Calculations of bacteriochlorophyll g primary donors in photosynthetic

Mark A. Thompson*4 and Jack Fajer*4. Molecular Science Research Center, Pacific Northwest Laboratory, Richland, Washington99352, and. Department of ...
0 downloads 0 Views 409KB Size
J . Phys. Chem. 1992,96, 2933-2935

2933

Calculations of Bacteriochlorophyll g Primary Donors in Photosynthetic Heliobacteria. How To Shift the Energy of a Phototrap by 2000 cm-' Mark A. Thompson*.+and Jack Fajer*** Molecular Science Research Center, Pacific Northwest Laboratory, Richland, Washington 99352, and Department of Applied Science, Brookhaven National Laboratory, Upton, New York I I973 (Received: December 18, 1991; In Final Form: February 6, 1992)

Heliobacteria are a recently uncovered class of photosynthetic bacteria comprised of a novel chromophore, bacteriochlorophyll (BChl) g. Only the substitution of a vinyl group for an acetyl group on ring I distinguishes the macrocycle of BChl g from that of the more common BChl b. The different substituents impart small differences of 30 nm or -500 cm-I in the Qy transitions of the chromophores in vitro but result in 2000-cm-' differences in the energies of the primary donors in reaction centers containing BChls b (960 nm) and BChls g (800 nm). INDO/s calculations are presented that consider whether this large spectral shift reflects a different mode of aggregation or architecture for the primary donor in Heliobacteria or whether the observed difference can be explained in terms of the dimers or special pairs found in organisms comprised of BChls b or a. Calculations based on the crystallographic coordinates of the BChls b in Rhodopseudomonas viridis with the acetyl groups replaced by vinyls yield good agreement with the observed Qy energies of BChl g monomers in vitro. Surprisingly little modification of the Rps. viridis dimer structure is required to model the optical features of Heliobacteria: for example, dimers of BChl g with an interplanar separation of 3.5 A and vinyl groups oriented -30' out of the porphyrin planes satisfactorily account for the differences between BChls g and b in vivo. The calculationsare also extended to predict the spectral properties of an as-yet undiscovered organism comprised of BChls "h", hypotheticalvinyl-substituted analoguesof BChls a. The calculations may also offer some guidelines for the considerable effort now devoted to chlorin-based artificial photosynthetic models. Introduction of acetyl functions or other polar substituents that can conjugate with the chlorin A system clearly provide simple synthetic avenues to modulate the optical properties of porphinoid monomers; effects that are further enhanced by dimerization, as evidenced by the large differences in the BChls g and b in vivo.

Introduction Bacteriochlorophylls g (BChl g ) are recently discovered chromophores that harvest light and transfer electrons in the photosynthetic Heliobacteria Heliobacterium chlorum, H . gestii, H. farciculum, and Heliobacillus mobilis.'J The molecular structure of BChl g is similar to that of a more common photosynthetic chromophore, BChl b, except for the substitution of a vinyl group for the acetyl function found on ring I of BChl b.3" (Acetyl groups also occur in the ubiquitous BChl a; see Figure 1.) Optical spectra of BChls b and g do not differ significantly in v i t r ~ : ~the - ~lowenergy Qy absorption maxima are found a t 791 and 762 nm in acetone, for b and g, re~pectively.~In vivo, however, reaction centers (RCs) of Rhodopseudomonas viridis, that contain BChls b, exhibit low-energy maxima at 960 nm6 whereas the first R C red bands of H. chlorum and Hb. mobilis,1.2comprised of BChls g, are found at 798-800 nm, a difference of 2000 cm-' in the energies of the phototraps or primary electron donors. Crystallographic results for reaction centers of Rps. viridis' and Rhodobacter ~phaeroides*-~ (comprised of BChls a ) have established that the primary donors in these organisms consist of dimeric or "special pairs" of BChls. Various theoretical treatments, based on the coordinates of the special pairs, readily rationalize the optical properties of the phototraps.I*l3 In particular, the shift of BChls b from 791 nm in vitro to 960 nm in vivo is easily explained in terms of dimer f0rmati0n.l~ The small spectral differences between BChls g in vitro and in vivo suggest, at first glance, that the phototraps of Heliobacteria simply consist of a monomeric BChl g,I4 unlike all other photosynthetic bacteria characterized to date. However, EPR results for triplets and oxidized primary donors of Heliobacteria2 imply the presence of dimers. We present here INDO/s calculation^'^^^^ that consider possible configurations of the BChl g phototrap derived from the dimeric structure of the BChl b dimer in Rps. viridis. Surprisingly little modification of the Rps. viridis dimer structure is required to model the optical features of Heliobacteria: dimers of BChl g with an interplanar separation of 3.5 A and vinyl groups oriented -30° out of the porphyrin planes satisfactorily account for the Pacific Northwest Laboratory.

