UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Structures of photosynthetic reaction centers with alternative cofactors Hardjasa, Amelia 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


24-ubc_2015_september_hardjasa_amelia.pdf [ 7.39MB ]
JSON: 24-1.0166379.json
JSON-LD: 24-1.0166379-ld.json
RDF/XML (Pretty): 24-1.0166379-rdf.xml
RDF/JSON: 24-1.0166379-rdf.json
Turtle: 24-1.0166379-turtle.txt
N-Triples: 24-1.0166379-rdf-ntriples.txt
Original Record: 24-1.0166379-source.json
Full Text

Full Text

Structures of PhotosyntheticReaction Centers WithAlternative CofactorsbyAmelia HardjasaB.Sc., The University of British Columbia, 2010B.A., The University of British Columbia, 2010A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinThe Faculty of Graduate and Postdoctorate Studies(Microbiology and Immunology)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)July 2015c© Amelia Hardjasa 2015AbstractRhodobacter sphaeroides is a model organism for the study of bacterial photosynthesis.The R. sphaeroides photosynthetic reaction centre (RC) is the primary site of electrontransfer, which is mediated by the photosynthetic pigments bacteriochlorophyll a (BChl)and bacteriopheophytin (BPhe). The substitution of key amino acid residues can changethe type of cofactors present in the RC. In particular, studies have shown that when theleucine residue in position 214 of the M subunit [(M)L214] is converted into a histidine,the BPhe normally present in the neighbouring position (HA) is replaced with a BChl.This study investigated the hypothesis that steric exclusion by the coordinatingresidue causes dechelation of the central magnesium ion in BChl, producing BPhe. Crys-tal structures of RCs where (M)L214 is substituted for glycine and alanine were deter-mined, which demonstrated that the presence of BPhe in the HA pocket is unchangeddespite decreasing the size of the residue in position (M)214. A crystal structure of anRC where (M)L214 is substituted for asparagine was also determined and showed thatthe replacement of BPhe with BChl at HA occurs if residue (M)214 includes an amidemoiety.In the R. sphaeroides ∆bchd strain, which lacks the ability to make BChl, it is believedthat the RC cofactor sites are populated exclusively with zinc-bacteriochlorophyll (Zn-BChl). The crystal structures of this Zn-BChl containing RC (Zn-RC) and a Zn-RC withthe (M)L214H substitution (Zn-β-RC) were solved for the first time. These structuresconfirmed the presence of Zn-BChl in every cofactor position and the tetracoordinationof the HA Zn-BChl in the Zn-β-RC, as well as revealing that the occupancy of the HBcofactor was much lower than that of all other cofactors.iiPrefaceThe majority of the work presented in this thesis is drawn from published literature.This work was made possible by extensive collaboration with Dr. J Thomas Beatty’slaboratory at the University of British Columbia. Since the collaboration provides crucialsupport for the findings, portions from the publications have been incorporated into thisthesis. Below are detailed the contributions of my collaborators.Saer, R.G., Hardjasa, A., Rosell, F.I., Mauk, A.G., Murphy, M.E.P., and Beatty, J.T.(2013) The role of Rhodobacter sphaeroides photosynthetic reaction center residue M214in the composition, absorbance properties, and conformations of HA and BA cofactors.Biochemistry, 52(13), 2206-2217. I performed the crystallization, data collection, andstructure solution and analysis, and wrote the first draft of the relevant sections of themanuscript. R. Saer performed the purification, spectroscopy, and photosynthetic growthstudies, and drafted all other sections of the manuscript. Drs. M. Murphy and J. Beattyedited the manuscript. The text and figures in the Materials and Methods, Results, andDiscussion sections of this manuscript were adapted and included in the correspondingsections of this thesis.Saer, R.G., Pan, J., Hardjasa, A., Lin, S., Rosell, F., Mauk, A.G., Woodbury, N.W.,Murphy, M.E.P., and Beatty, J.T. (2014) Structural and kinetic properties of Rhodobac-ter sphaeroides photosynthetic reaction centers containing exclusively Zn-coordinated bac-teriochlorophyll as bacteriochlorin cofactors. BBA Bioenergetics, 1837(3), 366-374. Iperformed the crystallization, data collection, and some of the structure solution. M.Murphy performed the remainder of the structure solution and analysis, and wrote theiiiPrefacetext in the corresponding sections of the manuscript. A. Chan created Figure 2 of themanuscript. R. Saer performed the purification and spectroscopy, and drafted all othersections of the manuscript. Drs. M. Murphy and J. Beatty edited the manuscript. Onlythe sections of this manuscript related to data I produced were adapted and included inthe corresponding sections of this thesis.The work in this thesis was approved by the University of British Columbia BiosafetyCommittee under the project title “Heme and Iron Uptake in Pathogenic Bacteria”,Certificate #B130096.ivTable of ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiPreface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiiTable of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vList of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiList of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiiAcknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Bacterial photosynthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Characteristics of R. sphaeroides . . . . . . . . . . . . . . . . . . . . . . 21.2.1 Photosynthesis in vivo . . . . . . . . . . . . . . . . . . . . . . . . 21.2.2 Cofactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31.3 The R. sphaeroides RC . . . . . . . . . . . . . . . . . . . . . . . . . . . 61.3.1 History of purification and X-ray crystallography . . . . . . . . . 61.3.2 Catalytic cycle of the RC . . . . . . . . . . . . . . . . . . . . . . 101.3.3 Spectroscopic characteristics of the RC . . . . . . . . . . . . . . . 111.4 Reaction center variants . . . . . . . . . . . . . . . . . . . . . . . . . . . 131.4.1 (M)L214H, the β mutant . . . . . . . . . . . . . . . . . . . . . . 141.4.2 The Zn-RC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151.5 Objective of the present study . . . . . . . . . . . . . . . . . . . . . . . 16vTable of Contents2 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182.1 RC purification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182.2 Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192.3 X-ray crystallography . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202.4 Data collection, processing, and structure solution . . . . . . . . . . . . 213 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233.1 Structural changes in variant RCs with a substitution in (M)L214 . . . . 233.1.1 Low-temperature spectroscopy . . . . . . . . . . . . . . . . . . . 233.1.2 X-ray crystal structures . . . . . . . . . . . . . . . . . . . . . . . 253.2 Structural characterization of Zn-BChl-containing-RCs . . . . . . . . . . 294 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354.1 Structural changes in variant RCs with a substitution in (M)L214 . . . . 354.1.1 Aliphatic non-polar residues (G,A) . . . . . . . . . . . . . . . . . 354.1.2 Amide residues (N) . . . . . . . . . . . . . . . . . . . . . . . . . 404.2 Structural characterization of Zn-BChl-containing-RCs . . . . . . . . . . 414.3 Conclusions and future directions . . . . . . . . . . . . . . . . . . . . . . 44Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46viList of Tables3.1 X-Ray Data Collection and Refinement Statistics for (M)L214G, (M)L214A,and (M)L214N RCs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263.2 X-Ray Data Collection and Refinement Statistics for the Zn and Zn-β RCs. 303.3 Anomalous map statistics for metals in the Zn-RC and Zn-β-RC X-raydiffraction datasets. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31viiList of Figures1.1 Structural formulae of BChl and BPhe. . . . . . . . . . . . . . . . . . . . 21.2 Biosynthetic pathway of BChl and chlorophyll a from protoporphyrin-IX. 41.3 Different types of esterifying alcohols found attached to BChl and BPheas tails in photosynthetic bacteria. . . . . . . . . . . . . . . . . . . . . . 51.4 Three-dimensional structure of the R. sphaeroides RC. . . . . . . . . . . 81.5 Arrangement of cofactors in the wild type R. sphaeroides reaction center. 91.6 Low-temperature (∼11 K) steady state absorption spectrum of wild typeRC. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121.7 Movement of the BA BChl macrocycle resulting from mutation of thetyrosine at (M)210 to a tryptophan. . . . . . . . . . . . . . . . . . . . . . 143.1 Low-temperature steady state absorption spectra of the nonpolar (M)214mutant series. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243.2 Low-temperature steady state absorption spectra of the (M)L214N and(M)L214H mutants. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253.3 Omit difference electron density map of the (M)L214G RC, wild-type RC,and the (M)L214A RC. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273.4 Omit difference electron density map of the (M)L214N RC. . . . . . . . . 333.5 Stick models and electron density of the HAA cofactor and axial (M)214residues from the crystal structures of the Zn-RC and Zn-β-RC. . . . . . 34viiiAcknowledgementsI acknowledge the financial support for this work provided by a CIHR Canada Grad-uate Scholarship. I also thank the Stanford Synchrotron Research Laboratory and theCanadian Light Source for synchrotron data collection.I am deeply indebted to my supervisor, Dr. Michael Murphy, for his immense andunflagging support, guidance, and understanding. This thesis would not have been pos-sible without his continual and generous efforts on my behalf. Under his supervision Ihave had unprecedented opportunities to grow both as a scientist and human being, allof which have been invaluable.I am also very grateful to Dr. Tom Beatty for his advice and support over severalyears. It was his guidance that first set me on the path to becoming a scientist. I wouldlike to thank my primary collaborator in the Beatty Lab, Dr. Rafael Saer, for his scientificdirection and contribution to the work published in this thesis.I would also like to thank all of the other members of the Murphy Lab and Beatty Labwho have supported me on my way: Paul Jaschke, Angele Arrieta, Anson Chan, JasonGrigg, Stephanie Pfaffen, Catherine Gaudin, Daniel Jun, Jeanette Beatty, my studentsJaimie and Fabian, and many others, all of whom have been not only trusted friends andsupports but also exemplary role models.I am immeasurably grateful to Darlene Birkenhead for her support and encourage-ment, without which this work would not have been completed.Finally, I am deeply thankful to all of my friends and family for their forbearanceand goodwill over the course of my graduate studies. Especially, I would like to thankIvan Zverev for his continual kindness, reassurance, and patience through all the ups anddowns of this chapter of my life.ixChapter 1Introduction1.1 Bacterial photosynthesisSunlight is one of the most freely available and bountiful energy sources on earth.Photosynthesis gives organisms the means whereby this abundant energy can be accessedfor growth, and is thus responsible for a vast portion of primary production, being themain metabolic mode for all plants, algae, cyanobacteria, and many other organisms.One well known group of prokaryotic photosynthesizers is the purple photosyntheticbacteria. These bacteria only perform photosynthesis under anoxic conditions, especiallythose with an abundance of hydrogen sulfide [46]. They are thus prolific in the depthsof aquatic environments including lakes, ponds, and lagoons. The oxygenic phototrophsthat live above the purple bacteria absorb most of the short-wavelength, high-energyincoming light with their chlorophylls [16]. Purple bacteria consequently have developedspecialized pigment-protein complexes which preferentially absorb light at 750 nm andabove, resulting in the purple colour from which they derive their name [64].The subject of this study, Rhodobacter sphaeroides, is a well-studied member of thepurple photosynthetic bacteria, as it possesses many features that make it ideal for ex-perimentation. It exhibits extraordinary metabolic diversity, including phototrophy andheterotrophy, as well as being relatively amenable to laboratory culture and genetic ma-nipulation [31]. Although light