Temperature Dependence of Picosecond Fluorescence Kinetics of a Cyanobacterial Photosystem I Particle

Photochemistry and Photobiology, Vol. 57, No. 1 , pp. 113-1 19, 1993 Printed in the United States. A11 rights reserved

003 1-8655/93 $05.00+0.00 0 1993 American Society for Photobiology

TEMPERATURE DEPENDENCE OF PICOSECOND FLUORESCENCE KINETICS OF A CYANOBACTERIAL PHOTOSYSTEM r PARTICLE* GERDSCHWEITZERand ALFREDR. HOLZWART& Max-Planck-Institutfur Strahlenchemie,Stiftstr. 34-36, D-4330 Muhlheim an der Ruhr, Germany

SANDRA TURCONI,

(Received 3 1 March 1992; accepted 30 July 1992) Abstract- Picosecond time-resolved fluorescence of photosystem I particles isolated from Synechococcus sp. was recorded in the wavelength range from 680 nm to 736 nm for temperatures of 6°C to 42°C and - 100°C using the single-photon-timing technique. By global analysis of the data we found four contributing lifetime components at the higher temperatures (7,= 12 ps, T~ = 35 ps, T~ = 65 ps, T~ = 1000 ps). We attribute T, to an energy transfer between two pigment pools, T* to the charge separation process in the reaction center, component T~ is assigned to

aggregate and T~ to uncoupled chlorophyll emission. The corresponding decay-associated spectra are presented. We also applied a target analysis procedure to fit parameters of a kinetic model directly to the data. The resulting rate constants and species-associated spectra are discussed. The data indicate substantial spectral heterogeneity in the antenna with at least three substantially different chlorophyll pools. The overall exciton decay kinetics (by charge separation)is trap-limited.

In contrast, van Grondelle and coworkers9J0 suggested that long-wavelength antenna pools serve for focusing excitations to the RC. By studying the picosecond time-resolved fluorescence and its temperature dependence as a function of emission wavelength one can obtain basic information on the energy transfer mechanisms and the pathways of energy migration to the traps. Special laser systems and fast multichannel plate detectors allow for better time resolution; single-photon timing (SPT) techniques in combination with advanced data analysis methods allow the resolution of complex kinetics.”J’ Furthermore, biochemical isolation procedures are available that allow the investigation of PS I particles with various well-defined antenna compositions and antenna sizes. Several studies on the temperature dependence of the energy transfer kinetics in PS I have already been reported.’”-” Most of these studies were camed out on large “native” antenna size preparations of PS I (=200 ChVP700). In several of these studies the time resolution was limited to about 50 PS. In this study, we investigated PS I particles of an antenna size of about 100 Chl/P700 with a time resolution better than 10 ps. We present an analysis of the fluorescence and energy transfer kinetics in PS I from the thermophilic cyanobacterium Synechococcus sp. measured as a function of temperature and under various external conditions. The data are discussed in terms of decay-associated spectra (DAS) and a kinetic model. A preliminary account of this study has been presented in a previous publication.16

INTRODUCTION

The transfer of the excitation energy to the photosynthetic reaction centers (RC)$ in photosystems (PS) I and I1 is performed by antenna chlorophyll (Chl) molecules whose spectral properties and spatial arrangement are not known in detail and are still a matter of discussion. The Chl antenna pigment molecules are organized in spectrally different pools,’J possibly resulting from different environments. The dynamics of energy transfer and trapping in PS I is still not understood in all aspects. The fluorescence yield of PS I at room temperature (r.t.) is small compared to the one of PS 11. A characteristic feature of PS 1 steady-state fluorescence is the band at 735 nm (F735) for higher plants, 720 nm (F720) for green algae and =7 15 nm for cyanobactena, which strongly increases in intensity upon lowering the temperature? Several tentative explanations for the long-wavelength fluorescence were proposed: Butler et a[.4assigned its origin to C-705, a chlorophyll species absorbing at 705 nm and transfemng excitation directly to the RC. In contrast, Kuang et assumed that the pigments bound to the 21 kDa polypeptide of the light-harvesting complex I (LHC I) were the origin of the long-wavelength fluorescence. Wittmershaus,6 however, assigned it partly both to the internal and the peripheral antennae. Mukeji and Sauer’ assumed that F720 arises from the internal antenna and that F735 is located in the peripheral antenna. Furthermore, the function of this long-wavelength pool is still a matter of discussion: W i t t m e r s h a d proposed that it could be an intermediate trap, while Mukeji and SaueF considered a protective role.

