Research Article

Primary processes in the bacterial reaction center probed by two-dimensional electronic spectroscopy

Andrew Niedringhaus, Veronica R. Policht, Riley Sechrist, Arkaprabha Konar, Philip D. Laible, David F. Bocian, Dewey Holten, Christine Kirmaier, and Jennifer P. Ogilvie
PNAS April 3, 2018 115 (14) 3563-3568; first published March 19, 2018 https://doi.org/10.1073/pnas.1721927115

  1. Edited by Arieh Warshel, University of Southern California, Los Angeles, CA, and approved March 1, 2018 (received for review December 16, 2017)

Significance

The remarkable near-unity quantum efficiency of photosynthetic charge separation has motivated decades of research to uncover the underlying design principles. Much of our current understanding of photosynthetic charge separation is rooted in studies of the bacterial reaction center (BRC). We present two-dimensional electronic spectroscopy of the BRC as it undergoes charge separation, resolving the energy-transfer and charge-separation processes with time and excitation frequency resolution. These measurements reveal the excitonic structure of the BRC, including a previously hidden exciton state. We present a multiexcitation 2D global analysis method that supports two-step sequential charge separation in the BRC without evidence for secondary charge separation pathways. We extract the spectral signatures of the charge-separation intermediates.

Abstract

In the initial steps of photosynthesis, reaction centers convert solar energy to stable charge-separated states with near-unity quantum efficiency. The reaction center from purple bacteria remains an important model system for probing the structure–function relationship and understanding mechanisms of photosynthetic charge separation. Here we perform 2D electronic spectroscopy (2DES) on bacterial reaction centers (BRCs) from two mutants of the purple bacterium Rhodobacter capsulatus, spanning the Qy absorption bands of the BRC. We analyze the 2DES data using a multiexcitation global-fitting approach that employs a common set of basis spectra for all excitation frequencies, incorporating inputs from the linear absorption spectrum and the BRC structure. We extract the exciton energies, resolving the previously hidden upper exciton state of the special pair. We show that the time-dependent 2DES data are well-represented by a two-step sequential reaction scheme in which charge separation proceeds from the excited state of the special pair (P*) to P+HA− via the intermediate P+BA−. When inhomogeneous broadening and Stark shifts of the B* band are taken into account we can adequately describe the 2DES data without the need to introduce a second charge-separation pathway originating from the excited state of the monomeric bacteriochlorophyll BA*.

Photosynthesis, the process by which sunlight is converted into chemical energy, is arguably the most important chemical reaction required to sustain the abundance of life on Earth. In addition to their ubiquitous role in supporting life, photosynthetic systems also offer important model systems for understanding the efficient conversion of photoexcitation into stable charge-separated states, potentially impacting the development of artificial light-harvesting devices (1). Despite decades of studies characterizing the structure and photoexcitation dynamics of the reaction centers where primary charge separation occurs, there remain many open questions regarding the relationship between the structure of reaction centers (RCs) and the functionality of the system as a whole (2).

The bacterial RC (BRC) from purple bacteria is a widely studied model system which forms the basis for much of our current understanding of photosynthetic charge separation (2). The BRC is composed of a pseudo-twofold-symmetric hexameric core of pigments (shown in SI Appendix, Fig. S1) that convert excitation energy to a stable charge-separated state. The near-unity quantum efficiency of charge separation, and the high specificity with which charge separation occurs along the “A” branch of the BRC structure, make it a remarkable system (2). The BRC pigments include a strongly coupled “special pair” (collectively P) of bacteriochlorophyll a (BChl), as well as two additional BChl (BA and BB) and bacteriopheophytin (HA and HB), labeled for their respective location on the A- and B branches. Compared with the Photosystem I and Photosystem II RCs in oxygenic photosynthesis, the BRC features distinct absorptions in the Qy region (Fig. 1), making it a simpler system for resolving the ultrafast processes of energy transfer and charge separation (3). Due to the ultrafast timescales of its initial photoinduced charge separation, BRCs have been studied extensively using time-resolved spectroscopy. Ultrafast transient absorption and time-resolved fluorescence experiments under a wide variety of excitation conditions, detection wavelengths, and temperatures have been performed on BRCs from Rhodobacter sphaeroides, Rhodobacter viridis, and Rhodobacter capsulatus and have been reviewed extensively (2, 4⇓–6).

Fig. 1.
Fig. 1.

Linear absorption spectrum (77 K) of the W(M250)V mutant in 50/50 buffer/glycerol (black), showing the underlying exciton states derived from multiexcitation 2D global fitting of the data. The P+*, P−*, HA*, and HB* exciton states are modeled with Gaussian profiles, while the BA* and BB* peaks are each constructed from a Gaussian distribution of Lorentzian profiles used to model the inhomogeneity in the 2D spectra. The fit to the linear spectrum is shown in red.

Two mechanisms considered in early studies of primary charge separation in the BRC were a two-step sequential process from the excited special pair P* to P+HA− via an intermediate P+BA− and one-step superexchange (P* → P+HA−), or a combination of the two, until P+BA− was resolved (4, 7). Several studies present evidence of alternative charge-separation pathways not involving excitation of the special pair, such as BA* → BA+HA− or B* → P+BA− (8⇓⇓⇓–12). Following the formation of P+HA−, electron transfer proceeds to QA and then across to QB. A second photoexcitation leads to formation of a doubly reduced QB. After acquiring two protons and departing the RC, the reduced quinone is oxidized by the cytochrome bc1, fueling the pumping of protons, leading to the transmembrane potential gradient used for ATP production (6).

