitting for one component of about 1.4 MHz. In the 1H-ENDOR spectrum only this hyperfine coupling component can be figured out unambiguously at 20 and 18.6 MHz, respectively. The graphical evaluation of the experimental HYSCORE spectrum according to Deligiannakis et al. [118] yields very similar values compared to the values in their work. In the HYSCORE simulation the signals that belong to these protons could be reproduced using the values obtained by Hofbauer et al. [89, 112]. No value for AZ at ≈19 MHz resulted. For the AX and AY values there is a good agreement in all works with the exception of Farrar et al. [110] and Deligiannakis et al [118]. The simulation of the contributions from these two protons yields features in the simulated HYSCORE spectrum that have a good resemblance to the respective part of the experimental spectrum. The hyperfine coupling values that were found for these protons are in accordance with the works of Rigby et al [111] and Hofbauer [112]. The deviation to the values found by Deligiannakis et al [118] who also used used HYSCORE spectroscopy is remarkable. The spectrum recorded by these workers is very similar to the spectrum presented here. The graphical evaluation of both spectra according to Deligiannakis et al.[118] yields similar results, that are neither reproduced in the HYSCORE simulation presented here nor by other groups. Not many hfc data exist for the protons at the positions 2 and 6, since their contributions to the 1H-ENDOR spectrum stretch into the matrix region or overlap with contributions from other protons. Nevertheless the line at 4 MHz has been assigned to one hfc component of the 2,6-ring protons. These protons deliver only weak contributions to the HYSCORE spectrum. The value of 4 MHz could not be reproduced in the HYSCORE simulation. Although the value for AX is in good agreement with the result of Deli- giannakis et al. [118] the values for AY and AZ have to be quite large with 5.5 MHz to reproduce the respective feature in the spectrum. In the ENDOR spectra the line orig- inally assigned to the 2,6-protons is very close to the proton matrix signals. In these 5.5. CONCLUSION 65 spectra there is intensity in the region of 5.5 MHz, that can account for a hyperfine value of the 2,6-protons.

5.5 Conclusion

The magnetic interactions of protons in YD• of PS II have been investigated using HYSCORE, and the measured spectra have been simulated using a program for the simulation of HYSCORE spectra. The acquired hyperfine values have been compared with data provided in the literature for YD• [89, 110, 111, 118]. In this work it was possible to disentangle the highly complicated coupling situation of the tyrosine radical in PS II. In the pulsed ENDOR spectrum the hyperfine coupling tensors superimpose each other and make it very difficult to elucidate their values. Also hyperfine couplings close to the proton matrix region are superimposed by other tensor components and therefore hardly accessible. Due to the two-dimensionality of the HYSCORE experiment and the excellent signal/noise ratio of the spectrum the whole set of hyperfine coupling values could be derived for the case of YD• in PS II. Owing to the fact that crosspeaks of anisotropic tensors lead to very prominent signals, further assessment of the spectrum is simplified. This is the case for the anisotropic α-protons at the positions 3 and 5 in YD• . The presented work also shows that the assignment of hyperfine values only based on a graphical evaluation leads to slightly different parameters. For example, the derivation of the hyperfine values with the method given by Deligiannakis et al [118] provides in the case of the very anisotropic 3,5- protons deviating results. Here the data derived from the 1ENDOR-spectrum are more precise. Therefore a simulation of the HYSCORE spectra seems necessary. Only here the full magnetic interaction of all protons of the molecule can be calculated and assessed. To find good starting parameters ENDOR is a valuable tool. Combination of these two methods, ENDOR and HYSCORE, provides complementary data for an understanding of the magnetic interactions of paramagnetic species with many hyperfine interactions. 1 66 CHAPTER 5. H-ENDOR AND HYSCORE OF YD• Chapter 6

1 Pulsed H-ENDOR of the S2-State and the S−2-State of the Manganese Cluster in Photosystem II

6.1 Motivation

The oxygen-evolving complex (OEC) of PS II, where the oxidation of water to molecular oxygen takes place, probably consists of a cluster of four manganese ions ligated by amino acids and other cofactors, for reviews see[5, 13]. Although much effort has been made to gain knowledge on structure and function of PS II direct information about the manganese cluster is scarce. The recent X-ray structure[8, 123] with a resolution of 3.8A-3.6˚ A˚ reveals a lot about protein folding and the structure of PS II but has not a sufficient resolution to obtain a detailed picture of the manganese cluster. As the OEC passes through the Kok-cycle it takes on oxidation states where it changes its electronic properties making some states accessible to EPR. These native paramagnetic states are the S2 state whose EPR-signal was discovered by Dismukes and Siderer [124] and the S0 state first measured with EPR by Messinger et al. and Ahrling˚ et al. [125, 126].

The S2 is still the easiest state to generate and therefore most of the EPR work until now has been done on it. However, for standard cw EPR various problems arise. The interpretation of the EPR spectrum of this state, characterized by a g=4.1 signal and a highly structured multiline signal (MLS) around g=2, is diﬃcult because of the complex spin coupling in the manganese clusters. Other paramagnetic species are suggested that originate from higher spin states [127]. The origin of these signals is under debate although

67 1 68 CHAPTER 6. H-ENDOR OF THE S 2 AND THE S2 STATE − agreement exists that the dominating contribution for the MLS comes from a mixed valent

Mn(III)Mn(IV) group leading to an Seff =1/2 ground state[128, 129]. The number of lines and the spectral width has been explained by several groups assuming different coupling schemes and hyperfine parameters [14, 130–136]. Due to the vast number of simulation parameters and the lack of reliable facts about the coupling situation and coupling parameters of contributing atoms these approaches did not lead to unique information. ENDOR provides the means to investigate the MLS in more detail. The MLS consists of hyperfine coupling contributions not only from 55Mn but also from 1H and probably other nuclei. Protons are ubiquitous in proteins. Additional to their presence in amino acids protons are also part of water molecules. These must intrude into the OEC to maintain the function of PS II, since they deliver the electrons required in the charge separation process. Proteins like CP 33 probably act as a sluice for water molecules and guide them to the OEC. They also shield the manganese cluster from surplus water. Water must come near to the manganese cluster since it has to interact directly with the manganese ions in the redox reactions of the Kok-cycle. This proximity could give rise to a coupling between the water protons and the manganese cluster in its paramagnetic states. 1 H-ENDOR measurements of the S2 state of PS II have been carried out earlier [137–140] but only proton hfc’s very close to the proton matrix region have been observed. It is not clear whether the observed signals are attributable to the MLS at all and not to the nearby EPR signal of YD• . Some of the couplings vanished when the samples were treated with heavy water D2O which showed that the respective protons are exchangable by deuterons. Because their coupling parameters are very small these protons are hardly in the proximity of the manganese ions of the OEC. If these protons belong to water molecules they would not be directly bound to manganese.

Within the Kok-cycle of the OEC of PS II the so called overreduced states of the manganese cluster do not occur. The most reduced state in this cycle is the S0 state. It is possible to reduce the OEC chemically, even to lower oxidation states than those that occur in the Kok-cycle. If the manganese cluster gets reduced chemically the reached states are thought to be the native states which are identical to the ones involved during the assembly of the OEC. The exploration of further reduced species of the OEC is of great interest in order to determine the lowest oxidation state of the manganese cluster without desintegration of the OEC. The most reduced state is thought to be accessible by treatment with reductants such as hydrazine, hydroxylamine or nitric oxide. This state would then be assumed to consist purely of manganese ions of oxidation state II. It is assumed that the overall oxidation number of the manganese cluster changes by one per turnover ﬂash. If it would be possible to reach this most reduced state chemically 6.1. MOTIVATION 69 and reenter the Kok-cycle by light ﬂashes one could derive the oxidation states of the manganese cluster in the Kok-cycle. Knowledge about the mechanism of water oxidation and particularly about the assembly of the manganese cluster during the photoactivation would be gained. This assembly is a light driven stepwise reaction which involves the binding of Mn2+ ions and their subsequent oxidation. It is characterized by two distinct light-driven steps sep- arated by a dark interval. First one Mn2+ ion binds under illumination to the apo-PS II reaction centre protein where it becomes photooxidized. This is followed by a step that requires no light where a Calcium ion is introduced. The binding and photooxidation of the next Mn2+ needs light [23, 143]. The kinetics of the two other Mn2+ ions which are supposedly required to form the functioning complex are bound later in a until now not understood way.

States of the OEC below the oxidation state of the S0 state have been accessible for quite a time now. The reductants hydroxylamine or hydrazine can be used to generate the

S 1 [144], the S 2 [145] and the S 3 state which is the most reduced state of the OEC − − − that could be unambiguously identiﬁed so far [146] by O2 electrode measurements. EPR signals have been reported and attributed to overreduced states of the OEC [147,

148]. The S 2 state of the OEC should be paramagnetic according to two one-electron − reductions coming from the S0 state which is reported to be paramagntic and to show an EPR signal with methanol added to the PS II suspensions [125, 126]. Since methanol binds very close to the manganese cluster [181] it is assumed that it changes the magnetic interaction of the manganese ions, and therefore a paramagnetic state detectable by EPR is induced. This state is also accessible by reduction with nitric oxide (NO) and shows a multiline EPR signal [149] only after this reduction method. The signal originates from the very slow interaction, which is presumably a reduction, of NO with PS II at −30◦C. Particular conditions are required to generate this signal by incubation with NO. The EPR signal is lost after a short time at room temperature but reappears after reincuba- tion for 30 min at −30◦C in the dark, even without further addition of NO. This signal was assigned to the S 2 state also by oxygen flash pattern measurements and is accessible − in PS II from different organisms [149, 151, 152]. The related EPR spectrum shows similarities to EPR spectra from dinuclear Mn(II)- Mn(III) model complexes and was assigned to the S=1/2 ground state of a magnetically isolated Mn(II)-Mn(III) dimer [149, 150]. It was shown that the reduction by NO occurs via the intermediate S 1 state [153]. − The EPR multiline signal of the S 2 state has been simulated assuming a Mn(II)-Mn(III) − dimer in the unoriented powder type as well as in oriented samples [150]. Here a set of hyperfine parameters assuming axial symmetry and collinear g and A tensors led to a good resemblance of the simulated with the experimental spectrum. It was suggested 1 70 CHAPTER 6. H-ENDOR OF THE S 2 AND THE S2 STATE − that the Mn(II)-Mn(III) dimer is one of the di-µ-oxo bridged units that are supposed to make up the manganese cluster, which is believed to be a tetramer. The simulation parameters of the S 2 multiline signal are similar to those of the Mn(II)-Mn(III) catalase, − which suggests structural similarities between these two species [149, 154].

The S 2 state mulitiline signal has also been measured on samples where exchangable − protons had been replaced by deuterons. This resulted in a change and narrowing of lines of the according multiline signal [152]. Exchangable protons should be in close proximity to the Mn(II)-Mn(III) dimer leading to this narrowing. 1H-ENDOR provides the tool to investigate the occurence of these protons near this dimer. 1 The ﬁrst H-ENDOR for the S 2 state of PS II in protonated as well as in deuterated − samples is reported here.

6.2 Experiment

6.2.1 Samples for the S2 state

The samples used in this work were prepared from fresh market spinach according to the method described by Berthold, Babcock and Yocum (“BBY”) [9] with the modiﬁca- tions described in [115]. Sucrose at a concentration of 0.4 M was used as cryoprotectant, methanol was added to get a maximum of the S2 state MLS after illumination. The ﬁnal concentration was adjusted to ≈ 15 mg(Chl)/ml.The samples were illuminated at 200 K for 20 min.

6.2.2 Samples for the S 2 state −

PS II membranes were prepared from market spinach by standard procedures described in [9, 155]. The samples were stored in a buﬀer containing 0.4 M sucrose, 15 mM NaCl,

5 mM MgCl2 and 40 mM MES in liquid nitrogen prior to the treatment with nitric oxide. For the NO incubation the samples were thawed and brought to a ﬁnal concentration of

5-6 mg(Chl)/ml. In the case of D2O samples the buffer was prepared with D2O (99.9 %, euriso-top), and the samples were washed twice with D2O buffer, then incubated for 3 h on ice and finally washed once more in D2O buffer.

NO treatment was carried out anaerobically in dim green light at 0 ◦C in 4 mm EPR quartz tubes by slowly bubbling NO through the samples. This procedure took ≈ 1 min, and the ﬁnal NO concentration in the samples was estimated on the basis of the known solubility 6.3. RESULTS 71 of NO in water (3 mM at 0 ◦C) to be 0.6 mM. The EPR tubes were then immediately sealed and transferred into liquid nitrogen. this was followed by an overnight incubation in a freezer at −30 2◦C. Then the samples were checked with cw EPR whether they showed the S 2 multiline signal. To remove the strong NO EPR signal the samples were − thawed again and bubbled with argon to get rid of the NO. This was again followed by incubation in a freezer for 5 h at −30 2◦C.

6.2.3 1H-Davies-ENDOR

Field swept ESE and pulsed ENDOR experiments were performed at T = 4.5 K at X band (9.5 GHz) microwave frequencies for both types of samples. A Bruker ESP 380E spectrometer, equipped with an ESP 360D-P pulsed ENDOR accessory and an ENI A-

500 RF ampliﬁer was used. Davies’ pulse sequence [47] with a tRF = 8µs RF pulse was applied.

6.3 Results

6.3.1 Field swept ESE of the S2 state

Figure 6.1 shows the field swept ESE spectrum of the S2 state at 4.5 K in a frozen solution of PS II membrane fragments with methanol added (3%). ENDOR-spectra had been recorded at the field positions marked with arrows in figure 6.1. The arrows numbered

1 and 3 indicate positions ≈14 mT away from the prominent EPR-signal of YD• . These

field positions had been chosen to avoid contributions from YD• in the ENDOR spectra but still gain a sufficient ESE amplitude. The arrow numbered 2 indicates a position directly on the YD• . The ENDOR-spectrum should therefore contain contributions from both paramagnetic species. The field position of arrow 4 is outside the field region of the MLS signal. The field-swept ESE spectrum for the MLS shows the typical width of ≈ 250 mT. This is in accordance with the width of cw spectra. Since the spectrum has been recorded in absorption the very prominent modulations for the hyperfine lines in the cw spectrum appear only as shallow wiggles superimposed on a broad Gaussian-type line. In the low- field part of the spectrum where the signal appears to fall to the baseline very sudden, whereas on the high-field side the end of the spectrum is difficult to discern. The two spikes at 405 and 415 mT are instrumental artefacts and have their origin in paramagnetic centers of the used sapphire dielectric resonator. 1 72 CHAPTER 6. H-ENDOR OF THE S 2 AND THE S2 STATE − Amplitude 2

200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 Magnetic field [mT] 1 3 Amplitude

200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 Magnetic field [mT]

Figure 6.1: Field swept ESE of the S2 state of the manganese cluster in spinach PS II. The ENDOR- positions are marked with arrows, arrow 4 lies outside the spectrum to measure the background signal. The two spikes at 405 and 415 mT are instrumental artefacts. Experimental conditions: T = 4.5 K, νMW=9.699 GHz, τ=300 ns, π/2 microwave pulse length = 32 ns.

Also here the large signal of YD• clearly dominates the spectrum at g = 2. Its size is approximately fourfold the maximum of the MLS. Usually this ratio is ≈ 1/50 but the YD• signal under these conditions is diminished by saturation eﬀects due to the low temperature and the short shot repetition time of the experiment. The concentration of free manganese appears to be very low in the sample since the typical Mn2+ signal consisting of six lines with a spacing of ≈ 9 mT is absent in this spectrum.

6.3.2 Field swept ESE of the S 2 state −

Figure 6.2 shows the ﬁeld swept ESE spectra of the S 2 state at 4.5 K in a frozen solution − of PS II membrane fragments in a buﬀer containing H2O and in a D2O exchanged sample.

It is clearly visible, that the prominent YD• EPR signal which dominates all other signals in the g = 2 region for PS II samples is essentially missing for the protonated sample. This is due to the strong reductive eﬀect of NO which also reduces this species to form the diamagnetic amino acid. In the D2O exchanged sample there is a narrow signal from NO that had not entirely been removed. It is unlikely that the origin of this signal is YD• since under these strong reducing conditions the tyrosine should be in its neutral diamagnetic form. 6.3. RESULTS 73

Amplitude 1 2

220 240 260 280 300 320 340 360 380 400 420 440 460 480 Magnetic field [mT]

Figure 6.2: Field swept ESE of the S−2 state of the manganese cluster in spinach PS II at T = 4.5 K for the protonated sample (bottom) and the D2O-exchanged sample (top). The ENDOR-positions are marked with arrows.

Furthermore, the absence of the very strong NO signal around g = 2 proves that the removal procedure for the NO was completely successful for the protonated sample. Al- though there is an EPR signal that stems from NO in the deuterated sample it is smaller by two orders of magnitude. Therefore it can be concluded that NO is absent in most of the PS II centers. 2+ Little free manganese (Mn ) EPR signals seem to be superimposed on the S 2 signal. − The spikes at 405 and 415 mT are instrumental artefacts stemming from the paramagnetic centres of the resonator. The overall width of the spectra is in accordance with the values found by others [149, 150]. Some of the hyperﬁne structure is visible particularly in the centre region of the spectra.

1 6.3.3 H-ENDOR on protonated samples in the S2 state

Figure 6.3 shows the ENDOR-spectra for the S2 state EPR signals in PS II membrane fragments. For these ENDOR-spectra a much larger frequency region than in previous investigations [137–140] has been chosen to cover the full range of possible hyperﬁne couplings. Particularly strong dipolar couplings accounting for substrate water protons near the manganese cluster would be interesting. The ENDOR spectrum recorded at a ﬁeld 1 74 CHAPTER 6. H-ENDOR OF THE S 2 AND THE S2 STATE −

B Amplitude

-10 -5 0 5 10 15 20 25 ν ν RF − H [MHz]

1 Figure 6.3: H-Davies-ENDOR of the manganese cluster in the S2-state. The spectra have been recorded at the following ﬁeld positions: A) : 333 mT; B) : 346 mT; C) : 357 mT and D) : 215 mT to have a comparison to background signals. Experimental conditions: T = 4.5K, νMW=9.5 GHz, tRF.

position outside the apparent MLS shown on the bottom of ﬁgure 6.3 shows no features protruding from the almost straight baseline. This accounts for the lack of instrumental artefacts. All other spectra possess a very strong unfeatured signal underlying the more pronounced signals. This background is smaller compared to the ENDOR signals recorded at the

EPR position of the YD• but since the electron spin echo together with the ENDOR effect for YD• are stronger in general this is just a matter of scaling. This background is not present in the ENDOR spectrum of YD• with the manganese cluster in the S1 state (see previous chapter). So it must clearly be attributed to the ENDOR of the MLS. Its origin is not clear. One possibility is that it is the superposition of many couplings of hydrogen bound to amino acids in the surrounding of the manganese cluster. In the spectrum it stretches over the whole range up to 17 MHz. This would account for hyperfine coupling values of up to 34 MHz and for protons very near to the manganese cluster. Around the proton matrix region some distinct lines are distinguishable. Lines corresponding to hyperfine couplings of ≈ 4.0 MHz, ≈ 2.2 MHz and ≈ 1.0 MHz are visible. If these lines really belong to the ENDOR of the MLS or if they are traces of YD• ENDOR features or originate from some magnetic interaction of the manganese cluster with YD• is not clear. It can be excluded that these lines stem directly from YD• since all other prominent features 6.3. RESULTS 75

far from the matrix region (for instance the line at ≈ 13 MHz) assigned hfc’s of YD• do not appear in the ENDOR spectra for the MLS.