MATERLALS AND METHODS

*Parts of this work will be included in the Ph.D. thesis of S.T. at the Heinrich Heine UniversitSt, Diisseldorf, Germany. ?To whom correspondence should be addressed. +lbbreviutions: Chl, chlorophyll; DAS, decay-associated spectrum; FWHM, fullwidth half maximum; LHC, light-harvestingcomplex; P700, primary electron donor in photosystem I; PS,photosystem; r.t., room temperature; RC, reaction center; SAS, species-associated spectrum; SB 12, N-dodecyl-N,N-dimethyl-ammonio-3-propane sulfonate; SPT, single-photon timing. 113

Photosystem 1particles were isolated from thermophilic Synechococcus sp. membrane preparations by an isolation procedure that was somewhat modified as compared to one already rep01ted.I~The membraneswere obtained as described previously’8but using a modified buffer (MMCS buffer, pH 6.5: 20 mM 2-[N-morpholino]ethanesulfonic acid, I0 mM MgCl,, 20 mMCaCl,, 20% sucrose). They were incubatedwith the zwitterionicdetergent N-dodecy1-N.Ndimethyl-ammonio-3-propane sulfonate (SBl2) (0.4%) for 20 min at room temperature and then centrifuged at 140 000 g for 45 min.

SANDRA TURCONI et al.

114

1

1,

+ . (d

=c .ffl K

a,

c

K .-

a,

0 K

a, 0

m

??

used giving a spectral resolution of 4 nm. An MCP type photomultiplier (Hamamatsu R2809U-05) with appropriate electronics served as detector. The overall system prompt response (full width half maximum [FWHMJ) was typically 70 ps. Data were analyzed either by single decay or by global analysis methodsz' over a complete set recorded at nine different emission wavelengths. In the global analysis we calculated decay-associated spectra (DAS) for each data set, Kinetic modeling was camed out applying a target analysis program in which a rate constant matrix corresponding to assumed models was fitted directly to the measured This program yielded species-associated spectra (SAS) data and an optimal set of rate constants for the tested kinetic model.

0

3 -

LL

RESULTS 0

1

J 650

700

750

800

1

850

Wavelength / nm Figure 1. Steady-state emission spectrum of photosystem I particles isolated from Synechocuccus sp. T = 20°C, excitation wavelength X = 435 nm. The pellet, containing PS I, was resuspended in MMCS buffer containing 0.4% SB12 and again centrifuged for 30 min at 140000 g. The pellet was again resuspended and incubated for 30 min in the same buffer with addition of 1% SB12 and recentrifuged. The resulting supernatant was layered on a 1040% sucrose gradient (MMCS buffer, 0.05% SB12) and centrifuged for 16 h in a swing-out rotor at 1 10 000 g.The intense PS I-containing green band floating on the bottom of the cuvette was collected and characterized in terms of chlorophyll content, Chl/P700 ratio and fluorescence stationary spectra. The Chl/P700 ratio was 100 +- 10 as determined by the method published by Markwell et a1.I' For all measurements, the samples were diluted to contain approximately 8 gg Chl/mL using MMCS buffer containing 0.05% SB12 without sucrose. For standard conditions, 10 mM ascorbate and 10 fiM phenazine methosulfate were freshly added, unless stated otherwise. Liquid samples were pumped through a Row measuring cuvette (1.5 x 1.5 mm2) generally at a flow rate of 12 mL/s, though measurements at a pumping rate I0 times higher yielded identical results. The sample was thennostated to the desired temperature t1"C. Steady-state fluorescence spectra of all samples were recorded before and after each set of time-resolved measurements to prove the stability of the samples. Low temperature measurements were camed out putting sample cells of 0.1 mm path length into a liquid nitrogen cryostat as described in Roelofs et a[." Time-resolved fluorescence data were measured applying the single-photon timing technique and using a synchronously pumped cavity-dumped dye laser with a repetition rate of 800 kHz as described previouslyz' with sulforhodamine 101 as the laser dye. A double monochromator (Jobin Yvon DHIO) with slits of 1 mm was

Spectra and decays in liquid samples A typical steady-state fluorescence spectrum o f the Synechococcus sp. PS I particles is shown i n Fig. 1. It exhibits a m a x i m u m a t A, = 7 12 nm a n d a shoulder at A,, = 695 n m and clearly indicates t h e high degree o f purification o f the PS I particles. For t h e time-resolved measurements fluorescence was excited a t A,, = 6 6 7 n m , i.e. a t the absorption m a x i m u m o f the sample, and t h e decays were recorded in the range 680-736 n m i n intervals of 7 nm. These measurements were carried out in t h e temperature range between + 6°C a n d +42"C. Each decay contained a t least 40 000 counts i n t h e peak channel. T h e data from different wavelengths were first analyzed by single decay analysis. Four lifetimes were generally necessary for a good fit, these were 7 ,= 6-1 3 ps, r2 = 28-37 ps, r3 = 46-60 ps, r4 = 0.75-1.0 ns. T h e lifetimes were n o t dependent on t h e emission wavelength within t h e error limits except for a slight increase i n 7 ,when going from t h e blue to t h e red end of t h e spectrum. This small difference is, however, close to t h e error limits in singledecay analysis.

