Dichotomous Disorder versus Excitonic Splitting of the B800 Band of

Therefore, we uphold our earlier interpretation, first reported in [A. Kell et al., J. Phys. Chem. B 2017, 121, 9999], that the two B800 sub-bands are...
1 downloads 0 Views 1MB Size
Download PDF
Letter pubs.acs.org/JPCL

Cite This: J. Phys. Chem. Lett. 2018, 9, 4125−4129

Dichotomous Disorder versus Excitonic Splitting of the B800 Band of Allochromatium vinosum Adam Kell,†,§ Anton Khmelnitskiy,† Mahboobe Jassas,† and Ryszard Jankowiak*,†,‡ †

Department of Chemistry and ‡Department of Physics, Kansas State University, Manhattan, Kansas 66506, United States

Downloaded via UNIV OF READING on July 24, 2018 at 23:38:57 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: The LH2 antenna complex of the purple bacterium Allochromatium vinosum has a distinct double peak structure of the 800 nm band (B800). Several hypotheses were proposed to explain its origin. Recent 77 K two-dimensional electronic spectroscopy data suggested that excitonic coupling of dimerized bacteriochlorophylls (BChls) within the B800 ring is largely responsible for the B800 split [M. Schröter et al., J. Phys. Chem. Lett. 2018, 9, 1340]. Here we argue that the excitonic interactions between BChls in the B800 ring, though present, are weak and cannot explain the B800 band split. This conclusion is based on holeburning data and modeling studies using an exciton model with dichotomous protein conformation disorder. Therefore, we uphold our earlier interpretation, first reported by Kell et al. [J. Phys. Chem. B 2017, 121, 9999], that the two B800 sub-bands are due to different siteenergies (most likely due to weakly and strongly hydrogen-bonded B800 BChls).

S

excitation; however, under these conditions light-induced photoconversion of B800R → B800B occurs (vide infra), which could mask a simultaneous bleach within the B800B band.1,2 However, it is more likely that, for a coupled dimer, the weak coupling strength (compared to reorganization energy) leads to dynamic localization of the transitions, and no excitonic response is present in the HB spectra. Our previous work provided clear evidence that the two B800 subbands (B800R and B800B) are largely due to different BChl site energies and these bands were assigned to weakly and strongly hydrogen-bonded BChl chromophores.1 Below, we examine the nature of the B800 band split in more detail and show that, while a weak excitonic coupling between pigments contributing to the B800R and B800B sub-bands is present, distinct siteenergies of weakly and strongly hydrogen-bonded BChls contributing to the B800 band are largely responsible for the observed B800 band split. Temperature-Dependent Light-Induced Photoconversion of B800 BChls. First, it must be emphasized that the photoconversion discussed in refs 1 and 2 does not lead to persistent holes at 77 K, i.e., the temperature at which 2DES frequency maps for Alc. vinosum were generated.5 This is illustrated in Figure 1, which shows three delta absorption spectra (b, c, and e) using laser excitation near 805 nm. Spectra a/b, c, and d/e were obtained at 77, 40, and 5 K, respectively. Spectra c and e are the HB spectra revealing B800R → B800B photoconversion at both 40 and 5 K, respectively; the bleach, however, is absent at 77 K (see curve b), independent of fluence and excitation energy (data not shown). This behavior is consistent with previous data showing that by 70 K the holes are annealed and

everal papers published recently have shown that the B800−850 LH2 antenna from the photosynthetic purple sulfur bacterium Allochromatium (Alc.) vinosum exhibits an unusual spectral splitting of the B800 absorption band at cryogenic temperatures.1−4 Two hypotheses concerning the origin of this splitting have been proposed: (i) two distinct B800 bacteriochlorophyll (BChl) site energies are involved, arising from different binding pockets of the B800 BChls;1,2 and (ii) the B800 molecules are arranged in alternating structural motifs leading to excitonically coupled dimers which, in addition to B800−B850 excitonic interactions, gives rise to a splitting of B800.3 The former was suggested based on holeburning (HB) spectroscopy and modeling studies, while the latter was originally suggested from single-complex spectroscopy and modeling studies. Recently Schröter et al.,5 claimed to provide unambiguous evidence that the B800 excitonic interactions (i.e., dimerization of B800 BChls) shapes the splitting of the B800 band. This conclusion was reached using two-dimensional electronic spectroscopy (2DES) performed at 77 K.5 Although weak excitonic coupling between B800 BChls occurs for LH2 complexes of all bacteria6−10 (see our proposed structure described in ref 2), it is unlikely that dimerization of B800 BChls is largely responsible for the ∼200 cm−1 peak splitting of the B800 absorption band in Alc. vinosum. Excitonic coupling between the B800 pigments was clearly demonstrated by the weak B800 cross peaks in the 77 K 2DES data,5 consistent with the dimer model proposed by Löhner et al., who argued that such interactions can only weakly contribute to the B800 band splitting.3 This agrees with earlier 77 K transient absorption experiments11 which did not clearly observe a simultaneous bleaching of both B800 sub-bands upon selective excitation of B800R . Recent 5 K HB experiments obtained for Alc. vinosum also did not reveal any obvious bleach of both B800 sub-bands upon selective © 2018 American Chemical Society