1 6.3.4 H-ENDOR on protonated and deuterated samples in the S 2 state −

Figure 6.4 shows the Davies ENDOR of protonated and deuterated samples poised in the

S 2 state, respectively, recorded at the magnetic field positions marked in figure 6.2. The − positions off the g = 2 region have been chosen to avoid any contribution from possibly existing features there. Clearly visible is the broad background in the spectra that had also been observed in the ENDOR spectra for the S2 state. Since it is observable in the ENDOR-spectrum of deuterated samples as well these protons must be non-exchangable protons. The g anisotropy of the multiline signal is too small to lead to pronounced orientation selection effects in X band ENDOR. Therefore, for the given difference of magnetic field positions, no differences for the two ENDOR spectra recorded are visible. For a given effective hyperfine coupling constant a and a nuclear Zeeman frequency νn, the ENDOR a resonance condition is νENDOR = |νn 2 |. Here, the proton Zeeman frequency is νn = 14.8 MHz. The spectral features corresponding to proton hyperfine couplings are therefore grouped symmetrically about this frequency.

One advantage to the S2 state ENDOR spectra is the lack of the strong EPR signal of

YD• due to the oxidized tyrosine radical having been reduced completely during the NO treatment (at least for the protonated samples). This signal and its interaction with the multiline signal of the manganese cluster lead to much confusion in the past about the origin of the observed proton couplings [137, 138]. The underlying broad background signal of these ENDOR spectra is similar to that observed in the ENDOR spectra for the MLS of the S2 state. Many features are at the same frequency positions in all spectra. This could indeed account for protons of the surrounding amino acids coupling to the manganese cluster. One striking point is that some ENDOR lines in the proton matrix region observed for the MLS in the S2 state are missing. There are no lines accounting for hfc’s at 1 and

4 MHz. Since the EPR signal of YD• is not present in these samples this point supports the idea that these couplings do not belong to the ENDOR of the MLS in the S 2 state. − Some very distinct and strong proton couplings are distinguishable in the spectrum for the protonated sample. At least two peaks with maxima at -3.8/3.8 MHz and -1.5/1.5 MHz dominate the spectrum, which would account for two hyperﬁne coupling components of 7.6 MHz and 3 MHz, respectively. Whether these features are rather isotropic or are just 1 76 CHAPTER 6. H-ENDOR OF THE S 2 AND THE S2 STATE −

357 mT

333 mT Amplitude

357 mT

333 mT

-14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 ν ν RF − H [MHz]

Figure 6.4: Davies’ ENDOR of the S−2 state of the manganese cluster in spinach PS II at T = 4.5 K for a protonated (top) and a deuterated (bottom) sample. The spectra have been recorded at the magnetic

field positions 333 mT (top) and 357 mT (bottom). Experimental conditions: νMW=9.65 GHz, . the middle components of rhombic hyperfine couplings tensors is hard to say. In the ENDOR spectra for the deuterated sample there are no signals at the positions, where ENDOR signals had been observed for the protonated sample. The protons giving rise to ENDOR lines in the spectra for the protonated sample are therefore exchangable with deuterons by treatment with D2O. Only the strong nonspecific background from unexchangable protons possibly bound to amino acid residues and a very diminished proton matrix peak can be seen here. For the spectrum recorded at 333 mT one coupling centered around the matrix peak is left with small intensity.

The coupling model based on a point-dipole approximation

For axially symmetric hf tensors, with small anisotropy, usually both principal values A and A can be obtained. These dipolar couplings have characteristic lineshapes and k ⊥ the hf tensor principal values are related to each other by A = −2A . Referring to a k ⊥ hydrogen bond, which shows a purely dipolar hf tensor, the length r of this bond can be calculated using a simple point-dipole approximation: c A(θ) = ∗ ρ ∗ (3cos2θ − 1) (6.1) r3 6.3. RESULTS 77

S2 state hfc’s Kawamori Zweygart Tang Fiege [MHz] Ref. [137] Ref. [142] Ref. [138] Ref. [139–141] 0.53 0.5 0.2-0.3 0.76 0.7 1.0 1.19 1.0 1.1 1.44 2.2 2.41 2.4 2.4 2.3 4.0 4.01 4.2 4.9

S 2 state hfc’s [MHz] − 3.0 7.6

1 Table 6.1: Hyperﬁne values observed for the H-ENDOR spectra of PS II in the S2 and the S−2 state, respectively. The data for the S2 state are compared to data from other groups.

3 Here A(θ) is the hf coupling in MHz, c = geµegnµn, which is 79 MHz/A˚ for protons, ρ is the unpaired electron spin density, θ is the angle between the applied magnetic ﬁeld B0 and the vector connecting the electron and the proton. The following considerations are partly adapted from Fiege et al [139–141] and are altered for the case of the magnetically coupled Mn(II)-Mn(III) dimer as it is the case for the S 2 − multiline EPR signal. The hyperﬁne coupling of a proton to a Mn(II)Mn(III) complex can be expressed in terms of the Hamilton formalism:

Hˆ = S1an1In + S2an2In. (6.2)

The intrinsic hyperfine tensors for a proton coupled to a Mn(II) and to a Mn(III) are here an1 and an2. In is the nuclear spin operator for the proton whereas S1 and S2 are the single spin operators for the electron spins of the manganese ions. The intrinsic hyperfine tensors are usually diagonal in different coordinate systems. Therefore the effective spin operator

Seﬀ is introduced for this kind of magnetically coupled systems. For antiferromagnetically coupled Mn(II) and Mn(III) with S(II) = 5/2 and S(III) = 2 this results in a total Seff of 1/2. The according Hamiltonian can now be rewritten:

ˆ S H = Seﬀ AnIn, (6.3) with S An = c1an1 + c2an2. (6.4) 1 78 CHAPTER 6. H-ENDOR OF THE S 2 AND THE S2 STATE − H ψ r1 r2

3+ Mn2+ Mn d

Figure 6.5: Deﬁnition of the parameters given in the text for the coupling model of a proton with a Mn(II)-Mn(III) dimer antiferromagnetically coupled having a net spin of 1/2. according to [139–141]

Here the parameters c1 and c2 are the spin projection factors for the manganese ions. In general the principal values are [139–141, 156]: −δ 0 0 S An = c  0 −Γ + δ/2 0  (6.5)  0 0 Γ + δ/2,      with the parameters δ and Γ as:

3 3 δ = c1 ∗ r1− + c2 ∗ r2− (6.6)

3 3 3 2 2 3 3 Γ = ∗ (c ∗ r− + c ∗ r− ) − 4(sin ψ) ∗ c ∗ r− ∗ c ∗ r− 2 1 1 2 2 1 1 2 2 q 3 2 6 2 6 3 3 = ∗ c ∗ r− + c ∗ r− + 2 ∗ c ∗ r− ∗ c ∗ r− ∗ cos(2ψ) (6.7) 2 1 1 2 2 1 1 2 2 q For a Mn(II)Mn(III) dimer the spin projection factors are c1(Mn(II))=7/3 and c2(Mn(III))=- 4/3. Therefore for δ and Γ the following equations apply [160]:

7 3 4 3 δ = r− − r− (6.8) 3 1 3 2 1 6 3 3 6 Γ = 49r− − 56r− r− cos(2ψ) + 16r− . (6.9) 2 1 1 2 2 q The parameters ri and ψ in this model are deﬁned by Fiege et al [139–141] as shown in ﬁgure 6.5.

6.3.5 Analysis of the ENDOR signals for the S 2 state −

1 As yet only H-ENDOR spectra for the S2 state of the OEC existed [137–139, 141]. The interpretation of these spectra has been extremely diﬃcult for the experimentators be- 6.3. RESULTS 79 cause of interferences of paramagnetic species and possibly invalid assumptions. The manganese cluster of PS II is agreed to consist of four manganese ions. Many proposi- tions have been made to explain the magnetic interaction of the electron spins as well as the nuclear spins of the ions. It has been concluded that the multiline signals of the

S2 and the S0 state originate from an S=1/2 ground state [128, 161]. This should also be the case for the S 2 MLS [148]. In the case of the S2 state this has been explained − by assuming particular oxidation states for the manganese ions in the S2 state such as

Mn(III)Mn3(IV) or Mn3(III)Mn(IV) and an antiferromagnetic coupling situation of the manganese ions assuming two ions coupling to give a net spin of 1/2. The simplest model assumes that the other two manganese ions having an equal oxidation state couple to end up with a net spin of zero. For the coupling of this net spin of 1/2 with the nuclei in its surrounding several models have been introduced. Basically it is still very diﬃcult to simulate the magnetic interactions of more than 2 manganese ions coupling to each other numerically. Most of the manganese model complexes designed to imitate the structural properties of the manganese cluster or even its function of water splitting have been restricted to the dinuclear case.

The coupling of protons with the net electron spin of the manganese cluster in the S2 state can not be explained using this simpliﬁed model. 1H-ENDOR studies did not support the existence of strongly coupled protons belonging to substrate water bound to the OEC.

For the S 2 state the problem seems to be easier. In simulations of the S 2 state EPR − − multiline signal satisfying results have been achieved assuming only one paramagnetic dinuclear Mn(II)-Mn(III) cluster [149].

At least two couplings of 7.6 and 3 MHz could be gained from the ENDOR spectra of the protonated S 2 state. In the dipolar coupling case the lines are assigned to the A − ⊥ component. At ﬁrst it shall be tried to apply the simple point-dipole approximation of equation 6.1. Assuming a spin density of 1 the distances of the coupling protons to the OEC are 2.2 A˚ and 3.0 A,˚ respectively. Manganese-manganese distances have been determined by EXAFS. By this method two distances of 2.7 A˚ and 3.3 A˚ could be established [162, 163]. From this it is clear that the simple point-dipole approximation is not applicable here since the calculated distances are in the range of the manganese-manganese distances. The application of this approximation has two drawbacks. If the two coupling partners came very close to each other, and their distance is in the range of the extent of the di- or even multipole this leads to deviations. The other much greater problem arises if the proton couples to an exchange-coupled system. Here the diﬀerences between the spin 1 80 CHAPTER 6. H-ENDOR OF THE S 2 AND THE S2 STATE −

−7.6MHz 2 7.6MHz 1

0 3+ Mn2+ Mn r [100 pm]

-1

-2

3.0MHz -3 −3.0MHz

-4 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 r [100 pm]

Figure 6.6: Resulting possible positions for the measured proton couplings in the ENDOR-spectrum for protonated samples in the S−2 state. Drawn in blue are the equipotential lines for the coupling component of 7.6 MHz, and in red the lines for the coupling component observed at 3 MHz. The manganese-manganese distance is 2.7 A˚ for this case.

projection factors of the exchange-coupled components can be vast, negative projection factors are often found. For large distances the proton encounters only the net spin. For distances in the range of the exchange-coupled components this leads to incorrect results. With the three main values of the hyperfine coupling tensor extracted from an ENDOR spectrum the coupling model by Fiege et al [139–141] allows to determine two possible positions of the respective proton with respect to the manganese ions. As shown there each component of the tensor defines a surface around the two manganese ions where the proton could be. This surface is rotationally symmetric to the line connecting the manganese ions. The knowledge of all three components is necessary. In the case dealt with here it is not possible to determine all components of the hfc tensor. It is likely that the missing components are broad and can not be distinguished from the background signal. The component observed should be the middle component S of An. According to Fiege et al [141] the component −δ is the smallest absolute value and therefore leads to the respective ENDOR line in the middle of the spectrum. Figures 6.6 and 6.7 shows the equipotential lines indicating the possible positions of the protons giving rise to the observed coupling components at 7.6 MHz and 3.0 MHz, respectively, for both manganese-manganese distances known from EXAFS calculations. The calculation of the equipotential lines for the manganese-manganese distances of 2.7 A˚ and 3.3 A˚ results in only small differences. Further discussion will focus on figure 6.6 regarding a manganese-manganese distance of 2.7 A.˚ The resulting picture much better visualizes the dependence of position of protons and their coupling parameters and provides maxima and minima of proton-manganese distances. The simple point-dipole approximation, that assumes the interaction of two point- 6.4. DISCUSSION 81

−7.6MHz 2 7.6MHz

3+ 0 Mn2+ Mn r [100 pm]

-1

-2

3.0MHz -3 −3.0MHz

-4 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 r [100 pm]

Figure 6.7: Resulting possible positions for the measured proton couplings in the ENDOR-spectrum for protonated samples in the S−2 state. Drawn in blue are the equipotential lines for the coupling component of 7.6 MHz, and in red the lines for the coupling component observed at 3 MHz. The manganese-manganese distance is 3.3 A˚ for this case. dipoles, fails to provide reasonable results here. For the coupling component of 7.6 MHz the upper limit for the proton-manganese distance is found for ψ = 0 and amounts

≈ 2.8 A˚ for a negative δ. The shortest possible distance on the other hand for ψ = 180◦ is ≈ 1.1 A.˚ In this case the proton should be located directly between the two manganese ions. Within this margin of 2.8 A˚ and 1.1 A˚ the proton with c ∗ δ = 7.6 MHz has its distance to one of the manganese ions. For the other possible proton with c ∗ δ = 3.0 MHz the lower limit is the same, the upper is approximately 1 A˚ larger.

6.4 Discussion

6.4.1 The S2 state

1 The H-ENDOR for the S2 state in a BBY sample partly corroborates the findings of other groups for the matrix region for protons. Lines have been found accounting for hfc’s at ≈ 4.0 MHz, ≈ 2.2 MHz and ≈ 1.0 MHz shown in table 6.1. Similar results have been obtained and interpreted in the past leading to different results for direct water ligation to the manganese ions of the OEC [137–141]. Compared with each other these results shown in table 6.1 exhibit slight differences in their values often because the signal-noise ratio was insufficient to determine them with high accuracy. Kawamori et al [137] in their ENDOR spectra of the proton matrix region found six lines grouped symmetrically around the free proton frequency. They assumed that the man- 1 82 CHAPTER 6. H-ENDOR OF THE S 2 AND THE S2 STATE − ganese spins are concentrated at the center of gravity. With this assumption in mind they applied the simple point-dipole approximation and calculated the proton-manganese cluster distances. Particularly for the larger couplings which lead to smaller distances and therefore hint direct water ligation this model is not applicable. Tang et al [138] measured 1H-ENDOR in combination with ENDOR-Induced-EPR. Thereby they identified the line at 4 MHz as belonging to the ENDOR spectrum of YD• . In contrast to all other works they observed a coupling at 4.9 MHz. The three largest couplings at 4.9 MHz, 2.4 MHz and 1.0 MHz have been assumed by these researchers to belong to a single rhombic hyperfine coupling tensor. They used a spin-coupled point-dipole model described in [158, 159]. Application of this model with the aforementioned assumption resulted in a minimal proton-manganese distance of 3.65A˚ for the case of an antiferromagnetically coupled Mn(III)Mn(IV) dimer. This led these workers to the conclusion that the manganese cluster has no water in its vicinity, at least none, that would be directly bound. Together with the low signal-to-noise ratio and the fact, that the largest coupling component could not be reproduced by other groups the assignment of these ENDOR signals to a single rhombic tensor seems arbitrary. Fiege et al [139] found couplings at 0.6 MHz, 1.1 MHz, 2.3 MHz and 4.0 MHz. The largest couplings have been reproduced by other groups too. Since these protons are probably not bound directly to the manganese ions dipolar coupling tensors with axial symmetry could be expected. Already for the model of a manganese dimer coupling antiferromagnetically to give a net spin of 1/2 the effective hyperfine tensor is not axial anymore but rhombic [139, 141, 156]. Most of the coupling models for the S2 state of the OEC assume this dimer where the other two manganese ions should also couple antiferromagnetically to give a net spin of zero. Fiege et al. used an extended point-dipole model which takes into account that the electron spin is distributed over two manganese ions forming a Mn(III)Mn(IV) dimer. This model has also been used in this work for the evaluation of the ENDOR spectrum for the S 2 state assuming a Mn(II)Mn(III) dimer. − With the help of this model and the two largest hyperfine values Fiege et al [139–141] calculated the maximum distances of protons to the manganese ions. They assumed that the observed lines belong to the middle components of the hfc tensor. For the line at 4.0 MHz they calculated distances of 2.5A˚ and 3.3A,˚ respectively for the distance of the proton to Mn(IV) and Mn(III). Hereby they came to the conclusion that the cluster is not necessarily dry taking into account that the oxygen atom of the water molecule binds directly. This would allow a longer distance of the water hydrogens to the manganese ions. At least one substrate water molecule should already be bound to the manganese cluster in the S2 state [157]. Protons bound to water which is ligated via the oxygen atom to a member at a spin-carrying centre in the simplest case exhibit an axial hfc tensor. Here 6.4. DISCUSSION 83 the extended dipole-dipole-coupling model of Fiege et al [139–141] would apply, possibly for more than two manganese ions coupling. Since the hyperfine values found in this work are similar to the ones reported by other groups one could and applying the extended point-dipole model from Fiege et al [139] this leads to similar results for the proton-manganese distances. No larger resolved hfc’s than the ones reported earlier could be gained from the ENDOR spectra of the S2 state MLS here. It is not clear if the point-dipole approximations introduced in the mentioned works are applicable at all since they assume only a manganese dimer (Mn(III)Mn(IV)) which the protons are supposed to couple to. The coupling situation could be much more complex. This could also be the reason why no distinctly larger couplings are observable in the ENDOR spectrum. It is well possible that they exist but are broadened due to a large anisotropy until they can not be observed anymore. Perhaps this is what the broad background signal that stretches up to 17 MHz and would account for hfc’s in the 34 MHz range. Protons with these coupling values would be very near to the manganese cluster. Simplification is to expect if really only a dimer of manganese ions would couple to the protons in the surrounding.

6.4.2 The S 2 state −

1 In the H-ENDOR spectra for protonated PS II samples in the S 2 state broad but − distinct ENDOR lines are observed. These lines vanish for the D2O treated samples. So they must have their origin in hydrogens that can be exchanged by deuterons. From the width of these lines it has been assumed that they are the middle components of rhombic tensors. Using the extended point-dipole coupling model of Fiege et al[139–141] but for a Mn(II)-Mn(III) dimer the equipotential lines referring to this coupling value could be calculated. The approximate distances to the manganese ions could be estimated. For the larger coupling at 7.6 MHz the upper distance limit is 2.8A˚ (for the 3.0 MHz coupling it is ≈ 3.8A).˚ There are several possibilities how hydrogen can come close to the manganese cluster.

First of all it is supposed to be exchangable to deuterons by treatment with D2O. It could then be in an amino group of an amino acid, in a hydroxylic group or in water. For the case of an amino acid it is hopeless to determine more than the distance because the spatial arrangement of amino acids with respect to the manganese cluster still remains unclear. Many geometrical arrangements are possible since there are few structural restrictions in terms of ligating amino acids. Hydroxylic groups could be bound to the manganese cluster too. Here one would assume 1 84 CHAPTER 6. H-ENDOR OF THE S 2 AND THE S2 STATE −

3 H

2 O H O 1

2+ 3+ 0 −7.6MHz Mn Mn 7.6MHz

r [100 pm] O

-1

O O

-2

H H 3.0MHz -3 −3.0MHz

-4 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 r [100 pm]

Figure 6.8: Possible positions of the hydroxylic groups bound to the manganese cluster in its S−2 state.