In the past two decades, the technique of 2D electronic spectroscopy (2DES) has been successfully applied to study a variety of photosynthetic systems (13⇓⇓–16). By spreading the frequency dependence of the optical response onto two axes (ωex, ωdet), 2DES reduces the spectral congestion that often makes interpretation of transient absorption experiments challenging for multichromophoric systems such as photosynthetic RCs, directly exposing electronic coupling and energy-transfer events as cross-peaks in the 2DES spectrum (14, 17). Homogeneous and inhomogeneous linewidths, which broadband transient absorption cannot resolve, can be distinguished in 2DES. To study excitation-dependent differences in kinetics in spectrally congested systems, transient absorption spectroscopy uses narrowband pulses to achieve selective excitation, at the expense of time resolution. Fourier transform 2DES decouples the time resolution and excitation-frequency selectivity (17). Performing 2DES with 10–12-fs pulses as we do here provides higher time resolution than previous BRC experiments (11, 18⇓–20).

There have been several previous 2DES studies of the BRC. Schlau-Cohen et al. (21) used 2DES to study the B* band in oxidized BRCs, resolving the BA and BB transitions and a sub-100-fs interaction between them. There have also been several groups that have recently studied coherent processes in the BRC using 2DES and two-color photon echo spectroscopy. Coherent effects have been of considerable recent interest in 2DES studies of photosynthetic antennae (22, 23) and RCs (24, 25). Westenhoff et al. (15) studied the oxidized BRC, resolving rapid energy transfer between H* and B*, and observing coherence lasting longer than the energy-transfer time. Using polarization-dependent 2DES they assigned the coherences to electronic and mixed electronic-vibrational origin. Ryu et al. (26) studied neutral, oxidized, and mutant BRCs using two-color photon echo spectroscopy, assigning the observed coherences to mixed vibrational-electronic origin. Flanagan et al. (27) studied coherences in oxidized preparations of several mutant BRCs, finding that the mutations affect energy gaps but not energy-transfer rates. Here we study the charge-separating BRC, using 2DES to resolve its excitonic structure and the mechanisms of energy transfer and charge separation. Some of the fastest kinetics we resolve occur on sub-100-fs timescales, when coherent effects could be difficult to disentangle. We observe coherent signals from the BRC that depend strongly on excitation and detection frequency. For the purposes of extracting the kinetics, we show in SI Appendix, Fig. S4 that the coherences produce relatively small modulations of the population kinetics (<10%) and are unlikely to have a large effect on the kinetics analysis we report here.

We present 2DES studies of the BRC from two mutants of R. capsulatus. The first mutant, W(M250)V, is a wild-type RC except for its lack of QA, which enables higher repetition rate experiments without the accumulation of the long-lived P+QA− and P+QB− product states (28). In the second mutant, DLL, the HA pigment is absent (29, 30) and the high oxidation potential of P precludes charge separation (31, 32). DLL was studied to provide insight into the charge-separation process in the W(M250)V mutant. To uncover reaction pathways, many transient absorption studies perform some type of global analysis (33⇓–35) and/or target analysis (33, 34, 36⇓–38). The general goal of such an approach is to find an appropriate kinetic model that describes the process of interest, extracting the spectral signatures of real chemical species and reaction intermediates, and successfully reproducing the time-dependent transient absorption signals. Global analysis of transient absorption data for the BRC can lead to nonunique results (39). Nonuniqueness problems can be somewhat mitigated by more extensive transient absorption datasets at multiple excitation wavelengths (40). By resolving excitation wavelengths spanning the pump-pulse bandwidth, 2DES can be considered to be a massively parallel transient absorption dataset, with higher time resolution afforded by the Fourier transform nature of the method. It therefore carries a wealth of kinetic information for testing kinetic models through global analysis.

Several approaches have been taken to extracting kinetics from 2D spectra. Myers et al. (16) performed exponential fits to each (ωex, ωdet) point in the 2D spectra of the Photosystem II RC, generating amplitude and lifetime maps to illustrate the 2D spectral signatures and corresponding timescales of kinetic processes. Global analysis of full 2D data has also been performed, where a small number of exponentials are fit to the data to extract 2D “decay-associated spectra” (41). As for global analysis of transient absorption data, this approach works well when the timescales of the different kinetic processes are well separated (39). The standard 2D global analysis approach ignores knowledge about the underlying excitonic structure and connectivity of the states that can be directly revealed from 2DES spectra. Thyrhaug et al. (42) have demonstrated that global fitting of polarization-dependent 2DES and linear absorption data can be used to directly retrieve energy-transfer rates and uncover the energy-transfer pathways in the Fenna–Matthews–Olson complex. Dostál et al. (43) have expanded on this work, showing that global fitting of 2D spectra in combination with linear absorption spectra can uniquely identify kinetic models under some conditions. Here we extend these previous approaches (42, 43) to include the intermediate charge-separation states, inhomogeneous broadening, and Stark shifts induced by the charge-separated product state. We test two different models of charge separation, and show that the 2DES data can be well described by a single pathway in which charge separation proceeds via a two-step sequence.

Results

The real absorptive 2DES spectra from W(M250)V with magic-angle polarization are shown in Fig. 2 for several representative population times. In accordance with other 2DES studies, and in contrast to transient absorption spectroscopy, we adopt the sign convention where ground-state bleach (GSB) and stimulated emission (SE) contributions are positive, while excited-state absorption (ESA) contributions are negative in absorptive spectra. At early population times, the 2D spectrum along the diagonal is predominantly a combination of SE and GSB components from the initial excitation, with features corresponding to the linear absorption spectrum. There are clearly discernible diagonal peaks corresponding to the P*, B*, and H* bands. The individual contributions from the BA* and BB* excitonic states are apparent from the broadening of the B*-band peak along the diagonal. This B*-band splitting is also visible as a shoulder on the red side of the B* band in the 77-K linear absorption spectrum in Fig. 1. Within the first 30 fs, there is already significant energy transfer from H* → B*, and from B* → P*, indicated by the H*/B* and B*/P* cross-peaks below the diagonal. The spectrum along the detection axis at 11,500 cm−1 excitation frequency corresponds to excitation of P*, and shows a weak ESA signal at 12,400 cm−1. By 100 fs, an H*/P* cross-peak appears, which indicates H* → P* energy transfer occurs on a 100-fs timescale (44). The diagonal B*-band peak shows inhomogeneous broadening corresponding to BA* and BB* excitations. At 500 fs, the H*-band diagonal peak has decayed almost entirely, and most of the H*/B* cross-peak amplitude has also diminished due to rapid energy transfer to P*. The B*- and P*-excitation bands show similar broad, negative ESA features above the diagonal, indicating the formation of a common product state.