energy can be scanty at the depths where R. sphaeroidesis found, the species is nonetheless is able to thrive, thanks in large part to the highquantum efficiency of its photosynthetic reaction center (RC) and light harvesting com-plexes; nearly 100 % of absorbed photons are converted into usable energy in the form11.2. Characteristics of R. sphaeroidesof a proton motive force [73].1.2 Characteristics of R. sphaeroides1.2.1 Photosynthesis in vivoUpon exposure to light and anaerobic or microaerophilic conditions, invaginations formin the R. sphaeroides cellular membrane known as chromatophores or intracytoplasmicmembranes (ICMs), which for the most part bud incompletely from the cellular mem-brane, although it is possible for them to exist as free-floating vesicles [25]. These ICMsare enriched with photosynthetic apparatus proteins, and their presence has the effect ofincreasing membrane surface area available for light capture without requiring a drasticjump in cell volume [18]. In the ICM, the photosynthetic reaction center complex canbe found surrounded by light harvesting antenna complexes (light harvesting complex 1(LH1) and 2 (LH2)), which assist in transferring captured photons to the reaction center,but are not a requirement for photosynthetic growth [37, 68].Figure 1.1: Structural formulae of BChl (A) and BPhe (B), with R representing thecofactor tail (usually phytyl). Reprinted from [64] with permission.21.2. Characteristics of R. sphaeroides1.2.2 CofactorsThe role of the reaction center is the conversion of trapped light energy to a trans-membrane proton gradient via electron transfer through the protein. This is mediated bya set of porphyrin-derived cofactors, sometimes referred to as “chlorins” or “pigments”,terms which will be used interchangeably throughout this text. The two cofactor typesfound natively in the R. sphaeroides reaction center are bacteriochlorophyll a (BChl)and bacteriopheophytin (BPhe) (Figure 1.1).Biosynthesis. BChl is derived from protoporphyrin-IX through the well-elucidatedbacteriochlorophyll biosynthetic pathway (Figure 1.2). Firstly, the enzymes encoded bythe bch operon genes bchHDI catalyze the chelation of a central magnesium ion by thepyrrole ring. After several subsequent reductions of the porphyrin, the final step is theaddition of a hydrophobic “tail” by means of an esterifying alcohol [9]. BPhe is identicalto BChl structurally, except for the lack of the central ion. It is believed that BPheis derived from BChl by the removal of the Mg2+ [2], although it has not yet beendetermined whether a specific “magnesium dechelatase” protein exists to perform thisrole, or whether some other means is responsible for removal of the central ion, such asthe environment of the cofactor in the reaction center itself.Hydrophobic tails. Little is known about the role of the hydrophobic tails in variouschlorins. In R. sphaeroides, the esterifying alcohol is phytol, resulting in a phytyl tail.Although most BChls and many chlorophylls (Chl) incorporate a phytyl tail, this is notuniversal; for instance, BChls in Rhodospirillum rubrum have tails esterified by the lesssaturated geranylgeraniol (Figure 1.3) [40].In 1999, Addlesee and Hunter identified the protein responsible for reduction of theesterifying alcohol prior to its attachment to the main BChl macrocycle [1]. It is believedthat a mutation in this gene is responsible for the variation in the R. rubrum tail [2]. A31.2. Characteristics of R. sphaeroidesFigure 1.2: Biosynthetic pathway of BChl and chlorophyll a from protoporphyrin-IX.Reprinted from [9] with permission.41.2. Characteristics of R. sphaeroidestransposon mutant of R. sphaeroides lacking this reductase was able to grow photosyn-thetically, although it had much lower levels of light harvesting complexes and an overalllower growth rate. The reaction center was not specifically investigated, however, so it isunknown whether electron transfer in the RC was impaired or whether the lack of LH2was solely responsible for the reduction in growth [10]. Wen et al. also investigated theeffects of removing the reductase in the green bacterium Chloroflexus aurianticus [72].Their study focused on the FMO complex, a trimeric complex containing 7 BChls permonomer that mediates electron transfer from the light harvesting complex to the C. au-rianticus RC. Similarly, they found that overall amounts of FMO protein had decreased,but they also noted that there was more pigment heterogeneity in the FMO proteinscontaining the altered BChls.Figure 1.3: Different types of esterifying alcohols found attached to BChl and BPhe astails in photosynthetic bacteria. R. sphaeroides cofactors have exclusively phytyl tails,but geranylgeranyl tails are found in, e.g., Rhodospirillum rubrum. Reprinted from [64]with permission.51.3. The R. sphaeroides RC1.3 The R. sphaeroides RC1.3.1 History of purification and X-ray crystallographyPurification. One large benefit of R. sphaeroides ’ diverse metabolism is its ability togrow in the dark heterotrophically. Otherwise, modifications that resulted in impairedphotosynthetic growth might be outcompeted by the wild type bacterium when grownphotosynthetically. In combination with the development of a simple method for obtain-ing large quantities of relatively pure RC protein by Goldsmith and Boxer in 1996 [26],this has led to the dominance of R. sphaeroides as the organism of choice for studyingbacterial photosynthetic reaction centers. Hence, the following structural details will fo-cus on the R. sphaeroides RC, although differences from reaction centers of other studiedpurple phototrophic bacteria will be noted when relevant.Crystallization. The bacterial photosynthetic reaction center has the distinction ofbeing the first membrane protein to have its structure determined by X-ray crystal-lography in 1985, for which the authors later received the Nobel Prize in Chemistry[17]. Although the first such structure determined was of the Blastochloris viridis (for-merly Rhodopseudomonas viridis) RC, subsequent study focused on the R. sphaeroidesRC thanks to its aforementioned greater utility in laboratory research. Since then, it hasbeen studied extensively, with almost a hundred structures deposited in the Protein DataBank (PDB). Despite this, only few high-resolution wild type structures are available,with only one structure determined to greater than 2.0 A˚ resolution (PDB entry 2J8C,1.8 A˚).The first crystallization of the R. sphaeroides RC by Allen was in somewhat compli-cated conditions derived from those used for R. viridis. The protein solution contained61.3. The R. sphaeroides RCRC at a concentration of 5.5 mg/ml, 0.1 % LDAO, 2.5 % 1,2,3-heptanetriol, 3 % triethy-lammonium phosphate, 1.2 M (NH4)2SO4, 10 mM Tris HCl (pH 8.0), 0.6 mM EDTA,and 0.1 % NaN3, over a reservoir of 2.1 M (NH4)2SO4, and crystals took 2-9 weeks to ap-pear [4]. The dark-coloured crystals grew in monoclinic (P2 ) and orthorhombic (C222 )space groups and were photosynthetically active, as demonstrated by spectroscopy. Thisfirst structure was solved to a resolution of 3.5 A˚. As purification techniques have im-proved, the complexity of required crystallization conditions has decreased correspond-ingly, typically requiring only a potassium phosphate buffer (1.0-1.8 M), the amphiphile1,2,3-heptanetriol (1.5 - 5 % weight by volume, depending on detergent concentration),and a solubilizing detergent. The RC has also served as a fertile testing ground fortechniques specific to membrane protein crystallization, being among the first proteinscrystallized by lipidic cubic and lipidic sponge phase methods [39]. These techniquesuse the lipid monoolein to form porous lipidic structures containing the protein, whichfacilitates crystallization, possibly by enhancing the long-term stability of the molecules[34].Organization of the RC. The RC comprises three subunits titled L, M, and H 1.4.The H subunit is a globular protein primarily located on the cytoplasmic side, with asingle transmembrane α-helix near the amino terminus. In contrast, the L and M subunitseach contain five transmembrane helices, which are the major transmembrane portion ofthe RC. The L and M subunits form an approximately symmetrical complex (the LMcomplex) around a rotational axis, with abundant contact between the two proteins. Onthe periplasmic side, there are five helical segments in each subunit lying approximatelyperpendicular to the transmembrane helices; in the cytoplasm, the LM complex contactsthe H subunit. The transmembrane region of the LM complex forms a cage-like structurearound eight electronically active cofactors and an Fe2+ ion.The cofactors comprise six chlorins and two ubiquinone-10s, divided along two nearly71.3. The R. sphaeroides RCFigure 1.4: Three-dimensional structure of the R. sphaeroides RC, from PDB entry 2J8C.The globular H subunit, which is situated in the cytoplasm, is shown in blue. The Land M subunits, coloured green and yellow respectively, form the major transmembraneportion of the protein complex with five transmembrane helices each. These helicessurround the electronically active cofactors, shown in grey.symmetrical branches termed the A and B branches. In the native reaction center,electron transfer occurs only through the A branch.The first cofactor site holds two closely associated BChls, known as the special pair.Below the special pair sit two accessory BChls, one on each branch, which are in turnadjacent to one BPhe each. Finally, a ubiquinone can be found below each BPhe, withthe non-heme iron between them. These four paired sets, and their positions, are referredto as the P, B, H, and Q sites respectively. A subscript A or B denotes in which branchthe cofactor is situated (see Figure 1.5).Structural characterization. The first RC structures confirmed many predictions onthe nature of the complex, such as its situation in the membrane and the special paircomprising two abutting BChls [5, 14, 52, 78]. Structures of the RC with cytochromec2 bound have shown its presence on the periplasmic side, congruent with its role in81.3. The R. sphaeroides RCFigure 1.5: Arrangement of cofactors in the wild type R. sphaeroides reaction center. Thecofactors are approximately symmetrical, with two branches labeled A and B. Electrontransfer only occurs down branch A in the native RC. Bacteriochlorins are found atthree sites: P, the special pair BChl; B, the accessory BChl; and H, the BPhe. SiteQ is occupied by quinones and a structural non-heme iron is located between the twobranches.regenerating the special pair to neutral state [6]. Another surprising discovery that wasnot suggested by spectroscopic data was the presence of a bilateral cofactor pathway,symmetrical to a r.m.s.d of 0.7 A˚ in the special pair; this was shortly followed by theobservation that, although both sets of cofactors are present, only one of the branches isactive in electron transfer[5]. The locations of the centers of the chlorin rings were well-conserved as compared to the previously characterized RC of R. viridis [3]. Curiously,although early reports indicated considerable conformational variation in the phytyl tailsof BA and BB, more recent structures have exhibited no such difference [5, 14, 20]. Itis possible that the poorer resolution of these older structures resulted in the mistakenassignation of electron density to a tail instead of a detergent molecule.Lipids and detergents. Purification and crystallization of the reaction center, as withmost membrane proteins, involves the solubilization of the protein in detergent. Through91.3. The R. sphaeroides RCneutron-scattering experiments, Roth et al. examined RC crystals and found that thesolubilizing detergent, either LDAO or BOG, formed elliptical micelles around the hy-drophobic transmembrane regions of the protein, as well as facilitating crystal contactsbetween multiple copies of the RC [59]. The majority of these detergent molecules werenot ordered and thus not visible in the X-ray structures. Although for the most partthese detergent arrangements are similar to what would be expected in a lipid bilayer,any given detergent molecule in the crystal structure may not be representative of thereaction center in vivo.Several studies have noted the presence of native (i.e., not present in the crystalliza-tion or solubilization media) lipid molecules on the surface of the RC, most frequentlycardiolipin (CDL), the glycolipid glucosylgalactosyl diacylglycerol (GGD), and the