Global anarysis Application of a global analysis method t o the decay d a t a resulted also i n four exponential components that were necessary for a n accurate fit as judged by global Xz-values and plots o f weighted residuals. T h e resulting lifetimes were 7 , = 9-13 ps, r2 = 29-35 PS, T~ = 59-65 ps, T~ = 1030-1 180 p s (the range i n values refers to different temperatures) a n d are summarized in Table 1. Figure 2a,b show DAS from decays

Table 1. Global analysis results from fluorescence decays of photosystem I particles as a function of temperature ~~~

Temperature ("C) Ratio A,/A, Al/A, Ai/A, A,/A,

6

12 1

.o

6.3 -0.8 -0.3 11 35 65 1050

0.7 3.1

-0.7 -3.8 13 31 59 1030

18

24

30

36

42

-loo*

0.6 7.5 -0.8 - 10.7

0.5 6.1 -0.8 - 10.7

0.3 3.6 -0.7 - 10.0

0.4 27.2 -0.6 -56.1

11 33 66 1010

10 31 64 1030

0.2 40.0 -0.6 - 126 10 29 143 940

14.9 49.7 - 1.0 -2.6 14 72 190 1710

11 29 62 1180

9 30 114 1170

In the upper part, the amplitude ratios A1/A2 and A1/A3 are shown for the two wavelengths 687 nm and 722 nm. In the lower part, the corresponding lifetimes r l to T~ are summarized. Errors in lifetimes are k 1 ps for T ] , t 2 ps for T ~ t, 5 ps for T~ and -+20 ps for T ~ . *Measurements done with the sample in the cryostat (see Materials and Methods).

Picosecond kinetics of photosystem I

30}

(a)

301

20 -

20

115

(a>

c

190 X 1710

a

680

690

a,

y

c ._ -

E

a

10-

0 -10

-

-20

-.

:

rn 35 A 65 X

I

A

1050 1

I

I

I

I

I

I

680

690

700

710

720

730

740

Wavelength / nm

I

,

700

710

I

720

730

740

Wavelength / nm I

-20 -

A

143

X

940 I

I

I

680

690

700

710

720

730

740

Wavelength / nm Figure 2. Decay-associated spectrum (DAS) of photosystem I particles isolated from Synechucuccus sp. (a) Temperature T = 6°C and (b) temperature T = 42°C. Note: both the amplitudes of the two longest lifetime components are approximately zero, so the symbols overlap. measured at + 6 T and at +42"C, respectively. These two DAS most clearly display the changes occumng due to the temperature variation. The decay kinetics is dominated by the two fastest components with the lifetimes 7-] and 7-* that contribute more than 90% to the total amplitude. The amplitude A , is positive for wavelengths below -700 nm, whereas for longer wavelengths it becomes negative, corresponding to a rise time in the kinetics. This characterizes the T , component as an energy transfer component. The DAS obtained at +6"C are quite similar to the ones we have reported previously,'6 although the preparation procedure and the detergent used has been different. No significant change is observed in the lifetime values over the temperature range examined except for the T~ component at the two highest temperatures (see Detergent effects below). The influence of the temperature on the amplitudes of the decay components is depicted in Table 1 as amplitude ratios for the three fastest components (the amplitude of the fourth one does not vary significantly).

Wavelength / nm Figure 3. (a) Decay-associated spectrum (DAS) and (b) speciesassociated fluorescence spectra (SAS) of photosystem I particles isolated from Synecchucoccus sp. at temperature T = - 100°C. temperature data were also submitted to the global analysis. The resulting DAS are shown in Fig. 3a. As compared to the data at 6°C and 42°C (Fig. 2a,b), a significant decrease in the amplitude A, (solid squares in Figs. 2 and 3a), especially at wavelengths below 7 10 nm, can be seen. Correspondingly, the spectrum of the short-lived component T , is increased in intensity and red-shifted by -20 nm at low temperature.