Received: May 21, 2018 Accepted: July 9, 2018 Published: July 9, 2018 4125

DOI: 10.1021/acs.jpclett.8b01584 J. Phys. Chem. Lett. 2018, 9, 4125−4129

Letter

The Journal of Physical Chemistry Letters

Figure 1. Spectra a/b and d/e are 77 and 5 K absorption/delta absorption spectra of B800 bands of Alc. vinosum. Photoconversion is observable at 40 K (see curve c) but is absent at 77 K (see curve b). Curves c and e reflect B800R →B800B photoconversion via a photochemical HB process that is only persistent at T < 60 K. Curves b/c and e were obtained with λB of 805.0 and 804.2 nm, and hole depth of 23.8% and 13.3%, respectively. Curves at each temperature are offset for clarity.

Figure 2. Four spectra (from top to bottom; T = 5 K) are the persistent, shallow (low fluence) holes burned at excitation wavelengths of 813.0, 811.0, 805.0, and 803.0 nm, respectively. The lowest spectrum is the 5 K absorption spectrum of the B800 band.

the photoconversion process does not produce persistent holes above this temperature.1 Absorption spectra (curves a and d) are shown here for an easy comparison. In the absence of photoconversion, the 2DES spectra5 could clearly reveal weak cross peaks showing that B800 pigments are excitonically coupled (and the excitation of one transition could indeed “bleach” the other); however, the intensity of the cross peak is very weak and a phrase like “strong coupling” (as concluded in ref 5, where the authors argued that ground-state bleaching cross-peaks occur only for strongly coupled transitions) does not mean much without a metric for what “strong” means. Therefore, it is argued below, in agreement with previously published results,1,2 that B800R and B800B are split primarily due to different site energies, most likely due to different binding pockets in the two different apoproteins, while excitonic coupling only weakly contributes to the B800 band splitting. Selective Light-Induced Photoconversion. Figure 2 (four top spectra) shows constant fluence (f = 1 J cm−2) resonant holes burned into B800R at excitation wavelengths of 813.0, 811.0, 805.0, and 803.0 nm, respectively (see solid arrows). Note that at these conditions, HB spectra, as expected, do not reveal any simultaneous bleaching of both B800 sub-bands upon selective excitation of B800R. That is, after excitation, a resonant bleach of B800R is accompanied only by the corresponding photoproduct due to the B800R → B800B phototransformation (see weak positive bands indicated by asterisks); for details, see refs 1 and 2. Modeling of Split B800. It is shown below that the main conclusion reached by Schröter et al.,5 i.e., that excitonic coupling is largely responsible for the observed B800 band split, can be easily refuted via excitonic calculation using two models: Model I, an excitonically coupled dimer with varying coupling matrix elements, V1,2, between neighboring BChls1,2; and Model II, in which the entire LH2 complex using a Ph. molischianum-based structure published recently for Alc. vinosum assuming C12 symmetry is used in our modeling studies.1,2 Both models are based on a non-Markovian Redfield theory with Monte Carlo disorder averaging of site energies.