lipidphosphatidylcholine (PC), although the orientations and positions of these lipids do notalways appear to be consistent between structures [13, 36, 48]. For example, in earlierstructures the density for the three tails of cardiolipin had been modeled with three LDAOmolecules, as the density was often weak and/or discontinuous. However, McAuley et al.showed in 1999 that continuous density could be fit for this lipid; furthermore, additionalsite-directed mutagenesis experiments observed that disrupting the putative binding sitesfor this lipid resulted in decreased RC stability [24, 48]. It is possible that other regionscurrently modeled by detergent molecules suffer similarly from misinterpreted electrondensity and that more such lipids may be found. The use of brominated lipids has shownsome promise in this direction [58].1.3.2 Catalytic cycle of the RCThe kinetics of electron transfer in the RC have been studied intensively over the pastfew decades [49, 55, 56, 73, 76, 77] and are briefly summarized below. Currently, thegenerally understood sequence of events is as follows: The catalytic cycle of the RC is101.3. The R. sphaeroides RCinitiated when the absorption of a photon excites the special pair P, with the excitedstate referred to as P*. Subsequently, P* decays via the transfer of an electron to theneighbouring BA cofactor, giving rise to the P+B−A state, with a time constant estimatedat approximately 3 ps. From this point, it is possible for backwards charge recombinationto occur, returning P+B−A to the neutral, unexcited ground state. However, in the wildtype reaction center the rate of this reaction is almost negligible. Instead, the electronis passed on to HA, forming P+H−A, in less than 1 ps, then to QA in about 200 ps. Q−Athen reduces QB, producing ubisemiquinone. The oxidized P+ meanwhile receives anelectron from a cytochrome c2 to return it to the P state. This transfer pattern occursonce more, with the second reduction of QB producing a quinol, QH2. QH2 leaves theQB site and diffuses through the membrane; meanwhile the QB site is refilled with anoxidized quinone, completing the acceptor quinone cycle. The quinol is oxidized by thecytochrome bc1 complex, which results in the release of two protons into the periplasm.Cytochrome c2 accepts one electron from the cytochrome bc1 complex and can thereforeserve as an electron donor to P+. This coupling of electron flow to proton translocationcreates a proton motive force which can eventually be used by an ATP synthase. Thiscycle is extremely efficient, with virtually every captured photon contributing successfullyto electron transfer.1.3.3 Spectroscopic characteristics of the RCInvestigation of the electronic characteristics of the R. sphaeroides RC can be per-formed using absorption spectroscopy, with greater detail visible at lower temperatures[63, 76]. The conjugated ring systems of the chlorin-derived cofactors in the reactioncenter give rise to two prominent absorption peaks, one along each axis, termed the Qxand Qy transitions. Qx transition peaks can be found in the 500-700 nm region, whileQy transition peaks are found in the 700-900 nm range. In the wild type reaction center,peaks are associated with the cofactors as seen in Figure 1.6 below.111.3. The R. sphaeroides RCFigure 1.6: Low-temperature (∼11 K) steady state absorption spectrum of wild typeRC. Peaks are labeled with contributing cofactor names. The dashed line divides thespectrum into two regions, Qx and Qy, which arise from transitions along the two axesof the cofactor macrocycle. Data provided by R. Saer.Although the relative peak positions remain the same, the precise wavelength of thepeak depends on the transition energy and therefore can change significantly dependingon the temperature at which the spectrum is recorded. Hence, in Figure 1.6, the specialpair Qy peak is centered around 895 nm, but is sometimes referred to in the literature asP865 (e.g. [52]) because of its wavelength at room temperature. Other factors that maycause a change in free energy include a change in the central ion, a change in ion coordi-nation, or alteration in the protein environment of the cofactor. The associated changein energy may result in red or blue shifts of a given cofactor’s Qx and/or Qy peaks, aswell as a change in relative breadth and height [73]. For example, a heterodimeric specialpair comprising a BPhe and a BChl can be produced by mutating (M)H200 to leucine[42]. The resulting absorption spectrum has a severely diminished, almost absent specialpair Qy transition peak, and a decreased peak height at the special pair Qx transitionpeak, as now only one BChl is contributing to these particular transitions (Figure 1,121.4. Reaction center variants[42]). The special pair BPhe appears to instead contribute to the HA,B transition peaks,which exhibit slight increases in height over the corresponding wild type peaks.1.4 Reaction center variantsThe R. sphaeroides reaction center is remarkably resilient to changes in its primary se-quence, maintaining its gross structure even in the presence of multiple point mutations.Several studies have investigated the effect of site-directed mutagenesis of residues sit-uated close to the electronically active cofactors on the positioning and environment ofthese cofactors, as well as subsequent effects on electron transfer [12, 23, 29, 66, 71, 73].For instance, McAuley et al. found that the substitution of a tryptophan for a tyrosineat position (L)210, close to BA, caused a decrease in electron transfer [47]. The crystalstructure of this variant was solved and it was determined that there had been a shiftin the angle of the BA BChl (Figure 1.7). Although the tilt was small (fewer than 3◦),it was nonetheless clearly visible in the electron density maps, highlighting the potentialof structural investigation to detect small but impactful changes in cofactor orientationwith associated effects on electron transfer.A great deal of work has also been done on investigating the reasons for the apparentinactivity of the B branch cofactors. A variant RC which performs electron transferalong the B branch has been created, requiring several point mutations in the L and Msubunits and an adjustment of the cofactor composition in the A branch [70]. From theseand other experiments [54, 76], it can be concluded that the configuration of the proteinscaffold surrounding the reaction center cofactors plays a critical role in fine-tuning thefree energy gaps between adjacent cofactors in the electron transfer pathway; these gapsin turn determine the direction of electron flow, as well as maintain the high efficiencyof energy conversion characteristic of the RC.131.4. Reaction center variantsFigure 1.7: Movement of the BA BChl macrocycle resulting from mutation of the tyrosineat (M)210 to a tryptophan. The cofactor in the mutant RC (light grey) is tilted fromthat of the WT (dark grey) by approximately 3◦.There have been several studies on means by which cofactor composition can be altered.Each chlorin cofactor in the RC has a residue axial to the plane of the macrocycle[11, 15, 27, 35]. In the wild type RC, this residue is always histidine when the cofactoris BChl, the imidazole group of which acts as a fifth ligand for the Mg2+ ion. For BPhe,the adjacent residue is a leucine. Mutating the histidines to a residue unable to ligate themagnesium, including phenylalanine as well as leucine, can result in the substitution ofa BPhe for the BChl [12]. Conversely, mutating the leucine to a liganding residue resultsin the substitution of a BChl for a BPhe. One example of this is of particular interest tothe present study: (M)L214H, the β mutant.1.4.1 (M)L214H, the β mutantIn 1991 Christine Kirmaier et al. [41] found that replacing the M subunit L214 residuewith a histidine resulted in an altered absorption spectrum for the RC, notably theappearance of a Qy peak at 785 nm and a Qx peak increase at 600 nm, accompanied by141.4. Reaction center variantsa decrease at 760 and 545 nm, the regions indicative of BPhe presence. Based on thisand other evidence, later supported by a crystal structure of the variant reaction center[15], Kirmaier concluded that the accessory HA BPhe had been replaced by a BChl. Thiswas denoted as a β BChl, giving rise to the common description of (M)L214H as “the βmutant”. After measuring electron transfer kinetics, it was found that although chargeseparation still occurred, the quantum yield of the reaction was only 60 % of wild typeefficiency. The authors hypothesized that the decrease in the free energy gap between BAand HA led to an increased rate of charge recombination (P+B−A → P) relative to that ofelectron transfer to the quinone (P+B−A → P+Q−A). The mechanism by which a BChl isincorporated into HA rather than a BPhe was not determined.1.4.2 The Zn-RCAnother notable variant reaction center was discovered in 2007 by Jaschke and Beatty[32]. The authors found that a R. sphaeroides mutant lacking the magnesium chelatasegene (∆bchD) had neither BChl nor BPhe cofactors present in the cell. In their place,the RC assembled with a cofactor structurally similar to BChl, but with a central Zn2+ion instead of Mg2+, henceforth referred to as “Zn-BChl”. These Zn-BChls occupy the6 spaces formerly occupied by BChl and BPhe cofactors in this RC, henceforth referredto as the Zn-RC. Surprisingly, although these cofactors are “unnatural” and moreover,the HA,B cofactors are identical to those in the BA,B site, Lin et al. [45] found that therate of electron transfer from P∗ to P+H−A was comparable to that of the WT RC. It wassuggested that this is possible because despite the change in absolute free energy values,the potential difference of the P∗ to P+H−A reaction is sufficiently similar to maintain thesame rate of forward electron transfer. Specifically, it was posited that Zn-BChls at theHA,B sites were tetracoordinated rather than pentacoordinated as in the P and B sites,and therefore more similar in reduction potential to the naturally occurring BPhe thanto BChl.151.5. Objective of the present studyAn analogue of the β mutant was made by mutating L214 to histidine in the ∆bchDbackground, with the goal of providing a fifth ligand for the HA Zn2+, causing it to bepentacoordinated. This will be referred to as the Zn-β-RC. When both RCs are referredto, the term “Zn-BChl-containing-RCs” will be used. Spectroscopic investigation of theZn-β-RC [51] found that the 560 nm band representing the Qx transition for tetra-coordinated H site Zn-Bchls was absent. This was unexpected, as one would expect thateven if the HA Zn2+ was pentacoordinated, the HB Zn2+ would remain tetra-coordinatedand therefore give rise to a small 560 nm Qx peak. The authors suggested several possibleexplanations, including: a) the HB cofactor is in fact absent from both the Zn and Zn-β-RCs, and so the 560 nm band represents the electronic contribution of HA only; b) bothZn-BChls in H sites are pentacoordinated in the Zn-β-RC and tetra-coordinated in theZn-RC.1.5 Objective of the present study1. Substitutions at residue (M)L214. To expand on previous work investigatingthe effects of substitutions at residue (M)L214, a series of mutants at this site was madeand their structures determined by X-ray crystallography. These mutants fell into twocategories: aliphatic non-polar residues (glycine and alanine) and amide residues (as-paragine). The former category was used to investigate whether the size of the residuehad an effect on the dechelation of the central magnesium ion to produce a BPhe at theHA site, as well as any other structural impacts. The latter was to investigate whether anamide oxygen could ligate the magnesium in the same manner as the histidine nitrogenand if this was sufficient to result in the presence of a BChl in the HA pocket.2. Zn-RC and Zn-β-RC. Heretofore, there have been no structures of Zn-BChl-containing RCs. To determine any structural impacts from the substitution of the naturalchlorins with Zn-BChls, crystallization experiments were undertaken with the goal of161.5. Objective of the present studysolving the structures of these two proteins. Additionally, structural investigation coulda) confirm the coordination of the HA Zn-BChl Mg2+ and b) suggest an explanation forthe absence of the HA,B 560 nm Qx band in the Zn-β RC.17Chapter 2Materials and Methods2.1 RC purificationReaction centers were purified using a modified version of the protocol of Goldsmithand Boxer [26]. The cells from 21 L of the