* Brookhaven National Laboratory.

differences between BChls g and b in vivo. The calculations are also extended to predict the spectral properties of an as-yet undiscovered organism comprised of BChls "h",the hypothetical vinyl-substituted analogues of BChls a (Figure 1).

Methods The calculations were performed with the INDO/s method introduced by Zerner and co-workers.15 The method consists of a ground-state self-consistent-field calculation followed by monoexcited configuration interaction (CI). Excitations were generated from an active space comprised of 14 (28) HOMOS and 14 (28) LUMOS for the monomers (dimers). Calculated state and transition dipole moments retained all one-center terms but omitted two-center contributions. The calculations were carried (1) Gest, H.; Favinger, J. L. Arch. Microbiol. 1983, 136, 1 1 . (2) For comprehensive recent reports see: Trost, J. T.; Blankenship, R. E. Biochemistry 1989, 28, 9898. Nitschke, W.; Setif, P.; Liebl, U.; Feiler, U.; Rutherford, A . W. Biochemistry 1990, 29, 11079. (3) Brockmann, H.; Lipinski, A . Arch. Microbiol. 1983, 136, 17. (4) Michalski, T. J.; Hunt, J. E.; Bowman, M. K.; Smith, U.; Bardeen, K.; Gest, H.; Norris, J. R.; Katz, J. J. Proc. Narl. Acad. Sci. U.S.A. 1987, 84, 2570. ( 5 ) Hoff, A.; Amesz, J. In Chlorophylls; Scheer, H., Ed.;CRC Press: Boca Raton, FL, 1991, p 723. (6) Davis, M. S.; Forman, A.; Hanson, L. K.; Thornber, J. P.; Fajer, J. J. Phys. Chem. 1979,83, 3325. (7) Deisenhofer, J.; Michel, H.Science 1989, 245, 1463. (8) Yeates, T. 0.; Komiya, H.; Chirino, A.; Reese, D. C.; Allen, J. P.; Feher, G. Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 7993. (9) Chang, C. H.; El-Kabbani, 0.;Tiede, D.; Norris, J. R.; Schiffer, M. Biochemistry 1991, 30, 5352. (10) Warshel, A.; Parson, W. W. J . Am. Chem. SOC.1987, 109, 6143. Parson, W. W.; Warshel, A. J . Am. Chem. SOC.1987, 109, 6152. ( 1 1 ) Friesner, R. A.; Won, Y . Biochim. Biophys. Acta 1989, 977, 99. (12) Scherer, P. 0. J.; Fischer,S. F. Chem. Phys. 1989,131, 115;J. Phys. Chem. 1989, 93, 1633. (13) Thompson, M. A.; Zerner, M. C.; Fajer, J. J . Phys. Chem. 1991,95, 5693. (14) Fajer, J. Barkigia, K. M.; Fujita, E.; Goff, D. A.; Hanson, L. K. Head, J . D.; Horning, T.; Smith, K. M.; Zerner, M. C. In Antennas and Reacrion Centers of Phorosynthetic Bacteria; Michel-Beyerle, M. E., Ed.; Springer.~ Verlag: Berlin, 1985; p 324. ( 1 5 ) Ridley, J.; Zerner, M. Theor. Chim. Acta (Berlin) 1973, 32, 1 1 1 ; 1976, 42, 223. Zerner, M.; Loew, G.; Kirchner, R.; Mueller-Westerhoff, U. J . Am. Chem. SOC.1980, 102, 589.