3 H

2 −7.6MHz H O 7.6MHz

0 2+ 3+ Mn Mn r [100 pm]

-1

-2

-3 3.0MHz −3.0MHz

-4 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 r [100 pm]

Figure 6.9: The possible position of a water molecule bound to the manganese cluster in its S−2 state. direct ligation of the oxygen. Due to its simple structure there are few restrictions. The distance of manganese to ligating oxygen lies in synthetic complexes in the range of 2.00.1 A˚ [164, 165]. Together with the bond length for a hydroxylic group of 0.95 A˚ five possible positions of the OH have been figured out and are shown in figure 6.8. Possible is also that the oxygen forms a µ-oxo bridge. The position would then be rather symmetrical between the manganese ions. Here the equipotential lines are very close to each other and the position of the hydrogen would be well-defined. Shown in figure 6.8 are the possible arrangements of hydroxylic groups taking into account the calculated distance and the bond lengths the of hydroxylic group and the manganese ions. Four of the five found possibilities result in a O-Mn-Mn angle of smaller than 90◦. The oxygen atom would in these cases be in a position where it forms a µ-oxo-bridge between the manganese ions.

Possibly the angle H-O-Mn could not be assumed 180◦ anymore. This is still in accordance with the found distances. It is not clear for the S 2 state whether water is bound to the − OEC or an hydroxylic group OH. If a water molecule is bound there should be also some signal for the second hydrogen atom. The second observed coupling would then account 6.4. DISCUSSION 85 for the second hydrogen. The bond length in water is 0.958 A,˚ and the bond angle 105◦. Therefore the hydrogen-hydrogen distance would be 1.5 A.˚ If the two observed couplings stem from the same water molecule with the calculated distances and the geometrical restrictions possible arrangements have been tried. Given in ﬁgure 6.9 is the only possible position for such a water molecule obeying the magnetical and molecular-geometrical restrictions. For the direct binding of water tight restrictions are valid. Only for water ligation where the O-Mn-Mn angle is smaller than 90◦ the water molecule could be ﬁt within these restrictions. From this also the sign of the couplings can be derived which would be 7.6 MHz and -3.0 MHz. In almost all arrangements found the oxygen is situated between the two manganese ions.

Protonated µ-oxo-bridges are possibly the case for the S 2 state of PS II. − 1 86 CHAPTER 6. H-ENDOR OF THE S 2 AND THE S2 STATE − Chapter 7

55 Pulsed Mn-ENDOR of the S2-State of the manganese cluster in Photosystem II

7.1 Motivation

The EPR mulitiline signal (MLS) of the manganese cluster of PS II is very complex. Given that four inequivalent 55Mn nuclei contribute to it, 1296 hyperﬁne lines overlap here (given an S=1/2). This results in the MLS showing 20-23 features with line spacings of 90 G due to spectral congestion and therefore single transitions can not be resolved.

Most of the spectra recorded for the MLS of the S2 state have been recorded at X-band. Only few have been recorded at Q-band but did not deliver new data [166, 167]. Therefore simulations of this EPR signal and also the signal at g = 4.1 only relying on data derived from the X-band MLS did so far not lead to satisfying results [129–133, 175]. A method to reduce the number of contributing transitions is at hand. The hyperﬁne and quadrupole couplings of these I=5/2 55Mn can be studied with Davies ESE-ENDOR. Only 40 allowed 55Mn nuclear spin transitions are possible for the regarded system. Taking this into account the 55Mn-ENDOR spectra will show fewer features and be easier to evaluate. In the past 55Mn-ENDOR has already been used to characterize enzymes containing only two manganese ions [154, 171] and also model compounds having a manganese dimer core to mimic the OEC of PS II [169, 171]. 55Mn-ENDOR for the MLS of PS II has already been measured [14, 169, 172]. The researchers here tried to simulate the acquired ENDOR spectrum together with the

87 55 88 CHAPTER 7. MN-ENDOR OF THE S2-STATE

EPR signal. This puts up tighter constraints to the choice of simulation parameters and apparently led to a more reliable result. The conclusion they made was that four eﬀective 55Mn hyperﬁne tensors A are required to simulate the experimental spectra, whose main components for methanol-treated samples were determined as follows (values in MHz): A1 = [−232, −232, −270]; A2 = [200, 200, 250]; A3 = [−311, −311, −270];

A4 = [180, 180, 240]. From the simulations they also derived the quadrupole parameters which are essential for the ENDOR-linepositions in such manganese complexes. In this chapter it will be shown that these results are reproducible although possibly not all information concerning the hyperﬁne values has been retrieved by Peloquin et al [14] from the ENDOR spectra.

7.2 55Mn-ENDOR spectra of exchange coupled complexes

ENDOR spectra for single manganese ions surrounded by strong or weak ligand fields exist. Their evaluation is relatively uncomplicated. ENDOR spectra for exchange coupled manganese dimers are more complex since they no longer consist of isotropic hyperfine lines. Due to asymmetric ligand fields induced by different ligands and often Jan-Teller- distortions quadrupolar interactions have to be taken into account. These often shift the position oft the ENDOR lines to a remarkable extent. The quadrupole parameters are not easily available. For antiferromagnetically coupled manganese complexes such as Mn(III)Mn(IV) or Mn(II)Mn(III) having in their ground state an effective electron spin of 1/2 the zero field splitting is not of importance yet has to be considered while calculating the intrinsic hyperfine coupling values. 55Mn nuclei have a small g-value of 1.3819. For this nucleus the nuclear frequency at X- band is much smaller than its hyperfine coupling. This means that the resonance positions are to be found centered around Ai’/2 with a distance of 2νn:

Ai0 νENDOR = νn (7.1) 2

Considering also the quadrupole interaction for a 55Mn-ENDOR spectrum the following formula exempliﬁed for the magnetic ﬁeld parallel to the z-axis (assuming the g and the quadrupole tensor to be collinear) the energy levels and therefore the transition frequencies result: 2 Ai0 3e qQ Em +1 − Em = νn + (2mi + 1) i i 2 20h

Ai0 = νn + 2P (2mi + 1) (7.2) 2 ||

7.2. 55MN-ENDOR SPECTRA OF EXCHANGE COUPLED COMPLEXES 89

The quadrupole tensor for the ith nucleus is commonly deﬁned as (with ηi non zero, the tensor gets rhombic):

−(1 − η ) 0 0 P i P = ||i  0 −(1 + η ) 0  (7.3) i 3 i  0 0 2     

This leads to shifts of the basic ENDOR lines depending on mi and also ms and is shown nicely in Sch¨afer et al [171]. Including all orientations of the quadrupole tensor lines arising from an isotropic hfc now appear axial or even rhombic in the ENDOR spectrum.

If νn is known the quadrupole tensor components can be directly read out. Further complications arise due to anisotropies of g and hyperﬁne tensors. Lines overlap particularly in the centre. Here the density of contributions from diﬀerent mi is greater and already slight anisotropies and linewidths make ENDOR spectra extremely complex. As shown in [171] it is by far better to choose an EPR-position at the rim of the spectrum to measure ENDOR since the ENDOR spectra expected are less complicated. Here the signal intensity is the limiting value. 55Mn-ENDOR spectra and valid simulations of them are a necessary complementary tool to simulate EPR-spectra, since here quadrupole interactions play a decisive role. Par- ticularly where EPR-spectra can not be recorded at several frequency bands which is unfortunately the case for the MLS of the S2 state of PS II ENDOR measurements can be fruitful.

It is difficult to explain the origin of the MLS of the S2 state by assuming only a dimer of manganese ions [133, 134]. The effective hyperfine tensors have to be assumed highly anisotropic and very large as well as the recommended quadrupole parameters to simulate the EPR spectrum at this width. With the help of four hyperfine tensors, each in the expected range for either Mn(III) or Mn(IV) it is possible to simulate the EPR-spectrum quite well but with deficits concerning the ENDOR spectra. This could mainly be due to no quadrupole parameters given by these researchers [132, 135, 136], which makes an ENDOR simulation only relying on hyperfine values quite futile. Peloquin et al [14] simulated both the EPR and the ENDOR spectrum with high resemblance to the recorded spectra. The oxidation states of the manganese ions are not known with certainty nor are the quadrupole parameters. The signal intensity of the MLS is so far only sufficient at the centre of the spectrum to record an ENDOR spectrum in a reasonable time. Many assumptions have to be made simulating the according ENDOR spectra. One point is that at least the measured ENDOR spectra should be trustworthy. 55 90 CHAPTER 7. MN-ENDOR OF THE S2-STATE

7.3 Experiment

7.3.1 Samples and sample treatment

The samples used in this work were prepared from spinach according to the method described by Berthold/Babcock/Yocum (“BBY”) [9] with the modiﬁcations described in [115]. Sucrose at a concentration of 0.4 M was used as cryoprotectant. Methanol was added to a concentration of 3%. The concentration of the samples in the EPR tubes was ≈15 mg(Chl)/ml. Illumination was done at 200 K for 30 min.

7.3.2 Pulsed 55Mn-ENDOR

ENDOR at the required frequency range brings with it some difficulties due to the use of a DICE system which has a power output for radiofrequencies that drops remarkably at frequencies beyond 100 MHz. Using this equipment only Mn(IV) resonances (≈200 MHz) would be accessible. How this problem has been overcome is decribed extensively in the PhD-thesis of Schäfer [171] and shall only be mentioned briefly here. A second frequency source with a fixed frequency was used (Rhode&Schwarz SMT02) and its output mixed with the DICE output, which can be sweeped. The DICE system itself was modified to deliver constant power levels beyond 100 MHz. Hereby frequencies with a sufficient output of up to 300 MHz are possible. Nevertheless problems always arise using this equipment by the still low sweep width of the DICE-box and the occurence of other mixing frequencies. These sweep at the same time and the according signals are superimposed on the spectrum acquired with the wanted frequency. This problem was tackled with a high-pass filter (MiniCircuits NHP-100, passband 90- 400 MHz). A new ENDOR coil designed for higher frequencies was build by Bruker/Rheinstetten which has a capacitance/inductance combination whose resonance does not interfere with the Mn ENDOR spectrum. This equipment was fitted to a Bruker ESP 380E spectrometer equipped with an ESP 360D- Pulsed ENDOR accessory and an ENI A-100 RF amplifier. 7.4. RESULTS 91

7.4 Results

Figure 7.1 shows the ﬁeld swept ESE (top) and the 55Mn-ENDOR for two frequency regions for the methanol-treated S2 state of the OEC. The arrow indicates the position where 55Mn-ENDOR has been done. The ﬁeld swept ESE compares very well to the other

field swept ESE spectra for the S2 state in this work. The width and the line spacings are identical. The pseudo-modulated first derivative of the spectrum shows a high resemblance to the cw spectrum of the S2 state MLS. Also the prominent EPR signal of YD• ox is visible. The ENDOR spectrum shown in the first frequency region ranging from 60-180 MHZ has three distinct features at ≈90 MHz, at ≈115 MHz and at ≈145 MHz partly superimposing each other. The second frequency region ranging from 240 to 360 MHz shows a single feature at ≈285 MHz. The peaks in the first frequency region definitely can be assigned to 55Mn nuclear transitions of the manganese ions of the OEC. In contrast to the result obtained by Peloquin et al [14] who did not see any signal beyond 175 MHz with certainty in the spectrum here there is a signal at 285 MHz with a similar intensity as for the signals at lower frequencies. This signal proved to be reproducible.

7.5 Discussion

In this work it was possible to measure the 55Mn-ENDOR for the MLS of the OEC. In the frequency region from 90 to 200 MHz the same spectrum could be reproduced as previously measured by Peloquin et al [14]. Peak positions, line spacings and intensity ratios resemble each other in these two spectra. The ENDOR spectrum together with the corresponding EPR spectrum has been simulated by these researchers. Certainly the existence of the ENDOR spectrum much more encircles the possible values for the effective hyperfine couplings. It is quite clear that previous simulations of the EPR signal without the ENDOR data neglecting the quadrupole interaction are not appropriate to reflect the true coupling situation. Nevertheless even with these ENDOR data at hand a simulation using effective hyperfine tensors of four inequivalent 55Mn nuclei and quadrupole parameters to tune the simulation of the spectrum is not straightforward. Four coupling nuclei measured at this EPR-position account for ≈16 transitions, which are broadened by quadrupole coupling and hyperfine anisotropy. A simulation at this stage seems not to be useful. Peloquin et al [14] state that they did not observe any ENDOR signal beyond 175 MHz. Taking this for correct, the ENDOR spectrum measured at this magnetic field position can already been taken to exclude some of the coupling models constructed in the past [132–136]. 55 92 CHAPTER 7. MN-ENDOR OF THE S2-STATE

200 250 300 350 400 450 500 Magnetic Field / mT

60 80 100 120 140 160 180

240 260 280 300 320 340 360 Frequency / MHz

Figure 7.1: Field swept ESE of the S2 state of the manganese cluster in spinach PS II at T = 4.5 K. The ENDOR-position is marked with an arrow. Below it the 55Mn-ENDOR is shown. 7.5. DISCUSSION 93

Our spectra show a difference to the ENDOR spectra measured by Peloquin et al [14]. There exists a signal at 285 MHz. It has a signal/noise ration that is comparable to the one of the signals in the lower frequency range. Its frequency would result in an effective hyperfine coupling value of 570 MHz (neglecting quadrupolar contributions). Assuming that this signal has its origin in the magnetic interaction of the manganese ions an assessment would be difficult. To reach such a even for 55Mn nuclei tremendous effective hyperfine value the manganese ion would be a Mn(III) that should be coupled extremely weak to the rest of the cluster with a strong zero field contribution (the D/J ratio) to account for a projection factor being able to push the effective hyperfine tensor into this region. The effective hyperfine tensor for this ion should be very anisotropic. There is not much one can say about quadrupolar interaction contributions. But since the high frequency signal is apparently restricted to a single line they should be rather small, at least not in the range proposed by Ahrling˚ et al [133, 134]. The ENDOR simulation done by Peloquin et al [14] relies on the fact that there are no additional features in the ENDOR spectrum beyond 175 MHz. They need four manganese ions with moderate anisotropy and quadrupole parameters also found for synthetic model complexes. Here a simulation using fewer than four manganese nuclei would not lead to a good simulation of the EPR signal with a sufficient width. Taking the resonance at 285 MHz it would be possible to simulate the ENDOR as well as the EPR spectrum with two or three participating manganese ions. The ENDOR-spectrum was recorded at EPR-positions near to the centre of the EPR- spectrum and therefore probably includes several nuclear transitions. Peloquin et al [14] moved away from the centre of the EPR-spectrum but did not observe qualitative changes in the ENDOR-spectra. This shows that probably the hyperfine tensors are very anisotropic. Only moving to the very edges of the spectrum would simplify the ENDOR-spectra.

The conclusion that can be made is that the signal of the MLS for the S2 state must be improved tremendously to be able to gain a better signal/noise ratio also for the ENDOR spectra. It is not reasonable to carry out simulations of the ENDOR spectra with the experimental spectrum recorded at these EPR-positions that close to the centre of the MLS. If the line at 285 MHz in the 55Mn ENDOR spectrum proved to be reproducible, and instrumental artefacts could be excluded, the coupling model and its conclusion presented by Peloquin et al [14] had to be reassessed.

One point that should be made here is to measure the MLS of the S 2 state. Here only a − manganese dimer is the origin of the EPR signal, and the signal intensity should be suf- ﬁcient to choose an EPR-position nearer to the rim of the spectrum and therefore catch fewer transitions in the ENDOR spectrum. The projection matrices can be calculated using the exchange coupling model successfully applied already to manganese model com- 55 94 CHAPTER 7. MN-ENDOR OF THE S2-STATE plexes. Measuring the quadrupole parameters for this species and assuming no change in the ligand sphere in the S 2-S2 transition could be a good starting point for future − simulations of the MLS of the S2 state. Chapter 8

Methylamine and its Eﬀects on the Oxygen Evolving Ability of PS II and the S2-state Multiline Signal

8.1 Motivation

Primary amines are known to reversibly inhibit water oxiadtion in PS II and modify the magnetic properties of the S2 state[5]. The inhibition is thought to occur through binding to a chloride anion (Cl−) and/or to a substrate water binding site at or near the tetranuclear manganese cluster. Modiﬁcations of the hyperﬁne structure of the S2 EPR 15 14 multiline signal and NH3 and NH3 ESEEM studies have been interpreted to show direct binding of ammonia to the manganese cluster, possibly via a µ2-imido bridge[173, 174].

While the binding of primary amines has been described for the S1, S2 and S3 states, not much is known about binding to the S0 state. Because amines can be regarded as water analogues, such information can be valuable in learning more about substrate water binding in the diﬀerent S-states. Recently the S0 EPR multiline signal was discovered[125,

126, 175]. This opens the possibility to study the interaction of amines with the S0 state in detail using EPR techniques. Methylamine is especially interesting because it is also an analogue of methanol which is required for the observation of the S0 EPR mulitiline signal. The objective of this work was ﬁrst to analyze the eﬀects of methylamine on the oxygen- evolving ability of PS II and on the magnetic properties of the OEC using the fairly easy accessible S2-state.

95 96 CHAPTER 8. METHYLAMINE AS INHIBITOR TO OXYGEN EVOLUTION

8.2 Interaction of water analogues with the OEC

Chemical compounds that resemble water chemically and possibly in size are reported to interact with the OEC in PS II. Many of these water analogues have been added to

PS II preparations such as H2S[178], which was found only to act as inhibitor, H2O2 [179] or even methanol [180] possessing one OH-group and compared to water one hydrogen exchanged to a methyl group. Methanol comes very close to the manganese cluster and alters its magnetic properties. It hereby prevents the formation of the g = 4 signal. It is the only alcohol apparently small enough to get in the vicinity of the manganese ions to show deuteron modulations in ESEEM spectra with the hydroxylic hydrogen exchanged to deuterium [181]. No reactions of these molecules proved by any redox reaction products have been reported yet. Another family of water analogues are ammonia and its derivatives. Particularly ammonia itself has been investigated to a large extent. It binds directly to the manganese cluster at two binding sites. One is occupied in the S1 state and the other during the S1 to S2 transition [182, 183]. The lineshape of the S2 mulitiline signal is altered signiﬁcantly [184– 186]. Proposals that ammonia replaces one water molecule have been corroborated by 2H 2 ESE modulation experiments in ammonia-treated samples incubated with H2O [187]. The molecule being chemically in between methanol and ammonia, is methylamine. It has been reported that any amines being larger than ammonia do not bind to the manganese cluster directly because of steric reasons [186]. Although a reduction of the OEC with substituted hydroxylamine has been reported [188], this does not necessarily mean that this molecule has to bind to the manganese cluster.