Additionally, Model II includes the recently described dichotomous model for static disorder.2,12 The dichotomous disorder introduces additional inhomogeneous broadening due to conformational changes of the protein environment (see Supporting Information for more details). The data below focus only on B800 and light-induced photoconversion. Figure 3 shows calculations obtained for Model I. Frame A shows the experimental absorption (curve a) and three calculated absorption spectra for the B800 region assuming different coupling matrix elements (VB800−B800 = −26 cm−1 as calculated via the TrESp methodology)13 for a homodimer, i.e., the energy difference between BChls contributing to the B800R and B800B sub-bands (Δ) is zero. Spectra b, c, and d were obtained for VB800−B800 of −26, −50, and −80 cm−1, respectively. It is obvious that even a homodimer (i.e., Δ = 0) with relatively strong coupling does not contribute significantly to the band splitting. Frame B of Figure 3 shows similar B800 absorption spectra (curves a−d) obtained for a heterodimer with the same coupling matrix elements (as in frame A) but different site energies (here Δ = 200 cm−1); note that such a photoproduct was observed experimentally.1,2 The corresponding calculated photochemical holes assuming a photoproduct blue-shifted by 200 cm−1 are shown as curves a′−d′. Although the absorption spectrum with VB800−B800 = −26 cm−1 and Δ = 200 cm−1 nicely describe the B800 sub-bands, the calculated HB spectra poorly describe the experimentally observed phototransformation. This is most likely due to simplifications made above. Therefore, the modeling in Figure 4 considers the entire B800−850 complex (see inset in Figure 3A) and a dichotomous model of static disorder (details given in the Supporting Information). Here, only the B800 absorption region is shown, and the parameters are analogous to those used for Model I in Figure 3B. That is, the dichotomous site energy shift (ΔE) is 200 cm−1 and both absorption and photochemical HB spectra are calculated for V = −26, −50, and −80 cm−1. In this case, the average site energies for BChls contributing to each band B800R and B800B (E1 and E2, respectively) are E800 − 100 and E800 + 100 cm−1, respectively 4126

DOI: 10.1021/acs.jpclett.8b01584 J. Phys. Chem. Lett. 2018, 9, 4125−4129

Letter

The Journal of Physical Chemistry Letters

Figure 3. Frame A: Curve a is the experimental absorption spectrum obtained at 5 K. Spectra b, c, and d are calculated absorption spectra for a homodimer (Δ = 0 cm−1) obtained with site energy/coupling matrix element (E/VB800−B800) of 12480/−26, 12510/−50, and 12530/−80 cm−1, respectively. Frame B: Calculations for a heterodimer with Δ = 200 cm−1. In this case site energies contributing to the B800R and B800B bands are indicated by E1 and E2, respectively. The following E1/E2/VB800−B800 parameters were used to calculate spectra b, c, and d: (curve b) 12445/12645/ −26 cm−1; (curve c) 12450/12650/−50 cm−1; (curve d) 12470/12670/−80 cm−1.

Figure 4. Experimental (black solid lines) and calculated absorption and HB spectra for different sets of parameters (see text for details) using an exciton model with dichotomous protein conformation disorder.

Figure 5. Curves b and c are the HB spectra obtained for λB = 791.5 nm obtained for fluence of 7 and 35 J/cm2, respectively. Curve a is the 5 K B800 absorption band shown for easy comparison of B800R → B800B photoconversion.

(see frame A). Frame B shows the corresponding HB spectra, i.e., the result of B800R → B800B phototransformation. Again, only a weak excitonic coupling of −26 cm−1 describes the experimental data very well (see red dashed lines compared with the solid black experimental spectra). Only this coupling properly describes the isosbestic point, supporting a two-site model, likely with strongly and weakly hydrogen-bonded B800 BChls. B800B → B800R Energy Transfer Time. In ref 1 it was shown that the homogeneous line width of the zero-phonon line (i.e., one-half the line width of the zero-phonon hole) burned at 785 nm (12739 cm−1), i.e., burned into the high-energy wing B800B, corresponded to a 0.9 ps B800B → B800R excitation energy transfer (EET) time. This time is mainly attributed to B800B → B800R EET as the bleach of the B850 band was negligibly small. Note that this excitation bleached the entire B800R sub-band via uncorrelated EET to B800R and

subsequent phototransformation into B800B. In Figure 5 we show HB spectra obtained at λB = 791.5 nm (the maximum of B800B) for two fluences of 7 and 35 J cm−2. Interestingly, in this case, the homogeneous line width is narrow and corresponds to somewhat slower EET time. Although the zero-phonon holes in Figure 5 are noisy, the estimated EET time of about 1.7 ± 0.3 ps appears to be slower than that obtained for 785.0 nm excitation. These values are in contrast with the 3.9 ps estimated for B800B → B800R and 1.0 ps obtained for B800B → B850.5 The latter two values were estimated via fitting of the 2DES data with a kinetic scheme as done in ref 5. However, our data cannot distinguish between EET times from B800B → B800R and B800 → B850, thus our directly measured EET times (via HB spectroscopy) are in reasonable agreement with the analysis provided in ref 5, although the 3.9 ps time constant for the B800B → B800R 4127