R. sphaeroides semiaerobic culture werecollected by centrifugation, and the cell paste was resuspended in 10 mM Tris-HCl (pH8.0), 150 mM NaCl,and 2 mM MgCl to a volume of 175 mL. A few crystals of DNase Iwere added to the suspension before the cells were lysed in a French press at 18,000 psi.Lysed cells were centrifuged to pellet debris and unbroken cells. The supernatant wasthen centrifuged at 66,300 g and 4◦C overnight to pellet chromatophores. Followingultracentrifugation, the supernatant was discarded and the remaining pellet was resus-pended in 10 mM Tris (pH 8.0) and 150 mM NaCl to a volume of 100 mL. Samples(1 mL) of this suspension were placed in 1.7 mL microcentrifuge tubes and brought toroom temperature, and lauryldimethylamine oxide (LDAO, Fluka) was added to eachtube at concentrations ranging from 0.5 to 2.75 % to determine the optimal concentra-tion of detergent for RC solubilization. These samples were rocked in the dark for 30 min,and 900 µL was centrifuged at 107,400 g for 30 minutes at 4◦C. The 875 nm absorbancemaximum of the supernatants was used as a measure of the degree of photosyntheticcomplex solubilization. In these experiments, the optimal concentration of LDAO forchromatophore solubilization varied between 1.25 and 1.75 %.Once the optimal LDAO concentration was found, the remaining chromatophore sus-pension was brought to this detergent concentration and stirred in the dark for 30 minat room temperature, and centrifuged at 117,700 g and 4◦C for 15 minutes. The super-182.2. Spectroscopynatant solution was collected, and imidazole (10 mM) and NaCl (200 mM) were addedto the indicated concentrations before the mixture was loaded onto a Ni-NTA column(Qiagen). The column was washed with 10 mM Tris (pH 8.0), 150 mM NaCl, and 0.1 %LDAO until the maximal absorbance over the range of 500-950 nm was less than 0.01.The RC was eluted from this column with 10 mM Tris (pH 8.0), 150 mM NaCl, 300mM imidazole, and 0.1 % LDAO, then dialyzed against 10 mM Tris (pH 8.0) and 0.1 %LDAO (TL buffer). His-tagged Zn-BChl-containing RCs were isolated in a similar fash-ion, except that Zn-BChl-containing RCs were dialyzed against 10 mM Tris, 0.025 %LDAO (Zn-RC-TL buffer). To obtain sufficient amounts of Zn-BChl-containing RCs, thecell culture volume was scaled up to 72 L because of the low yield of these RCs from∆bchD mutant strains.For crystallization, RCs were purified further by anion exchange chromatography withan AKTA Explorer FPLC system (GE Healthcare) equipped with a SourceQ column andelution with a NaCl gradient (0 to 200 mM), with the majority of RCs eluting at 120 mMNaCl. The eluted RCs were desalted by being passed through a 10DG column (Bio-Rad)and concentrated with an Amicon 30 kDa centrifugal ultrafilter (Millipore). RCs usedfor crystallization experiments were never frozen and crystallization drops were set upimmediately following purification.2.2 SpectroscopyLow-temperature absorption spectroscopy. Cryogenic absorption spectra (0.6 nmbandwidth) were obtained at ∼10–13 K using a Cary 6000 spectrophotometer (Agilent)equipped with a closed-cycle helium cryostat (Omniplex OM-8, ARS Inc.) as describedby Lin et al.[45]. RC samples and Zn-BChl-containing-RC samples, in TL buffer and Zn-RC-TL buffer respectively, were concentrated to an A802 of ∼40 by centrifugation using a30 kDa cutoff Centricon ultrafilter (Millipore). Samples were supplemented with sodium192.3. X-ray crystallographyascorbate (final concentration of 1 mM) and diluted 1:1 with spectroscopic grade glycerolprior to being frozen between two quartz plates separated by an ∼25 µm polycarbonatespacer. Where necessary, a monotonic increase in background absorbance (500 to 730nm) resulting from light scattering was corrected by subtracting a linear function.2.3 X-ray crystallographyFresh solutions of the RC, purified as described in Section 2.1, were used to set up hangingdrop crystallization trials. Instead of mixing a reservoir solution with the protein, a sepa-rate precipitant solution of 1 M potassium phosphate (pH 7.4), 3.5 % 1,2,3-heptanetriol,and 0.1 % LDAO was prepared and mixed with the protein solution on cover slips beforesealing over the reservoir. Various ratios of precipitant and protein solution were foundto result in crystals of widely varying quality. Diffraction quality crystals of (M)L214G(∼0.3 mm × 0.3 mm × 0.3 mm) grew at a ratio of 2 µL of RC solution (7 mg/mL, or 20.6OD802/mL) to 1 µL of precipitant solution at 298 K, protected from light, over a reservoirof 1.52 M potassium phosphate buffer (pH 7.4). Crystals appeared after approximately72 h. Repeated exposure of the crystals to light appeared to diminish diffraction quality,so immediately after the initial exposure to light, crystals were transferred to a solutionof 1 M potassium phosphate buffer (pH 7.4) and 30 % glycerol before being flash-frozen inliquid nitrogen. All other crystals used in this work were produced and frozen in a similarfashion, with a concentration of protein solution between 3-9 mg/mL, protein:precipitantratios between 1:1 to 3:1, and potassium phosphate buffer (pH 7.4) from 1.5-1.6 M.202.4. Data collection, processing, and structure solution2.4 Data collection, processing, and structuresolution(M)L214 RC Mutants. Diffraction data from a single (M)L214G crystal were col-lected at the Canadian Light Source on beamline ID-1 and processed using AUTOXDS,Pointless, and Scala to 2.2 A˚ resolution [38, 75]. Data sets for (M)L214A and (M)L214Ncrystals were collected at the Stanford Synchrotron Radiation Lightsource on beamline7-1 and processed using HKL2000 [53] for (M)L214A, or using Mosflm, Pointless, andScala for (M)L214N [43, 75]. All three mutant RCs crystallized in space group P3121,with one molecule in the asymmetric unit. The solvent content was approximately 75 %,which is typical for a membrane protein crystal. All data sets were isomorphous withthe wild-type RC (PDB entry 2J8C, which was chosen as the starting model because itis the highest-resolution (1.85 A˚) RC structure in the PDB). The (M)214 mutant struc-tures initially underwent five cycles of rigid body refinement with Refmac5 [50]; the inputmodel used was an edited version of PDB entry 2J8C, with residue (M)L214 changedto the appropriate side chain. Subsequently, models were improved first by alternatingbetween inspection of 2Fo–Fc and Fo–Fc electron density maps and manual editing inCoot [19], and followed by restrained refinement with Refmac5. Manual editing consistedprimarily of adding and removing solvent molecules (primarily water, but also glycerol,1,2,3-heptanetriol, and LDAO). Water molecules were added using >3 σ peaks in theFo–Fc difference map with plausible hydrogen bonding geometry. The restrained refine-ment used a geometry weighting term of 0.25, and the standard CCP4 library files wereused for refinement of all cofactors [69]. To generate omit difference density maps, theoccupancies of the phytyl tail atoms from CGA to C20, as well as LDAO and/or glyc-erol, were set to 0, and the model was refined by eight cycles of restrained refinement inRefmac5. Unweighted omit maps were generated using the FFT program in the CCP4suite [75]. Data collection and refinement statistics are listed in Table Data collection, processing, and structure solutionZn-BChl-containing-RCs. Diffraction data were collected at Stanford SynchrotronLight Source on beamline 7.1 using a wavelength of 1.127 09 A˚ and processed with Mosflmand Aimless to 2.85 A˚ resolution [43, 75]. Anomalous differences were detected to 6 A˚resolution. As with the (M)L214 RCs, all data sets were isomorphous with wild type RC,which was used as the starting point for refinement with removal of the Mg2+ ions and,for the Zn-β-RC, the side chain atoms of (M)214. Initial phases were obtained by limitedrefinement with Refmac5 and Fo–Fc and anomalous difference maps were computed withCoot and FFT, respectively. Based on the peaks observed in both maps, Zn-BChl wasmodeled at each of the P, B, and H cofactor sites in both the A and B branches. Densityfor a histidine side chain was observed at residue (M)214 in the Zn-β-RC and modeledaccordingly using Coot. Refinement of the structures was continued with Refmac5 andthe removal of solvent atoms with unrealistic B-factors.22Chapter 3Results3.1 Structural changes in variant RCs with asubstitution in (M)L2143.1.1 Low-temperature spectroscopyLow-temperature (∼11 K) spectroscopy was performed on the (M)L214G and (M)L214Amutants and their spectra were compared with each other as well as with the wild-typeRC. Several notable differences were observed, as shown in Figure 3.1. Most obviousare the changes in the Qx HA transition peak near 545 nm (dashed line in Figure 3.1B);in both (M)L214G and (M)L214A, this peak exhibits a blue shift such that it mergespartially with the 535 nm HB Qx transition. The Qy HA transition peak (first dashedline in Figure 3.1A) similarly exhibits a blue shift in the two variants, although this ef-fect is much more pronounced in (M)L214G. Regarding transitions associated with BChlrather than BPhe, the special pair Qx peak is minimally affected, while the Qy peak(third dashed line in Figure 3.1A) evinces a slight blue shift, again to a greater extent in(M)L214G than in (M)L214A. A greater change is seen in the accessory BChl Qy tran-sition (second dashed line in Figure 3.1A): the long-wavelength side shoulder, which isthought to be associated with the inactive BB BChl, appears to have increased in ampli-tude in both variants, and the primary peak to have decreased substantially in amplitudein (M)L214G.The 11 K absorption spectra of the RC variants containing a polar residue at (M)214differ significantly from the spectrum of the wild-type protein, resembling that of the233.1. Structural changes in variant RCs with a substitution in (M)L214Figure 3.1: Low-temperature (∼11 K) steady state absorption spectra of the (A) Qy and(B) Qx transitions of the nonpolar (M)214 mutant series. Spectra were normalized withrespect to the Qy absorption maximum of the dimer (P) BChl peaks at 895 nm. Thedashed lines mark the Qx and Qy absorption peak wavelengths of HA, as well as the peakwavelengths for the Qy transition of P and BA/BB, in the spectrum of the wild-type RC.Taken from [60].“β-type” RC variant (M)L214H in important ways (Figure 3.2). For example, in thespectrum of the (M)L214N mutant, a new absorbance band was observed at 785 nm(denoted β), and the absorbance that remained at 756 nm is decreased relative to theamplitude of the P BChl absorbance band. In other words, it appears that the absorbanceintensity associated with the Qy transition of a BPhe group has been replaced with thecorresponding absorbance of a new BChl group. In addition, there is an increase in theintensity of the BA BChl absorbance band of the (M)L214H and (M)L214N mutants,where the β and BA absorbance bands overlap.The Qx transition spectra of the (M)L214N and (M)L214Q mutants confirm the con-clusion that these RCs are β-type (Figure 3.2B). As in the (M)L214H control, the peakaround 540 nm is absent from the absorbance spectra of the (M)L214N and (M)L214Q243.1. Structural changes in variant RCs with a substitution in (M)L214mutants, indicative of a loss of the HA BPhe from these RCs. There is a correspondingincrease in the (M)L214H/Q RC absorbance around 595 nm, indicating the presenceof BChl that was absent from the wild-type RC. In the (M)L214H mutant, the BChlQx region has an increased amplitude at ∼595 nm while retaining the ∼605 nm shoul-der. This Qx absorbance profile is thought to be representative of the additional BChl,and the corresponding loss of BPhe, at HA [29, 41]. However, in the amide-containing(M)L214N and -Q RCs, this ∼595 nm peak is blue-shifted relative to those of the wild-type and (M)L214H mutant RCs. The shift is roughly 5 nm in (M)L214N and ∼1-2 nmin (M)L214Q.Figure 3.2: Low-temperature (∼11 K) steady state absorption spectra of the (A) Qy and(B) Qx transitions of the (M)L214N and (M)L214H mutants. Spectra were normalizedwith respect to the Qy absorption maximum of the dimer (P) BChl peaks at ∼895 nm.The dashed line indicates the absorption peak wavelength of the BChl Qx transition inthe spectrum of the wild-type RC. Taken from [60].3.1.2 X-ray crystal structuresTo identify structural changes in the RC that might account for these spectroscopicshifts, such as changes in the positions of HA and BA, the crystal structures of these253.1. Structural changes in variant RCs with a substitution in (M)L214three mutant RCs were determined to resolutions of 2.20 to 2.85 A˚ (Table 3.1). Theprotein backbones of all three mutant RCs were almost identical to that of the wild-typeprotein (PDB entry 2J8C, r.m.s.d. of <0.15 A˚ for all CA atoms). The overall qualityis reasonable, with all three structures at the 50th percentile or better in PDB metrics(Rfree, clash score, Ramachandran outliers, sidechain outliers, and real-space R-valueZ-score [RSRZ]) when compared to other structures of similar resolution [8]. There are3–4 Ramachandran outliers in each structure, three of which are found consistently inother RC structures. The data completeness is high and coordinate error is below 0.2 A˚(Table 3.1).Table 3.1: X-Ray Data Collection and Refinement Statistics for (M)L214G, (M)L214A,and (M)L214N RCs.