8.3 Experiment

8.3.1 Samples

PS II membranes were prepared from fresh spinach leaves [9] by a 2 min incubation of the isolated thylakoids with the detergent Triton X-100 [115]. The samples were then resus- pended to a chlorophyll (Chl) concentration of 3.5 mg(Chl)/ml in betaine buﬀer (pH 6.5,

1 M betaine, 50 mM MES, 15 mM NaCl, 5 mM MgCl2, 5 mM CaCl2).

For the O2 measurements, buﬀers and methylamine solutions (buﬀered) with respective pH values were used. Oxygen evolution rates were determined at 25◦C utilizing a Clark- type electrode (Hansatech) and 500 µM K3[Fe(CN)6] and 400 µM DCBQ as acceptors. For EPR measurements the PS II membranes were thawed in the dark on ice, incubated at 8.4. ENZYME KINETICS AND THE EFFECT OF INHIBITORS 97

3.5 mg(Chl)/ml with 1% (0.32 M) of methylamine chloride (buﬀered) for about 1 minute and then transferred into EPR tubes. The samples were then concentrated by centrifugation to about 15 mg(Chl)/ml, frozen in an ethanol/dry ice bath at -80◦C and then transferred into liquid nitrogen. Continuous illumination of the EPR samples was performed at 200 K for 15 min.

8.3.2 cw-EPR at X-band cw-EPR measurements were carried out on a Bruker ESP 300 machine using a standard

Bruker TE102 cavity with an Oxford ESR 9 cryostat. Spectra were recorded at 8.5 K.

8.4 Enzyme kinetics and the eﬀect of inhibitors

Any substance that reduces the velocity of an enzyme-catalyzed reaction can be considered an “inhibitor”. The inhibition of enzyme activity is one of the major regulatory devices of living cells, and one of the most important diagnostic procedures to achieve knowledge concerning the kinetics and the mechanism of the specific enzyme. Inhibition studies often tell something about the specificity of an enzyme, the physical and chemical architecture of the active site. In the following analysis it is assumed that only a single substrate is involved in the reaction, as this assumption is valid for the “water splitting complex”, and that only one type of inhibitor, in this case methylamine, is present at any time. Methylamine can be assumed to be competing in binding to the enzyme with the substrate water itself or with chloride. The presence of chloride is of vital importance for the function of the OEC. In the first case methylamine would act as a competitive inhibitor. This means that the inhibitor and the substrate are mutually exclusive. A competitive inhibitor might be a nonmetabolizable analogue or derivative of the true substrate. There are many examples of competitive inhibition by compounds that bear no structural relationship to the substrate. The combination of the inhibitor with the enzyme causes a change in the conformation (tertiary or quarternary structure) of the enzyme that distorts the substrate site and thereby prevents the substrate from binding. The equilibria describing competitive inhibition are shown below (E : enzyme, S : substrate, I : inhibitor):

Ks k E + S )* ES −→p E + P K E + I )*i EI. 98 CHAPTER 8. METHYLAMINE AS INHIBITOR TO OXYGEN EVOLUTION with

Ki = [E][I]/[EI] (8.1)

Ks = [E][S]/[ES], (8.2) and kp being the rate constant for the breakdown of ES to E + P.

The maximal initial velocity in the presence of the competitive inhibitor equals Vmax (the maximum initial velocity in the absence of inhibitor). The apparent Km (measured as 1 [S] required for 2 Vmax) will increase in the presence of a competitive inhibitor because at any inhibitor concentration, a portion of the enzyme exists in the EI form which has no aﬃnity for S. Let v be the actual velocity of the enzyme reaction v = kp.

An expression relating v, Vmax, [S], Km, [I] and Ki in the presence of a competitive inhibitor can be derived from either rapid equilibrium or steady-state assumptions. The velocity equation for competitive inhibition in reciprocal form is: 1 K [I] 1 1 = m × 1 + × + . (8.3) v Vmax Ki ! [S] Vmax

Thus the slope of the plot 1/v vs. [S] increases by the factor (1 + [I]/Ki)∗Km/Vmax, but the 1/v axis intercept remains 1/Vmax. This equation can be rearranged to: 1 K 1 K = m × [I] + × 1 + m . (8.4) v Vmax × [S] × Ki Vmax [S] ! This allows in the experiment to change the concentration of the inhibitor while leaving the substrate concentration, which is water in the case of PS II, constant. Due to the 1 vulnerability of PS II it is not possible to decrease the water concentration. If the plot v vs. [I] obeys the rate equation for a competitive inhibitor the values of the rate constants can be derived. The value for Vmax can be measured separately in the absence of inhibitor.

8.4.1 Oxygen measurements in the presence of methylamine

+ Methylamine is a Lewis base gets protonated when dissolved in water (CH3NH2 + H2O )* CH3NH3 + OH−). The protonated base probably has no inhibitory effect and therefore only the free base methylamine should be regarded. Changing the pH is an effective measure to vary the concentration of free methylamine. A minor drawback is that the oxygen evolution activity is also influenced by the pH[177]. The concentration of unprotonated methylamine can be calculated using the dissociation 4 constant KDiss = 4.38 ∗ 10− mol/l for methylamine: + [CH3NH3 ] ∗ [KW] KDiss = + (8.5) [CH3NH2] ∗ [H ] 8.4. ENZYME KINETICS AND THE EFFECT OF INHIBITORS 99

10 Vmax/v

0 0 0.0002 0.0004 0.0006 0.0008 concentration of unprotonated methylamin (mol/l)

Vmax Figure 8.1: Plot of v vs. the methylamine concentration at pH 7.5, buﬀered with HEPES. The concentration of unprotonated methylamine has been calculated using formula 8.7. A Dixon plot has been ﬁtted to the measured values.

3.5

2.5 Vmax/v

1.5

1 0 2e-05 4e-05 6e-05 8e-05 concentration of unprotonated methylamine [mol/l]

Vmax Figure 8.2: Plot of v vs. the methylamine concentration at pH 6.5, buﬀered with MES. The concentration of unprotonated methylamine has been calculated according to formula 8.7. A Dixon plot has been ﬁtted to the measured values. 100 CHAPTER 8. METHYLAMINE AS INHIBITOR TO OXYGEN EVOLUTION

pH 6.5 pH 7.5

Vmax [µmol(O2)/(mg(Chl))×h)] 770 870

Km[mol/l] 14.2 5 12.3 4 6 6 5 5 Ki[mol/l] 8.84 ∗ 10− 4.7 ∗ 10− 1.93 ∗ 10− 1.0 ∗ 10−

Table 8.1: Values for Vmax. The values for Km and Ki have been determined according to the Dixon- plots in ﬁgure 8.1 and 8.2, the substrate concentration was assumed to be constant.

14 2 2 KW is 1.27*10− mol /l for water at 25◦C. The actual concentration of the unprotonated form outside the protein can be calculated as follows: [total] [CH3NH2] = K [H+] (8.6) Diss∗ + 1 KW + with [total] = [CH3NH2] + [CH3NH3 ]. In the pH-range from 6.5 to 7.5 the quotient K [H+] Diss∗ is much larger than 1. The equation can be written as: KW

[total] ∗ KW [CH3NH2] = + (8.7) KDiss ∗ [H ] This means that if the pH is changed from 6.5 to 7.5 the concentration of unprotonated methylamine rises tenfold. The figures 8.1 and 8.2 show Dixon plots for competitive inhibition of methylamine at Vmax different pH values. Here v is symbolized by the quotient between the maximum oxygen evolution Vmax and the oxygen evolution activity per time dependent on the concentartion of inhibitor. Shown are two graphs for two different pH values thus effectively changing the actual concentration of free methylamine. In both figures 8.1 and 8.2 linear graphs could be fitted. This means that methylamine acts as a competitive inhibitor at lower concentrations. Unfortunately it is not possible to vary the concentration of water as substrate to further confirm this assumption. Evaluating this graph according to formula 8.4 and knowlegde of Vmax, which is 770µmol(O2)/(mg(Chl))× h) for pH 6.5 and 870µmol(O2)/(mg(Chl))× h) for pH 7.5, leads to the values for Km and

Ki, which are listed in table 8.1. For both pH values the constant Km calculated from the ﬁtted values stays the same within the error. This is also the case for Ki in both experiments.

In the former experiment the concentration of chloride rose together with the concentration of methylamine. To eliminate the inﬂuence of this anion, which plays a decisive although not yet fully understood role in the catalytic function of the OEC, the concentration of chloride was kept constant at 0.1 M in the next experiment. Figure 8.3 shows a plot of the oxygen evolving activity vs. the methylamine concentration. A curve having 8.4. ENZYME KINETICS AND THE EFFECT OF INHIBITORS 101

800

700

600 mol(O2]/(mg(Chl)*h) µ 500

400

oxygen evolution activity in 300

200 0 2e-05 4e-05 6e-05 8e-05 concentration of methyl amine [%]

Figure 8.3: Plot of the oxygen evolution activity vs. the concentration of methyl amine keeping the concentration of chloride constant by addition of magnesium chloride. The pH has been adjusted to 6.5 by using a MES buﬀer. The concentration of unprotonated methylamine has been calculated using equation 8.7.

the parameters of an exponential decay could be fitted to this plot. It could only be A1 x fit assuming a formula y = A0 × e− × + A2 with a finite value for A2, which means, that even at an infinite concentration of the inhibitor methylamine there would still be a certain amount of oxygen evolution activity of ≈ 270µmol(O2)/(mg(Chl))×h), which is one third of the initial enzyme activity with no addition of inhibitor. Unfortunately this result can not be verified by increasing the concentration of methylamine much higher because of its reactions with the added electron acceptors. In figure 8.4 the normalized inverse oxygen evolution activity is plotted vs. the concentration of methylamine,which has been corrected to the concentration of unprotonated methylamine according to equation 8.7. One plot is depicted for the concentration of chloride rising with the concentration of methylamine (from 0.034 M to 0.1 M), for the other the concentration of chloride was kept constant (0.1 M). Dixon plots have been fitted to both data sets. Clearly there is not much of a difference comparing both curves within the error margins. The two Dixon fits have virtually the same fit parameters. This leads to the conclusion, that the inhibition by methylamine occurs without the influence of chloride, which means that methylamine is not competing with chloride. 102 CHAPTER 8. METHYLAMINE AS INHIBITOR TO OXYGEN EVOLUTION

2.5

1.5 normalized inverse oxygen evolution activity [V(max)/v]

1 0 5e-06 1e-05 1.5e-05 2e-05 2.5e-05 concentration of unprotonated methyl amine (mol/l)

Figure 8.4: Comparative Dixon plot for methylamine inhibition. In one experiment (red diamonds) the concentration of chloride is rising proportional to the concentration of methylamine (0.034 M to 0.1 M), in the second experiment (black circles) it is constant at 0.1 M. Dixon plots have been ﬁtted to the measured values.

Figure 8.5: S2-S1-diﬀerence spectra covering the whole range of light-induced signals. The upper spectrum has been recorded with PS II membranes with 1% methylamine added, the lower spectrum is from PS II membranes without additions. Spectrometer conditions: P = 20 mW, ν = 9.65 GHz. 8.4. ENZYME KINETICS AND THE EFFECT OF INHIBITORS 103

Figure 8.6: S2-S1-diﬀerence spectrum of the multiline region. The upper spectrum has been recorded with PS II membranes with 1% methylamine added, the lower spectrum is from PS II membranes without additions. Spectrometer conditions: P = 20 mW, ν = 9.65 GHz. 104 CHAPTER 8. METHYLAMINE AS INHIBITOR TO OXYGEN EVOLUTION

8.4.2 cw-EPR

The S2 state multiline signal

Figure 8.6 shows the light-minus-dark S2 state multiline signals for reasons of resolution. The dark-adapted and then frozen (liquid nitrogen) samples had been illuminated at 200 K for 10 min. Apart from a few changes on the wings of hyperfine peaks on the high field side the spectrum is comparable to the S2 multiline in untreated samples. Also the line spacing of the hyperfine lines remains unchanged with respect to the multiline spectrum of untreated samples.

Both S2 state signals of untreated and methylamine-treated PS II samples

Figure 8.5 shows a comparison of light-dark S2 state multilines for native untreated membrane fragments (bottom) and methylamine-treated samples, with the amine added in the S1 state before freezing (top). Clearly the intensity of the g = 4.1 signal gained in amplitude on the expense of the g = 2 multiline signal. This has also been observed for additions of NH4Cl [186]. The linewidth and g value of the EPR signal at g = 4.1 in the spectra with CH3NH2 added in ﬁgure 8.5 are indistiguishable from those of untreated sampled and g = 4.1 signals reported in the literature [25, 189]. In untreated PS II membrane samples the g = 4.1 EPR signal formed after illumination decays quantitatively at ≈210 K, forming the S2 state multilne signal in its place [190]. The increase in the intensity of the g = 4.1 EPR signal concomitant with a decrease in the mulitiline EPR signal when adding methylamine is increased due to a stabilization of the S2 state g = 4.1 signal species relative to the S2 state multiline EPR signal species.

8.5 Discussion

There are several possibilities how amines can bind in the OEC. To which site the respective amine binds, depends on the amine itself, the temperature and the S-state of the OEC. Ammonia seems to show the widest variety of binding possibilities. It is the only amine able to bind to the manganese cluster directly in a way that its nitrogen atom can be detected by an ESEEM experiment [173]. Although amines can bind in the S1 state of

PS II the binding seems to be more eﬀective in the S2 state [184]. Binding of NH3 in the

S2 state leads to an alteration of the S2 state multiline signal by means of a reduction in line spacing. 8.5. DISCUSSION 105

Primary amines behave in a different manner than ammonia. In terms of enzyme kinetics both species clearly inhibit the function of PS II that is the oxygen evolution activity. Sandusky et al [191–193] found the existence of two binding sites for ammonia to the OEC. One of them is supposed to be the binding site for chloride. The binding of ammonia to this site has a similar inhibitory effect as the removal of chloride [194]. In this case there are clear signs that ammonia competes with chloride for the site. Methylamine as a primary amine does not bind to the chloride binding site as shown in this work probably due to steric effects. Although the addition of methylamine clearly shows to inhibit the oxygen evolution in PS II the change of the concentration of chloride has no effect on the degree of inhibition. This is shown in figure 8.4. The type of inhibition seems to be a competitive inhibition according to the Dixon behaviour shown in figure 8.2. It can therefore be assumed that the binding site for this primary amine is identical to the second binding site for ammonia. It is not the chloride binding site. The molecule to which methylamine is considered to be the competitor is water. The second site to which methylamine should bind, is therefore the binding site of one of the water molecules. The access of substrate water to the OEC is restricted by extrinsic proteins. To avoid side reactions and destruction of this vulnerable cofactor (due to its high redox potential) only few, perhaps only two water molecules are allowed to enter the ligand sphere of the manganese cluster. Here it binds to specific binding sites. These two sites are chemically different and get possibly occupied at different states of the Kok-cycle [195, 196].

Ammonia and also methylamine affect the S2 EPR signals. Both species have an effect on the ratio of the relative amplitudes of the g = 4.1 and the multiline signal but only ammonia changes the appearance of the multiline signal. The g = 4.1 signal amplitude is enhanced already, if the addition of ammonia/amine is made in the dark, where the S1 state prevails, the sample is frozen and illuminated at 200 K to enter the S2 state. The shape of the multiline signal remains unchanged in this case. There are two possible explanations for this phenomenon. Assuming that the g = 4.1 signal arises from a different conformation of the OEC and the manganese cluster therein methylamine should act as a stabilizer in binding to the OEC. In this case the g = 4.1 signal would originate from the same species only differing by conformational changes. The change from the g = 4.1 signal would proceed kinetically inhibited due to the low temperature. The other possibility is that the g = 4.1 signal and the multiline signal arise from two different sites which by electron transfer between them are brought into paramagnetism. Here the methylamine should act as a modulator in changing the probability of electron transfer between the two species or even block it. This possibility has been excluded by experiments transforming the g = 4.1 signal into the multiline signal. Here the g = 4.1 signal was first generated alone by illuminating ammonia-treated samples at 130 K. After this the sample was heated to 210 K where the multiline signal could be generated at the 106 CHAPTER 8. METHYLAMINE AS INHIBITOR TO OXYGEN EVOLUTION same amount as in untreated samples [25, 184]. This was taken as a prove that amines do not interfere with the electron transfer in PS II at low temperatures Binding of amines to the OEC is supposed to occur at relatively high temperatures. Ex- periments by Beck et al [184] showed, that NH4Cl-treated samples illuminated at 210 K showed only a small g = 4.1 signal which rose dramatically with respect to and on cost of the g = 2 multiline signal. This was interpreted to be due to an enhanced binding of

NH3 in the S2. However, in this work the g = 4.1 signal is enhanced even though the illumination took place at 200 K.

Ammonia can occupy a second binding site when the sample is illuminated at 0 ◦C and hereby changes the magnetic properties of the manganese cluster, which then shows a diﬀerent multiline signal. This site is not accessible to primary amines. It may be the chloride binding site close to the manganese cluster. The inhibition of oxygen evolution by addition of methylamine is not aﬀected by changes in the concentration of chloride, whereas ammonia competes with chloride for its binding site. Putting this together the binding site for primary amines could be one of the binding sites for substrate water. They do not bind to the binding site for chloride which is only accessible for ammonia possibly because of its small size. ESEEM measurements are necessary to show the proximity of a methylamine molecule to the manganese cluster. Chapter 9

Summary and Outlook

The objective of this thesis was the investigation of some paramagnetic species at the donor site of Photosystem II by means of different EPR techniques. Measurements have been carried out using TR-EPR of the triplet state of the primary donor with the aim to find the location and orientation of the respective triplet molecule within the PS II- 1 complex. H-ENDOR and HYSCORE have been measured of the stable radical YD• , and the HYSCORE spectrum has been simulated to gain the hyperfine coupling parameters of this molecule. 1H-ENDOR on protons surrounding the OEC at different oxidation states led to information on the coordination of water or one of its derivatives. 55Mn-ENDOR of the OEC has been measured. cw-EPR was used as a supporting tool for observing the inhibited enzyme kinetics of PS II. All these experiments showed the versatility of EPR techniques if one is confronted with a specific problem. The results shall be mentioned shortly in the following:

The Light-induced Triplet State in PS II

Using light-excitation on PS II crystals from Synechococcus elongatus in which the electron transfer beyond the pheophytin has been blocked it could be shown that the recombination triplet state is conﬁned to a monomeric chlorophyll due to its D and E values. The analysis of the orientation-dependent EPR spectra in the single crystals with a self- written simulation program yielded precise data for the orientation of the D tensor in the unit cell. Since the D tensor is closely related to the geometrical structure of the porphyrin ring of the chlorophyll, the assignment and full orientation in space for this molecule could be derived. Together with structural information from X-ray crystallogra-

107 108 CHAPTER 9. SUMMARY AND OUTLOOK phy relating symmetry elements of the unit cell with molecular coordinates the number of possible chlorophyll molecules hosting the triplet state could be diminished to two. The angles of the D tensor principal axes with the pseudo C2 axis have been determined to be

302◦ for Z, 622◦ for X and 892◦ for Y. With the help of the first value it turned out that the triplet state should reside on one of the “accessory” chlorophylls ChlD1 or ChlD2. The X and Y axes define the in-plane orientation of the chlorophyll, an information that has so far not been available from the X-ray crystallographic data. Hereby it was possible to localize the fifth isocyclic ring of the chlorophyll. The orientation of the molecule is the same as for its structural counterpart in bRC’s although the triplet state is localized on a different chlorophyll. Assuming similar singlet-triplet splittings for all Chl molecules 3 within the PS II rc the localization of the triplet state “ P680” at cryogenic temperatures on an “accessory” chlorophyll implies that the same molecule should also be the trap for the singlet excitation and initiate the charge separation process.