-

Detergent and redox effect In order to further characterize the origin of the 7-2 and in particular the T~ components (the latter had not been present Table 2. Results from global analysis of decay data sets from photosystem I particles acquired under various conditions at T = 12°C Condition (PS) (PSI 7 3 (PSI 7-4 (PS) 71

7-2

Standard 13 31

Dithionite 12 27

59

54

1030

1620

Fast pump I1 29 52 1110

Femcyanide I1

31 60 890

Low temperature data

A further set of measurements was carried out at a temperature of - 100°C under conditions as described above. The sample was frozen quickly in liquid nitrogen. The low

For standard conditions see Materials and Methods; reducing conditions (dithionite);sample pumping rate of 120 mL/min (fast pump); oxidizing conditions (femcyanide). For details refer to text.

SANDRA TURCONI et al.

116 I

I

Pool 1

1

I

k12

,

Pool 2

4

r

k22

k2 1

.

Figure 4. Kinetic scheme to which our data sets were fitted. For discussion of the rates involved as well as their interpretation refer to the text.

in our previous data'?, we camed out several additional sets of measurements under varied conditions: Firstly, we increased the pumping rate of the sample to 120 mL/min in order to minimize the possibility of P7OO photooxidation by the exciting light. Secondly, we kept the sample at low redox potential (below -600 mV)z4by adding aliquots of dithionite. We also recorded the kinetics under oxidizing conditions after adding femcyanide (final concentration 4 mM) and shining red light from a slide-projector (A > 630 nm) onto the sample before it entered the cuvette. Each of these measurements was carried out at several wavelengths, and the data were analyzed by global analysis. None of these variations in conditions had a significant influence upon the kinetics, as can be seen from Table 2 summarizing the lifetimes.

0

Pool 1 Pool 2

m 680

690

700

710

720

730

740

Wavelength / nm

(b) 0

4!

3.1

!

3.2

3.3

3.4

3.5

3.6

'

1/ T ( x 10- K - )

Figure 6. Arrhenius plot of rate constants from Table 2. .--k,_,; A-k22;

@--k,_,.

For studying further the origin of the component T ~ we , conducted another set of measurements in which the concentration of the detergent (SB 12) in the sample buffer was vaned over the range 0.05-0.1Yo. As a result of this series, we could show that adjusting the SB12 concentration to 0.085% yielded fluorescence data that could be fitted very well by a sum of only three exponential decays (not shown), corresponding to components T,,T~ and T, in Table 1. These lifetimes did not depend on the detergent concentration. Decreasing the detergent concentration created the component T~ with substantial amplitude as shown in Fig. 2a ( c j also Table 1). Thus, we attribute the T~ lifetime to aggregates of PS I particles forming at low detergent concentrations, This effect also explains the change in the r,-values at the higher temperatures, because there the corresponding amplitude approaches zero and thus the error in the calculation of the lifetime becomes very large. Measurements at higher SB12 concentrations gave no further changes in the amplitudes of the short lifetimes, but an increase of the amplitude A, of the slowest component. We attribute this long-lived component to functionally uncoupled chlorophylls in our sample, which is also supported by its DAS, which has a maximum at short wavelength (below 680 nm). However, we should like to note that an alternative explanation for the 7, component might be a heterogeneity in antenna sizes. This would show a similar dependence on detergent concentration. Such possibilities are presently under further investigation. Kinetic modeling

Pool 1 Pool2

I

I

I

I

I

680

690

700

710

720

730

740

Wavelength / nm Figure 5. Species-associated spectrum (SAS) of photosystem I particles isolated from Synechococcus sp. (a) Temperature T = 6°C; (b) temperature T = 42°C.

The parameters of a kinetic model were further fitted directly to the data by a global target analysis procedure. In this procedure the quality of the fits, as judged by Xz-values and weighted residual plots, shows whether the applied model is consistent with the data. Three a priori feasible models were checked in our target analysis. The only one that describes the data in a physically reasonable way is the one depicted in Fig. 4. It considers two different antenna pools connected by a fast equilibration process (defined by two rate constants: k, -z and k2-,), and a further process (kZz)depopulating pool 2. The fourth rate constant in the matrix system, k,,, represents the decay to the ground state by either radiative or nonradiative processes and was kept constant at a reasonable value (0.3 nss') in order to obtain a unique sol ~ t i o n ; ~the ' . ~results, ~ however, are not effected by varying

Picosecond kinetics of photosystem I

117

Table 3. Rate constants and AG-values from target analysis of fluorescence decays for photosystem 1 particles as a function of temperature for the model shown in Fig. 4 T W) k,-, (ns-'1 k , - 2 (ns-I) k22(ns-9 AG (cm-I) M (cm-I) N,N