(M)L214G (M)L214A (M)L214NPDB Entry 4IN5 4IN6 4IN7unit cell parameters (A˚)a = b 139.11 139.03 139.14c 184.69 184.34 185.14resolution (A˚)a 60.24-2.20 60.00-2.70 38.54-2.85(2.32-2.20) (2.75-2.70) (3.00-2.85)unique reflections 105008 56265 48933Ramerge 0.085(0.590) 0.154(0.545) 0.098(0.498)I/σIa 11.3(2.8) 25.7(3.3) 17.2(4.3)multiplicitya 5.5(5.4) 5.9(3.9) 7.2(7.3)completenessa(%) 99.9(100.0) 98.5(97.8) 99.9(99.9)Rwork 0.191 0.189 0.179Rfree 0.212 0.225 0.207Wilson B factor (A˚2) 38.8 53.2 60.9overall B factor (A˚2) 40.8 41.2 42.5coordinate errorb (A˚) 0.08 0.16 0.16rmsd for bond lengths (A˚) 0.018 0.019 0.020rmsd for bond angles (◦) 1.91 2.18 2.27aNumbers in parentheses reflect statistics for the highest-resolution shell. bCoordinateerror is the estimated standard uncertainty from maximum likelihood refinement.In the crystal structures of the (M)L214G and (M)L214A mutant RCs, no differencefrom the wild-type RC in the location of the HA BPhe was observed within coordinateerror. The HA chlorin was well-defined but lacked the electron density that would indicate263.1. Structural changes in variant RCs with a substitution in (M)L214Figure 3.3: Omit difference electron density maps of (A) the (M)L214G RC, (B) thewild-type RC (PDB entry 2J8C), and (C) the (M)L214A RC. These maps were gener-ated by setting the occupancy of the phytyl tail and the nearby LDAO (if present) tozero, followed by eight cycles of refinement. Discontinuous density is visible in the rightconformation of the (M)L214G variant, and continuous density for the left conforma-tion can also be seen. Electron density meshes are contoured at 3.0 σ and carved in a2.5 A˚ radius. Oxygen atoms are coloured red and nitrogen atoms blue. Carbon atomsare coloured orange for LDAO, gray for glycerol and HA, yellow for residue (M)214, andgreen, cyan, and magenta for the (M)L214G, wild-type, and (M)L214A RCs, respectively.273.1. Structural changes in variant RCs with a substitution in (M)L214a metal chelated in the macrocycle, consistent with the interpretation that the HA pocketcontains a BPhe. However, there were clear differences in electron density in the vicinityof (M)214 in addition to the mutations themselves. These differences were clearest inthe highest-resolution structure of the (M)L214G mutant RC. In this structure, electrondensity around the phytyl tail of BA was discontinuous in the region where the tail bends,near the C4 methyl group (Figure 3.3A), implying that the BA tail is disordered. Notably,the bend in the phytyl tail is discernible and in the proximity of residue (M)L214 in thewild-type RC (Figure 3.3B).The volume made available by the deletion of the leucine side chain in the (M)L214Gmutant RC is occupied in part by the rotation around the phytyl tail bend. By inspectionof difference omit maps, the BA phytyl tail was modeled in two conformations withequal occupancy. The “right” conformer depicted in Figure 3.3A is equivalent to theconformation observed in the wild-type R. sphaeroides RC (Figure 3.3B). In the “left”conformation, also shown in Figure 3.3A, the BA phytyl tail is directed away from theHA macrocycle such that a portion of HA is exposed to the solvent. In this conformer,the phytyl chain and a glycerol (cryoprotectant) molecule displace the LDAO moleculethat occupies this space in the structure of the wild-type RC (compare panels A and Bof Figure 3.3).The two conformers and glycerol molecule do not completely account for the differencedensity present in sigma-weighted and unweighted maps at the mutation site or in theregion of the two phytyl tail conformations, suggesting that partially occupied solventmolecules and perhaps additional conformations of the phytyl tail are present. Thedifference maps derived from the (M)L214A crystal diffraction data are consistent withthe two conformations of the phytyl tail (Figure 3.3C), although the lower quality ofthe (M)L214A data restricts structural comparisons with the (M)L214G structure to themutation site.283.2. Structural characterization of Zn-BChl-containing-RCsIn the (M)L214N crystal structure, positive difference electron density was obvious atthe center of the chlorin macrocycle in the HA position (Figure 3.4). This density wasmodelled as a Mg2+ ion, confirming that the chlorin at this position is a BChl group. Aftersubsequent refinement, the B factor associated with the Mg2+ ion was commensuratewith the BChl ring, and no residual difference density was present. At residue (M)214,adjacent to the Mg2+, density for an asparagine side chain was present and modeledaccordingly. The (M)N214 OD2 atom is located approximately 2.0 A˚ from the Mg2+; theMg2+ to OD2 bond angle is approximately perpendicular to the plane of the macrocycle,and the Mg2+ ion is displaced ∼0.4 A˚ from the plane formed by the tetrapyrrole N atomstoward the new ligand, supporting the assignment of a metal–ligand bond. A similar0.4 A˚ displacement of Mg2+ from the plane was observed in the equivalent N–Mg2+ bondbetween (L)H153 and the BA BChl. The ND1 atom of (M)N214 is within H-bond distance(3.1 A˚) of the main chain carbonyl of (M)Y210. The alternative amide conformation ofthe asparagine side chain with Mg2+ coordination by the ND1 atom of Asn214 is unlikelybecause of the inability of the OD2 atom to establish a hydrogen bond with this mainchain carbonyl group, and because of the requirement for the metal to displace a protonof the NH2 group to form a metal—N ligand. Furthermore, the lone electron pair ofthe amide nitrogen atom should participate in a resonance structure of the amide group,resulting in a partial positive charge on the nitrogen.3.2 Structural characterization ofZn-BChl-containing-RCsX-ray crystallography and anomalous scattering. To elucidate the type, coordi-nation state and occupancy of the bacteriochlorin cofactors in Zn-BChl-containing RCs,the crystal structures of the Zn-RC and Zn-β-RC were solved to a resolution of 2.85 A˚.Completeness is somewhat lower for these structures than for the series of (M)L214293.2. Structural characterization of Zn-BChl-containing-RCsTable 3.2: X-Ray Data Collection and Refinement Statistics for the Zn and Zn-β RCs.Zn-RC Zn-β-RCPDB Entry 4N7K 4N7Lunit cell parameters (A˚)a = b 139.45 139.59c 184.18 184.05resolution (A˚)a 73.23-2.85 69.79-2.85(2.90-2.85) (2.90-2.85)unique reflections 46994 42741Ramerge 0.105(0.972) 0.098(0.705)I/σIa 10.5(1.5) 12.6(2.4)multiplicitya 6.2(4.7) 7.0(6.2)completenessa(%) 96(95) 87.4(85.6)Rwork 0.182 0.167Rfree 0.242 0.216Wilson B factor (A˚2) 53.1 47.0overall B factor (A˚2) 58.2 54.8coordinate errorb (A˚) 0.21 0.17rmsd for bond lengths (A˚) 0.011 0.013rmsd for bond angles (◦) 2.6 2.7aNumbers in parentheses reflect statistics for the highest-resolution shell. bCoordinateerror is the estimated standard uncertainty from maximum likelihood refinement.mutants, but nonetheless both structures are at the 50th percentile or better in PDBmetrics, compared against structures of similar resolution [8]. The Zn-RC and Zn-β-RCstructures had 7 (0.8 %) and 6 (0.7 %) Ramachandran outliers respectively, slightly morethan in the (M)L214 series structures. The refinement statistics can be seen in Table 3.2.Overall, the structures resemble that of the native RC (PDB entry 2J8C). In both theZn-RC and Zn-β-RC structures, the iron, carotenoid and all the bacteriochlorin cofac-tors are present. An electron density map was computed from the anomalous differencedata collected above the Zn-edge of the two RC structures to reveal the location of Zn2+ions within the two structures (Table 3.3). A peak greater than 4 σ was observed inthe anomalous maps at the center of all bacteriochlorin-type cofactors, indicative of thepresence of Zn2+ ions at these sites. At the wavelength used for data collection, theanomalous signal from Mg2+ is weak and would not significantly contribute to the den-303.2. Structural characterization of Zn-BChl-containing-RCssity observed in the map. The signal from iron can be used as an internal standard, andindeed, peaks at 8.7 σ and 10 σ were observed at the iron site in the Zn-RC and Zn-β-RC structures, respectively. In the Zn-RC structure, five of the bacteriochlorin cofactorsappear to be bound at full occupancy based on refined Zn2+ ion B-factors of (58 A˚2) orless and anomalous map peaks >12 σ. In contrast, the larger B-factor of the Zn atomat HB site (86 A˚2), in combination with the anomalous map peak (7σ), implies a loweroccupancy and/or a higher degree of disorder of this cofactor. The HB cofactor in theZn-β-RC structure exhibits even lower occupancy and/or greater disorder (Table 3.3).Table 3.3: Anomalous map statistics for metals in the Zn-RC and Zn-β-RC X-raydiffraction datasets.Site Zn-RC Zn-β-RCOmitFo-Fcpeak(σ)Anom.mappeak(σ)Zn B-factor(A˚2)Leu/His-Zn dis-tance(A˚)OmitFo-Fcpeak(σ)Anom.mappeak(σ)Zn B-factor(A˚2)Leu/His-Zn dis-tance(A˚)PA 25 13 50 2.1 25 16 43 2.2PB 25 13 50 2.1 24 14 44 2.2BA 25 14 44 2.4 25 15 39 2.3BB 24 14 39 2.4 25 16 37 2.5HA 20 12 58 3.2 26 15 43 2.3HB 13 7 86 3.6 11 4 113 3.9Fe 24 9 45 – 24 10 41 –These values for the HB cofactor in both the Zn-RC and Zn-β-RC, relative to thevalues for the other bacteriochlorin cofactors, indicate that the HB pocket is not well-occupied, and perhaps less well-occupied in the Zn-β-RC than in the Zn-RC. Interestingly,the B-factors of the amino acid residues surrounding the HB cofactor were similar tothose in the rest of the protein, indicating that the observed disorder was centered onthe Zn-BChl molecule itself, and not the surrounding protein. In comparison to the HBsite, the BB, PA/B, and A-branch Zn-BChls of the Zn-β-RC and Zn-RC appear to berelatively well-ordered and fully occupied, based on the anomalous signal intensity andB-factor values, which are similar to the mean values of each respective structure (Table313.2. Structural characterization of Zn-BChl-containing-RCs3.3). A side by side comparison of the HA cofactors for the Zn-RC and Zn-β-RC can beseen in Figure 3.5. For Zn-BChl in sites with a His residue available for coordination,the imidazole ring is observed 2.1 to 2.4 A˚ from the Zn2+ ion. At the Zn-RC HA site(with a leucine present), the side chain is separated by 3.2 A˚ from the Zn2+ ion of theZn-BChl. Although at 2.85 A˚ resolution for these structures the error estimate of thesedistances is ±0.3 A˚, the electron density maps indicate the absence of an externally-derived metal ligand, such as water, at all of the Zn-BChls. Therefore, it appears thatthe major structural difference between the A-branch electron transfer carriers in theZn-RC and Zn-β-RC is the coordination state of the Zn2+ ion in the HA Zn-BChl, whichis 4-coordinate in the Zn-RC and 5-coordinate in the Zn-β-RC.323.2. Structural characterization of Zn-BChl-containing-RCsFigure 3.4: Omit difference electron density map of the (M)L214N RC, illustrating pos-itive difference density for the Mg2+ ion at the center of the chlorin ring of BChl in theHA site. The electron density mesh is contoured at 4.0 σ and carved in a 2.0 A˚ radius.Oxygen atoms are coloured red and nitrogen atoms are blue. HA carbon atoms arecoloured purple and (M)L214N carbon atoms are yellow.333.2. Structural characterization of Zn-BChl-containing-RCsFigure 3.5: Stick models and electron density (mesh) of the HAA cofactor and axial(M)214 residues from the crystal structures of the (A) Zn-RC and (B) Zn-β-RC. Thedistances from the axial residues to the macrocycle centers are 3.9 and 2.1 to 2.4 forthe Zn-RC and Zn-β-RC, respectively. Colour code: teal, carbon atoms; blue, nitrogenatoms; red, oxygen atoms. Reprinted from [61] with permission.34Chapter 4Discussion4.1 Structural changes in variant RCs with