0022-365419212096-2933%03.00/0 0 1992 American Chemical Society

2934

The Journal of Physical Chemistry, Vol. 96, No. 7 , 1992 n

Letters TABLE I: Calculated Q y Transitions dihedral anale. dea

species BChl BChl BChl BChl

b (L) b (M) b dimer

g (L) BChl g (M) BChl g dimer BChl g dimer BChl g dimer BChl g dimer BChl g dimer BChl g dimer BChl g dimer BChl g dimer BChl g dimer histidines

-1 747 -1 567 -174, -156 ((3) -174 -156 120, 120 150, 150 180, 180 -174, -156 (0)

P P

B

+

150, 150 150, 150

interplanar seuaration. A

-3.3' 3.3 3.3 3.3 3.3 3.5 3.6 3.7 3.5 3.5

Qy, cm'l 12 484" 12789" 1078613 12922 13471 12072 11931 11702 11505 12150 12351 12497 12495 12676

Qy2," cm-I

1234513 13376 13176 12899 12726 13090 13197 13273 13435 13706

Upper Qy exciton component in the dimers.

a

"h"

Figure 1. Structural formulas of BChls a, b, g, and "h". The esterifying alcohols R are C20 chains in BChls a and b and C15 in BChl g.4 They have been truncated to methyl groups in the calculations.

out with the program A R G U S ' ~ on - ~ ~the Cray Y-MP at Florida State University, Tallahassee. For BChls b, the calculations utilized the coordinates of the heavy atoms of Rps. viridis at 2.3-A resolution7 which were obtained from the Brookhaven Protein Data Bank. Hydrogens were placed at standard distances. For BChl g calculations, the ring I acetyl groups in the Rps. viridis structure were replaced by vinyl groups at several orientations which were checked to avoid unacceptable steric interactions. For BChl "h" calculations, the coordinates used for BChl g were further modified by replacing the ethylidene group on ring I1 by an ethyl group and a proton.

Results and Discussion Previous INDO/s calculations of monomeric bacteriochlorins agree well with experiment, particularly for the energies of the Qy bands that reflect the frontier orbitals of the chromophores. For example, calculations for methyl bacteriopheophorbide a i 7 and Zn tetraphenylbacteriochlorin,18based on high-resolution single-crystal structures, match the Qy transitions observed in solution within 350 cm-l. Calculations of BChls b, based on the coordinates of the two monomer subunits of the special pair in Rps. viridis, yield Qy values of 12484 (801 nm) and 12789 cm-I (782 nm).I3 The differences are due to variations in the conformations of the porphyrin cores and to different orientations of the acetyl g r o ~ p s . ~The J ~ calculated Qy transitions of the two BChl b conformers bracket the value observed in acetone soluti~n,~ 12 642 cm-I or 79 1 nm, and the average of the calculated values, 12637 cm-I, equals the experimental number. For BChl g calculations, the BChl b coordinates in the Rps. uiridis special pair were modified by replacing the ring I acetyl groups with vinyl groups and maintaining the same dihedral angles relative to the porphyrin plane. The Qy values obtained for the BChls g are 12922 (774 nm) and 13471 cm-I (742 nm) for an average value of 13 197 cm-I or 758 nm, in good agreement with the experi~ noted mental value in acetone of 13 123 cm-l (762 r ~ m ) .As before, the calculated Qy transitions are sensitive to the ruffling of the macrocycles and to the orientation of the acetyl (or vinyl) (16) Thompson, M. A.; Zerner, M. C. J . Am. Chem. SOC.1991,113, 8210. (17) Gudowska-Nowak, E.; Newton, M. D.; Fajer, J. J . Phys. Chem. 1990, 94, 5795. (18) Barkigia, K. M.; Miura, M.; Thompson, M. A,; Fajer, J. Inorg. Chem. 1991, 30, 2233.