6

18

24

30

36

42

11.7 67.0 37.5 338 738 1.9

11.7 62.0 44.1 337 658 6.0

11.0 57.0 49.9 340 692 6.8

16.3 64.0 53.5 288 71 1 7.5

22.6 82.4 47.8 278 692 6.9

17.5 63.8 55.0 283 775 9.5

- loo*

0.0 13

68.3 13.8 1030 249 663

In the last two rows, the spectroscopic energy difference of the two maxima in the SAS (cf:Figs. 5a and 3b) and the calculated antenna ratio is listed. AG was calculated according to AG = -k,T In(k,-,/k,-,) k , , was kept constant during target analysis at 0.3 ns-I. *Measurementsdone with the sample in the cryostat (see Materials and Methods).

this value over a large range. The SAS for this model are given in Fig. 5a,b for +6"C and +42"C, respectively. The temperature dependence of the rate constants resulting from this analysis along with the calculated values for the free energy differencesbetween the excited pools according to the equation

AG

=

-kbTlntc)

are given in Table 3 and in Fig. 6, which shows an Arrhenius plot for the rate constants.

DISCUSSION

The global analysis procedure showed that four lifetime components are necessary to fit the data adequately at all temperatures. As can be seen from Table I , the values for the lifetimes T , to T~ as well as for the amplitudes vary only moderately with changing the temperature in the range +6"C to +42"C. The shortest lifetime component, T , , exhibits a positive amplitude representing a decay at wavelengths below about 700 nm and changes to negative values representing a rise term at longer wavelengths. This behavior is a strong indication for a fast energy transfer process. We assign 7,to an energy transfer between a short-wavelength-emitting pigment pool and a long-wavelength-emitting pigment pool. The lifetime component T* apparently represents the overall decay of the excited states of the antenna. The most reasonable interpretation of this process would be the charge separation in the RC. These two lifetimes and their DAS are not affected if strongly reducing or oxidizing conditions are established. They also do not depend significantlyon the flow rate of the sample or on the detergent concentration. The low amplitude components T~ and T~ do not correspond to typical intact and separated PS I particles. Lifetime T~ apparently belongs to aggregates that can be destroyed at higher temperatures and at higher detergent concentrations albeit at the expense of an increased amplitude of the long T~ component, which on the basis of its spectrum (maximum at X < 6 8 0 nm, see above) must be assigned to functionally uncoupled Chl.

Kinetic model We propose a straightforward kinetic model accounting for the T , (equilibration process) and 7* (overall excitation decay) as depicted in Fig. 4 . This model fits the data very well, at least at high temperatures. The SAS indicate emission maxima of X % 690 nm (F690) and X = 725 nm (F725) for the short- and long-wavelength-emittingpools, respectively. The SAS of the short-wavelength-emittingpool 1 (cf: Fig. 5a) shows a pronounced shoulder at X = 7 1 0 nm. This may represent in fact the emission spectrum of a mixture of two pools that might in turn be an indication for the existence of still unresolved faster equilibration processes within the initially excited pool 1 . This hypothesis is substantiated by the data at - 100°C (cf: Fig. 3b), where the emission maximum of pool 1 seems to be shifted to about 710 nm. We can estimate the ratio N,/N2 of the sizes (number of chlorophylls in ground state) of pools 1 and 2 by assuming a thermodynamic equilibrium according to

where h is Planck's constant, k, is the Boltzmann constant, cis the velocity oflight and X y x is the wavelength ofemission maximum of pool i as taken from SAS. If only two pools were present and thermodynamic equilibrium were established, we would expect a temperatureindependent constant ratio of N,/N2.As can be seen from Table 3, we calculate a value of about 7 for NJN,, which is nearly constant for the high temperature region. At - 1OOOC, however, N,M2is totally different. Additionally, the data at all temperatures indicate substantial spectral heterogeneity in the PS I antenna. In particular there are indications for a pool 3 emitting near 7 10 nm (F7 10). The transfer from pool 1 (F690) to pool 3 (F7 10) seems to be faster than our present time resolution of 7 ps. This would be consistent with transient absorption data of Causgrove et aLZ7who reported substantial amplitudes of 1-6 ps components in small PS I particles from spinach. The negative amplitude in the T~ component proves the first resolved energy transfer process between two different Chl pools in PS I at high temperature. We have also found a similar component in PS I stroma thylakoid particles from

118

SANDRATuRcoNi et al.

