Kinetic Determination of the Fast Exchanging Substrate Water

Research School of Biological Sciences, The Australian National University, ... site in the S3 state undergoes biphasic kinetics although the fast pha...
0 downloads 0 Views 112KB Size
16908

Biochemistry 1998, 37, 16908-16914

Kinetic Determination of the Fast Exchanging Substrate Water Molecule in the S3 State of Photosystem II†,‡ Warwick Hillier,*,§ Johannes Messinger,| and Tom Wydrzynski†,§ Research School of Biological Sciences, The Australian National UniVersity, Canberra, ACT 0200, Australia, and Physical Biosciences DiVision, Lawrence Berkeley National Laboratory, Berkeley, California 94720 ReceiVed April 3, 1998; ReVised Manuscript ReceiVed September 8, 1998

ABSTRACT: In a previous communication we showed from rapid isotopic exchange measurements that the exchangeability of the substrate water at the water oxidation catalytic site in the S3 state undergoes biphasic kinetics although the fast phase could not be fully resolved at that time [Messinger, J., Badger, M., and Wydrzynski, T. (1995) Proc. Natl. Acad. Sci. U.S.A. 92, 3209-3213]. We have since improved the time resolution for these measurements by a further factor of 3 and report here the first detailed kinetics for the fast phase of exchange. First-order exchange kinetics were determined from mass spectrometric measurements of photogenerated O2 as a function of time after injection of H218O into spinach thylakoid samples preset in the S3 state at 10 °C. For measurements made at m/e ) 34 (i.e., for the mixed labeled 16,18O product), the two kinetic components are observed: a slow component with k ) 2.2 ( 0.1 s-1 2 1 (t1/2 ∼ 315 ms) and a fast component with k2 ) 38 ( 4 s-1 (t1/2 ∼ 18 ms). When the isotopic exchange is measured at m/e ) 36 (i.e., for the double labeled 18,18O2 product), only the slow component (k1) is observed, clearly indicating that the substrate water undergoing slow isotopic exchange provides the ratelimiting step in the formation of the double labeled 18,18O2 product. When the isotopic exchange is measured as a function of temperature, the two kinetic components reveal different temperature dependencies in which k1 increases by a factor of 10 over the range 0-20 °C while k2 increases by only a factor of 3. Assuming simple Arrhenius behavior, the activation energies are estimated to be 78 ( 10 kJ mol-1 for the slow component and 39 ( 5 kJ mol-1 for the fast component. The different kinetic components in the 18O isotopic exchange provide firm evidence that the two substrate water molecules undergo separate exchange processes at two different chemical sites in the S3 state, prior to the O2 release step (t1/2 ∼ 1 ms at 20 °C). The results are discussed in terms of how the substrate water may be bound at two separate metal sites.

One of the most important processes in nature is the photosynthetic oxidation of water to molecular oxygen by the chlorophyll/protein complex photosystem II (PSII).1 The catalytic site for water oxidation involves a cluster of 4 Mn ions and a redox active tyrosine (YZ) in a domain called the water oxidizing complex (WOC) (1, 2). A fundamental property of the WOC is that O2 is released with a periodicity of four upon illumination with single turnover light flashes (3). To explain this behavior, Kok and co-workers (4) proposed that the reaction cycles through five so-called Sn states (for n ) 0, 1, 2, 3, 4). Beginning in S0 and traversing to S4, each S state is advanced by a single quantum event at the photochemical reaction center. Upon reaching the S4 * To whom correspondence should be addressed: Hillier@RSBS. ANU.EDU.AU, [email protected]. † This work was supported in part by a postdoctoral fellowship from the Australian Research Council and by a grant from the Deutsche Forschungsgemeinschaft (Me 1629/1-1) to J. M. ‡ Some of the results from this work were presented at the Symposium in Honor of Ken Sauer and Mel Klein held at the University of California, Berkeley, CA, January 7-8, 1998. § The Australian National University. | Lawrence Berkeley National Laboratory. 1 Abbreviations: EPR, electron paramagnetic resonance; PSII, photosystem II; Sn, redox states of the water oxidizing complex with n ) 0, 1, 2, 3, 4; WOC, water oxidizing complex.