DOI: 10.1021/acs.jpclett.8b01584 J. Phys. Chem. Lett. 2018, 9, 4125−4129

The Journal of Physical Chemistry Letters

■

transfer seems to be too slow, especially since the authors in ref 5 claim that the B800 split is excitonic in origin. Interestingly, two different time constants revealed by HB spectroscopy within B800B could suggest variable B800B downward relaxation. Preliminary calculations of the dichotomous model suggest that the lifetimes of the excitonic states contributing to B800R and B800B are on the order of 1−10 ps at both 5 and 77 K. We uphold our earlier interpretation, first proposed in ref 1, that the two sub-bands (B800R and B800B) are mostly due to weakly and strongly hydrogen-bonded B800 BChl chromophores, which can undergo conformational changes caused by proton dynamics and occurring under illumination.1,2 That is, we conclude that a two-site model with strongly and weakly hydrogen-bonded B800 BChls is largely responsible for the observed B800 band split,1 and white light illumination and/or selective laser excitations into B800R or B800B lead to B800R → B800B photoconversion. Selective excitation into B800B leads to uncorrelated EET to B800R and subsequent photoconversion of B800R → B800B. However, the bleach and subsequent photoproduct absorbance due to B800R → B800B photoconversion is not observed in HB spectra at temperatures above 60 K, independent of fluence and/or laser excitation (data not shown). Note that the cross peak in the 77 K 2D spectra in ref 5 occurs near 12620 and 12420 cm−1 for the excitation and probe, respectively. Persistent HB data cannot observe any simultaneous bleach of both B800R and B800B bands, indicating that the excitonic coupling is weaker than the reorganization energy and the electronic transitions are not delocalized over the entire B800 band. Of course, direct excitation into B800R leads to EET (∼2 ps) to B850, while selective excitation into B800B leads to transfer to both B850 and B800R bands.1,2,5,11 While similar EET rates were observed by HB1,2 and pump−probe experiments,11 the B800B → B800R EET time of 3.9 ps obtained via analysis of recent 2DES data is, in our opinion, too slow if the two bands, as suggested in ref 5, are split because of strong excitonic interactions. The latter value, however, could be overestimated, as the cross peaks are distorted by excited state absorption.5 Our modeling studies, however, using a simple dimer model and/or the dichotomous model excludes the possibility that excitonic interactions shape the B800 band, as suggested in ref 5. In summary, our data do not support the main conclusion of Schröter et al.5 who, while admitting that the large B800 peak splitting may be in part due to a heterodimeric structure, suggested that such splitting is largely of excitonic origin. That is, the phototransformation of B800R leads to a ∼ 200 cm−1 average blue-shift of transition energies, i.e., B800R changes into B800B. It appears that BChls contributing to B800R and B800B differ in site energies, most likely in the position of the proton in the BChl carbonyl−protein hydrogen bond, i.e., proton dynamics along the hydrogen bond is likely the major mechanism of the observed photoconversion.

■

Letter

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]; Phone: 785-532-6785. ORCID

Ryszard Jankowiak: 0000-0003-3302-9232 Present Address §

Department of Chemistry and Chemical Biology, University of California, Merced, California 95343, United States Author Contributions

R.J. designed the research and wrote the manuscript. Experiments and modeling were performed by A.Kh., A.K. and M.J. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Award Number DE-SC-0006678 (to R.J.). The authors acknowledge Drs. Kirsty Hacking and Richard J. Cogdell from the Institute of Molecular Cell and Systems Biology, College of Medical, Veterinary and Life Sciences, University of Glasgow, Glasgow, Scotland for providing us with the LH2 samples. R.J. acknowledges useful discussions with Dr. Tonu Reinot. E-mail exchange regarding the 2DES experiments with Dr. Donatas Zigmantas is also highly appreciated.