Pulsed 1H-ENDOR and HYSCORE of the Stable Tyrosine Radical Y D• ox

Great deviations exist in the literature concerning the assignment and the values of the 1 hyperfine coupling parameters for YD• ox in spinach PS II. Combination of H-ENDOR and HYSCORE data was successful in solving this problem. From the 1H-ENDOR the hfc’s could be derived for the spinach PS II that had before turned out to be good simulation parameters for EPR spectra of YD• ox in Synechococcus elongatus. Due to overlap of couplings at the proton matrix region of the ENDOR spectrum not all hfc’s could be derived solely from it, the application of HYSCORE proved to be complementary. The hyperfine data from ENDOR were taken as starting parameters for the simulation of the HYSCORE spectrum. The experimental HYSCORE spectrum could be simulated, and hereby the whole set of all hyperfine parameters the the tyrosine radical was gained. It turned out that a graphical evaluation of hfc’s only from the experimetal HYSCORE spectrum can lead to incorrect values. Together with the numerical values also the direction cosines for the respective hyperfine tensors with respect to the g tensor axes could be obtained by simulations of the HYSCORE spectrum which are in accordance with the values for YD• in PS II of Synechococcus elongatus obtained by Hofbauer et al. [89, 112]. 109

1 Pulsed H-ENDOR of the S 2 and the S2 State of the OEC −

1 The H-ENDOR spectra for the S2 State in spinach PS II partly showed similar signals close to the proton matrix region as observed by others [137–141]. No larger proton couplings could be observed for this state. The spectra showed a large, highly unstructured background, which could be seen for all ENDOR spectra shown in the respective chapter. It can not be excluded that very anisotropic rhombic hyperfine tensors are hidden in this background. They can not be resolved due to severe broadening. This broadening might be due to the protons coupling to the four manganese ions. 1 For the H-ENDOR spectra of the S 2 state two distinct lines accounting for coupling − values of 7.6 MHz and 3.0 MHz were obtained. The lines vanished in samples treated with D2O which indicates exchangability of the respective protons. Assumptions that only two manganese ions of the OEC contribute to the EPR spectrum are corroborated by a simulation of this signal [150]. The hyperfine coupling data obtained from the ENDOR spectra could be evaluated in this work using an extended point-dipole coupling model where protons couple to a Mn(II)Mn(III) dimer. With this model the distance limits as well as the distance lines of equal hyperfine coupling of the protons to the two manganese ions could be calculated. Assuming different hydrogen-containing groups with geometrical restrictions, possible positions of these with respect to the manganese dimer could be shown. Regarding hydroxyl and water ligation to the manganese cluster it is probable, that the oxygen forms a µ-oxo-bridge which is protonated.

55 Mn-ENDOR of the S2 state of the OEC

55 The Mn-ENDOR measured for the S2 state corroborated the ﬁndings of Peloquin et al. [14] for the lower frequency region. Within error margins the spectra in the frequency region ranging from 80 to 160 MHz are similar. The 55Mn-ENDOR spectrum in this work shows one feature at 280 MHz which is absent in the spectra recorded by Peloquin et al. [14]. This could be an indication for a composition of the manganese cluster diﬀerent to the one postulated by the authors. It is stated that with the experimental data available it is not possible to unambiguously determine the composition of the manganese cluster nor the oxidation states of the single manganese ions. 110 CHAPTER 9. SUMMARY AND OUTLOOK

Methylamine as an Inhibitor to Oxygen Evolution

The molecule methylamine acts as inhibitor to oxygen evolution. Oxygen evolution measurements have been carried out with addition of methylamine. The evaluation of the data shows a Dixon-dependence, which accounts for competitive inhibition. Linear re- gression of the resulting graphs provided the enzyme kinetic rate constant Km and the rate of inhibition KI. It is assumed in the literature that methylamine competes with chloride. This ﬁnding could not be corroborated, the Dixon-graphs for methylamine for constant and varying chloride concentrations showed to be almost identical.

EPR-measurements of the S2 state signals showed that although the addition of methylamine alters the ratio of the amplitudes of the the g = 4 and the MLS it does not alter the MLS if added to PS II samples in the S1 state,which are then frozen and illuminated at 200 K to form the S2 state. Methylamine does not compete with chloride for its binding site. The molecule it is more likely competing with a water binding site.

Outlook

The variety of EPR techniques could be used to answer questions that concern the structure and function of Photosystem II. TR EPR could be used to limit the possible places of the beginning of the charge separation to the two “accessory” chlorophylls. To be able to ﬁnally decide between the two possible chlorophylls mutants lacking one “accessory” chlorophyll or at least preventing electron transfer to it are necessary. Also the change of the ligands to the Chl’s is a possibility.

The hyperfine coupling parameters of YD• could be investigated with HYSCORE and EN- DOR measurements. Although the original function of this amino acid is still unclear, the knowledge of the electronic structure and spatial arrangement of the molecule will help to solve this problem. There is still a lot a work to do in terms of the overreduced states of the OEC. Further measurements on the S 2 state of PS II have to be done to improve the signal-to-noise − ration of the ENDOR spectra. Then it would be possible to gain all hyperfine tensor components and therefore nail down the precise position of the protons in the proximity of the OEC. 55 In order to obtain unique data sets the Mn-ENDOR of the S2 state MLS should be measured rather at the outer flanks to be able to distinguish between different manganese species. For such measurements the S/R of the MLS EPR-signal must, however, be im- 55 proved at least by one order of magnitude. Mn-ENDOR measurements on the S 2 state − 111 should be done, since here the signal-to-ratio is much better than for the S2 state. If the EPR-signal really originates from a Mn dimer the 55Mn-ENDOR will be easier to interpret and the electronic structure of the OEC in this oxidation state could be concluded. EPR methods that are able to detect nitrogen nuclei are necessary to decide whether methylamin is bound close to the manganese cluster. ESEEM measurements have been done but no modulations accounting for nitrogen close to the manganese cluster could be observed so far. 112 CHAPTER 9. SUMMARY AND OUTLOOK Chapter 10

Appendix

10.1 The Program Used for the Simulation of the Triplet Spec- tra

The program has been written in Octave. (www.octave.org)

# THIS PROGRAM SIMULATES TRIPLET SPECTRA RECORDED FROM A CRYSTAL OF THE # SPACE GROUP P212121 ROTATED IN AN EXTERNAL FIELD. # EACH SITE CONSISTS OF TWO MONOMERS RELATED BY A NON-CRYST. AXIS 1; clear #*************************************************************************** # TRANSFORMS ANGLES FROM DEGREE INTO RAD function y = rad(x) y=pi/180*x; endfunction #*************************************************************************** # GENERATION OF A ROTATION MATRIX AROUND THE THREE EULER ANGLES # DEFINITION: PSI AROUND Z → THETA AROUND X → PHI AROUND Z function y = euler(phi,theta,psi)

# ROTATION AROUND PHI e phi=[cos(phi) -sin(phi) 0; sin(phi) cos(phi) 0; 0 0 1];

# ROTATION AROUND THETA e theta=[1 0 0; 0 cos(theta) -sin(theta); 0 sin(theta) cos(theta)];

# ROTATION AROUND PSI e psi=[cos(psi) -sin(psi) 0; sin(psi) cos(psi) 0; 0 0 1];

# RESULTING TOTAL ROTATION MATRIX y=e phi*e theta*e psi;

113 114 CHAPTER 10. APPENDIX

endfunction #*************************************************************************** #CALCULATION OF THE REAL ORIENTATION OF THE D-TENSOR BY APPLYING #OF ALL SYMMETRY OPERATIONS function B = spektrum (Orientierung) #*************************************************************************** #INPUT OF THE ZERO FIELD PARAMETERS D AND E

DD=2.87*2.99792458e8; EE=0.43*2.99792458e8;

#RESULTING D-TENSOR

D=[-2*DD/3 0 0 0 DD/3+EE 0 0 0 DD/3-EE ]/1e6; #*************************************************************************** # ANGLES OF THE D-TENSOR WITH RESPECT TO THE CRYSTAL AXES SYSTEM # INPUT IN DEGREE (PHI,THETA,PSI)

D or grad=[132,60-2,73]; #*************************************************************************** #ANGLES OF THE CRYSTAL AXES WITH RESPECT TO THE MAGNETIC FIELD VECTOR #INPUT IN DEGREE (PHI,THETA,PSI) kris or grad=[2;90;-4]; #************************************************************************** # SYMMETRY OPERATIONS OF THE SPACE GROUP P212121 # INPUT IN DEGREE (PHI,THETA,PSI)

R sym 1 grad=[0;0;0];

R sym 2 grad=[0;0;180];

R sym 3 grad=[0;180;0];

R sym 4 grad=[0;180;180]; #*************************************************************************** # INPUT OF FURTHER PARAMETERS

# MIKROWAVE FREQUENCY global nue; nue=9.827e9-0.0169e9;

# PLANCK’S NUMBER, BOHR’S MAGNETON, G-VALUE h=6.6262e-34; mue bohr=927.400899e-26;g eﬀ=2.002319304386; #*************************************************************************** # TRANSFORMATION OF ALL ANGLES

D or=rad(D or grad); kris or=rad(kris or grad); R sym 1=rad(R sym 1 grad); R sym 2=rad(R sym 2 grad); R sym 3=rad(R sym 3 grad); R sym 4=rad(R sym 4 grad); #*************************************************************************** # INPUT OF THE NON-CRYSTALLOGRAPHIC DIMER ROTATION MATRIX

Ma dimer= [ -0.840997429145, +0.314390761750, -0.440320080284;

+0.314390761750, -0.378365075855, -0.870630988595;

-0.440320080284, -0.870630988595, +0.219362504999 ]; #*************************************************************************** # GENERATION OF THE ROTATION MATRICES

# ROTATION MATRIX OF THE D-TENSOR WITH RESPECT TO THE CRYSTAL SYMMETRY 10.1. THE PROGRAM USED FOR THE SIMULATION OF THE TRIPLET SPECTRA 115

R D=euler(D or(1),D or(2),D or(3));

# ROTATION MATRICES FOR THE SYMMETRY OPERATIONS OF THE P212121 SPACE GROUP

R l 1=euler(R sym 1(1),R sym 1(2),R sym 1(3)); R l 2=euler(R sym 2(1),R sym 2(2),R sym 2(3)); R l 3=euler(R sym 3(1),R sym 3(2),R sym 3(3)); R l 4=euler(R sym 4(1),R sym 4(2),R sym 4(3)); #************************************************************************** # INPUT OF THE PAULI SPIN MATRICES FOR AN (S=1) SYSTEM

X=[1,0,0;0,0,0;0,0,-1]; Y=[0,1/sqrt(2),0;1/sqrt(2),0,1/sqrt(2);0,1/sqrt(2),0]; Z=[0,-i/sqrt(2),0;i/sqrt(2),0,-i/sqrt(2);0,i/sqrt(2),0]; #************************************************************************** # CALCULATION OF THE RESONANCE POSITIONS FOR ALL SITES

# FIRST SITE, FIRST MONOMER global schliemann; global HZ; for j=1:length(Orientierung)

R=euler(kris or(1)+rad(Orientierung(j)),kris or(2),kris or(3)); #*************************************************************************** # FIELD PART OF THE HAMILTONIAN

HZ = g eﬀ*mue bohr/h*X;

#*************************************************************************** # TOTAL ROTATION MATRIX

D 1=R*R l 1*R D*D*R D’*R l 1’*R’; #*************************************************************************** # HAMILTONIAN WITHOUT FIELD schliemann = X*X*D 1(1,1) + Y*X*D 1(2,1) + Z*X*D 1(3,1) + X*Y*D 1(1,2) + Y*Y*D 1(2,2) + Z*Y*D 1(3,2) + X*Z*D 1(1,3) + Y*Z*D 1(2,3) + Z*Z*D 1(3,3); schliemann = schliemann*1e6; # VALUE IN Hz #*************************************************************************** # CALCULATION OF THE MAGNETIC FIELD RESONANCE POSITIONS (SUBROUTINE)

B1=fsolve(resonanzabweichungpos, 0.35); B2=fsolve(resonanzabweich35); #************************************************************************** # CALCULATION FOR ALL OTHER ELEMENTS WITHIN THE UNIT CELL

# FIRST SITE, SECOND MONOMER

D 1kris=R*R l 1*Ma dimer*R D*D*R D’*Ma dimer’*R l 1’*R’; schliemann = X*X*D 1kris(1,1) + Y*X*D 1kris(2,1) + Z*X*D 1kris(3,1) + X*Y*D 1kris(1,2) + Y*Y*D 1kris(2,2) + Z*Y*D 1kris(3,2) + X*Z*D 1kris(1,3) + Y*Z*D 1kris(2,3) + Z*Z*D 1kris(3,3); schliemann = schliemann*1e6; B3=fsolve(resonanzabweichungpos, 0.35); B4=fsolve(resonanzabweichungneg,0.35);

# SECOND SITE, FIRST MONOMER

D 2=R*R l 2*R D*D*R D’*R l 2’*R’; schliemann = X*X*D 2(1,1) + Y*X*D 2(2,1) + Z*X*D 2(3,1) + X*Y*D 2(1,2) + Y*Y*D 2(2,2) + Z*Y*D 2(3,2) + X*Z*D 2(1,3) + Y*Z*D 2(2,3) + Z*Z*D 2(3,3); schliemann = schliemann*1e6; B5=fsolve(resonanzabweichungpos, 0.35); B6=fsolve(resonanzabweichungneg,0.35); 116 CHAPTER 10. APPENDIX

# SECOND SITE, SECOND MONOMER

D 2kris=R*R l 2*Ma dimer*R D*D*R D’*Ma dimer’*R l 2’*R’; schliemann = X*X*D 2kris(1,1) + Y*X*D 2kris(2,1) + Z*X*D 2kris(3,1) + X*Y*D 2kris(1,2) + Y*Y*D 2kris(2,2) + Z*Y*D 2kris(3,2) + X*Z*D 2kris(1,3) + Y*Z*D 2kris(2,3) + Z*Z*D 2kris(3,3); schliemann = schliemann*1e6; B7=fsolve(resonanzabweichungpos, 0.35); B8=fsolve(resonanzabweichungneg,0.35);

# THIRD SITE, FIRST MONOMER

D 3=R*R l 3*R D*D*R D’*R l 3’*R’; schliemann = X*X*D 3(1,1) + Y*X*D 3(2,1) + Z*X*D 3(3,1) + X*Y*D 3(1,2) + Y*Y*D 3(2,2) + Z*Y*D 3(3,2) + X*Z*D 3(1,3) + Y*Z*D 3(2,3) + Z*Z*D 3(3,3); schliemann = schliemann*1e6; B9=fsolve(resonanzabweichungpos, 0.35); B10=fsolve(resonanzabweichungneg,0.35);

# THIRD SITE, SECOND MONOMER

D 3kris=R*R l 3*Ma dimer*R D*D*R D’*Ma dimer’*R l 3’*R’; schliemann = X*X*D 3kris(1,1) + Y*X*D 3kris(2,1) + Z*X*D 3kris(3,1) + X*Y*D 3kris(1,2) + Y*Y*D 3kris(2,2) + Z*Y*D 3kris(3,2) + X*Z*D 3kris(1,3) + Y*Z*D 3kris(2,3) + Z*Z*D 3kris(3,3); schliemann = schliemann*1e6; B11=fsolve(resonanzabweichungpos, 0.35); B12=fsolve(´resonanzabweichungneg`,0.35);

# FOURTH SITE, FIRST MONOMER

D 4=R*R l 4*R D*D*R D’*R l 4’*R’; schliemann = X*X*D 4(1,1) + Y*X*D 4(2,1) + Z*X*D 4(3,1) + X*Y*D 4(1,2) + Y*Y*D 4(2,2) + Z*Y*D 4(3,2) + X*Z*D 4(1,3) + Y*Z*D 4(2,3) + Z*Z*D 4(3,3); schliemann = schliemann*1e6; B13=fsolve(resonanzabweichungpos, 0.35); B14=fsolve(resonanzabweichungneg,0.35);

# FOURTH SITE, SECOND MONOMER

D 4kris=R*R l 4*Ma dimer*R D*D*R D’*Ma dimer’*R l 4’*R’; schliemann = X*X*D 4kris(1,1) + Y*X*D 4kris(2,1) + Z*X*D 4kris(3,1) + X*Y*D 4kris(1,2) + Y*Y*D 4kris(2,2) + Z*Y*D 4kris(3,2) + X*Z*D 4kris(1,3) + Y*Z*D 4kris(2,3) + Z*Z*D 4kris(3,3); schliemann = schliemann*1e6; B15=fsolve(resonanzabweichungpos, 0.35); B16=fsolve(resonanzabweichungneg,0.35); #*************************************************************************** # WRITING OF THE MATRIX OF RESULTS

B(j,:)=[Orientierung(j), B1,B2,B3,B4,B5,B6,B7,B8,B9,B10,B11,B12,B13,B14,B15,B16]; endfor endfunction #*************************************************************************** # CALCULATION OF FREQUENCY RESONANCE POSITIONS PER ELEMENT (SUBROUTINE)

# ABSORPTIVE function err pos=resonanzabweichungpos(B) global nue; global schliemann; global HZ; 10.1. THE PROGRAM USED FOR THE SIMULATION OF THE TRIPLET SPECTRA 117

# TOTAL HAMILTONIAN

H=schliemann+B*HZ;

# CALCULATION OF THE EIGEN VALUES

[v,E]=eig(H);

# EIGEN VALUE VECTOR huf=sort(real(diag(E))); nu pos= abs(huf(2)-huf(1)); nu halb=abs(huf(3)-huf(1));

# RESONANCE FREQUENCY err pos=(nu pos-nue)*abs(nu halb-nue); endfunction;

# EMISSIVE function err neg=resonanzabweichungneg(B) global nue; global schliemann; global HZ;

H=schliemann+B*HZ;

[v,E]=eig(H); huf=sort(real(diag(E))); nu neg= abs(huf(3)-huf(2)); nu halb=abs(huf(3)-huf(1)); err neg=abs(nu neg-nue)*abs(nu halb-nue); endfunction; #*************************************************************************** # GENERATION OF A VECTOR WITH THE EXPERIMENTAL ROTATION ANGLES for n=1:37

Orientierung(n) = n*5-5; endfor #*************************************************************************** # GENERATION OF THE STICK SPECTRA lines=spektrum(Orientierung); lines(:,2:17)=lines(:,2:17)*10000;

# FIELDAXIS anfang=3095 ; ende=3895; abstand= (ende-anfang)/255;

# GENERATION OF THE BASIC MATRIX strichspek=zeros(256,38);

# FIRST COLUMN ==>FIELD for i=1:256 strichspek(i,1)=anfang + (i-1)*abstand; 118 CHAPTER 10. APPENDIX

endfor

# CALCULATION OF THE STICK SPECTRA WITH THE CALCULATED VALUES for k=2:38 for i=1:256 for j=2:17

v=abs(lines((k-1),j)-strichspek(i,1)); w=v/abstand; if w <= 0.5 & & ﬂoor(j/2)==(j/2) strichspek(i,k)=strichspek(i,k)+1; elseif w <= 0.5 & & ﬂoor(j/2)<(j/2) strichspek(i,k)=strichspek(i,k)-1; endif endfor endfor endfor strichspek; #************************************************************************** # GAUSSFOLDING OF THE STICK SPECTRA function y=gaussfaltung(fusel,linienbreite) for i=1:256 if i<=128 gausskurve(i)=10/sqrt(2*pi)/linienbreite*exp(-(i)2/linienˆ breite2/2);ˆ endif if i>128 gausskurve(i)=10/sqrt(2*pi)/linienbreite*exp(-(i-256)2/linienˆ breite2/2);ˆ endif endfor; zw=gausskurve;