Fig. 2.
Fig. 2.

(Top Row) Absorptive 77-K 2DES spectra from the W(M250)V mutant BRC in 50/50 buffer/glycerol with magic-angle polarization at different T waiting times. (Bottom Row) Two-dimensional spectra reconstructed from excitation-dependent global fits to the single-pathway model. The dashed oval in the 30-fs measured spectrum indicates the location of the P+* state, which shows a weak diagonal signal and cross-peak from internal conversion to P−*. A magnified view of this spectral region is given in SI Appendix, Fig. S3.

Between 500 fs and 2 ps, we observe a growth in the amplitude of the negative ESA peaks present at 11,500 and 12,400 cm−1 excitation frequencies, as well as a pronounced splitting of the B*-band peaks near the diagonal, evolving from the inhomogeneously broadened diagonal peak at early times to two horizontal bands at detection frequencies ∼12,300 and 12,500 cm−1. The cross-section of the 2 ps 2DES spectra with B*-band excitation frequency resembles the P+BA− difference spectrum reported in a number of transient absorption studies (33, 34, 36), which is consistent with the initiation of charge separation on a picosecond timescale. After the first several picoseconds, the 2D spectrum exhibits little change; the B*-band splitting vanishes, and the detected spectra from P*, B*, and H* excitations all resemble the same product state. Since W(M250)V lacks QA, we can assign this final state to P+HA−, which decays on a 20-ns timescale (45), slower than we can accurately resolve in our 1-ns scan.

Here we aim to determine an appropriate model that can represent the entire 2DES data set, at all excitation frequencies, as a sum of a single set of basis spectra with time-dependent concentrations representing the actual mixtures of each state for a given excitation. We first used a single-pathway model, in which downhill energy transfer H* → B* → P* on the A- and B branches is followed by the charge-separation sequence P* → P+BA− → P+HA−. This model can be represented by seven time constants as shown in Fig. 3A. We simultaneously fit the linear absorption spectrum and magic-angle 2DES data using a multiexcitation 2D global analysis approach as detailed in SI Appendix. The linear absorption spectrum was decomposed into six excitonic states: the upper and lower excitons P+* and P−* of the special pair, with the remaining states BA*, BB*, HA*, and HB* named for their dominant contributions from the other four A- and B-side pigments. The result of the multiexcitation 2D global analysis for the single-pathway model was the set of “species associated difference spectra” (SADS) shown in Fig. 3B, representing the six excitonic states and two charge-separated states P+BA− and P+HA−. We note that the bandwidth of our probe spectrum restricts the spectral range of the SADS. We employed several constraints to the fitting process to avoid unphysical solutions as discussed in detail in SI Appendix. These included using the mutant DLL to distinguish kinetic processes associated with charge separation and to corroborate our assignment of the P−* basis spectra which is in reasonable agreement with other reports for P−* in the literature (36, 46). The P+* SADS was determined from fits to the 2D data at an excitation frequency of 11,900 cm−1 to capture this weak and short-lived intermediate. The basis spectra for HA* and HB* were constrained to the same Gaussian profiles used for the linear absorption fit in Fig. 1.

Fig. 3.
Fig. 3.

(A) Single charge-separation pathway model with extracted time constants from the multiexcitation 2D global analysis. Initial excitation of HA*, HB*, BA*, P+*, and P−* populates the homogeneous P+HA− state represented by a single basis spectrum, while excitation of BB* populates a mixture of inhomogeneous P+HA−(I) states. (B) Optimized basis spectra. HB*, HA*, BA*, and BB* have the same form as the linear absorption fit in Fig. 1. BA(I)* and BB(I)* show a subset of the inhomogeneous Lorentzian lineshapes. The P+* and P−* spectra were constrained based on fits to the data at an excitation frequency of 11,900 cm−1 as described in SI Appendix. The P+HA− and P+HA−(I) states were derived from the BA(I)* and BB(I)* Lorentzian bands as described in the text, and the P+BA− difference spectrum is calculated by linear least squares.

Our initial global fitting neglected the obvious inhomogeneous broadening of the BA* and BB* bands, and failed to reproduce the persistent B*-band diagonal elongation at long waiting times. These fits are discussed in detail in SI Appendix, along with additional details of the fitting process. To better reproduce the B*-band structure, we modeled the inhomogeneous broadening of the BA* and BB* bands in the following way: The BA* and BB* linear absorption spectra shown in Fig. 1 were each constructed from a Gaussian distribution of Lorentzian lineshapes representing inhomogeneous and homogeneous linewidths, respectively. For each excitation frequency of the 2D spectrum, we calculate the spectral overlap of the pump with each Lorentzian to determine the initially excited populations of the inhomogeneous substates comprising BA* and BB*. The time dependences of the BA* and BB* concentrations are then treated in the same way as the other states using a single set of rate constants. This simplification neglects any inhomogeneity in the population kinetics, producing effective rate constants for the distribution of BA* and BB* states, but it allows us to capture the inhomogeneous linewidths apparent in the 2D spectrum without computing populations for hundreds of states in each iteration of the optimization. The BA* and BB* basis spectra are the sum of their Lorentzian peaks (shown as BA(I)* and BB(I)* in Fig. 3), weighted by the initial spectral overlap at each excitation frequency. The homogeneous and inhomogeneous widths of BA* and BB* are each allowed to vary as free parameters in the global fit, and are simultaneously used for the linear absorption and 2DES fitting.