■

REFERENCES

(1) Kell, A.; Jassas, M.; Hacking, K.; Cogdell, R. J.; Jankowiak, R. On Light-Induced Photoconversion of B800 Bacteriochlorophylls in the LH2 Antenna of the Purple Sulfur Bacterium Allochromatium vinosum. J. Phys. Chem. B 2017, 121, 9999−10006. (2) Kell, A.; Jassas, M.; Acharya, K.; Hacking, K.; Cogdell, R. J.; Jankowiak, R. Conformational Complexity in the LH2 Antenna of the Purple Sulfur Bacterium Allochromatium vinosum Revealed by HoleBurning Spectroscopy. J. Phys. Chem. A 2017, 121, 4435−4446. (3) Löhner, A.; Carey, A.-M.; Hacking, K.; Picken, N.; Kelly, S.; Cogdell, R.; Köhler, J. The Origin of the Split B800 Absorption Peak in the LH2 Complexes from Allochromatium vinosum. Photosynth. Res. 2015, 123, 23−31. (4) Carey, A.-M.; Hacking, K.; Picken, N.; Honkanen, S.; Kelly, S.; Niedzwiedzki, D. M.; Blankenship, R. E.; Shimizu, Y.; Wang-Otomo, Z.-Y.; Cogdell, R. J. Characterisation of the LH2 Spectral Variants Produced by the Photosynthetic Purple Sulphur Bacterium Allochromatium vinosum. Biochim. Biophys. Acta, Bioenerg. 2014, 1837, 1849−1860. (5) Schröter, M.; Alcocer, M. J. P.; Cogdell, R. J.; Kühn, O.; Zigmantas, D. Origin of the Two Bands in the B800 Ring and Their Involvement in the Energy Transfer Network of Allochromatium vinosum. J. Phys. Chem. Lett. 2018, 9, 1340−1345. (6) Sundström, V.; Pullerits, T.; van Grondelle, R. Photosynthetic Light-Harvesting: Reconciling Dynamics and Structure of Purple Bacterial LH2 Reveals Function of Photosynthetic Unit. J. Phys. Chem. B 1999, 103, 2327−2346. (7) Cheng, Y. C.; Silbey, R. J. Coherence in the B800 Ring of Purple Bacteria LH2. Phys. Rev. Lett. 2006, 96, 028103. (8) Novoderezhkin, V.; Wendling, M.; van Grondelle, R. Intra- and Interband Transfers in the B800−B850 Antenna of Rhodospirillum molischianum: Redfield Theory Modeling of Polarized Pump−Probe Kinetics. J. Phys. Chem. B 2003, 107, 11534−11548. (9) Krueger, B. P.; Scholes, G. D.; Fleming, G. R. Calculation of Couplings and Energy-Transfer Pathways between the Pigments of LH2 by the ab Initio Transition Density Cube Method. J. Phys. Chem. B 1998, 102, 5378−5386.

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpclett.8b01584. Experimental methods and Monte Carlo disorder averaging and dichotomous conformation disorder in Alc. vinosum (PDF) 4128

DOI: 10.1021/acs.jpclett.8b01584 J. Phys. Chem. Lett. 2018, 9, 4125−4129

Letter

The Journal of Physical Chemistry Letters (10) Zigmantas, D.; Read, E. L.; Mancal, T.; Brixner, T.; Gardiner, A. T.; Cogdell, R. J.; Fleming, G. R. Two-Dimensional Electronic Spectroscopy of the B800-B820 Light-Harvesting Complex. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 12672−12677. (11) Niedzwiedzki, D. M.; Bina, D.; Picken, N.; Honkanen, S.; Blankenship, R. E.; Holten, D.; Cogdell, R. J. Spectroscopic Studies of Two Spectral Variants of Light-Harvesting Complex 2 (LH2) from the Photosynthetic Purple Sulfur Bacterium Allochromatium vinosum. Biochim. Biophys. Acta, Bioenerg. 2012, 1817, 1576−1587. (12) Rancova, O.; Sulskus, J.; Abramavicius, D. Insight into the Structure of Photosynthetic LH2 Aggregate from Spectroscopy Simulations. J. Phys. Chem. B 2012, 116, 7803−7814. (13) Renger, T.; Müh, F. Theory of Excitonic Couplings in Dielectric Media: Foundation of Poisson-TrEsp Method and Application to Photosystem I Trimers. Photosynth. Res. 2012, 111, 47−52.

4129

DOI: 10.1021/acs.jpclett.8b01584 J. Phys. Chem. Lett. 2018, 9, 4125−4129