# FOURIERTRANSFORMATION OF THE GAUSSIAN CURVE fougauss=fft(zw); y1=fft(fusel); falt=fougauss.*y1; y=real(ifft(falt)); endfunction #*************************************************************************** # INTRODUCTION OF THE LINEWIDTH PARAMETER linienbreite=4; #************************************************************************************* # FOURIERTRANSFORMATION OF THE STICK SPEKTRA FROM STRICHSPEK for i=2:38 simspek(:,1)=strichspek(:,1); simspek(:,i)=gaussfaltung(strichspek(:,i),linienbreite); endfor #*************************************************************************** # NORMALIZATION OF THE THEORETICAL SPECTRA 10.1. THE PROGRAM USED FOR THE SIMULATION OF THE TRIPLET SPECTRA 119

for tar=2:38 simspek(:,tar)=simspek(:,tar)/sqrt(sumsq(simspek(:,tar))); endfor #************************************************************************** # LOADING OF THE EXPERIMENTAL SPEKTRA load -force /home/kongo/micha/wust/spektrum2001/p680trip2/gnu/octave/ps2 160701 0 -ascii; . . . . . load -force /home/kongo/micha/wust/spektrum2001/p680trip2/gnu/octave/ps2 160701 180 -ascii; #************************************************************************** # GENERATION OF THE MATRIX OF RESULTS matrix=zeros(256,38); anf=3095; ende=3895; bereich = (ende-anf) abstand=bereich/255;

# FIRST COLUMN ==>FIELD for j=1:256 spekex(j,1)= anf+abstand*(j-1); endfor #*************************************************************************** # EXTRACTION, NORMALIZATION OF THE SPECTRA, WRITING INTO THE MATRIX for i=1:37 zack=eval(sprintf(ps2` 160701 % d`,(5*i-5))); zack(:,2)=zack(:,2)/sqrt(sumsq(zack(:,2))); spekex(:,(i+1))=zack(:,2); endfor #************************************************************************** # OPTIONAL SWITCH TO GENERATE A POSTSCRIPT FILE

# gset term post portrait color # gset output rotation.ps #*************************************************************************** # OUTPUT clearplot hold on gset nokey gset xrange[3095:3895]; gset yrange[-10:190]; gset ytics 10 gset xlabel Magnetic Field [G] gset ylabel Angle

for j=2:38 plot(simspek(:,1),simspek(:,j)*15+(j-2)*5,’3’) plot(spekex(:,1),spekex(:,j)*15+(j-2)*5,’1’) endfor hold oﬀ 120 CHAPTER 10. APPENDIX

gset term x11 #*************************************************************************** # CALCULATION OF THE SUM OF THE SQUARE OF THE DIFFERENCE OF THE SIGNALS # BETWEEN THE CORRESPONDING SIMULATED AND EXPERIMENTAL FIELD POSITIONS # SIMULATED: simspek, EXPERIMENTAL: spekex sumstaab=0; for i=2:38 for j=1:256 staab=(spekex(j,i)-simspek(j,i))2;ˆ sumstaab=sumstaab+staab; endfor endfor sumstaab

10.2 Calculation of equipotential lines for coupling nuclei

#Dieses Programm berechnet die Aequipotentiallinien #eines Kerns, welcher mit zwei austauschgekoppelten Atomen #die ihrerseits antiferromagnetisch zu einem s=1/2 koppeln, koppelt, #anhand der mittleren Komponente des hfc Tensors. #Es benutzt den Octave-Code, www.octave.org global psi; global kopplung; global abstand; global spinprojektion1; global spinprojektion2; kopplung=7.6; #Kopplung in MHz abstand=2.7; #Abstand der Kerne in Angstrom˚ spinprojektion1=7/3; #Spinprojektionsfaktor Kern1 spinprojektion2=-4/3; #Spinprojektionsfaktor Kern2 function y=f(x) global psi; global kopplung; global abstand; global spinprojektion1; global spinprojektion2; y(1) = x(1)2ˆ + x(2)2ˆ - 2*x(1)*x(2)*cos(psi) - abstand2;ˆ y(2) = spinprojektion1*x(1)ˆ(-3) + spinprojektion2*x(2)ˆ(-3) - kopplung/79; endfunction #Berechnung fuer positive Kopplung matrix=zeros(361,6); #Matrix:Winkel psi,r1,r1,konvergenz,Koordinaten x,y for winkel=1:361 psi=winkel/180*pi; [x, info] = fsolve (”f”, [4; 4]); matrix(winkel,1)=winkel; matrix(winkel,2)=x(1); matrix(winkel,3)=x(2); matrix(winkel,4)=info; alpha=pi-psi-asin(x(1)/2.7*sin(psi)); matrix(winkel,5)=cos(alpha)*x(1); matrix(winkel,6)=sin(alpha)*x(1); endfor #Berechnung fuer negative Kopplung matrixa=zeros(361,6); kopplung=kopplung*(-1); for winkel=1:361 psi=winkel/180*pi; [x, info] = fsolve (”f”, [2.5; 2]); matrixa(winkel,1)=winkel; matrixa(winkel,2)=x(1); matrixa(winkel,3)=x(2); matrixa(winkel,4)=info; beta=pi-psi-asin(x(2)/2.7*sin(psi)); matrixa(winkel,5)=cos(beta)*x(2); matrixa(winkel,6)=sin(beta)*x(2); 10.2. CALCULATION OF EQUIPOTENTIAL LINES FOR COUPLING NUCLEI 121

endfor #Graphische Darstellung der Potentiallinien clearplot gset nokey gset xlabel ”r [100 pm]” gset ylabel ”r [100 pm]” #gset term post landscape color #gset output ”hyperﬂae.ps” hold on plot(matrix(:,5)-matrix(180,2),matrix(:,6),’3’); plot(-matrixa(:,5)+matrix(180,3),matrixa(:,6),’3’); hold oﬀ 122 CHAPTER 10. APPENDIX Bibliography

[1] Stryer, L., Biochemistry, W.H. Freeman and Co., New York, 3rd edition, 1988.

[2] Calvin, M., & Benson, A.A., The path of carbon in photosynthesis, Science, 107, 476-480, 1948.

[3] Walker, J.C.G., Klein, C., Schidlowski, M., Schopf, J.W., Stevenson, D.J., & Walter, M.R., The environmental evolution of the archean-early proteozoic earth, In Earth’s Earliest Biosphere; its Origin and Evolution, Schopf, J.W., ed., Chapter 11, Princeton University Press, Princeton, 260- 290, 1983.

[4] Barber, J., Introduction to Photosynthesis, Homepage, Wolfson Laboratories, Department of Bio- logical Sciences, Imperial College London, 2003.

[5] Debus, R.J., The manganese and calcium ions of photosynthetic oxygen evolution, Biochim. Bio- phys. Acta, 1102, 269-352, 1992.

[6] Renger, G., Photosynthetic water oxidation to molecular oxygen: Apparatus and mechanism, Biochim. Biophys. Acta, 1503, 210-228, 2001.

[7] Kuwabara, T., & Murata, N., Chemical and physicochemical characterization of the proteins involved in the oxygen evolution system, In Advances in Photosynthesis Research, Sybesma, C. ed., Volume 1, Martinus Nijhoﬀ Publishers, Dordrecht, 371-374, 1984.

[8] Zouni, A., Witt, H.T., Kern, J., Fromme, P., Krauß, N., Saenger, W., & Orth, P., Crystal structure of Photosystem II from synechococcus elongatus at 3.8A˚ resolution, Nature, 409, 739-743, 2001.

[9] Berthold, D.A., Babcock, G.T., & Yocum, C.F., A highly resolved, oxygen-evolving Photosystem II preparation from spinach thylakoid membranes, FEBS Lett., 134, 231-234, 1981.

[10] Joliot, P, Barbieri, G, & Chabaud, R., Un nouveau modele des centres photochimique du system II, Photochem. Photobiol., 10, 309-329, 1969.

[11] Kok, B., Forbush, B., & McGloin, M., Cooperation of charges in photosynthetic oxygen evolution, Photochem. Photobiol., 11, 457-475, 1970.

[12] Rutherford, A.W., Photosystem II, the water-splitting enzyme, TIBS, 14, 227-232, 1989.

[13] Britt, R.D., Oxygen evolution, in Oxygenic Photosynthesis: The light reactions, Ort, D.R. and Yocum, C.F. eds, Kluwer Academic Publishers, Dordrecht, 137-164, 1996.

123 124 BIBLIOGRAPHY

[14] Peloquin, J.M., Campbell, K.A., Randall, D.W., Evanchik, M.A., Pecoraro, V.L., Armstrong, 5 W.H., & Britt, R.D., 5Mn ENDOR of the S2 state multiline EPR signal of Photosystem II: Implications on the structure of the tetranuclear Mn cluster, J. Am. Chem. Soc., 122, 10926-10942, 2000.

[15] Yachandra, V.K., Sauer, K., & Klein, M.P., Manganese cluster in photosynthesis: Where plants oxidize water to dioxygen, Chem. Rev., 96, 2927-2950, 1996.

[16] Diner, B.A., & Babcock, G.T., Structure, dynamics, and energy conversion eﬃciency in Photosys- tem II, In Oxygenic Photosynthesis: The Light Reactions, Ort, D.R., & Yocum, C.F. eds., Kluwer Academic Publishers, Dordrecht, 213-247, 1996.

[17] Prasil, O., Adir, N., & Ohad, I., Dynamics of Photosystem II: Mechanism of photoinhibition and recovery processes, In The Photosystems: Structure, Function and Molecular Biology, Topics in Photosynthesis, Vol. II, Elsevier Science Publishers, Amsterdam, 295-348, 1992.

[18] Michel, H., & Deisenhofer, J., Relevance of the photosynthetic reaction center from purple bacteria to the structure of Photosystem II, Biochemistry, 27, 1-7, 1988.

[19] Satoh, K., Isolation and properties of the Photosystem II reaction center, In The Photosynthetic Reaction Center, Deisenhofer, J., & Norris, J.R. eds., Vol. I, Academic Press, San Diego, 289-318, 1993.

[20] Fowler, C.F., & Kok, B., Proton evolution associated with the photooxidation of water in photosynthesis, Biochim. Biophys. Acta, 357, 299-307, 1974.

[21] Blankenship, R, & Sauer, K., Manganese in photosynthetic oxygen evolution. Electron paramagnetic resonance study of the environment of Mn in Tris-washed chloroplasts, Biochim. Biophys. Acta, 357, 252-266, 1974.

[22] Coleman, W.J., & Govindjee, A model for the mechanism of chloride activation of oxygen evolution in Photosytem II, Photosynth. Res., 13, 199-223, 1987.

[23] Tamura, N., & Cheniae, G., Photoactivation of the water-oxidizing complex in Photosystem II membranes depleted of Mn and extrinsic proteins. I. Biochemical and kinetic characterization, Biochim. Biophys. Acta, 890, 179-194, 1987.

[24] Messinger, M., Towards understanding the chemistry of photosynthetic oxygen evolution: dynamic structural changes, redox states and substrate water binding of the Mn cluster in Photosystem II, Biochim. Biophys. Acta, 1459, 481-488, 2000.

[25] Casey, J.L., & Sauer, K., EPR detection of a cryogenically photogenerated intermediate in photosynthetic oxygen evolution, Biochim. Biophys. Acta, 767, 21-28, 1984.

[26] Dismukes, G.C., & Siderer, Y., EPR spectroscopic observation of a manganese center associated with water oxidation in spinach chloroplasts, FEBS Lett., 121, 78-80, 1980.

[27] Barry, B.A., & Babcock, G.T., Tyrosine radicals are involved in the photosynthetic oxygen-evolving system, Proc. Natl. Acad. Sci USA, 84, 7099- 7103, 1987. BIBLIOGRAPHY 125

[28] Boerner, R.J., & Barry, B.A., Isotopic labeling and EPR spectroscopy show that a tyrosine radical is the terminal electron donor, Z, in manganese-depleted Photosystem II preparations, J. Biol. Chem., 268, 17151-17154, 1993.

[29] Debus, R.J., Barry, B.A., Sithole, I, Babcock, G.T., & McIntosh, L., Directed mutagenesis indicates + that the donor to P680 is tyrosine-161 of the D1 polypetide, Biochemistry, 27, 9071-9074, 1988.

[30] Gilchrist, M.L., Ball, J.A., Randall, D.W., & Britt, R.D., Proximity of the manganese cluster of

Photosystem II to the redox active tyrosine YZ, Proc. Natl. Acad. Sci. USA, 92, 9545-9549, 1995.

[31] Pecoraro, V.L., Gelasco, A., & Baldwin, M., Modeling the chemistry and properties of multin- uclear manganese enzymes, In Bioinorganic chemistry, Kessissoglu, D.P. ed., Kluwer Academic Publishers, 287-298, 1995.

[32] Babcock, G.T., The photosynthetic oxygen-evolving process, In Photosynthesis, Amesz, J. ed., Vol. 15, Elsevier, 125-158, 1987.

[33] Tommos, C., Tang, X.S., Warncke, K., Hoganson, C.W., Styring, S., McCracken, J., Diner, B.A., Babcock, G.T., Spin-density distribution, conformation and hydrogen bonding of the redox-active

tyrosine YZ in PS II from multiple electron paramagnetic resonance spectroscopies: Implications for photosynthetic oxygen evolution, J. Am. Chem. Soc., 117, 10325-10335, 1995.

[34] Hoganson, C.W., Lydakis-Simantiris, N., Tang, X.S., Tommos, C., Warncke, K., Babcock, G.T., Diner, B.A., McCracken, J., & Styring, S., A hydrogen-atom abstraction model for the function of

YZ in photosynthetic oxygen-evolution, Photosynth. Res., 46, 177-184, 1995.

[35] Faller, P., Debus, R.J., Brettel, K., Suguira, M, Rutherford, A.W., & Boussac, A., Rapid formation of the stable tyrosyl radical in Photosystem II, Proc. Natl. Acad. Sci. (USA), 98, 14368-14373 2001.

[36] Magnuson, A., Rova, F., Mamedov, F., Fredriksson, P.-O., & Styring, S., Involvement of Cytochrome-b559 and tyrosine-D in protection against photoinhibition during in vivo photoactivation of Photosystem II, Biochim. Biophys. Acta, 1411, 180-191, 1999.

[37] Styring, S.A., & Rutherford, A.W., In the oxygen-evolving complex of Photosystem II the S0 state

is oxidized to the S1 state by D+ (Signal IIslow), Biochemistry, 26, 2401-2405, 1987.

[38] Babcock, G.T., Barry, B.A., Debus, R.J., Hoganson, C.W., Atamian, M., McIntosh, L, Sithole, I., & Yocum, C.F., Water oxidation in Photosystem II: from radical chemistry to multielectron chemistry, Biochemistry, 28, 9557-9565, 1989.

[39] Atherton, N.M., Principles of electron spin resonance, Ellis Horwood PTR Prentice Hall, New York, 1993.

[40] Carrington, A., & McLachlan, A.D., Introduction to magnetic resonance with applications to chemistry and chemical physics, Chapter 9, Harper & Row, 1967.

[41] Abragam, A., Principles of nuclear magnetism, Oxford Science Publications, Oxford, 1961.

[42] Poole, C.P., & Farach, H.A., Theory of magnetic resonance, Wiley & Sons, 1987. 126 BIBLIOGRAPHY

[43] Abragam, A., & Bleaney, B., Electron paramagnetic resonance of transition ions, Oxford University Press, London, 1970.

[44] Mohr, P.J., & Taylor, B.N., CODATA recommended values for the fundamental physical constants, Rev. Mod. Phys., 72, 351-495, 2000.

[45] Feher, G., Observation of nuclear magnetic resonances via the electron spin resonance line, Phys. Rev., 103, 834-835, 1956.

[46] Schweiger, A. Pulsed electron spin resonance spectroscopy: Basic principles, techniques, and examples of applications. Angew. Chem. Int. Ed. Engl., 30, 265, 1991.

[47] Davies, E.R., A new pulse ENDOR technique, Phys. Lett., 47 A, 1-2, 1974.

[48] Mims, W.B., Pulsed ENDOR experiments, Proc. R. Soc., A 283, 452-457, 1965.

[49] Schweiger, A., Puls-Elektronenspinresonanz-Spektroskopie: Grundlagen, Verfahren und Anwen- dungsbeispiele, Angewandte Chemie, 103, 223-250, 1991.

[50] Deligiannakis, Y., Louloudi, M., & Hadjiliadis, N., Electron spin echo envelope modulation (ES- EEM) spectroscopy as a tool to investigate the coordination environment of metal centers, Coord. Chem. Rev., 204, 1-112, 2000.

[51] Dikanov, S.A., & Tsvetkov, Y.D., Electron spin echo envelope modulation (ESEEM), CRC Press, Boca Raton, 1992.

[52] Dikanov, S.A., & Astashkin, A.V., In Advanced EPR: Applications in Biology and Biochemistry, Hoﬀ, A.J. ed., Elsevier, 59-93, 1989.

[53] Mims, W.B., & Peisach, J., In Advanced EPR: Applications in Biology and Biochemistry, Hoﬀ, A.J. ed., Elsevier, 1-57, 1989.

[54] Mims, W.B., Envelope modulation in spin echo experiments, Phys. Rev. B, 5, 2409, 1972.

[55] Mims, W.B., Amplitudes of superhyperﬁne frequencies in the electron spin-echo envelope, Phys. Rev. B, 6, 3543, 1972.

[56] H¨ofer, P., Grupp, A., Nebenfuhr,¨ H., & Mehring, M., Hyperﬁne sublevel correlation (HYSCORE) spectroscopy: A 2D ESR investigation of the squaric acid radical, Phys. Lett., 132, 279, 1986.

[57] Shane, J.J., H¨ofer, P., Reijerse, E.J., & de Boer, E., Hyperﬁne sublevel correlation (HYSCORE) spectroscopy of disordered solids, J. Magn. Res., 99, 596, 1992.

[58] H¨ofer, P., Distortion-free electron spin echo envelope modulation spectra of disordered solids obtained from two and three dimensional HYSCORE experiments, J. Magn. Res., 111, 77-86, 1994.

[59] Tyryshkin, A.M., Dikanov, S.A., & Goldfarb, D., Sum combination harmonics in four-pulse ES- EEM spectra. Study of the ligand geometry in Aqua-Vanadyl complexes in polycrystalline and glass matrices, J. Magn. Res., 105, 271-283, 1993. BIBLIOGRAPHY 127

[60] Gemperle, C., Aebli, G., Schweiger, A., & Ernst, R.R., Phase cycling in pulse EPR, J. Magn. Res., 88, 241, 1990.

[61] D. R. Ort, & C. F. Yocum, Eds., In Advances in Photosynthesis. Oxygenic photosynthesis: The light reactions, (Kluwer, Dordrecht, The Netherlands), Vol. 4,1-9, 1996.

[62] J. Nugent, ed., Photosynthetic water oxidation, Biochim. Biophys. Acta special edition, 1503, 1-260, 2001.