We also observe significant inhomogeneity in the B* band after formation of P+HA−. The derivative lineshape of the P+HA− spectrum in this region has been attributed to Stark shifts of the B* band (47). To capture this persistent diagonal elongation in the B* band at long waiting times, we constructed a set of inhomogeneous P+HA− basis spectra by applying Stark shifts to the same Lorentzian peaks used for the BA* and BB* states. This provided a self-consistent model relating the linear absorption, BA*, BB*, and P+HA− spectra, which depends only on physically meaningful parameters (i.e., peak positions, widths, dipole strengths, and Stark shifts of BA* and BB*). Using this approach, we obtained the best agreement with the measured 2D spectrum when only the P+HA− spectra excited from BB* exhibit inhomogeneity, while the P+HA− spectrum excited by way of BA* was the ensemble average of the BA* and BB* Stark shifts. In other words, the P+HA− Stark shift only shows excitation frequency dependence upon excitation of BB*. This can be seen directly from the 2D spectra in Fig. 2 by noting the reduced amplitude above the diagonal on the red side of the B* band at 2 and 50 ps. Cross-sections of the 2D spectra at 50 ps (SI Appendix, Fig. S15) show that the P+HA− spectrum excited at frequencies corresponding to P*, BA*, and H* are well-matched, while excitation at BB* shows a clear shoulder at ∼12,400 cm−1 from the BB* Stark shift. The loss of correlation between excitation frequency and the BA* Stark shift may be attributed to large perturbations in the protein environment accompanying electron transfer on the A branch. To obtain satisfactory fits to the B* band at long waiting times we also found it necessary to include Stark shifts of the BA* and BB* bands upon formation of P+HA−. We obtained Stark shifts of 207 and 135 cm−1 for the BA* and BB* transitions, respectively. The exciton energies and dipole strengths extracted from the global analysis are given in SI Appendix, Table S1.

The fitted time constants for the energy-transfer and charge-separation processes, and the basis spectra used in the multiexcitation 2D global fit are shown in Fig. 3. Additional information about the global fit and comparisons between the fit and the 2DES data are given in SI Appendix, Figs. S13–S16, showing the quality of the fit throughout the 2DES data set. We also performed multiexcitation 2D global analysis with a two-pathway model that included both P* → PA+BA− → P+HA− and the possibility of charge separation via BA* → BA+HA− → P+HA−. The model and resulting fits are discussed in SI Appendix.

Discussion

The multiexcitation 2D global analysis reveals a detailed, high time-resolution picture of the energy-transfer and charge-separation processes in the BRC. We find that significant HA* → B* energy transfer occurs within the first 50 fs, which is faster than previously observed (11, 48). Our HB* energy-transfer time is in agreement with the 2D study of Westenhoff et al. (15). The exciton energies we obtain are in reasonable agreement with other values in the literature for studies of R. sphaeroides (44). We clearly resolve the upper exciton P+*, which is directly observable as a weak diagonal peak in the 2DES data, with coupling and internal conversion to P−* evident in the below-diagonal cross-peak at T= 30 fs (see dashed oval in Fig. 2 and SI Appendix, Fig. S3). We find that P+* has a frequency of 11,900 cm−1 (840 nm) at 77 K, which is red-shifted compared with previous estimates. In R. sphaeroides, theoretical work has predicted P+* to have a frequency of 12,346 cm−1 (810 nm) and 12,285 cm−1 (814 nm) at 77 K and room temperature, respectively (44). Other theoretical work has suggested that charge-transfer states mix strongly with the special pair states, causing significant energy shifts (49). Experiments on R. sphaeroides at 1.5–10 K assign P+* to a location between 12,225 and 12,821 cm−1 (818–780 nm) (50⇓⇓–53), or at 12,121 cm−1 (825 nm) at room temperature (54). The difference between our P+* assignment of 11,900 cm−1 (840 nm) and that of previous work may be the result of differences between species studied or solvent mixture which has been shown to cause spectral shifts (55). The low oscillator strength of P+*, its coupling and proximity to other exciton states, and its rapid internal conversion have made it historically difficult to detect. By spreading the signal along two dimensions, 2DES has the advantage over other methods of high time resolution and the ability to resolve weak transitions in congested spectra. We observe P+* → P−* internal conversion with a 25-fs time constant, somewhat faster than previous estimates from lower time resolution measurements (4).

Using our multiexcitation 2D global-fitting approach, we find a minimal representation of the 2D spectrum using a common set of basis spectra and rate constants for all excitation frequencies. Our approach utilizes the entire 2D spectrum, the linear absorption spectrum, and additional constraints including a direct comparison of the kinetics of the DLL and W(M250)V mutants. We find that we can satisfactorily reproduce the 2DES data using an eight-state model that considers the single P−* → P+BA− → P+HA− charge-separation pathway, in addition to inhomogeneous broadening and Stark shifts of the B* band. The model constrains the basis spectra and rates to account for the interdependence of different excitation frequencies of the 2D spectrum. The low-frequency tails of the P−* and P+HA− SADS shown in Fig. 3B are quite similar due to the bandwidth of our probe which cuts off near 11,100 cm−1 (Fig. 1). This prevents observation of the expected larger Stokes shifted emission from the P−* state (46). Given the bandwidth limitations of our measurement, we find reasonable agreement with other reported SADS for P−* (36, 46). The 2DES data clearly reveal inhomogeneous broadening in the B* band that persists at long waiting times. We found that without explicitly including inhomogeneous broadening of the BA* and BB* transitions, the single-pathway (P−* → P+BA− → P+HA−) model could not capture the persistent inhomogeneous B*-band structure of the 2DES data, as described in SI Appendix. Interestingly, the BA* transition displays spectral diffusion on the timescale of charge separation, while the BB* inhomogeneity persists for >1 ns. We attribute the difference between the spectral diffusion on the A- and B side to the asymmetric charge separation that occurs on the A branch. The rapid spectral diffusion of BA* is consistent with the transient reduction of BA giving rise to significant perturbation and reorganization of the BA electrostatic environment, causing a loss of correlation between the initial excitation and final detection frequencies. In contrast, on the B branch, where no transient reduction occurs, the correlation is maintained, and in combination with the Stark shift of BB* gives rise to the persistent elongation in the 2DES data at long T delays.