spinach.I6 At low temperature such components were resolved in large (-200 ChVP700) PS I particles by Mukerji and Sauer’ and WittmershauP who reported a 12 ps lifetime. studied PS I from Synechococcus sp. ( 5 5 or Owens et 122 Chl/P700) and also from C h l u m y d o m o n u s mutant^.'^ They did not observe a fast component with features comparable to our T~ component. Instead their lifetimes (29 ps and 38 ps for the small and large Synechocaccus particles, respectively) were very similar to those of the T, component reported here. These authors noted some emission wavelength dependence in the lifetimes, which probably derives from a similar, but unresolved, fast component as reported here. Wittmershaus and coworkers measured cyanobacterial PS I and found only one fast component which is interme~ case ~ the fludiate between our lifetimes T , and T ~ In.that orescence was not wavelength resolved. A similar result was obtained by Sparrow et a1.j’ for another cyanobacterial PS I. They suggested that their fast component should in fact originate from two different unresolved excited species. Owens et ul.-” and HolzwarthJZ suggested a linear dependence of the trapping time on the antenna size. An exact comparison of the overall exciton trapping time ( T ~component) found by us and by other groups is difficult, however, in view of the lifetime modifications introduced by unresolved kinetic components. Taking such lifetime averaging effects into account, our value for the T, component ofabout 33 ps at high temperatures seems to fit reasonably well into the range of trapping times spanned by small PS I particles ( 16 ps for 24 Chl/P700)zs and large ones ( 3 8 ps for 120 Chl/ P700).z8 Surprisingly, other data from spinach do not fit well into this variation (100 ChVP700, T = 90-100 PS).~Clearly the dependence ofthe trapping time on the antenna size needs further elucidation. The rate constant k,, represents the apparent charge separation process. It is slightly temperature dependent, indicating a temperature-activated charge separation process. This is a consequence of the presence of a long-wavelength antenna pool at lower energy than P700. In agreement with other groups we find that P700+ is an equally efficient quencher of excitons as P700*.z83’ Reducing the acceptor side also had no effect on the lifetimes, which is at variance with an increase in lifetime by a factor of 2 as reported in Owens et d z 8 Interestingly, we find little evidence for lifetimes, which could be indicative of a charge recombination as reported for PS Ij3 and for PS I L Z 6 . j 4 There occurs no significant change in the long lifetimes when going from normal to strongly reducing conditions (redox potential < -600 mV). This can only be explained, if the first radical pair decays very rapidly by charge stabilization and further, if the first long-lived radical pair has a large free energy gap to the P700*. A decay of the primary radical pair in less than 50 ps has indeed been suggested.j5 Our data indicate an equilibration of the excitons in the antenna over the various spectral forms with a lifetime of about 10 ps. This lifetime apparently is the slowest antenna equilibration component in the system. The data, particularly SAS at low temperatures, indicate additional and still faster, but kinetically unresolved, equilibration components. With an overall charge separation time of > 30 ps there exists at least a 3 : I ratio of exciton migration to charge separation. This clearly fulfills the conditions of a trap-limited exciton kinetics.j6 This finding is auite in contrast to the diffusion-

limited kinetics suggested by Owens and c o ~ o r k e r for s~~~~~ similar-sized particles. This fast antenna equilibration is also only in partial disagreement with the fairly slow single-step times on the order of a few picoseconds suggested by Lyle and S t r ~ v e and ’ ~ with the data from hole-burning expenments.j8 The more or less constant N,/N, ratio and the fairly linear Arrhenius plot of rate constants over the temperature range 6°C to +42”C indicate that the equilibrated two-pool model provides a reasonable description. However, the - 100°C data yield quite a different ratio (NJN, > 600). The drastic change in N,/N, along with the substantial change in the SAS provides a clear indication that the two-pool model is no longer good at low temperatures. At least one additional pool is required, having its emission maximum near 7 10 nm (pool 3). However, at the higher temperature we could not kinetically resolve this pool. Also at - 100°C we could not resolve any transfer from the main pool 1 (F690) to pool 3 . At present we cannot make a clear conclusion about the actual role of the long-wavelength pool (pool 2). Whether it serves to concentrate energy near the RC, as suggested for bacterial antenna9 and PS I , I u is not obvious. At least at lower temperatures the presence of the F720 pool slows down charge separation substantially, as indicated by rate constant k,, and its temperature dependence (see Fig. 6).

+

REFERENCES

1. Thornber, J. P., R. S. Alberte, F. A. Hunter, J. A. Shiozawaand K. S. Kan (1977) The organisation of chlorophyll in the plant photosynthetic unit. Brookhaven Symp. Biol. 28, 132-148. 2. Holzwarth, A. R. (1 991) Excited-statekinetics in chlorophyll systems and its relationship to the functional organization of the photosystems. In Chlorophylls (Edited by H. Scheer), pp. 1 125-1 151. CRC Press, Boca Raton, FL. 3. Butler, W. L. (1 96 1) A far-red absorbing form of chlorophyll, in vivo. Arch. Biochem. Biophys. 93, 4 13-422. 4. Butler, W. L., C . J. Tredwell, R. Malkin and J. Barber (1979)

5.