asubstitution in (M)L2144.1.1 Aliphatic non-polar residues (G,A)The above data suggest that the size of the adjacent residue is not the determiningfactor in the presence of BChl or BPhe in the RC HA pocket. Although it was plausiblethat this site could dechelate a BChl Mg2+ ion through steric exclusion via the bulky(M)L214 side chain, the crystal structures show that the chlorin composition of the P,B, and H sites of the (M)L214G and (M)L214A RCs is the same as in the wild-typeprotein, despite the lower side chain volumes. Rather, the data suggest that the typeof cofactor found in the HA pocket is determined solely by the presence or absence ofan axial residue capable of Mg2+ coordination. These results are complementary to andextend previous studies of “cavity mutants” of the R. sphaeroides RC, in which His axialligands to the P BChls were changed to Gly residues, and proposed to introduce a cavityin the vicinity of the P BChls. Although crystal structures were not presented, Starkspectroscopy studies and pigment analysis demonstrated that those RCs assemble with anative pigment composition. Furthermore, water molecules were proposed to act as a fifthligand to the P BChls in the cavity mutants, and these waters appeared to be switchedto other coordinating small molecules by incubation of the RC in the appropriate solute[27].354.1. Structural changes in variant RCs with a substitution in (M)L214The results with mutations at HA enrich our understanding of the complexity in pig-ment discrimination within RCs by indicating that not all binding pockets conform toa single set of rules when it comes to populating these cofactor binding sites. Althoughadventitious ligands such as water may compensate for the loss of an imidazole groupin the P BChls, such polar ligands do not appear to be suitable as BChl Mg2+ ligandswhen the (M)214 Leu side chain is absent from the HA pocket. This difference may stemfrom the fact that the HA pocket is embedded more deeply in the membrane bilayer thanthe P region and hence less likely to incorporate water. Instead, the void is occupied inpart by an alternative conformation of the phytyl tail of BA. An analogous structuralplasticity in the P region of the RC was found in “heterodimer” mutations in which oneof the P BChls was changed to BPhe by substitution of Leu for a His ligand to a BChlMg2+ [12, 42]. In an (M)L214H background, however, a heterodimer failed to assemble,and the resultant RC incorporated two BChls at P for unknown reasons. Those resultsdemonstrated that even when the coordinating axial (M)H202 residue is substituted witha Leu residue, the P binding pocket is capable of incorporating BChl. Although it is con-ceivable that an adventitious ligand was incorporated into the binding pocket of the PBChls in the (M)H202/(M)L214H double mutant, those results raise additional questionsabout the nature of pigment selectivity in RCs as well as in the biogenesis of (B)Phe.The biological production of a pheophytin (Phe) from a Chl is best understood in senes-cent plant materials. Shioi and co-workers partially purified a heat-stable “magnesiumdechelating substance” that catalyzed the conversion of chlorophyllide to pheophorbide[65]. Although this as yet poorly defined low-molecular weight substance was implicatedas functioning during leaf senescence, and therefore Chl degradation, the possibility thatsuch a substance could function during Chl synthesis, to yield Phe for incorporation intoRCs, cannot be excluded.364.1. Structural changes in variant RCs with a substitution in (M)L214Other work on plant material, in this case the source of Phe in the photosystem II RC[30], points to a pre-Chl origin of Phe in which a branch point in the Chl biosyntheticpathway appeared to yield Phe in etiolated leaves. The authors suggested that the mag-nesium chelatase enzyme itself was responsible for Mg2+ removal. Such a branch pointhas not been reported in the BChl biosynthesis pathway of purple phototrophic bacte-ria despite extensive mutagenic analyses [74]. Although it is possible that a dedicateddechelating enzyme exists for the synthesis of RC BPhe, a mutant strain deficient in thisactivity has not been discovered thus far. Moreover, BPhe is functionally present onlyin RCs, and free BPhe is not found in nonsenescent photosynthetic organisms. Alterna-tively, it is possible that the local environment of the BPhe-binding pockets within RCs,in general, results in the loss of Mg2+ from BChl unless a coordinating ligand specificfor the central metal is present. Because of the tendency for this ion to exist only in thepenta- or hexacoordinated state in BChl, it is conceivable that the Mg2+ is removed fromthe macrocycle in the HA site in part because of the absence of a fifth coordinate.An interesting consequence of substituting (M)L214 with small-volume, nonpolar sidechain residues in the R. sphaeroides RC is the shifting of the Qx transition correspondingto the HA BPhe to a higher energy. It was previously suggested by Bylina et al. that ananalogous blue shift may be attributed to changes in the hydrogen bonding interactionbetween the HA BPhe macrocycle and a nearby Glu at position (L)104 [11]. Specifi-cally, when (L)E104 was changed to a weaker proton donor (Gln) or to another residueincapable of hydrogen bonding to the BPhe (Leu), the Qx absorbance maximum of HAbecame more similar to that of HB. In the case of an RC (L)E104L mutant, the Qxtransitions of HA and HB were almost overlapping. A similar blue shift was observed inthe (M)L214G and (M)L214A mutants (Figure 3.1B), but the crystal structures showthat these shifts are independent of changes in macrocycle hydrogen bonding interac-tions. Potentially, the changes in the Qx transition of HA observed in the (M)L214G374.1. Structural changes in variant RCs with a substitution in (M)L214and (M)L214A mutants are induced by the disorder in the BA phytyl tail, and relatedchanges in the binding of detergent and possibly disordered water molecules.The X-ray structures of the (M)L214G and (M)L214A mutants reveal that this residueaffects the orientation of the BA phytyl tail (Figure 3.3) and offer an explanation for thechanges in the low-temperature absorbance spectra of the RCs with these low volume sidechains. Although the influence of the phytyl tail in the direct tuning of the electrochemicalmidpoint potential of a chlorin is not well understood and considered to be negligible, thetail was thought to affect the packing of nearby protein residues and other chlorins [62].The crystal structures indicate that, other than the substituted side chain itself, thereis not a direct effect of the (M)214 side chain substitution on the electronic propertiesof HA (i.e., there appear to be no changes in the distance between the HA macrocycleand protein atoms that would give rise to changes in the electronic properties of the HABPhe). Therefore, the absorbance changes in the Qx and Qy transitions of HA appear tostem predominantly from changes in the BA phytyl tail and perhaps disordered water, asan indirect effect of the substitution of (M)L214 with smaller, nonpolar moieties. Thesemutational changes are also accompanied by a broadening of the Qy absorbance bandof BA in the (M)L214G mutant (Figure 3.1), perhaps reflecting a population of mutantRCs that assume a range of alternative BA phytyl tail configurations, as indicated by thecrystal structure.Although the X-ray diffraction data from (M)L214A RC crystals indicate that somedegree of BA phytyl tail disorder is present (Figure 3.3C), this disorder appears to beless than in the (M)L214G RC crystal, as electron density for the right conformation iscontinuous whereas electron density for the alternative conformation is weaker. It couldbe argued that this difference in electron density arises solely from the lower resolutionof the (M)L214A data relative to the resolution of the data for the (M)L214G structure,but in the absorbance spectra, the broadening of the BChl BA Qy band is seen most384.1. Structural changes in variant RCs with a substitution in (M)L214prominently in the (M)L214G mutant (Figure 3.1). To a lesser extent, the BChl BA Qyband is also broadened in the (M)L214A RC without a corresponding increase in theamplitude of the band. In combination with the similarities between the (M)L214A and(M)L214G Qx transitions, these data indicate some disorder or motion in the (M)L214ABA phytyl tail, although it appears to be insufficient to allow the adoption of alternativeconformations to the same degree as in the (M)L214G mutant.The RC contains the electronically active cofactors within well-insulated tunnelingpathways that prevent molecules such as intramembrane quinones from interacting withRC pigments and thus potentially affecting rates of electron transfer. In vivo, the RCis surrounded by LH1, which may act as a barrier between the quinone pool and theRC [7, 44, 57]. However, it was shown that 25-30 % of the native quinone pool isretained in an isolated RC/LH1/PufX preparation, evidently because quinones bind tothese proteins and may be located within the space between the RC and LH1. In additionto the insulating shield provided by the RC protein matrix, the long, aliphatic phytyltails esterified to the BChl and BPhe macrocycles may also serve a similar role, at leastin vitro.Although it is plausible that these cofactor hydrophobic tails spontaneously adoptalternative configurations, there has been little account of the mechanisms by which thesetails assume their native positions, or what factors might govern their three-dimensionalconfigurations.The in vivo function of R. sphaeroides RCs containing these structural perturbationstestifies to the robustness of this protein in the catalysis of light-driven excitation energytransduction. However, kinetic analysis of electron transfer reactions in purified RCsreveals that the (M)L214G RC is significantly impaired in ET from HA to QA, as is the(M)L214A RC to a lesser extent [54].394.1. Structural changes in variant RCs with a substitution in (M)L2144.1.2 Amide residues (N)The mutant RCs containing Asn and Gln amide side chains at (M)214 incorporatedBChl into the HA pocket. These proteins resemble the (M)L214H (β-type) mutant RCdescribed by Kirmaier et al. [41] in terms of the in vivo activity indicated by similarlyimpaired phototrophic growth kinetics under a low light intensity [60]. However, theabsorbance band shifts induced by these three β-type mutations were very different fromeach other (Figure 3.2). Because the Qy band shifts of the (M)L214N and (M)L214QRCs differed more from each other than from that of the (M)L214H RC (Figure 3.2),it appears that the absorption spectra of HA BChls are not greatly affected by whetherthe Mg2+-ligating atom is an imidazole nitrogen or an amide oxygen. In these cases,the possibility that the amide nitrogen atom acts as a ligand is excluded because of theparticipation of the nitrogen’s lone electron pair in a resonance structure of the amidegroup, the location of the ND1 atom of (M)N214 within H-bonding distance of the mainchain carbonyl of (M)Y210, and the need for loss of a proton of the NH2 group to forma Mg2+—N ligand bond.The side chain of the residue replacing the Leu residue in the (M)L214N, (M)L214Q,and (M)L214H RCs differs in length and volume. The Asn side chain is shorter thanthe Gln side chain, but it offers equivalent atoms to coordinate the Mg2+ ion. The(M)L214N and (M)L214Q RC HA Qy absorbance bands differed by ∼25 nm (Figure3.2), so the length and/or volume of the Mg2+-coordinating side chain appears to havea major effect on the energy of the HA Qy absorbance band, with the longer Gln sidechain resulting in a higher-energy HA band in the Qy region. Indeed, studies of BChlsin a variety of organic solvents demonstrated that the Qy transition is sensitive to thenature of the axial ligand to the pigment. Specifically, the Qy transition shifts to a higherenergy when the Mg2+ ion approaches the plane of the macrocycle because of a strongermetal—macrocycle interaction (this interaction being strongest in an in-plane Mg2+, as404.2. Structural characterization of Zn-BChl-containing-RCsis the case in hexacoordinated BChls), and vice versa [21, 28]. The distance betweenthe HA Mg2+-coordinating atom in the (M)L214H side chain and the BChl Mg2+ maybe intermediate between the distances in the (M)L214N and (M)L214Q RCs; if so, thisdifference would account for the intermediate position of the HA Qy absorbance bandrelative