Stark shifts have been previously reported in BRC studies (56⇓–58). From our global analysis we obtained Stark shifts of 207 and 135 cm−1 for the BA* and BB* transitions, respectively. These values are in reasonable agreement with the 15-K report by Vos et al. (57), which found a Stark shift of 220 cm−1 for BA*, but which did not report a value for BB*. Our values differ from the room-temperature measurements made by Guo et al. (58), who obtained respective shifts for BA* and BB* of 104 and 130 cm−1. Stark shifts in the BRC are known to be temperature dependent, but it is not clear that the different temperatures of the experiment of Guo et al. (58) can fully account for the difference between our measurements.

We also performed multiexcitation 2D global analysis of a second model that included a parallel charge-separation pathway originating from BA*. In this model BA* → BA+HA− → P+HA− occurs in parallel with P−* → P+BA− → P+HA−. This second model was also able to nicely reproduce the 2DES data when the B*-band inhomogeneity and Stark shifts were taken into account. This is hardly surprising given that this model adds an additional basis state and two additional time constants to the single-pathway model (SI Appendix, Figs. S17 and S20 and Table S3). Most of the time constants appear reasonable, with the exception of a 93-ps timescale for the initial charge separation P−* → P+BA−, which is more than an order of magnitude longer than the timescale reported in other studies and obtained by us with the single-pathway model. The basis spectrum for BA+HA− also appears to largely consist of a linear combination of the P+BA− and P+HA− basis spectra, calling into question the need for this additional state. We note that fits to the second model that neglect the B*-band inhomogeneity and Stark shifts also failed to satisfactorily reproduce the B*-band structure at long waiting times, and produced linearly dependent basis spectra for the charge-separation intermediates. They also gave an unreasonably long P+BA− → P+HA− timescale as detailed in SI Appendix. Thus, we conclude that the information contained solely in the 2DES spectrum spanning the 700–900-nm region of the Qy bands does not support the presence of additional charge-separation pathways beyond the single P−* → P+BA− → P+HA− pathway model. Compared with previous work that has supported the existence of a secondary pathway involving charge separation from BA*, here we include inhomogeneous broadening and Stark shifts of the B* band. Previous work reporting the BA* pathway has employed transient absorption spectroscopy, which is less suited than 2DES to exposing inhomogeneity. It may be that the need to account for B*-band inhomogeneity and Stark shifts is at the root of the additional charge-separation pathways introduced in other work. It is also possible that the inclusion of data outside the Qy region, as used in several studies that support the existence of the BA* pathway (9), may provide support for additional charge-separation pathways that were not resolved in our study.

Conclusions

We have presented a broadband 2DES study of two BRC mutants from R. capsulatus spanning the Qy region on timescales ranging from 10 fs to 1 ns at 77 K. This multidimensional dataset contains a wealth of information about the energy- and charge-transfer processes, with time and excitation frequency resolutions not achievable with transient absorption spectroscopy. We developed and applied a multiexcitation 2D global analysis technique to disentangle the internal conversion, energy-transfer and charge-separation kinetics, and extract physically meaningful time constants and spectral signatures of charge-transfer intermediates. We have observed rapid downhill energy transfer from H* → B* and B* → P* on ∼100-fs timescales. We have resolved the upper exciton P+* of the special pair, and its rapid internal conversion to the lower exciton P−*. We find that a single-pathway model, in which charge separation proceeds in two sequential steps from P−* to P+HA− through the P+BA− intermediate, describes our 2DES data well when inhomogeneity and Stark shifts of the B* band are taken into account. We find little support for a secondary charge-separation pathway initiated from BA*. We hope that the extensive 2D dataset presented here will stimulate theoretical work to develop atomistic modeling of photosynthetic charge separation in the BRC.

Materials and Methods

We employed the 2DES setup described previously (59), with some modifications to probe the broad near-IR absorption bands of the R. capsulatus. Details regarding the experimental setup, sample preparation, and data analysis are presented in SI Appendix.

Acknowledgments

The authors thank Yin Song for assistance with acquiring data on the DLL mutant. The authors also thank Steven Boxer and Jessica (Chuang) Seeliger for the gift of the DLL mutant sample that was prepared with support from the NSF (Grant MCB-0416623). V.R.P., A.N., A.K., and J.P.O. gratefully acknowledge the support of the National Science Foundation (NSF) through Grant PHY-1607570 and instrumentation Grant CHE-1428479. R.S. acknowledges funding from the NSF Graduate Fellowship program. The US Department of Energy, Office of Science, Office of Basic Energy Sciences supported C.K. and D.H. (Grant DE-SC0002036) and P.D.L. (associated Field Work Proposal) for the W(M250)V mutant. C.K. and D.H. acknowledge the NSF (Grant MCB-0314588) for work on the DLL mutant.

Footnotes

  • ↵1To whom correspondence should be addressed. Email: jogilvie{at}umich.edu.

  • Author contributions: D.F.B., D.H., C.K., and J.P.O. designed research; A.N. and V.R.P. performed research; P.D.L., D.H., and C.K. contributed new reagents/analytic tools; A.N., V.R.P., R.S., and A.K. analyzed data; A.N., V.R.P., R.S., A.K., P.D.L., D.F.B., D.H., C.K., and J.P.O. discussed data, analysis, and interpretation; and A.N. and J.P.O. wrote the paper.

  • The authors declare no conflict of interest.

  • This article is a PNAS Direct Submission.

  • This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1721927115/-/DCSupplemental.

Published under the PNAS license.