[63] B. Svensson, C. Etchebest, P. Tuﬀery, P. van Kan, J. Smith, & S. Styring, A model for the Pho-

tosystem II reaction center core including the structure of the primary donor P680, Biochemistry, 35, 14486-14502, 1996.

[64] C. R. D. Lancaster, U. Ermler, & H. Michel, The structure of photosynthetic reaction centers from Purple Bacteria as revealed by X-ray crystallography, In Anoxygenic Photosynthetic Bacteria, R. E. Blankenship, M. T. Madigan, C. E. Bauer, eds., Kluwer Academic Publishers, Dordrecht, The Netherlands, Vol. 2, 503-526, 1995.

[65] P. Jordan, P. Fromme, H. T. Witt, O. Klukas, W. Saenger, & N. Krauss, Three-dimensional structure of cyanobacterial Photosystem I at 2.5 A˚ resolution, Nature, 411, 909-917, 2001.

[66] J. Amesz, & A. J. Hoﬀ., Biophysical techniques in photosynthesis, In Advances in Photosynthesis, Govindjee, ed., Kluwer Academic Publishers, Dordrecht, The Netherlands, Vol. 3, Part 2, Magnetic Resonance, 426 pp, 1996.

[67] W. Lubitz, F. Lendzian, & R. Bittl, Radicals, radical pairs and triplet states in photosynthesis, Acc. of Chem. Res., 35, 313-320, 2002.

[68] J. R. Durrant, L. B Giorgi, J. Barber, D. R. Klug, & G. Porter, Characterisation of triplet states in isolated Photosystem II reaction centres: oxygen quenching as a mechanism for photodamage, Biochim. Biophys. Acta, 1017, 167-175, 1990.

[69] R. J. Cogdell, T. D. Howard, R. Bittl, E. Schlodder, I. Geisenheimer, & W. Lubitz, How carotenoids protect bacterial photosynthesis, Philos. Trans. R. Soc. London B, 355, 1345-1349, 2000.

[70] H. Levanon, Time-resolved electron paramagnetic resonance spectroscopy. Principles and applications, In Advances in photosynthesis, J. Amesz, and A. J. Hoﬀ, eds., Kluwer Academic Publishers: Dordrecht, Vol. 3, 211-233, 1996.

[71] D. E. Budil, & M. C. Thurnauer, The chlorophyll triplet state as a probe of structure and function in photosynthesis, Biochim. Biophys. Acta, 1057, 1-41, 1991.

[72] P. L. Dutton, J. S. Leigh, & M. Seibert, Primary processes in photosynthesis: in situ ESR studies on the light induced oxidized and triplet state of reaction center bacteriochlorophyll, Biochem. Biophys. Res. Commun., 46, 406-413, 1971.

[73] H. A. Frank, M. B. McLean, & K. Sauer, Triplet states in Photosystem I of spinach chloroplasts and subchloroplast particles, Proc. Natl. Acad. Sci., 76, 5124- 5128, 1979. 128 BIBLIOGRAPHY

[74] A. W. Rutherford, D. R. Paterson, & J. E. Mullet, A light-induced spin-polarized triplet detected by EPR in Photosystem II reaction centers, Biochim. Biophys. Acta, 635, 205-214, 1981.

[75] J. R. Norris, D. E. Budil, P. Gast, C. H. Chang, O. El-Kabbani, & M. Schiﬀer, Correlation of paramagnetic states and molecular structure in bacterial photosynthtis reaction centers: the symmetry of the primary electron donor in Rhodopseudomonas viridis and Rhodobacter sphaeroides R-26, Proc. Natl. Acad. Sci. USA, 86, 4335-4339, 1989.

[76] F. J. E. van Mieghem, K. Satoh, & A. W. Rutherford, A chlorophyll tilted 30◦ relative to the membrane in the Photosystem II reaction centre, Biochim. Biophys. Acta, 1058, 379-385, 1991.

[77] A. Kamlowski, L. Frankem¨oller, A. van der Est, & D. Stehlik, & A. R. Holzwarth, Evidence 3 for delocalization of the triplet state P680 in the D1D2cyt559-complex of Photosystem II ,Ber. Bunsenges. Phys. Chem., 100, 2045-2051, 1996.

[78] M. Di Valentin, C. W. M. Kay, G. Giacometti, & K. M¨obius, A time-resolved electron nuclear double resonance study of the photoexcited triplet state of P680 in isolated reaction centers of Photosystem II, Chem. Phys. Lett., 248, 434-441, 1996.

[79] F. Lendzian, R. Bittl, & W. Lubitz, Comparative pulsed ENDOR study of the photo-excited triplet

states of BChl a and the primary donor P865 in Rb. sphaeroides R-26 reaction centers, Photosynth. Res., 55, 189-197, 1998.

[80] F. Lendzian, R. Bittl, A. Telfer, J. Barber, & W. Lubitz, Time-resolved ENDOR of the triplet state of P680 in PS II reaction centers, In Photosynthesis: Mechanisms and Eﬀects, Garab, G., ed., Kluwer Academic Publishers: Dordrecht, Vol. II, 1057-1060, 1998.

[81] N. W. Woodbury, & J. P. Allen, The pathway, the kinetics and thermodynamics of electron transfer in wild-type and mutant reaction centers of Purple Nonsulfur Bacteria, In Anoxygenic Pho- tosynthetic Bacteria, R. E. Blankenship, M. T. Madigan, & C. E. Bauer, eds., Kluwer Academic Publishers: Dordrecht, 527-557, 1995.

[82] W. W. Parson, Photosynthetic reaction centers, In Protein electron transport, D. Bendall, ed., Bios Scientiﬁc Publ., Oxford, 125-160, 1996.

[83] B. A. Diner, E. Schlodder, P. J. Nixon, W. J. Coleman, F. Rappaport, J. Lavergne, W. F. J. Ver- maas, & D. A. Chisholm, Site-directed mutations at D1-His198 and D2-His197 of Photosystem II in Synechocystis PCC 6803: Sites of primary charge separation and cation and triplet stabilization, Biochemistry, 40, 9265-9281, 2001.

[84] T. Noguchi, Dual role of triplet localization on the accessory chlorophyll in the Photosystem II reaction center: photoprotection and photodamage of the D1 Protein, Plant Cell Physiol., 43, 1112-1116, 2002.

[85] S. G. Zech, J. Kurreck, H. J. Eckert, G. Renger, W. Lubitz, & R. Bittl, Pulsed EPR measurement +• −• of the distance between P680 and QA in Photosystem II, FEBS Lett., 414, 454-456, 1997.

[86] S. G. Zech, J. Kurreck, G. Renger, W. Lubitz, & R. Bittl, Determination of the distance between OX• −• YZ and QA in Photosystem II by pulsed EPR spectroscopy on light-induced radical pairs, FEBS Lett., 442, 79-82, 1999. BIBLIOGRAPHY 129

[87] A. W. Rutherford, Orientation of EPR signals arising from components in Photosystem II membranes, Biochim. Biophys. Acta, 807, 189-201, 1985.

[88] A. W. Rutherford, & S. Acker, Orientation of the primary donor in isolated Photosystem II reaction centers studied by electron paramagnetic resonance, Biophys. J., 49, 101-102, 1986.

[89] W. Hofbauer, A. Zouni, R. Bittl, J. Kern, P. Orth, F. Lendzian, P. Fromme, H. T. Witt, & W. Lubitz, Photosystem II single crystals studied by EPR spectroscopy at 94 GHz: the tyrosine · radical YD, Proc. Natl. Acad. Sci., 98, 6623-6628, 2001.

[90] J. P. Dekker, E. J. Boekema, H. T. Witt, & M. Rögner, Refined purification and further characterization of oxygen-evolving and Tris-treated photosystem II particles from the thermophilic cyanobacterium Synechococcus sp., Biochim. Biophys. Acta, 936, 307-318, 1988.

[91] A. Zouni, C. Luneb¨ erg, P. Fromme, M. D. Schubert, W. Saenger, & H. T. Witt, Characterisation of single crystals of Photosystem II from the thermophilic cyanobacterium Synechococcus elongatus, In Photosynthesis: Mechanisms and Eﬀects, G. Garab, ed., Kluwer Academic Publishers: Dordrecht, Vol. II, 925-928, 1998.

[92] B. A. Barry, R. J. Boerner, & L. C. de Paula, The use of cyanobacteria in the study of the structure and function of Photosystem II, In The molecular biology of cyanobacteria, D. A. Bryant, ed., Kluwer Academic Publishers: Dordrecht, 217-257, 1994.

[93] K. C. Han, R. J. Shen, M. Ikeuchi, & Y. Inoue, Chemical crosslinking studies of extrinsic proteins in cyanobacterial Photosystem II, FEBS Lett., 355, 121-124, 1994.

[94] A. Zouni, R. Jordan, E. Schlodder, P. Fromme, & H. T. Witt, First Photosystem II crystals capable of water oxidation, Biochim. Biophys. Acta, 1457, 103-105, 2000.

[95] F. Lendzian, R. Bittl, A. Telfer, & W. Lubitz, Hyperﬁne structure of the photoexcited triplet 3 state P680 in plant PS II reaction centers as determined by pulse ENDOR spectroscopy, Biochim. Biophys. Acta, 2002, submitted.

[96] M. C. Thurnauer, J. J. Katz, & J. R. Norris, The triplet state in bacterial photosynthesis; possible mechanisms of the primary photo-act, Proc. Natl Acad., 72, 3270-3274, 1975.

[97] M. C. Thurnauer, & J. R. Norris, The ordering of the zero ﬁeld triplet spin sublevels in the chlorophylls. A magnetophotoselection study, Chem. Phys. Lett, 47, 100-105, 1977.

[98] Biesadka, J., & Saenger, W., personal communication.

[99] D. E. Budil, S. S. Taremi, P. Gast, J. R. Norris, & H. A. Frank, Single crystal electron spin resonance studies of the photochemical reaction center from Rhodobacter sphaeroides wild type strain 2.4.1., Isr. J. Chem., 28, 59-66, 1988.

[100] F. Lendzian, M. Huber, R.A. Isaacson, B. Endeward, M. Plato, B. Bnigk, K. M¨obius, W. Lubitz, & G. Feher, The electronic structure of the Primary Donor cation radical in Rhodobacter sphaeroides R-26: ENDOR and TRIPLE resonance studies in single crystals of reaction centers, Biochim. Biophys. Acta, 1183, 139-160, 1993. 130 BIBLIOGRAPHY

[101] Stubbe, J., Protein radical involvement in biological catalysis, Annu. Rev. Biochem., 58, 257-285, 1989.

[102] Sj¨oberg, B.-M., & Gr¨aslund, A., Ribonucleotide reductase, In Advances from Inorganic Biochem- istry, Theil, E.C., Eichhorn, G.L., & Marzilli, L.G., eds., Vol. 5, Elsevier, 87-110, 1983.

[103] Smith, W.L., Eling, T.E., Kulmacz, R.J., Marnett, L.J., & Tsai, A.L., Tyrosyl radicals and their role in hydroperoxide-dependent activation and inactivation of prostaglandin endoperoxide synthase, Biochemistry, 31, 3-7, 1992.

[104] Whittaker, M.M., & Whittaker, J.W., A tyrosine-derived free radical in apogalactose oxidase, J. Biol. Chem., 265, 9610-9613, 1990.

[105] DeFilippis, M.R., Murthy, C.P., Faraggi, M., & Klapper, M.H., Pulse radiolytic measurement of redox potentials: the tyrosine and tryptophan radicals, Biochemistry, 28, 4847-4853, 1989.

[106] DeFilippis, M.R., Murthy, C.P., Broitman, F., Weintraub, D., Faraggi, M., & Klapper, M.H., Electrochemical properties of tyrosine phenoxy and tryptophan indolyl radicals in peptides and amino-acid analogs, J. Phys. Chem., 95, 3416-3419, 1991.

[107] Hoganson, C.W., Lyadkis, Simantiris, N., Tang, X.S., Tommos, C., Warncke, K., Babcock, G.T., Diner, B.A., McCracken, J., & Styring, S., A hydrogen-atom abstraction model for the function of

YZ in photosynthtic oxygen evolution, Photosynth. Res., 46, 177-184, 1995.

[108] Hoganson, C.W., & Babcock, G.T., Protein-tyrosyl interactions in Photosystem II studied by electron spin resonance and electron nuclear double resonance spectroscopy: Comparison with ribonucleotide reductase and in vitro tyrosine, Biochemistry, 31, 11874-11880, 1992.

[109] Dorlet, P., Rutherford, A.W., & Un, S., Orientation of the tyrosyl D, pheophytin anion, and •− semiquinone QA radicals in Photosystem II determined by high-ﬁeld electron paramagnetic resonance, Biochemistry, 39, 7826-7834, 2000.

[110] Farrar, C.T., Gerfen, G.J., Griﬃn, R.G., Force, D.A., & Britt, R.D., Electronic structure of the YD tyrosyl radical in Photosystem II: A high frequency electron paramagnetic resonance spectroscopic and density functional theoretical study, J. Phys. Chem. B, 101, 6634-6641, 1997.

[111] Rigby, S.E.J., Nugent, J.H.A., & O’Malley, P.J.O., The dark stable tyrosine radical of Photosys- tem II studied in three species using ENDOR and EPR spectroscopies, Biochemistry, 33, 1734-1742, 1994.

[112] Hofbauer, W., Electron paramagnetic resonance at 94 GHz: Methodological developments and studies of photosynthetic reaction centers, PhD thesis, Technische Universit¨at, Berlin, 2001.

[113] Britt, R.D., Electron spin echo methods in photosynthesis research, in Biophysical Techniques in Photosynthesis, Amesz, J, & Hoﬀ A.J. eds., Kluwer Academic Publishers, 235-253, 1996.

[114] Warncke, K., Babcock, G.T., & McCracken, J., Structure of the YD tyrosine radical in Photo- system II as revealed by 2H electron echo envelope modulation (ESEEM) spectroscopic analysis of hydrogen hyperﬁne interactions, J. Am. Chem. Soc., 116, 7332-7340, 1994. BIBLIOGRAPHY 131

[115] Kuwabara, T. & Murata, Inactivation of photosynthetic oxygen evolution and concomitant release of three polypeptides in the Photosytem II particles of spinach chloroplasts, N. Plant. Cell. Physiol., 23, 533-539, 1982.

[116] O’Malley, P.J., & Ellson, D., 1H,13C and 17O isotropic and anisotropic hyperﬁne coupling prediction for the tyrosyl radical using hybrid density functional methods, Biochim. Biophys. Acta, 1320, 65- 72, 1997.

[117] Himo, F., & Gr¨aslund, A., & Eriksson, L.A., Density functional calculations in model tyrosine radicals, Biophys. J., 72, 1556-1567, 1997.

[118] Deligiannakis, Y., Ivancich, A., & Rutherford, A.W., 2D-hyperﬁne sublevel correlation spectroscopy of tyrosyl radicals, Spectrochimica Acta Part A, 58, 1191-1200, 2002.

[119] Dikanov, S.A., & Bowman, M.K., Crosspeak lineshape of two-dimensional ESEEM spectra in disordered S=1/2, I=1/2 spin systems, J. Magn. Res. A, 116, 125-128, 1995.

[120] Dikanov, S.A., & Tyryshkin, A.M., & Bowman, M.K., Intensity of cross-peaks in HYSCORE spectra of S=1/2, I=1/2 spin systems, J. Magn. Res., 144, 228-242, 2000.

[121] Evelo, R.G., Dikanov, S.A., & Hoﬀ, A.J., , Chem. Phys. Lett., 157, 25, 1989.

[122] van Gastel, M., Coremans, J.W.A., Jeuken, L.J.C., Canters, G.W., & Groenen, E.J.J., Electron spin-echo envelope modulation spectrum of azurin at X-band, J. Phys. Chem., 102, 4462-4470, 1998

[123] Zouni, A., Kern, J., Loll, B., Fromme, P., Witt, H.T., Orth, P., Krauß, N., Saenger, W., & Biesadka, J., Biochemical characterization and crystal structure of the water oxidizing Photosys- tem II from Synechococcus elongatus, in 12th International congress on photosynthesis, S05-003, CSIRO Publishers, Melbourne, Australia, Brisbane.

[124] Dismukes, G.C., & Siderer, Y., Intermediates of a polynuclear manganese center involved in photosynthetic oxidation of water, Proc. Natl. Acad. Sci. USA, 78, 274-278, 1981.

[125] Messinger, J., Robblee, J.H., Yu, W.O., Sauer, K., Yachandra, V.K., & Klein,M.P., The S0 state of the oxygen-evolving complex in Photosystem II is paramagnetic: Detection of an EPR multiline signal, J. Am. Chem. Soc., 119, 11349-11350, 1997.

[126] Ahrling,˚ K.A., Peterson, S., & Styring, S., An oscillating manganese electron paramagnetic res-

onance signal from the S0 state of the oxygen evolving complex in Photosystem II, Biochemistry, 36, 13148-13152, 1997.

[127] Boussac, A., Un, S., Horner, O., & Rutherford, A.W., High-spin states (S ≥ 5/2) of the Photo- system II manganese complex, Biochemistry, 37, 4001-4007, 1998.

[128] Britt, R.D., Lorigan, G.A., Sauer, K., Klein, M.P., & Zimmermann, J.-L., The g = 2 multiline

EPR signal of the S2 state of the photosynthetic oxygen-evolving complex originates from a ground spin state, Biochim. Biophys. Acta, 1140, 95-101, 1992. 132 BIBLIOGRAPHY

[129] Ahrling,˚ K., & Pace, R., Simulation of the S2 state multiline electron paramagnetic resonance signal of Photosystem II: A multifrequency approach, Biophys. J., 68, 2081-2090, 1995.

[130] Zheng, M., & Dismukes, G.C., Photosynthetic water oxidation: What have we learned from the multiline EPR signals? in Research in Photosynthesis, Vol II, Murata, N. ed., Kluwer Academic Publishers, 305-308, 1992.

[131] Kusonoki, M., A new paramagnetic hyperﬁne structure eﬀect in manganese tetramers. The origin

of ‘multiline’ EPR signals from an S2 state of a photosynthetic water-splitting enzyme, Chem. Phys. Lett., 197, 108-116, 1992.

[132] Hasegawa, K., Kusonoki, M, Inoue, Y., & Ono, T.-A., Simulations of the S2 state multiline EPR signal in oriented Photosystem II membranes: Structural implications for the manganese cluster in an oxygen-evolving complex, Biochemistry, 37, 9457-9465, 1998.

[133] Ahrling,˚ K.A., Smith, P.J., & Pace, R.J., Nature of the Mn centers in Photosystem II. Modeling and behavior of the g=4 resonances and related signals, J. Am. Chem. Soc., 120, 13202-13214, 1998.

[134] Ahrling,˚ K., & Pace, R.J., Simulation of the S2 state multiline electron paramgnetic resonance signal of Photosystem II: A multifrequency approach, Biophys. J., 68, 2081-2090, 1998.

[135] Zheng, M., & Dismukes, G.C., Orbital conﬁguration of valence electrons, ligand ﬁeld symmetry

and manganese oxidation states of the photosynthetic water oxidizing complex: Analysis of the S2 state multiline EPR signals, Inorg. Chem., 35, 3307-3319, 1996.