to the HA Qy bands of the (M)L214N and (M)L214Q RCs, but the coordinatesof the (M)L214H mutant RC are unavailable.The available data suggest that the distance between the metal-coordinating atom ofthe RC protein and the Mg2+ ion of (B)Chls, and therefore the position of the Mg2+relative to the plane of the (B)Chl macrocycle, is the major determinant of the Qyband energy in absorbance spectra, rather than the identity of the coordinating atom(i.e., a His nitrogen atom or an oxygen atom from Asn or Gln). Longer distances ofcoordinating atoms from the Mg2+, as is the case with Asn, result in a Qy absorbancemaximum at a longer wavelength, whereas shorter distances (as with His or Glu) resultin correspondingly shorter wavelengths of Qy absorbance peaks.Because the Qy band shifts of the (M)L214N, (M)L214Q, and (M)L214H RCs mustreflect differences in βA excitation energies, these changes should be reflected in changesin ET rates, as in the original (M)L214H mutant RC [41]. Ongoing time-resolved spec-troscopy and structural studies of these new RC variants should yield a deeper under-standing of connections between protein side chain composition, pigment content, andcatalytic activity in this tractable experimental system.4.2 Structural characterization ofZn-BChl-containing-RCsAs mentioned above, in the Zn-RC, the B-factor for the protein component surroundingthe HB Zn-BChl is comparable to the average of the structure. This is consistent with414.2. Structural characterization of Zn-BChl-containing-RCsthe idea that the RC is a relatively rigid protein, and does not rearrange its folding dueto changes in cofactor composition or even when a cavity is created by the lack of acofactor, as was observed in pigment exclusion experiments on the HB BPhe [35, 71].With regard to the cofactors, however, the crystal structures show that, relative tothe WT-RC, the Zn-BChl in the HB pocket of the Zn-RC is bound at less than fulloccupancy or is highly disordered, and perhaps the occupancy is even lower or disordergreater in the Zn-β-RC (Table 3.3). These conclusions are based on weaker density inomit difference maps of the HB Zn-BChl, the large B-factors associated with the Zn2+in the HB pocket, and the low anomalous signal intensities for that Zn2+ atom (comparethe values for HB to those of the other cofactors in Table 3.3).Interestingly, the average B-factor of the HB BPhe from the highest resolution RCstructure currently available in the PDB database is greater than all of the other cofactors,perhaps indicating weaker binding to this pocket in the WT RC as well. Additionalevidence for relatively weak binding of BPhe in the WT RC HB pocket comes fromexperiments involving substitution of BPhe with plant Phe, as well as a variety of otherpigments. These experiments demonstrated that substitution is preferential in the HBover the HA site, consistent with a weaker binding of BPhe in the HB pocket [22, 49]. Invitro, the binding pockets for HA,B are selective for pheophytins, as incubation of RCs ina molar excess of metal-containing bacteriochlorins (the standard procedure for pigmentexchange) resulted in substitution of pigments at the BA,B binding pockets instead ofHA,B [35, 63, 67]. The greater hydrophobicity of BPhe over BChl and Zn-BChl, asindicated by the longer retention time for BPhe in reversed-phase HPLC using a C18column [32], may be part of the reason why binding of metal-containing bacteriochlorinsto the HA,B pockets is weak, although it is not clear why the HB Zn-BChl may be lostmore readily or less efficiently inserted into the RC than the HA Zn-BChl. Alternatively,it is possible that the Zn-RCs contain partially occupied HB sites due to an insufficient424.2. Structural characterization of Zn-BChl-containing-RCsquantity of available Zn-BChls in the cell. One important consequence of the bchDmutation yielding Zn-BChl-producing strains of R. sphaeroides is the severely impairedability to synthesize the pigment [32, 33]. It is presumed that the inability of the organismto grow photoheterotrophically is because of a paucity in the cellular content of light-harvesting and RC complexes.Unlike the HB cofactors of the Zn-RC and Zn-β-RC, the HA cofactors appear lessdisordered, suggesting there is no issue with the occupancy of the pocket. From a struc-tural perspective, the HA cofactors appear to assume a similar configuration in the twoZn-BChl-containing RCs (Figure 3.5), with a possible macrocycle displacement of 0.6 A˚between the two RCs. Given the modest resolution of the data obtained for these twoRCs, a firm conclusion cannot be made on this geometric displacement between the co-factors; however, the small degree of displacement is consistent with the interpretationthat the differences in the functional characteristics of the two Zn-BChl-containing RCsstem from changes in HA cofactor coordination state.These results suggest that RCs have a degree of functional tolerance for different co-factors, and that the protein component is a key instrument in the design of an efficientcharge separation pathway. The Zn-RC studied here is a case in point, as it contains acofactor arrangement supportive of a faster rate of ground state recombination from theP+I−A state, but is nevertheless capable of attaining a relatively high efficiency of elec-tron transfer [45]. This is likely because the protein scaffold of the Zn-RC has remainedunchanged from its WT-RC counterpart, granting this RC the same favourable cofac-tor geometries and tunnelling pathways available to the WT-RC. From an evolutionaryperspective, this could be advantageous to photosynthetic organisms during periods ofenvironmental change, such as a change in the wavelengths of light available, for it wouldallow for the continued utilization of existing protein machinery when the production of anew type of chlorin would provide a selective advantage. In such a case, the new pigment434.3. Conclusions and future directionscould be incorporated into a protein “template” with suitable geometries, which wouldgrant the organism some degree of charge separation capability, enabling photosyntheticgrowth. Mutations in protein side chains may later provide further energetic optimiza-tion, resulting in a high-efficiency RC that assembles with a new pigment. For this tohappen, however, the photosynthetic organism must be able to survive for some time withan inefficient RC to allow beneficial mutations to accumulate. Previous results show thatLH2-lacking R. sphaeroides strains containing RCs with small nonpolar side chains at HAretain their ability to grow at low photon fluxes, despite the relatively low efficiencies ofthese RCs compared to the WT-RC, indicating that purple photoheterotrophs are indeedquite capable of surviving with lower efficiency RCs [60].4.3 Conclusions and future directionsHigh-resolution structures of the R. sphaeroides RC provide insight into the interactionsbetween the protein environment and photosynthetic cofactors, especially in cases wheresteady-state absorption spectroscopy provides insufficient detail. In this study, althoughthe precise mechanism for dechelation of BChl in R. sphaeroides remains elusive, ourunderstanding of the factors that lead to either a BChl or a BPhe occupying HA sitehas been broadened. Structures of (M)L214 variants in addition to those describedhere could provide even more clarification on the role this important residue plays, andperhaps clarify the causes of peak shifts in their spectra. Moreover, the structures thatwere determined suggest an additional avenue of investigation: the phytyl tails of thechlorins, which have so far received relatively little attention. Further investigation ofpossible mobility in these tails could be valuable, perhaps by obtaining and characterizingRCs where the chlorins incorporate a different esterifying alcohol, such as geranylgeraniolor farnesol.Similarly, the structures of Zn-containing RCs not only provided support for previous444.3. Conclusions and future directionshypotheses on the nature of their cofactors, but also illuminated the intriguing fact thatthe occupancy of the HB site is surprisingly low, in not only the Zn-RC but also potentiallyin the WT RC as well. Kinetic studies and structures of variant RCs where electrontransfer occurs through the B-branch cofactors could shed some light on the universality(or lack thereof) of this discovery, and its impact on photosynthetic efficiency. Thesefirst-ever structures of Zn-containing RCs can also serve as a foundation for deeper studyof this unique variant.45Bibliography[1] Addlesee, H.A. and Hunter, C.N. (1999) Physical mapping and functional assign-ment of the geranylgeranyl-bacteriochlorophyll reductase gene, bchP, of Rhodobactersphaeroides. J. Bacteriol., 181(23), 7248–7255.[2] Addlesee, H.A. and Hunter, C.N. (2002) Rhodospirillum rubrum possesses a vari-ant of the bchP gene, encoding geranylgeranyl-bacteriopheophytin reductase. J.Bacteriol., 184(6), 1578–1586.[3] Allen, J., Feher, G., Yeates, T., Rees, D., Deisenhofer, J., Michel, H., and Huber, R.(1986) Structural homology of reaction centers from Rhodopseudomonas sphaeroidesand Rhodopseudomonas viridis as determined by X-ray diffraction. Proc. Natl. Acad.Sci. U.S.A., 83(22), 8589–8593.[4] Allen, J.P. and Feher, G. (1984) Crystallization of reaction center from Rhodopseu-domonas sphaeroides: preliminary characterization. Proc. Natl. Acad. Sci. U.S.A.,81(15), 4795–4799.[5] Allen, J.P., Feher, G., Yeates, T.O., Komiya, H., and Rees, D.C. (1987) Structureof the reaction center from Rhodobacter sphaeroides R-26: the cofactors. Proc. Natl.Acad. Sci. U.S.A., 84(16), 5730–5734.[6] Axelrod, H.L., Abresch, E.C., Okamura, M.Y., Yeh, A.P., Rees, D.C., and Feher, G.(2002) X-ray structure determination of the cytochrome c2: reaction center electrontransfer complex from Rhodobacter sphaeroides. J. Mol. Biol., 319(2), 501–515.[7] Bahatyrova, S., Frese, R.N., Siebert, C.A., Olsen, J.D., Van Der Werf, K.O.,Van Grondelle, R., Niederman, R.A., Bullough, P.A., Otto, C., and Hunter, C.N.46Bibliography(2004) The native architecture of a photosynthetic membrane. Nature, 430, 1058–1062.[8] Berman, H., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T., Weissig, H.,Shindyalov, I., and Bourne, P. (2000). The Protein Data Bank. [Online; accessed4-May-2015].[9] Bollivar, D.W. (2006) Recent advances in chlorophyll biosynthesis. Photosynth. Res.,89(3), 1–22. doi:10.1007/s11120-006-9076-6.[10] Bollivar, D.W., Wang, S., Allen, J.P., and Bauer, C.E. (1994) Molecular geneticanalysis of terminal steps in bacteriochlorophyll a biosynthesis: characterization ofa Rhodobacter capsulatus strain that synthesizes geranylgeraniol-esterified bacteri-ochlorophyll a. Biochemistry, 33(43), 12763–12768.[11] Bylina, E.J., Kirmaier, C., Mcdowell, L., Holten, D., and Youvan, D.C. (1988)Influence of an amino-acid residue on the optical properties and electron transferdynamics of a photosynthetic reaction centre complex. Nature, 336, 182–184.[12] Bylina, E.J. and Youvan, D.C. (1988) Directed mutations affecting spectroscopicand electron transfer properties of the primary donor in the photosynthetic reactioncenter. Proc. Natl. Acad. Sci. U.S.A., 85(19), 7226–7230.[13] Camara-Artigas, A., Brune, D., and Allen, J.P. (2002) Interactions between lipidsand bacterial reaction centers determined by protein crystallography. Proc. Natl.Acad. Sci. U.S.A., 99(17), 11055–11060. doi:10.1073/pnas.162368399.[14] Chang, C.H., El-Kabbani, O., Tiede, D., Norris, J., and Schiffer, M. (1991) Structureof the membrane-bound protein photosynthetic reaction center from Rhodobactersphaeroides. Biochemistry, 30(22), 5352–5360.47Bibliography[15] Chirino, A.J., Lous, E.J., Huber, M., Allen, J.P., Schenck, C.C., Paddock, M.L.,Feher, G., and Rees, D.C. (1994) Crystallographic analyses of site-directed mutantsof the photosynthetic reaction center from Rhodobacter sphaeroides. Biochemistry,33(15), 4584–4593.[16] Cogdell, R.J., Gall, A., and Ko¨hler, J. (2006) The architecture and function ofthe light-harvesting apparatus of purple bacteria: from single molecules to in vivomembranes. Q. Rev. Biophys., 39(03), 227–324.[17] Deisenhofer, J., Epp, O., Miki, K., Huber, R., and Michel, H. (1985) Structure ofthe protein subunits in the photosynthetic reaction centre of Rhodopseudomonasviridis at 3 A˚ resolution. Nature, 318, 618–624.[18] Drews, G. and Golecki, J.R. (1995) Structure, molecular organization, and biosyn-thesis of membranes of purple bacteria. In Anoxygenic Photosynthetic Bacteria, pp.231–257. Springer.