References

  1. ↵
    1. Scholes GD,
    2. Fleming GR,
    3. Olaya-Castro A,
    4. van Grondelle R
    (2011) Lessons from nature about solar light harvesting. Nat Chem 3:763–774.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Parson WW,
    2. Warshel A
    (2009) Mechanism of charge separation in purple bacterial reaction centers. The Purple Phototrophic Bacteria, eds Hunter CN, Daldal F, Thurnauer MC, Beatty JT (Springer Science, Dordrecht, The Netherlands), pp 355–377.
  3. ↵
    1. Yoder LM,
    2. Cole AG,
    3. Sension RJ
    (2002) Structure and function in the isolated reaction center complex of photosystem II: Energy and charge transfer dynamics and mechanism. Photosynth Res 72:147–158.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Zinth W,
    2. Wachtveitl J
    (2005) The first picoseconds in bacterial photosynthesis–Ultrafast electron transfer for the efficient conversion of light energy. ChemPhysChem 6:871–880.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Allen JP,
    2. Williams JC
    (1998) Photosynthetic reaction centers. FEBS Lett 438:5–9.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Blankenship RE
    (2002) Molecular Mechanisms of Photosynthesis (Blackwell Science, Oxford), pp 1–321.
  7. ↵
    1. Kakitani Y,
    2. Hou A,
    3. Miyasako Y,
    4. Koyama Y,
    5. Nagae H
    (2010) Rates of the initial two steps of electron transfer in reaction centers from Rhodobacter sphaeroides as determined by singular-value decomposition followed by global fitting. Chem Phys Lett 492:142–149.
    OpenUrl
  8. ↵
    1. van Brederode ME, et al.
    (1998) On the efficiency of energy transfer and the different pathways of electron transfer in mutant reaction centers of Rhodobacter sphaeroides. Photosynth Res 55:141–146.
    OpenUrlCrossRef
  9. ↵
    1. Huang L,
    2. Ponomarenko N,
    3. Wiederrecht GP,
    4. Tiede DM
    (2012) Cofactor-specific photochemical function resolved by ultrafast spectroscopy in photosynthetic reaction center crystals. Proc Natl Acad Sci USA 109:4851–4856.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    1. Lin S,
    2. Jackson J,
    3. Taguchi AKW,
    4. Woodbury NW
    (1998) Excitation wavelength dependent spectral evolution in Rhodobacter sphaeroides R-26 reaction centers at low temperatures: The Q(y) transition region. J Phys Chem B 102:4016–4022.
    OpenUrl
  11. ↵
    1. Vos MH,
    2. Breton J,
    3. Martin JL
    (1997) Electronic energy transfer within the hexamer cofactor system of bacterial reaction centers. J Phys Chem B 101:9820–9832.
    OpenUrlCrossRef
  12. ↵
    1. Zhou HL,
    2. Boxer SG
    (1998) Probing excited-state electron transfer by resonance Stark spectroscopy. 1. Experimental results for photosynthetic reaction centers. J Phys Chem B 102:9139–9147.
    OpenUrl
  13. ↵
    1. Brixner T, et al.
    (2005) Two-dimensional spectroscopy of electronic couplings in photosynthesis. Nature 434:625–628.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Lewis KLM,
    2. Ogilvie JP
    (2012) Probing photosynthetic energy and charge transfer with two-dimensional electronic spectroscopy. J Phys Chem Lett 3:503–510.
    OpenUrlCrossRef
  15. ↵
    1. Westenhoff S,
    2. Palecek D,
    3. Edlund P,
    4. Smith P,
    5. Zigmantas D
    (2012) Coherent picosecond exciton dynamics in a photosynthetic reaction center. J Am Chem Soc 134:16484–16487.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Myers JA, et al.
    (2010) Two-dimensional electronic spectroscopy of the D1-D2-cyt b559 photosystem II reaction center complex. J Phys Chem Lett 1:2774–2780.
    OpenUrlCrossRef
  17. ↵
    1. Jonas DM
    (2003) Two-dimensional femtosecond spectroscopy. Annu Rev Phys Chem 54:425–463.
    OpenUrlCrossRefPubMed
  18. ↵
    1. King BA,
    2. McAnaney TB,
    3. deWinter A,
    4. Boxer SG
    (2000) Excited state energy transfer pathways in photosynthetic reaction centers. 3. Ultrafast emission from the monomeric bacteriochlorophylls. J Phys Chem B 104:8895–8902.
    OpenUrlCrossRef
  19. ↵
    1. Stanley RJ,
    2. King B,
    3. Boxer SG
    (1996) Excited state energy transfer pathways in photosynthetic reaction centers. 1. Structural symmetry effects. J Phys Chem 100:12052–12059.
    OpenUrlCrossRef
  20. ↵
    1. Jonas DM,
    2. Lang MJ,
    3. Nagasawa Y,
    4. Joo T,
    5. Fleming GR
    (1996) Pump-probe polarization anisotropy study of femtosecond energy transfer within the photosynthetic reaction center of Rhodobacter sphaeroides R26. J Phys Chem 100:12660–12673.
    OpenUrlCrossRef
  21. ↵
    1. Schlau-Cohen GS,
    2. De Re E,
    3. Cogdell RJ,
    4. Fleming GR
    (2012) Determination of excited-state energies and dynamics in the B band of the bacterial reaction center with 2D electronic spectroscopy. J Phys Chem Lett 3:2487–2492.
    OpenUrlCrossRef
  22. ↵
    1. Engel GS, et al.
    (2007) Evidence for wavelike energy transfer through quantum coherence in photosynthetic systems. Nature 446:782–786.
    OpenUrlCrossRefPubMed
  23. ↵
    1. Tiwari V,
    2. Peters WK,
    3. Jonas DM
    (2013) Electronic resonance with anticorrelated pigment vibrations drives photosynthetic energy transfer outside the adiabatic framework. Proc Natl Acad Sci USA 110:1203–1208.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Fuller FD, et al.