[136] Lakshmi, K.V., Eaton, S.S., Eaton, G.R., & Brudvig, G.W., Orientation of the tetranuclear man-

ganese cluster and tyrosine Z in the O2-evolving complex of Photosystem II: An EPR study of the • S2Yz state in oriented acetate-inhibited Photosytem II membranes, Biochemistry, 38, 12758-12767, 1999.

[137] Kawamori, A., Inui, T., Ono, T., & Inoue, Y., ENDOR study on the position of hydrogen close to

the manganese cluster in S2 state of Photosystem II, FEBS Lett., 254, 219-224, 1989.

[138] Tang, X.-S., Sivaraja, M., & Dismukes, G.C., Protein and substrate coordination to the manganese 15 1 cluster in the photosynthetic water oxidizing complex: N and H ENDOR spectroscopy of the S2 state multiline signal in the thermophilic cyanobacterium Synechococcus elongatus, J. Am. Chem. Soc., 115, 2382-2389, 1993.

[139] Fiege, R., Zweygart, W., Irrgang, K.-D., Adir, N., Geiken, M., Renger, G., & Lubitz, W., EPR/ENDOR studies of the water oxidizing complex in Photosystem II, in Photosynthesis: From light to biosphere, Mathis, P. ed., Vol. II, Kluwer Academic Publishers, Dordrecht, 369-372, 1995.

[140] Fiege, R., Zweygart, W., Bittl, R., Adir, N., Renger, G., & Lubitz, W., EPR and ENDOR studies of the water oxidizing complex of Photosystem II, Photosynthesis Research, 48, 1-2, 1996.

[141] Fiege, R., Paramagnetische Metallzentren in Photosystem II untersucht mit EPR/ENDOR Tech- niken, Thesis, Technische Universit¨at Berlin, 1997. BIBLIOGRAPHY 133

[142] Zweygart, W., Weyhermuller,¨ T., Renger, G., Wieghardt, K., & Lubitz, W., EPR and ENDOR studies of the manganese clusters in the water-oxidizing complex, In Research in Photosynthesis, Murata, N. ed., Kluwer Academiv Publishers, Volume II, 289-292, 1992.

[143] Ananyev, G.M., Zaltsman, L., Vasko, G.C., & Dismukes, G.C., The inorganic biochemistry of photosynthetic oxygen evolution/water oxidation, Biochim. Biophys. Acta, 1503, 52-68, 2000.

[144] Bouges, B., Action de faibles concentrations d’hydroxylamine sur l’emission d’oxygene des algues Chlorella et des chloroplastes d’epinard, Biochim. Biophys. Acta, 234, 103-112, 1971.

[145] Messinger, J., & Renger, G., The reactivity of hydrazine with Photosystem II strongly depends on the redox state of the water oxidizing system, FEBS Lett., 277, 141-146, 1990.

[146] Messinger, J., Seaton, G., Wydrzynski, T, Wacker, U, & Renger, G., S−3 state of the water oxidase in Photosystem II, Biochemistry, 36, 6862-6873, 1997.

[147] Mei, R., & Yocum, C.F., Comparative properties of hydroquinone and hydroxylamine reduction of the calcium-stabilized oxygen-evolving complex of Photosystem II: reductant-dependent manganese(2+) formation and activity inhibition, Biochemistry, 31, 8449-8454, 1992.

[148] Goussias, C., Ioannidis, N., & Petrouleas, V., Low-temperature interactions of NO with the S1

and S2 states of the water-oxidizing complex of Photosystem II. A novel Mn-multiline EPR signal

derived from the S1 state, Biochemistry, 36, 9261-9266, 1997.

[149] Sarrou, J., Ioannidis, N., Deligiannakis, Y., & Petrouleas, V., A Mn(II)-Mn(III) EPR signal arises

from the interaction of NO with the S1 state of the water-oxidizing complex of Photosytem II, Biochemistry, 37, 3581-3587, 1998.

[150] Hanley, J., Sarrou, J., & Petrouleas, V., Orientation of the Mn(II)-Mn(III) dimer which results from the reduction of the oxygen-evolving complex of Photosystem II by NO: An electron param- ganetic resonance study, Biochemistry, 39, 15441-15445, 2000.

[151] Schansker, G., Goussias, C., Petrouleas, V, & Rutherford, A.W., Reduction of the Mn Cluster of

the water-oxidizing enzyme by nitric oxide: Formation of the S−2 state, Biochemistry,

[152] Sarrou, J., Isgandarova, S., Kern, J., Zouni, A., Renger, G., Lubitz, W., & Messinger, J., Nitric

oxide-induced formation of the S−2 state in the oxygen-evolving complex of Photosystem II from Synechococcus elongatus, Biochemistry, 42, 1016-1023, 2003.

[153] Ioannidis, N., Sarrou, J., Schansker, G., & Petrouleas, V., NO reversibly reduces the water-oxidizing

complex of Photosytem II through S0 and S−1 to the state characterized by the Mn(II)-Mn(III) multiline EPR signal, Biochemistry, 37, 16445-16451, 1998.

[154] Zheng, M., Khangulov, S.V., Dismukes, G.C., & Barynin, V.V., Electronic structure of dimanganese(II,III) and dimanganese(III,IV) complexes and dimanganese catalase enzyme: A general EPR spectral simulation approach, Inorg. Chem., 33, 382-387, 1994.

[155] Ford, R.C., & Evans, M.C.W., Isolation of a Photosystem II preparation with highly enriched oxygen evolution activity, FEBS Lett., 160, 159-163, 1983. 134 BIBLIOGRAPHY

[156] Sch¨afer, K.-O., Bittl, R., Zweygart, W., Lendzian, F., Haselhorst, G., Weyhermuller,¨ T., Wieghardt, K., & Lubitz, W., Electronic structure of antiferromagnetically coupled dinuclear manganese (MnIII MnIV ) complexes studied by magnetic resonance techniques, J. Am. Chem. Soc., 120, 13104-13120, 1998.

[157] Hillier, W, & Wydrzynski, T., The aﬃnities for the two substrate water binding sites in the O2 evolving complex of Photosystem II vary independently during S-state turnover, Biochemistry, 39, 4399-4405, 2000.

[158] Dismukes, G.C., Tang, X.-S., Khangulov, S.V., Sivaraja, M, &, Pessiki, P., The tetramanganese cluster of water oxidation: Electronic structure, protein coordination and relation to manganese catalase, In Research in Photosynthesis, Murata, n. ed., Volume II, Kluwer Academic Publishers, 257-264, 1992.

[159] Khangulov, S., Sivaraja, M., Barynin, V.V., & Dismukes, G.C., The dimanganese(III,IV) oxidation state of catalase from Thermus thermophilus: Electron nuclear double resonance of water and protein ligands in the active site, Biochemistry, 32, 4912-4924, 1993.

[160] Bittl, R., Personal communication.

[161] Ahrling,˚ K.A., Peterson, S., & Styring, S., The S0 state EPR signal from the Mn cluster in Photosystem II arises from an isolated S=1/2 ground state, Biochemistry, 37, 8115-8120, 1998.

[162] Yachandra, V.K., Guiles, R.D., McDermott, A.E., Britt, R.D., Dexheimer, S.L., Sauer, K., & Klein, M.P., The state of manganese in the photosynthetic apparatus. 4. Structure of the manganese

complex in Photosystem II studied using EXAFS spectroscopy. The S1 state of the O2-evolving Photosystem II complex from spinach, Biochem. Biophys. Acta, 850, 324-332, 1986.

[163] George, G.N., Prince, R.N., & Cramer, S.P., The manganese site of the photosynthetic water- splitting enzyme, Science, 243, 789-791, 1989.

[164] Weyhermuller,¨ T., Synthese und Charakterisierung oxo-verbrückter Mangankomplexe mit Aza- makrozyklen. Modelle für das Photosystem II?, PhD thesis, Ruhr-Universität, Bochum, 1995.

[165] Haselhorst, G., Spektroskopische Untersuchungen von homo- und heterovalenten strukturellen Mod- ellen f¨ur das Photosytem II, PhD thesis, Ruhr-Univerist¨at, Bochum, 1995.

[166] Smith P.J., Ahrling˚ K.A. und Pace R.J., Nature of the S2 state electron paramagnetic resonance signals from the oxygen-evolving complex of Photosystem II: Q-Band and oriented X-Band studies, J. Chem. Soc. Faraday Trans., 89, 2863-2868 1993.

[167] Kammel, M., Diploma thesis, Technische Universit¨at Berlin, 1998.

[168] Hasegawa, K., Ono, T.-A., Inoue, Y., & Kusonoki, M., Spin-exchange interactions in the S2 state manganese tetramer in the photosynthetic oxygen-evolving deduced from the g=2 EPR signal, Chem. Phys. Lett., 300, 9-19, 1999.

[169] Randall, D.W, Sturgeon, B.E., Ball, J.A., Lorigan, G.A., Chan, M.K., Klein, M.P., Armstrong, W.H., & Britt, R.D., 55Mn ESE-ENDOR of a mixed valence Mn(III)Mn(IV) complex: Comparison BIBLIOGRAPHY 135

with the Mn cluster of the photosynthetic oxygen-evolving complex, J. Am. Chem. Soc., 117, 11780- 11789, 1995.

[170] Bencini, C., & Gatteschi, D., EPR of exchange coupled systems, Springer Verlag, Berlin, 1990.

[171] Sch¨afer, K.O., Exchange coupled manganese complexes: Model sytems for the active centres of redoxproteins investigated with EPR techniques, PhD thesis, Technische Universit¨at Berlin, 2002.

[172] Peloquin, J.M., Campbell, K.A., & Britt, R.D., 55Mn Pulsed ENDOR demonstrates that the Pho- tosystem II “split” EPR signal arises from a magnetically coupled mangano-tyrosyl complex, J. Am. Chem. Soc., 102, 6840-6841, 1998.

[173] Britt, R.D., Zimmermann, J.-L., Sauer, K., & Klein, M.P., Ammonia binds to the catalytic manganese of the oxygen-evolving complex of Photosystem II. Evidence by electron spin-echo envelope modulation spectroscopy, J. Am. Chem. Soc., 111, 3522-3532, 1989.

[174] Dau, H., Andrews, J.C., Roelofs, T.A., Latimer, M.J., Liang, W., Yachandra, V.K., Sauer, K., & Klein, M.P., Structural consequences of ammonia binding to the manganese center of the photosynthetic oxygen-evolving complex: An X-ray absorption spectroscopy study of isotropic and oriented Photosystem II particles, Biochemistry, 34, 5274-5287, 1995.

∗ [175] Messinger, J., Nugent, J.H.A., & Evans, M.C.W., Detection of an EPR multiline signal for the S0 state in Photosystem II, Biochemistry, 36, 11055-11060, 1997.

[176] Porra, R.J., Thompson, W.A., & Kriedemann, P.E., Determination of accurate extinstion coeffi- cients and simultaneous equations for assaying chlorophylls a and b extracted with four different solvents: verification of the concentration of chlorophyll standards by atomic absorption spectroscopy, Biochim. Biophys. Acta, 975, 384-394, 1989.

[177] Messinger, J., Thesis, Technische Universit¨at Berlin, Germany,

[178] Sivaraja, M., Hunziker, D., & Dismukes, G.C., The reaction of H2S with the photosynthetic water- oxidizing complex and its lack of reaction with the primary electron acceptor in spinach, Biochim. Biophys. Acta, 936, 228-235, 1988.

[179] Mano, J., Takahashi, M., & Asada, K., Oxygen evolution from hydrogen peroxide in Photosystem II: Flash-induced catalytic activity of water-oxidizing Photosystem II membranes, Biochemistry, 26, 2495-2501, 1987.

[180] Zimmermann, J.L., & Rutherford, A.W., Electron paramagnetic resonance properties of the S2 state of the oxygen-evolving complex of Photosystem II, Biochemistry, 25, 4609-4615, 1986.

[181] Force, D.A., Randall, D.W., Lorigan, G.A., Clemens, K.L., & Britt, R.D., ESEEM studies of alcohol binding to the manganese cluster of the oxygen evolving complex of Photosystem II, J. Am. Chem. Soc., 120, 13321-13333, 1998.

[182] Britt, R.D., Zimmermann, J.L., Sauer, K., & Klein, M.P., Ammonia binds to the catalytic manganese of the oxygen-evolving complex of Photosystem II. Evidence by electron spin-echo envelope modulation spectroscopy, J. Am. Chem. Soc., 111, 3522-3532, 1989. 136 BIBLIOGRAPHY

[183] Boussac, A., Rutherford, A.W., & Styring, S., Interaction of ammonia with the water splitting enzyme pf Photosystem II, Biochemistry, 29, 24-32, 1990.

[184] Beck, W.F., & Brudvig, G.W., Binding of amines to the O2-evolving center of Photosystem II, Biochemistry, 25, 6749-6756, 1986.

[185] Beck, W.F., Brudvig, G.W., Ligand-substitution reactions of the O2-evolving center of Photosys- tem II, Chemica Scripta, 28 A, 93-98, 1988.

[186] Beck, W.F., de Paula, J.C., & Brudvig, G.W., Ammonia binds to the manganese site of the

O2-evolving complex of Photosystem II in the S2 state, J. Am. Chem. Soc., 108, 4018-4022, 1986.

[187] Britt, R.D., DeRose, V.J., Yachandra, V.K., Kim, D.K., Sauer, K., & Klein, M.P., Pulsed EPR studies of the manganese center of the oxygen-evolving complex of Photosystem II, in Current Research in Photosynthesis, Baltscheﬀsky, M. ed., Vol. II, Kluwer Academic Publishers, Dordrecht, 769-772, 1990.

[188] Radmer, Ollinger, FEBS Lett., 152, 39-43, 1983.

[189] Zimmermann, J.L., & Rutherford, A.W., EPR studies of the oxygen-evolving enzyme in Photosys- tem II, Biochim. Biophys. Acta, 767, 160-167, 1984.

[190] de Paula, J.C., & Brudvig, G.W., Magnetic properties of manganese in the photosynthetic oxygen- evolving complex, J. Am. Chem. Soc., 107, 2643-2648, 1985.

[191] Sandusky, P.O., & Yocum, C.F., The mechanism of amine inhibition of the photosynthetic oxygen evolving complex: Amines displace functional chloride from a ligand site on manganese, FEBS Lett., 162, 339-343, 1983.

[192] Sandusky, P.O., & Yocum, C.F., The chloride requirement for photosynthetic oxygen evolution: Analysis of the eﬀects of chloride and other anions on amine inhibition of the oxygen-evolving complex, Biochim. Biophys. Acta, 766, 603-611, 1984.

[193] Sandusky, P.O., & Yocum, C.F., The chloride requirement for photosynthetic oxygen evolution: Factors aﬀecting nucleophilic displacement of chloride from the oxygen-evolving complex, Biochim. Biophys. Acta, 849, 85-93, 1986.

[194] Ono, T.-A., Zimmermann, J.-L., Inoue, Y., & Rutherford, A.W., EPR evidence for a modiﬁed S- state transition in chloride-depleted Photosystem II, Biochim. Biophys. Acta, 851, 193-201, 1986.

[195] Messinger, J., Badger, M., & Wydrzynski, T., Detection of one slowly exchanging substrate water

molecule in the S3 state of Photosystem II, Proc. Natl. Acad. Sci. USA, 92, 3209-3213, 1995.

[196] Hillier, W, Messinger, J., & Wydrzynski, T., Kinetic determination of the fast exchanging substrate

water molecule in the S3 state of Photosystem II, Biochemistry, 37, 16908-16914, 1998. Danksagung

Am Ende dieser Doktorarbeit möchte ich mich bei denen bedanken, die mir geholfen haben, dort anzukommen, wo ich jetzt bin. Prof. Dr. Lubitz hat es mir ermöglicht, diese Arbeit zu beginnen. Unter seiner Agide¨ habe ich mir die Voraussetzungen erworben, diese Arbeit zu ihrem Ende zu fuhren.¨ Dafur¨ bin ich dankbar, es wird mir immer im Gedächtnis bleiben. Ich möchte mich bei meinem zweiten Berichter Robert Bittl bedanken. Er hat nicht nur im wissenschaftlichen Bereich fur¨ die Impulse gesorgt, die letztendlich in der vorliegenden Arbeit mundeten.¨ Mein Dank gilt weiteren Mitgliedern meiner Arbeitsgruppe und des Max-Volmer-Insti- tuts/Laboratoriums. Fur¨ die Motivation, das Reden und die praktische Hilfe bin ich Marc (Hannes) Brecht dankbar. Christian Teutloff ist die Hilfsbereitschaft in Person und immer fur¨ einen guten Scherz zu haben. Wulf Hofbauer ist eine imposante Mixtur aus physikalis- chem Enthusiasmus, Selbstlosigkeit und einer mir völlig fremden kulturellen Identität. Jan Kern war nicht nur Lieferant der Kristalle des Photosystems, sondern in seinem nicht- fachlichen Interesse immer eine Erfrischung. Celine Elsässer ist, obwohl manchmal etwas schreckhaft, ein Mensch, mit dem man rechnen kann. Markus Galander half mir speziell bei der Einfuhrung¨ in “Octave”. Kai Oliver Schäfer war der Mann fur¨ Computerprob- leme. Maurice van Gastel half mir bei der Simulation der HYSCORE-Spektren. Iosifina

Sarrou fertigte einige der S 2-Proben und fuhrte¨ mich in die Präparation ein. Johannes − Messinger saß mir im Buro¨ gegenub¨ er. Aus unserem Institut möchte ich mich weiter bei Friedhelm Lendzian, Athina Zouni, Eberhard Schlodder und Steffen Sarstedt bedanken. Finanziell wurde ich ub¨ er die Jahre von der TU Berlin ausgestattet. Ich danke meinen Eltern,meinen Schwestern und meinen Neffen ..., kurz: meiner Familie. Andrea, ohne Dich existierte diese Arbeit nicht. All den hier erwähnten Menschen sei alles Gute fur¨ die Zukunft gewunsc¨ ht!

137 138 BIBLIOGRAPHY Lebenslauf

Name Michael Kammel Geburtsort Gustro¨ w (Mecklenburg-Vorpommern) Eltern Hannelore Kammel und Friedrich Kammel

10.06.1970 Geburt

Ausbildung

September 1977 - Juli 1985 Besuch der POS “Clara Zetkin” in Laage September 1985 - Juli 1989 Besuch der ESOS “Georg Thiele” in Kleinmachnow Abschluss: Abitur der Spezialoberschule naturwissenschaftlich -technischer Richtung Oktober 1990-Juli 1993 Grundstudium der Chemie an der TU Berlin Abschluss: Vordiplom 1993 Klaus-Koch-Preis der TU Berlin September 1993-Juli 1994 Studium der Chemie an der University of Kent at Canterbury (UK) Abschluss: Diploma Oktober 1994-1998 Hauptstudium der Chemie an der TU Berlin Abschluss: Diplom 1998-Gegenwart Anfertigung der Dissertation auf dem Gebiet der Elektronenparamagnetischen Resonanz am Max-Volmer-Institut (seit April 2001: Max-Volmer-Laboratorium) der TU Berlin

T¨atigkeiten

November 1989-März 1990 Wehrdienst in der NVA April 1990 - September 1990 Zivildienst 1994-1998 Krankenpfleger in der Haus- und Schwerstkrankenpflege 1998-2003 Wissenschaftlicher Mitarbeiter an der TU Berlin mit Lehrtätigkeit 139