state, O2 is released (within 1-2 ms) and the S0 state is regenerated to begin the cycle again. To account for the observed damping in the period four oscillations, a miss parameter (R) and a double hit parameter (β) were introduced to allow for the mixing of the Sn state populations during a flash sequence. Since the peak O2 yield occurs on the third flash in dark-adapted samples, the S1 state was concluded to be dark-stable. In the original Kok hypothesis it was implied that the two substrate water molecules entered the reaction sequence during the last step, just prior to O2 release (4). More modern models invoke substrate water binding to the WOC at the beginning of the S state cycle (5, 6) with the possible involvement of a peroxyl intermediate in the S3 state (7). However, definitive experimental evidence to show how and at which step the O-O bond is formed has remained elusive. Upon the discovery of a low-temperature, multiline EPR signal that is associated with the S2 state and arises from the catalytic manganese cluster (8), many attempts were made to identify manganese-water interactions through the use of isotopically labeled substrate analogues (e.g., H217O, D2O, 15NH ) (9-14). Although it is generally concluded from 3 these studies that substrate water is indeed bound to the catalytic Mn in the S2 state, some of the experimental

10.1021/bi980756z CCC: $15.00 © 1998 American Chemical Society Published on Web 11/11/1998

Substrate Exchange in the S3 State of Photosystem II evidence is not easily justified (15-19). One difficulty in this experimental approach is to unequivocally discriminate between the isotopic exchange by the substrate water and the possible isotopic exchange by other water, oxygen, or proton ligands that may be in the first coordination sphere of the catalytic Mn. Furthermore, there is not yet complete agreement in the literature as to the origin of the S2 state EPR signals, whether they arise from magnetic couplings between two (20) or four (2) Mn ions. If the EPR signals arise from only part of the catalytic Mn cluster, then perhaps only part of the manganese-water interactions will be observable. In another approach to study the binding interactions between the substrate water and the catalytic site, isotopic exchange measurements can be made of photogenerated O2 upon incubation of a sample in 18O-labeled water using mass spectrometric techniques (21-24). In the first studies using this approach, it was concluded that there was no nonexchangeable water bound to the WOC prior to the S4 state (21, 22). However, in these studies an “open” sample chamber system had to be used in which the long mixing and instrument stabilization times needed upon the addition of isotopically labeled water to the sample limited the kinetic resolution of the measurements to g30 s. Recently, we were able to reduce the mixing/stabilization times considerably, by a factor of about 1000, through the use of a “closed” chamber system. In the first report of our findings, we showed that the WOC in the S3 state exhibits two distinct kinetic phases in the 18O isotopic exchange, a slow component with a rate constant of 2.2 s-1 at 10 °C and an unresolved fast component (24). We have since been able to improve the mixing/stabilization times of our experimental setup by another factor of 3, and we report in this communication the first detailed kinetics for the fast exchanging component. MATERIALS AND METHODS Thylakoid membranes were prepared by standard procedures from hydroponically grown spinach and suspended in a medium containing 50 mM HEPES/NaOH (pH 6.8), 400 mM sucrose, 15 mM NaCl, and 5 mM MgCl2. The samples were then frozen as small beads in liquid N2 and stored at -70 °C. After thawing the thylakoids at room temperature in the dark, they were given a preflash and a 10 min dark period in order to enrich the S1 population and then kept on ice until measurement. The thylakoid suspension medium was pH calibrated at each temperature used in the various measurements. The flash-induced O2 produced by the sample was measured at m/e ) 34 and ) 36 with an in-line mass spectrometer (Vacuum Generation MM6, Winsford, UK) for the single and double 18O-labeled O2, respectively. A closed, stirred cuvette similar to the one used in our earlier work (24) but with a smaller internal volume (160 µL) was connected to the mass spectrometer via a dry ice/ethanol bath. A silicon membrane was used to separate the liquid phase from the mass spectrometer inlet, to allow for the passage of gases. Rapid injection of H218O was achieved with a Hamilton CR-200 syringe that was triggered with a computeractuated solenoid. Samples were illuminated with saturating light flashes (∼8 µs) from a xenon lamp (FX-193 lamp, 4

Biochemistry, Vol. 37, No. 48, 1998 16909 µF @ 1 kV capacitor, EG & G, Salem MA) which was positioned directly in front of the sample chamber window. A chlorophyll concentration of 0.9 mg mL-1 was used for optimal S/N. The flash and injection sequence was controlled via a computer, and accurate timing intervals were established with a digital oscilloscope. The exact 18O isotopic enrichment in the sample chamber was determined from the ratio of 18O incorporation into CO2 measured at m/e ) 44, 46, and 48, that is,