    (2014) Vibronic coherence in oxygenic photosynthesis. Nat Chem 6:706–711.
    OpenUrl
  25. ↵
    1. Romero E,
    2. Novoderezhkin VI,
    3. van Grondelle R
    (2017) Quantum design of photosynthesis for bio-inspired solar-energy conversion. Nature 543:355–365.
    OpenUrl
  26. ↵
    1. Ryu IS,
    2. Dong H,
    3. Fleming GR
    (2014) Role of electronic-vibrational mixing in enhancing vibrational coherences in the ground electronic states of photosynthetic bacterial reaction center. J Phys Chem B 118:1381–1388.
    OpenUrl
  27. ↵
    1. Flanagan ML, et al.
    (2016) Mutations to R. sphaeroides reaction center perturb energy levels and vibronic coupling but not observed energy transfer rates. J Phys Chem A 120:1479–1487.
    OpenUrlCrossRefPubMed
  28. ↵
    1. Laible PD, et al.
    (2003) Quinone reduction via secondary B-branch electron transfer in mutant bacterial reaction centers. Biochemistry 42:1718–1730.
    OpenUrlCrossRefPubMed
  29. ↵
    1. Robles SJ,
    2. Breton J,
    3. Youvan DC
    (1990) Partial symmetrization of the photosynthetic reaction center. Science 248:1402–1405.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Breton J,
    2. Martin JL,
    3. Lambry JC,
    4. Robles SJ,
    5. Youvan DC
    (1990) Ground state and femtosecond transient absorption spectroscopy of a mutant of Rhodobacter capsulatus which lacks the initial electron acceptor bacteriopheophytin. Structure and Function of Bacterial Photosynthetic Reaction Centers, ed Michel-Beyerly ME (Springer, New York), pp 293–302.
  31. ↵
    1. Chuang JI,
    2. Boxer SG,
    3. Holten D,
    4. Kirmaier C
    (2006) High yield of M-side electron transfer in mutants of Rhodobacter capsulatus reaction centers lacking the L-side bacteriopheophytin. Biochemistry 45:3845–3851.
    OpenUrlCrossRefPubMed
  32. ↵
    1. Chuang JI,
    2. Boxer SG,
    3. Holten D,
    4. Kirmaier C
    (2008) Temperature dependence of electron transfer to the M-side bacteriopheophytin in Rhodobacter capsulatus reaction centers. J Phys Chem B 112:5487–5499.
    OpenUrlPubMed
  33. ↵
    1. Holzwarth AR,
    2. Müller MG
    (1996) Energetics and kinetics of radical pairs in reaction centers from Rhodobacter sphaeroides. A femtosecond transient absorption study. Biochemistry 35:11820–11831.
    OpenUrlCrossRefPubMed
  34. ↵
    1. van Stokkum IHM,
    2. Beekman LMP,
    3. Jones MR,
    4. van Brederode ME,
    5. van Grondelle R
    (1997) Primary electron transfer kinetics in membrane-bound Rhodobacter sphaeroides reaction centers: A global and target analysis. Biochemistry 36:11360–11368.
    OpenUrlCrossRefPubMed
  35. ↵
    1. Haffa ALM, et al.
    (2002) The dependence of the initial electron-transfer rate on driving force in Rhodobacter sphaeroides reaction centers. J Phys Chem B 106:7376–7384.
    OpenUrlCrossRef
  36. ↵
    1. van Brederode ME,
    2. van Mourik F,
    3. van Stokkum IHM,
    4. Jones MR,
    5. van Grondelle R
    (1999) Multiple pathways for ultrafast transduction of light energy in the photosynthetic reaction center of Rhodobacter sphaeroides. Proc Natl Acad Sci USA 96:2054–2059.
    OpenUrlAbstract/FREE Full Text
  37. ↵
    1. Holzapfel W, et al.
    (1990) Initial electron-transfer in the reaction center from Rhodobacter sphaeroides. Proc Natl Acad Sci USA 87:5168–5172.
    OpenUrlAbstract/FREE Full Text
  38. ↵
    1. Van Brederode ME,
    2. Jones MR,
    3. Van Mourik F,
    4. Van Stokkum IHM,
    5. Van Grondelle R
    (1997) A new pathway for transmembrane electron transfer in photosynthetic reaction centers of Rhodobacter sphaeroides not involving the excited special pair. Biochemistry 36:6855–6861.
    OpenUrlCrossRefPubMed
  39. ↵
    1. van Stokkum IHM,
    2. Larsen DS,
    3. van Grondelle R
    (2004) Global and target analysis of time-resolved spectra. Biochim Biophys Acta 1657:82–104.
    OpenUrlCrossRefPubMed
  40. ↵
    1. Romero E,
    2. van Stokkum IHM,
    3. Novoderezhkin VI,
    4. Dekker JP,
    5. van Grondelle R
    (2010) Two different charge separation pathways in photosystem II. Biochemistry 49:4300–4307.
    OpenUrlCrossRefPubMed
  41. ↵
    1. Milota F, et al.
    (2013) Vibronic and vibrational coherences in two-dimensional electronic spectra of supramolecular J-aggregates. J Phys Chem A 117:6007–6014.
    OpenUrlPubMed
  42. ↵
    1. Thyrhaug E,
    2. Žídek K,
    3. Dostál J,
    4. Bína D,
    5. Zigmantas D
    (2016) Exciton structure and energy transfer in the Fenna-Matthews-Olson complex. J Phys Chem Lett 7:1653–1660.
    OpenUrl
  43. ↵
    1. Dostál J,
    2. Benešová B,
    3. Brixner T
    (2016) Two-dimensional electronic spectroscopy can fully characterize the population transfer in molecular systems. J Chem Phys 145:124312.
    OpenUrl
  44. ↵
    1. Jordanides XJ,
    2. Scholes GD,
    3. Fleming GR
    (2001) The mechanism of energy transfer in the bacterial photosynthetic reaction center. J Phys Chem B 105:1652–1669.
    OpenUrlCrossRef
  45. ↵
    1. Haberle T, et al.
    (1996) Ultrafast electron and excitation energy transfer in modified photosynthetic reaction centers from Rhodobacter sphaeroides. The Reaction Center of Photosynthetic Bacteria, ed Michel-Beyerle ME (Springer, Berlin), pp 239–254.