6.

7.

8. 9.

The relationship between the lifetime and yield of the 735 nm fluorescence of chloroplasts at low temperatures. Biochim. Biophys. Acta 545, 309-3 15. Kuang, T.-Y., J. H. Argyroudi-Akoyunoglou, H.Y. Nakatani, J. Watson and C . J. Arntzen (1984) The origin of the long-wavelength fluorescence emission band (77-degrees-K)from photosystem I. Arch. Biochem. 235, 618-627. Wittmershaus, B. P. (1987) Measurements and kinetic modeling of picosecond time-resolved fluorescence from photosystern I and chloroplasts. In Progress Photosynthesis Research, 1 (Edited by J. Biggins), pp. 75-82. Nijhoff Publishers, Dordrecht. Mukeqi, I. and K. Sauer (1989) Temperature-dependent steadystate and picosecond kinetic fluorescence measurements of a photosystem I preparation from spinach. In Photosynthesis. Plant Biology, Vol. 8 (Edited by W. R. Brigs), pp. 105-122. Alan R. Liss, New York. Mukerji, I. and K. Sauer (1990) A spectroscopic study of a photosystem I antenna complex. Curr. Res. Photosynth. 0 , 32 1324. van Grondelle, R., H. Bergstrom, H. Sundstrom,R. J. van Dorssen, M. Vosand C . N. Hunter ( 1 988) Excitationenergytransfer in the light-harvesting antenna of photosynthetic purple bacteria: the role of the long-wavelength absorbing pigment B896. In Photosynthetic Light-harvesting Systems. Organization and Function (Edited by H. Scheer and S. Schneider),pp. 519-530.

de Gruyter, Berlin. 10. van Grondelle, R. and V. Sundstrom (1988) Excitation energy transfer in photosynthesis. In Photosynthetic Light-Harvesting Systems. Organization and Function (Edited by H. Scheer and S. Schneiderl

DD.

403-438. de Gruvter. Berlin.

Picosecond kinetics of photosystem I 1 1. Holzwarth, A. R. (1987) Picosecond fluorescence spectroscopy

12.

13.

14. 15. 16.

17.

18.

19. 20.

21.

22. 23.

24.

and energy transfer in photosynthetic antenna pigments. In The Light Reactions. Topics in Photosynthesis, Vol. 8 (Edited by J. Barber), pp. 95-1 57. Elsevier, Amsterdam. Pellegrino, F., A. Dagen, P. Sekuler and R. R. Alfano (1983) Temperature dependence of the 735 nm fluorescence kinetics from spinach measured by picosecond laser-streak camera system. Photobiochem. Photobiophys. 6, 1 5-23. Searle, G. F. W., R. Tamkivi, A. van Hoek and T. J. Schaafsma (1 988) Temperature dependence of antenna chlorophyll fluorescence kinetics in photosystem I reaction center protein. J. Chem. SOC.Farad. Trans. I I 84, 315-327. Lyle, P. A. and W. S. Struve (1991) Temperature dependence of antenna excitation transport in native photosystem I particles. J. Phys. Chem. 95,4152-4158. Werst, M., Y. W. Jia, L. Mets and G. R. Fleming (1 992) Energy transfer and trapping in the photosystem I core antenna. A temperature study. Biophys. J. 61, 868-878. Holzwarth, A. R., W. Haehnel, R. Ratajczak, E. Bittersmann and G. H. Schatz (1990) Energy transfer kinetics in photosystern I particles isolated from Synechococcus sp. and from higher plants. In Current Research in Photosynthesis, 2 (Edited by M. Baltscheffsky),pp. 6 1 1-6 14. Kluwer Academic Publishers, Dordrecht. Witt, I., H. T. Witt, S. Gerken, W. Saenger, J. P. Dekker and M. Rogner (1987) Crystallization of reaction center I of photosynthesis. Low-concentration crystallization of photoactive protein complexes from the cyanobacterium Synechococcus sp. FEBS Lett. 221, 260-264. Miyairi, S. and G. H. Schatz (1983) Oxygen-evolvingextracts from a thermophilic cyanobacterium Synechococcus sp. Z. Naturforsch. C 38, 4 4 4 8 . Markwell, J. P., J. P. Thornber and M. P. Skrdla (1980) Effect of detergents on the reliability of a chemical assay for P-700. Biochim. Biophys. Acta 591, 391-399. Roelofs, T. A., M. Gilbert, V. A. Shuvalov and A. R. Holzwarth (1991) Picosecond fluorescencekinetics ofthe DI-D2-cyt-b559 photosystem I1 reaction center complex. Energy transfer and primary charge separation processes. Biochim. Biophys. Acta 1060, 237-244. Wendler, J. and A. R. Holzwarth (1987) State transitions in the green alga Scenedesmus obliquus probed by time-resolved chlorophyll fluorescence spectroscopy and global data analysis. Biophys. J. 52, 717-728. Beechem, J. M., M. Ameloot and L. Brand (1985) Global and target analysis of complex decay phenomena. Anal. Instrum. 14, 379-402. Miiller, M. G., H. Riimbeli, W. Sidler, H. Zuber and A. R. Holzwarth (1 992) Studies on chromophore coupling in isolated phycobiliproteins. IV. Energy transfer kinetics in phycoerythrocyanin trimers characterized by decay-associated and species-associated spectra. Biophys. J. (In press) Kok, B. (196 1) Partial purification and determination of oxidation reduction potential of the photosynthetic chlorophyll complex absorbing at 700 nm. Biochim. Biophys. Acta 48,527533.