44:46:48 ) (1 - )2:2(1 - ):2 ) 100%

(1)

where  is the 18O% concentration. The typical 18O enrichment value was  ) 12.0% ( 0.25%. The H218O isotopic equilibration with CO2 is a consequence of the normal hydration/dehydration reactions of CO2 in the suspension medium. The 18O isotopic exchange was measured for the S3 state at 20, 15, 10, 5, and 0 °C. The thylakoids were loaded in the dark, and the cuvette was degassed for 10-12 min. The injection/flash protocol employed to measure the isotopically labeled O2 is illustrated below where ∆t is the time between the H218O injection and the third flash:

The dark time between the second and third flashes (t2,3) was held constant as ∆t was varied in order to maintain the same extent of S state deactivation within a data set (t2,3 was typically 10 s). Signals from the mass spectrometer were recorded on an x-t plotter. Due to the slow response of the instrument because of the gas diffusion path length, the O2 yield per flash had to be determined by extrapolating to the time of flash excitation (see Figure 2 in ref 24). The O2 background signal caused by the injection of the labeled water (Yinj) was determined by performing separate injections under the same conditions but without flashing. To reduce the size of the Yinj, small quantities of glucose, glucose oxidase, and catalase were added to the labeled water prior to injection. This caused a reproducible reduction in Yinj without interfering with the photogenerated O2. Typically the Yinj was kex(MnIII-OH2) > kex(MnIV-OH2), as the ionic radius of the metal center decreases (38). On the basis of a comparison with Fe and Ru complexes (39-41), the water ligand exchange rates for Mn ions may be expected to slow by ∼10-4 s-1 for each formal oxidation state increase (Table 2). Similarly, a comparison with CrIII (42) can provide a clue to the magnitude of the water ligand exchange rate for MnIV since the two ions are isoelectronic with a stable d3 configuration (Table 2). In general, therefore, we may expect that the water ligand exchange rates are in the range of 10-3 to 10-1 s-1 for MnIII and 10-5 to 10-7 s-1 for MnIV.

Hillier et al. Table 2: Ligand Exchange Rates for Various Hydrated Metal Complexes S3 state of PS II slow phase (k1 at 10 °C) fast phase (k2 at 10 °C) hydrated M2+ ions [Ca(H2O)6]2+ (dimagnetic) [Fe(H2O)6]2+ (high spin) [Ru(H2O)6]2+ (low spin) hydrated M3+ ions [Fe(H2O)6]3+ (high spin) [Ru(H2O)6]3+ (low spin) [Cr(H2O)6]3+ (high spin) terminal oxo ligands [TiO(H2O)5]2+ [VO(H2O)5]2+ µ-oxo bridges [FeIII2] - ribonucleotide reductase [MoV2 O4 (OH2)6]2[MoIV3 O4 (OH2)9]4-

exchange rate (s-1)

reference

2 × 100 4 × 101

this work work

3 × 108 4 × 106 2 × 10-2

37 39 41

2 × 102 4 × 10-6 2 × 10-6

40 41 42

2 × 104 1 × 10-5

43, 44 45

8 × 10-4

49

3 × 10-6 kex(MsOH) > kex(MdO), where protonation of the bound water ligand will depend on the pKa of the metal complex, which in turn is a function of not only the redox state of the metal center but also of the type(s) of other ligands. This behavior is illlustrated in Table 2 by the large difference in the rate constants for oxo ligand exchange in a TidO complex (43, 44) and in a VdO complex (45). In this case, it is believed that the TidO complex is very easily protonated (43). The importance of protonation is also observed with the horseraddish peroxidase FeIVdO complex which undergoes 18O exchange at pH 7 but not at pH 9 (46). Consequently, although higher oxidation state Mn ions will tend to polarize bound water favoring hydroxyl or oxo ligands, the pKa of the Mn complex as influenced by the other ligands bound to it could have a strong effect on the water ligand exchange rate. The pKa’s for various Mn complexes have been reported elsewhere (47). It has also been proposed that a bridging di (Mn)2 µ-oxo intermediate may be involved in the water oxidation chemistry (48). However, in this case the exchange process should be very slow. As shown in Table 2, the exchange rate for the µ-oxo bridge in ribonucleotide reductase (49) and in a [MoV2 O4(OH2)6]2+ complex (50) is