groups.I7 Rolling the latter into the plane of the porphyrin (0 = 180' or Oo)I9 will maximize conjugation and result in a red shift whereas rolling the groups out of plane (e = 90') will minimize their effect and cause blue shifts. (These theoretically predicted trends are supported by results with synthetic m0de1s.I~) The major conclusion here, however, is that the Rps. viridis coordinates yield reasonable values for the BChl b and g monomers. Crystallographic results of RCs comprised of BChls b or a reveal a common molecular architecture with rings I of the monomers overlapping with interplanar separations of 3.3 A in Rps. viridis7 and 3.5 A in Rb. ~phaeroides.~~~ INDO/s calculations of the BChl b special pair of Rps. viridis, based solely on the crystallographic coordinates, afford good agreement with the experimental Qy transitions, linear dichroism and positions of the exciton bands.I3 Similar calculations for the BChls a in Rb. sphaeroides also match experimental re~u1ts.l~We thus consider here whether a simple substitution of vinyl groups for acetyl groups in the BChl b special pair is sufficient to account for the phototrap shift from 960 nm in Rps. viridis to 798 nm in H. chlorum. The fundamental question is thus whether a dimer is still a viable possibility in organisms that utilize a different window in the solar spectrum or whether Heliobacteria have abandoned the use of the dimeric phototraps found in all photosynthetic bacteria comprised of BChls a or b. Initial calculations of BChl g dimers started with the architecture of the BChl b special pair in Rps. uiridis with the acetyl groups replaced with vinyl groups. The latter were oriented as in Rps. viridis,' Le., the dihedral anglesIg are -174' and -156' for the L and M monomers,20 respectively, and they point into the space between the two monomer subunits. This is labeled the /3 configuration, for convenience. In addition, other orientations of the vinyl groups into and out of the porphyrin planes were considered. These calculations were prompted by crystallographic,21-23NMR,24and EPR25,26data for chlorophylls, pheophytins, and protoporphyrins in vitro and in vivo which clearly show that the vinyl groups of the chlorins and porphyrins adopt a wide range of orientations relative to the macrocycle planes. (19) The dihedral angle is defined by the plane of the vinyl group relative to the pyrrole ring; 180' places the vinyl group in the ring plane and 90' perpendicular to that plane. (20) L and M refer to the protein subunits in the RC with which each BChl is associated.' (21) Chow, H . C.; Serlin, R.; Strouse, C. E. J . Am. Chem. Soc. 1975.97, 7230. Serlin, R.; Chow, H . C.; Strouse, C. E. Ibid. 1975, 97, 7237. (22) Fischer, M. S . ; Templeton, D. H.; Zalkin, A,; Calvin, M. J . Am. Chem. SOC.1972, 94, 3613. (23) Takano, T. J. Mol. Bioi. 1977, 110, 569. (24) Lee, K. B.; Jun, E.; LaMar, G. N.; Rezzano, I . N . ; Pandey, R. K.; Smith, K. M.;Walker, F. A.; Buttlaire, D. H . J . Am. Chem. SOC.1 9 9 1 , 113, 3576. (25) Forman, A,; Renner, M. W.; Fujita, E.; Barkigia, K. M.; Evans, M. C. W.; Smith, K. M.; Fajer, J. I s r . J . Chem. 1989, 29, 57. (26) Barkigia, K. M.; Chantranupong, L.; Kehres, L. A,; Smith, K. M.; Fajer, J . In Photochemical Energy Conversion;Norris, J. R., Meisel, D., Eds.; Elsevier: New York, 1989; p 21 1 .

2935

J. Phys. Chem. 1992, 96, 2935-2937 Steric constraints limit rotation of the vinyl groups into the dimer. However, the groups will swivel out of the dimer to dihedral angles of 120° without crowding. Table I presents the results of the BChl g dimer calculations with the 6 configuration and with the vinyl groups at 180°, 150°, and 120O. As observed for monomers, moving the vinylic substituents further out of conjugation leads to larger blue shifts of the Qy bands. Note that simply introducing vinyl groups and changing their orientations yields BChl g dimers with Qy bands that are blue shifted 700-1300 cm-' relative to the BChl b dimer which is calculated at 10786 cm-' and observed at 10417 cm-'. Previous calculations of special pairs predictedi0J'J3,26that increasing the interplanar separation in BChl dimers would cause blue shifts, and the different absorption of Rb. sphaeroides relative to that of Rps. viridis has been attributed to the crystallographic observed interplanar differences of -0.2 A in the two organi s m ~ . Calculations ~ ~ , ~ ~ for the BChl g dimer as a function of interplanar separation are also listed in Table I. The results illustrate the blue shifts that can be induced by widening the spacin within the dimers. Starting with the 6 configuration a t -3.3 , increasing the separations by 0.4 A to 3.7 yields a blue shift of 1000 cm-l within the BChl g series to 12 497 cm-l or 800 nm. Clearly, quite reasonable combinations of interplanar separations and vinyl rotations can yield BChl g dimers with the desired spectral properties. A separation of 3.5 A and vinyl orientations of 150° also result in a Qy transition at 800 nm. Furthermore, inclusion of the histidines found as Mg axial ligands in Rps. viridisl and Rb. s p h a e r ~ i d e scauses ~ * ~ a further blue shift to 789 nm. In other words, quite "pedestrian" BChl g dimers, with only small variations on the Rps. viridis configuration and its ligands, are entirely compatible with the observed optical (and EPR) properties of Heliobacteria. Recall that the concept of special pairs was originally proposed to explain the properties of P700, the primary electron donor of photosystem I in green plants and algae that contain chlorophyll