119

25. Roelofs, T. A. and A. R. Holzwarth (1990) In search ofa putative long-lived relaxed radical pair state in closed photosystem 11. Kinetic modeling of picosecond fluorescence data. Biophys. J. 57, 1141-1153. 26. Schatz, G. H., H. Brock and A. R. Holzwarth (1988) A kinetic and energetic model for the primary processes in photosystem 11. Biophys. J. 54, 397405. 27. Causgrove, T. P., S. M. Yang and W. S. Struve (1988) Electronic excitation transport in core antennae of enriched photosystem I particles from spinach chloroplasts. J. Phys. Chem. 92, 61214124. 28. Owens, T. G., S. P. Webb, R. S. Alberte, L. Mets and G. R. Fleming (1988) Antenna structure and excitation dynamics in photosystem I. I. Studies of detergent-isolated photosystem I preparations using time-resolved fluorescence analysis. Biophys. J. 53, 733-745. 29. Owens, T. G., S. P. Webb, L. Mets, R. S. Alberte and G. R. Fleming (1 989) Antenna structure and excitation dynamics in photosystem I. 11. Studies with photosynthetic mutants of Chlamydomonas reinhardtii lacking photosystem 11. Biophys. J. 56, 95-106. 30. Wittmershaus, B. P., D. S. Berns and C. Huang (1987) Picosecond time-resolved fluorescence from detergent-free photosystem I particles. Biophys. J. 52, 829-836. 31. Sparrow, R., R. G. Brown, E. H. Evans and D. Shaw (1990) Time-resolved spectroscopy of photosynthetic systems. 3: Photosystem I preparations from the cyanobacterium Chlorogloea frilschii. J. Phorochem. Photobiol. B 5, 445-455. 32. Holzwarth, A. R. (1987) Pikosekundenstudien zum Energietransfer in photosynthetischen Antennenpigmenten. HabiLSchrij Univ.Marburg, 1-208. 33. Shuvalov, V. A., A. M. Nuijs, H. J. van Gorkom, H. W. J. Smit and L. N. M. Duysens (1986) Picosecond absorbance changes upon selective excitation of the primary electron donor P-700 in photosystem I. Biochim. Biophys. Acta 850, 319-323. 34. Krause, G. H. and E. Weis (1991) Chlorophyll fluorescence and photosynthesis: the basics. Annu. Rev. Plant Physiol. Plant Mol. Biol. 42, 3 13-349. 35. Nuijs, A. M., V. A. Shuvalov, H. J. Van Gorkom, J. J. Plijter and L. N. M. Duysens (1986) Picosecond absorbance difference spectroscopy on the primary reactions and the antennaexcited states in photosystem I particles. Biochim. Biophys. Acta 850, 310-318. 36. Beauregard, M., I. Martin and A. R. Holzwarth (199 1) Kinetic modelling of exciton migration in photosynthetic systems. (1) Effects of pigment heterogeneity and antenna topography on exciton kinetics and charge separation yields. Biochim. Biophys. Acta 1060, 271-283. 37. Owens, T. G., S. P. Webb, L. Mets, R. S. Alberte and G. R. Fleming (1 987) Antenna size dependence offluorescence decay in the core antenna of photosystem I: estimates of charge separation and energy transfer rates. Proc. Natl. Acad. Sci. USA 84, 1532-1 536. 38. Gillie, J. K., G. J. Small and J. H. Golbeck (1989) Nonphotochemical hole burning of the native antenna complex of photosystem I (PSI-200). J. Phys. Chem. 93, 1620-1627.