  46. ↵
    1. Faries KM, et al.
    (2017) Manipulating the energetics and rates of electron transfer in Rhodobacter capsulatus reaction centers with asymmetric pigment content. J Phys Chem B 121:6989–7004.
    OpenUrl
  47. ↵
    1. Vos MH, et al.
    (1992) Femtosecond spectral evolution of the excited state of bacterial reaction centers at 10 K. Proc Natl Acad Sci USA 89:613–617.
    OpenUrlAbstract/FREE Full Text
  48. ↵
    1. Lin S,
    2. Taguchi AKW,
    3. Woodbury NW
    (1996) Excitation wavelength dependence of energy transfer and charge separation in reaction centers from Rhodobacter sphaeroides: Evidence for adiabatic electron transfer. J Phys Chem 100:17067–17078.
    OpenUrlCrossRef
  49. ↵
    1. Mancal T,
    2. Valkunas L,
    3. Fleming GR
    (2006) Theory of exciton-charge transfer state coupled systems. Chem Phys Lett 432:301–305.
    OpenUrl
  50. ↵
    1. Breton J,
    2. Vermeglio A
    (1988) The Photosynthetic Bacterial Reaction Center, NATO ASI Series (Plenum Press, New York), Vol 149, pp 59–69.
  51. ↵
    1. Meech SR,
    2. Hoff AJ,
    3. Wiersma DA
    (1986) Role of charge-transfer states in bacterial photosynthesis. Proc Natl Acad Sci USA 83:9464–9468.
    OpenUrlAbstract/FREE Full Text
  52. ↵
    1. Hartwich G,
    2. Scheer H,
    3. Aust V,
    4. Angerhofer A
    (1995) Absorption and ADMR studies on bacterial photosynthetic reaction centers with modified pigments. Biochim Biophys Acta Bioenerg 1230:97–113.
    OpenUrl
  53. ↵
    1. Rancova O,
    2. Jankowiak R,
    3. Kell A,
    4. Jassas M,
    5. Abramavicius D
    (2016) Band structure of the Rhodobacter sphaeroides photosynthetic reaction center from low-temperature absorption and hole-burned spectra. J Phys Chem B 120:5601–5616.
    OpenUrl
  54. ↵
    1. Arnett DC,
    2. Moser CC,
    3. Dutton PL,
    4. Scherer NF
    (1999) The first events in photosynthesis: Electronic coupling and energy transfer dynamics in the photosynthetic reaction center from Rhodobacter sphaeroides. J Phys Chem B 103:2014–2032.
    OpenUrlCrossRef
  55. ↵
    1. Kirmaier C,
    2. Holten D,
    3. Parson WW
    (1985) Picosecond-photodichroism studies of the transient states in Rhodopseudomonas sphaeroides reaction centers at 5 K. Effects of electron transfer on the six bacteriochlorin pigments. Biochim Biophys Acta Bioenerg 810:49–61.
    OpenUrl
  56. ↵
    1. Steffen MA,
    2. Lao K,
    3. Boxer SG
    (1994) Dielectric asymmetry in the photosynthetic reaction center. Science 264:810–816.
    OpenUrlAbstract/FREE Full Text
  57. ↵
    1. Vos MH,
    2. Rischel C,
    3. Jones MR,
    4. Martin JL
    (2000) Electrochromic detection of a coherent component in the formation of the charge pair P(+)H(L)(-) in bacterial reaction centers. Biochemistry 39:8353–8361.
    OpenUrl
  58. ↵
    1. Guo Z,
    2. Lin S,
    3. Woodbury NW
    (2013) Utilizing the dynamic Stark shift as a probe for dielectric relaxation in photosynthetic reaction centers during charge separation. J Phys Chem B 117:11383–11390.
    OpenUrl
  59. ↵
    1. Fuller FD,
    2. Wilcox DE,
    3. Ogilvie JP
    (2014) Pulse shaping based two-dimensional electronic spectroscopy in a background free geometry. Opt Express 22:1018–1027.
    OpenUrlCrossRefPubMed
View Abstract
Back to top
Article Alerts
Email Article
Citation Tools
Request Permissions
Primary processes in the bacterial reaction center probed by two-dimensional electronic spectroscopy
Andrew Niedringhaus, Veronica R. Policht, Riley Sechrist, Arkaprabha Konar, Philip D. Laible, David F. Bocian, Dewey Holten, Christine Kirmaier, Jennifer P. Ogilvie
Proceedings of the National Academy of Sciences Apr 2018, 115 (14) 3563-3568; DOI: 10.1073/pnas.1721927115
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo
Proceedings of the National Academy of Sciences: 115 (14)
Table of Contents

Submit

Article Classifications

Jump to section

You May Also be Interested in

Thus far, researchers have mostly just monitored the body’s microbiome. Now they’re starting to modify it to treat disease. Image credit: Science Source/Paul Gunning.
News Feature: Editing the microbiome
Thus far, researchers have mostly just monitored the body’s microbiome. Now they’re starting to modify it to treat disease.
Image credit: Science Source/Paul Gunning.
Greenhouse gas emissions in the first decades of the 21st century are likely to affect global sea level for centuries and limiting long-term sea level rise might require stringent near-term emission reductions. Image courtesy of Lucy Reading-Ikkanda.
Sea-level rise and Paris Agreement pledges
Greenhouse gas emissions in the first decades of the 21st century are likely to affect global sea level for centuries, and limiting long-term sea-level rise might require stringent near-term emission reductions.
Image courtesy of Lucy Reading-Ikkanda (artist).
A single formulation combining two gene therapies treated an array of age-related disorders in mice. Image courtesy of Pixabay /Sipa.
Multidisease gene therapy in mice
A single formulation combining two gene therapies treated an array of age-related disorders in mice.
Image courtesy of Pixabay/Sipa.

Similar Articles