x-

a.27 In vitro, photoisomerization of the ethylidene group of BChl g leads directly to Chl a f ~ r m a t i o n . Intriguingly, ~ extensive

analogies between Heliobacteria RCs and photosystem I components are observed.2 (Preliminary results indicate that Chl a dimers with configurations similar to those considered here for the BChls g would have Qy transitions of -680-690 nm2*.) Last, we consider the possibility that there may exist an as-yet undiscovered (or engineered) organism with a hypothetical BChl "h" in which the acetyl group of BChl a has been changed to a vinyl as in the BChl b to g transformation (Figure 1). Calculations for such a BChl "h" dimer with the same configuration as described above for BChl g predict a Qy transition at 12515 cm-l, a value comparable to that of the BChls g. In a related context, we note that considerable efforts are now devoted to chlorin-based artificial photosynthetic models.26~30~3i Introduction of acetyl functions or other polar substituents that can conjugate with the chlorin r system clearly provide simple synthetic avenues to modulate the optical properties of porphinoid monomers; effects that are further enhanced by dimerization, as evidenced by the large differences in the BChls g and b in vivo and by BChl dimers and aggregates in v i t r ~ . ~ ~ . ~ '

Acknowledgment. We thank Drs. L. Chantranupong and L. K. Hanson for preliminary calculations on the BChls g. This work was supported by the Division of Chemical Sciences, US.Department of Energy, under Contracts No. DE-AC02-76CH00016 at BNL and DE-AC06-76RLO1830 at PNL. (27) Norris, J. R.; Uphaus, R. A.; Crespi, H. L.; Katz, J. J. Proc. Narl. Acad. Sci. U.S.A. 1971, 68, 625. (28) Thompson, M. A.; Hanson, L. K.; Fajer, J. Unpublished results. (29) The compound has been made synthetically: Struck, A,; Cmiel, E.; Katheder, I.; Scheer, H. FEBS Lett. 1990, 268, 180. (30) Wasielewski, M. R. Photoinduced Electron Transfer; Fox, M. A., Chanon, M., Eds.; Elsevier: Amsterdam, 1988; Part A, Chapter 1.4: Photochem. Photobiol. 1988, 47, 923. (31) Rosenbach-Belkin,V.;Fisher, J. R. E.; Scherz, A. J . Am. Chem. SOC. 1991, 113, 616.

Intermolecular Triplet Excimers of Aromatic Molecules with Permanent Dipole Moments: Carbazole, Dibenzofuran, and Dibenzothiophene Jianjian Cai and E. C. Lim*.' Department of Chemistry, The University of Akron, Akron, Ohio 44325-3601 (Received: December 23, 1991; In Final Form: January 31, 1992)

We report here the formation of intermolecular triplet excimers, in fluid solution of carbazole,dibenzofuran,and dibenzothiophene, which could very well be a general phenomenon for aromatic molecules with permanent dipole moments.

In contrast to the extensive documentation that exists on singlet excimers of aromatic hydrocarbons, the formation of the triplet excimer from fluid solutions of aromatic compounds has been substantiated only for a handful of diarylalkanes.' The dearth of information on the triplet excimer, as well as the existing controversy concerning the triplet excimer of n a ~ h t h a l e n e ,is~ ? ~ largely due to the small binding energy of the triplet excimer (relative to that of the corresponding singlet excimer) which is stabilized by van der Waals (electrostatic, induction, and dispersion forces) as opposed to the large exciton and charge resonance stabilization of the singlet excimer. For aromatic molecules, with no permanent dipole moment, the leading contribution to the classical electrostatic interactions comes from weak quadrupolequadrupole interactions. As a consequence, the formation of intermolecular triplet excimers of aromatic hydro-

'Holder of the Goodyear Chair in Chemistry at The University of Akron. 0022-3654/92/2096-2935$03.00/0

SCHEME I

1cc

t

hv" Phos.

carbons is expected to be much less efficient than that involving aromatic molecules with permanent dipole moments. 0 1992 American Chemical Society