[19] Emsley, P. and Cowtan, K. (2004) Coot: model-building tools for moleculargraphics. Acta Crystallogr., Sect. D: Biol. Crystallogr., 60(12), 2126–2132. doi:10.1107/S0907444904019158.[20] Ermler, U., Fritzsch, G., Buchanan, S.K., and Michel, H. (1994) Structure of thephotosynthetic reaction centre from Rhodobacter sphaeroides at 2.65 A˚ resolution:cofactors and protein-cofactor interactions. Structure, 2(10), 925 – 936. doi:http://dx.doi.org/10.1016/S0969-2126(94)00094-8.[21] Fiedor, L., Kania, A., Mysliwa-Kurdziel, B., Orzel, L., and Stochel, G. (2008) Under-standing chlorophylls: central magnesium ion and phytyl as structural determinants.Biochim. Biophys. Acta, Bioenerg., 1777(12), 1491–1500.[22] Franken, E.M., Shkuropatov, A.Y., Francke, C., Neerken, S., Gast, P., Shu-valov, V.A., Hoff, A.J., and Aartsma, T.J. (1997) Reaction centers of Rhodobacter48Bibliographysphaeroides R-26 with selective replacement of bacteriopheophytin by pheophytin a:I. Characterisation of steady-state absorbance and circular dichroism, and of the P+Q−A state. Biochim. Biophys. Acta, Bioenerg., 1319(2), 242–250.[23] Frolov, D., Marsh, M., Crouch, L.I., Fyfe, P.K., Robert, B., van Grondelle, R.,Hadfield, A., and Jones, M.R. (2010) Structural and spectroscopic consequences ofhexacoordination of a bacteriochlorophyll cofactor in the Rhodobacter sphaeroidesreaction center. Biochemistry, 49(9), 1882–1892.[24] Fyfe, P.K., Isaacs, N.W., Cogdell, R.J., and Jones, M.R. (2004) Disruption of aspecific molecular interaction with a bound lipid affects the thermal stability of thepurple bacterial reaction centre. Biochim. Biophys. Acta, Bioenerg., 1608(1), 11–22.[25] Gabrielsen, M., Gardiner, A.T., and Cogdell, R.J. (2009) Peripheral complexes ofpurple bacteria. In The Purple Phototrophic Bacteria, pp. 135–153. Springer.[26] Goldsmith, J.O. and Boxer, S.G. (1996) Rapid isolation of bacterial photosyntheticreaction centers with an engineered poly-histidine tag. Biochim. Biophys. Acta,Bioenerg., 1276(3), 171–175.[27] Goldsmith, J.O., King, B., and Boxer, S.G. (1996) Mg coordination by amino acidside chains is not required for assembly and function of the special pair in bacterialphotosynthetic reaction centers. Biochemistry, 35(7), 2421–2428.[28] Hanson, L.K. (1988) Theoretical calculations of photosynthetic pigments. Pho-tochem. Photobiol., 47(6), 903–921.[29] Heller, B.A., Holten, D., and Kirmaier, C. (1995) Characterization of bacterial re-action centers having mutations of aromatic residues in the binding site of the bac-teriopheophytin intermediary electron carrier. Biochemistry, 34(15), 5294–5302.49Bibliography[30] Ignatov, N.V. and Litvin, F.F. (1994) Photoinduced formation ofpheophytin/chlorophyll-containing complexes during the greening of plant leaves.Photosynth. Res., 42(1), 27–35.[31] Imhoff, J.F. (1995) Taxonomy and physiology of phototrophic purple bacteria andgreen sulfur bacteria. In Anoxygenic Photosynthetic Bacteria, pp. 1–15. Springer.[32] Jaschke, P.R. and Beatty, J.T. (2007) The photosystem of Rhodobacter sphaeroidesassembles with zinc bacteriochlorophyll in a bchd (magnesium chelatase) mutant.Biochemistry, 46(43), 12491–12500. doi:10.1021/bi701407k. PMID: 17910480.[33] Jaschke, P.R., Hardjasa, A., Digby, E.L., Hunter, C.N., and Beatty, J.T. (2011)A bchD (magnesium chelatase) mutant of Rhodobacter sphaeroides synthesizes zincbacteriochlorophyll through novel zinc-containing intermediates. J. Biol. Chem.,286(23), 20313–20322. doi:10.1074/jbc.M110.212605.[34] Johansson, L.C., Wo¨hri, A.B., Katona, G., Engstro¨m, S., and Neutze, R. (2009)Membrane protein crystallization from lipidic phases. Curr. Opin. Struct. Biol.,19(4), 372–378. doi:10.1016/j.sbi.2009.05.006.[35] Jones, M.R. (2008) Structural plasticity of reaction centers from purple bacteria. InThe Purple Phototrophic Bacteria, pp. 295–321. Springer.[36] Jones, M.R., Fyfe, P.K., Roszak, A.W., Isaacs, N.W., and Cogdell, R.J. (2002)Protein-lipid interactions in the purple bacterial reaction centre. Biochim. Biophys.Acta, Biomembr., 1565(2), 206–214.[37] Jones, M.R., Visschers, R.W., Van Grondelle, R., and Hunter, C.N. (1992) Construc-tion and characterization of a mutant of Rhodobacter sphaeroides with the reactioncenter as the sole pigment-protein complex. Biochemistry, 31(18), 4458–4465. doi:10.1021/bi00133a011.50Bibliography[38] Kabsch, W. (2010) Integration, scaling, space-group assignment and post-refinement. Acta Crystallogr., Sect. D: Biol. Crystallogr., 66(2), 133–144.[39] Katona, G. (2003) Lipidic Cubic Phase Crystal Structure of the PhotosyntheticReaction Centre from Rhodobacter sphaeroides at 2.35 A˚Resolution. J. Mol. Biol.,331(3), 681–692. doi:10.1016/S0022-2836(03)00751-4.[40] Katz, J., Strain, H., Harkness, A., Studier, M., Svec, W., Janson, T., and Cope,B. (1972) Esterifying alcohols in the chlorophylls of purple photosynthetic bacteria.A new chlorophyll, bacteriochlorophyll (gg), all-trans-geranylgeranyl bacteriochloro-phyllide a. J. Am. Chem. Soc., 94(22), 7938–7939.[41] Kirmaier, C., Gaul, D., DeBey, R., Holten, D., and Schenck, C.C. (1991) Chargeseparation in a reaction center incorporating bacteriochlorophyll for photoactivebacteriopheophytin. Science, 251(4996), 922–927.[42] Kirmaier, C., Holten, D., Bylina, E.J., and Youvan, D.C. (1988) Electron transferin a genetically modified bacterial reaction center containing a heterodimer. Proc.Natl. Acad. Sci. U.S.A., 85(20), 7562–7566.[43] Leslie, A.G. and Powell, H.R. (2007) Processing diffraction data with mosflm. InEvolving Methods for Macromolecular Crystallography, pp. 41–51. Springer.[44] Lilburn, T.G., Prince, R.C., and Beatty, J.T. (1995) Mutation of the Ser2 codonof the light-harvesting B870 alpha polypeptide of Rhodobacter capsulatus partiallysuppresses the pufX phenotype. J. Bacteriol., 177(16), 4593–4600.[45] Lin, S., Jaschke, P.R., Wang, H., Paddock, M., Tufts, A., Allen, J.P., Rosell, F.I.,Mauk, A.G., Woodbury, N.W., and Beatty, J.T. (2009) Electron transfer in theRhodobacter sphaeroides reaction center assembled with zinc bacteriochlorophyll.Proc. Natl. Acad. Sci. U.S.A., 106(21), 8537–8542. doi:10.1073/pnas.0812719106.51Bibliography[46] Madigan, M.T. and Jung, D.O. (2009) An overview of purple bacteria: systematics,physiology, and habitats. In The Purple Phototrophic Bacteria, pp. 1–15. Springer.[47] McAuley, K.E., Fyfe, P.K., Cogdell, R.J., Isaacs, N.W., and Jones, M.R. (2000)X-ray crystal structure of the YM210W mutant reaction centre from Rhodobactersphaeroides. FEBS Lett., 467(2-3), 285–290.[48] McAuley, K.E., Fyfe, P.K., Ridge, J.P., Isaacs, N.W., Cogdell, R.J., and Jones,M.R. (1999) Structural details of an interaction between cardiolipin and an integralmembrane protein. Proc. Natl. Acad. Sci. U.S.A., 96(26), 14706–14711.[49] Meyer, M. and Scheer, H. (1995) Reaction centers of Rhodobacter sphaeroides R26containing C-3 acetyl and vinyl (bacterio)pheophytins at sites HA,B. Photosynth.Res., 44(1-2), 55–65. doi:10.1007/BF00018296.[50] Murshudov, G.N., Vagin, A.A., and Dodson, E.J. (1997) Refinement of macromolec-ular structures by the maximum-likelihood method. Acta Crystallogr., Sect. D: Biol.Crystallogr., 53(3), 240–255.[51] Neupane, B., Jaschke, P., Saer, R., Beatty, J.T., Reppert, M., and Jankowiak, R.(2012) Electron transfer in Rhodobacter sphaeroides reaction centers containing Zn-bacteriochlorophylls: a hole-burning study. J. Phys. Chem. B, 116(10), 3457–3466.doi:10.1021/jp300304r.[52] Olson, T., Williams, J., and Allen, J. (2014) The three-dimensional structures ofbacterial reaction centers. Photosynth. Res., 120(1-2), 87–98.[53] Otwinowski, Z. and Minor, W. (1997) Processing of X-ray diffraction data collectedin oscillation mode. In J. Charles W. Carter (Ed.), Macromolecular CrystallographyPart A, vol. 276 of Methods Enzymol., pp. 307 – 326. Academic Press. doi:http://dx.doi.org/10.1016/S0076-6879(97)76066-X.52Bibliography[54] Pan, J., Saer, R.G., Lin, S., Guo, Z., Beatty, J.T., and Woodbury, N.W. (2013) Theprotein environment of the bacteriopheophytin anion modulates charge separationand charge recombination in bacterial reaction centers. J. Phys. Chem. B, 117(24),7179–7189. doi:10.1021/jp400132k.[55] Parson, W.W. and Warshel, A. (2009) Mechanism of charge separation in purple bac-terial reaction centers. In The Purple Phototrophic Bacteria, pp. 355–377. Springer.[56] Pawlowicz, N.P., van Grondelle, R., van Stokkum, I.H., Breton, J., Jones, M.R., andGroot, M.L. (2008) Identification of the first steps in charge separation in bacterialphotosynthetic reaction centers of Rhodobacter sphaeroides by ultrafast mid-infraredspectroscopy: electron transfer and protein dynamics. Biophys. J., 95(3), 1268–1284.[57] Qian, P., Bullough, P.A., and Hunter, C.N. (2008) Three-dimensional reconstructionof a membrane-bending complex: The RC-LH1-PufX core dimer of Rhodobactersphaeroides. J. Biol. Chem., 283(20), 14002–14011.[58] Roszak, A.W., Gardiner, A.T., Isaacs, N.W., and Cogdell, R.J. (2007) Brominatedlipids identify lipid binding sites on the surface of the reaction center from Rhodobac-ter sphaeroides. Biochemistry, 46(11), 2909–2916.[59] Roth, M., Lewit-Bentley, A., Michel, H., Deisenhofer, J., Huber, R., and Oester-helt, D. (1989) Detergent structure in crystals of a bacterial photosynthetic reactioncentre. Nature, 340, 659–662.[60] Saer, R.G., Hardjasa, A., Rosell, F.I., Mauk, A.G., Murphy, M.E.P., and Beatty, J.T.(2013) The role of Rhodobacter sphaeroides photosynthetic reaction center residueM214 in the composition, absorbance properties, and conformations of HA and BAcofactors. Biochemistry, 52(13), 2206–2217. doi:10.1021/bi400207m.[61] Saer, R.G., Pan, J., Hardjasa, A., Lin, S., Rosell, F., Mauk, A.G., Woodbury,N.W., Murphy, M.E., and Beatty, J.T. (2014) Structural and kinetic properties53Bibliographyof Rhodobacter sphaeroides photosynthetic reaction centers containing exclusivelyZn-coordinated bacteriochlorophyll as bacteriochlorin cofactors. Biochim. Biophys.Acta, Bioenerg., 1837(3), 366 – 374.[62] Scheer, H. (2006) An overview of chlorophylls and bacteriochlorophylls: biochem-istry, biophysics, functions and applications. In B. Grimm, R.J. Porra, W. Ru¨diger,and H. Scheer (Eds.), Chlorophylls and Bacteriochlorophylls, vol. 25 of Advancesin Photosynthesis and Respiration, pp. 1–26. Springer Netherlands. doi:10.1007/1-4020-4516-6 1.[63] Scheer, H. and Hartwich, G. (1995) Bacterial reaction centers with modified tetrapyr-role chromophores. In Anoxygenic Photosynthetic Bacteria, pp. 649–663. Springer.[64] Senge, M.O. and Smith, K.M. (1995) Biosynthesis and Structures of the Bacteri-ochlorophylls. In R.E. Blankenship, M.T. Madigan, and C.E. Bauer (Eds.), Anoxy-genic Photosynthetic Bacteria, vol. 2 of Advances in Photosynthesis and Respiration,pp. 137–151. Springer Netherlands. doi:10.1007/0-306-47954-0 8.[65] Shioi, Y., Tomita, N., Tsuchiya, T., and Takamiya, K. (1996) Conversion of chloro-phyllide to pheophorbide by Mg-dechelating substance in extracts of Chenopodiumalbum. Plant Physiol. Biochem., 34(1), 41–47.[66] Shochat, S., Arlt, T., Francke, C., Gast, P., van Noort, P.I., Otte, S.C., Schelvis,H.P., Schmidt, S., Vijgenboom, E., Vrieze, J., Zinth, W., and Hoff, A.J. (1994)Spectroscopic characterization of reaction centers of the (M)Y210W mutant of thephotosynthetic bacterium Rhodobacter sphaeroides. Photosynth. Res., 40(1), 55–66.[67] Struck, A. and Scheer, H. (1990) Modified reaction centers from Rhodobactersphaeroides R26: exchange of monomeric bacteriochlorophyll with 132-hydroxy-bacteriochlorophyll. FEBS Lett., 261(2), 385–388.54Bibliography[68] Sturgis, J.N. and Niederman, R.A. (2009) Organization and assembly of light-harvesting complexes in the purple bacterial membrane. In The Purple PhototrophicBacteria, pp. 253–273. Springer.[69] Vagin, A.A., Steiner, R.A., Lebedev, A.A., Potterton, L., McNicholas, S., Long, F.,and Murshudov, G.N. (2004) REFMAC5 dictionary: organization of prior chemicalknowledge and guidelines for its use. Acta Crystallogr., Sect. D: Biol. Crystallogr.,60(12), 2184–2195.[70] Wakeham, M. and Jones, M. (2005) Rewiring photosynthesis: engineering wrong-way electron transfer in the purple bacterial reaction centre. Biochem. Soc. Trans.,33(4), 851–857.[71] Watson, A.J., Fyfe, P.K., Frolov, D., Wakeham, M.C., Nabedryk, E., van Gron-delle, R., Breton, J., and Jones, M.R. (2005) Replacement or exclusion of the B-branch bacteriopheophytin in the purple bacterial reaction centre: the HB cofactor isnot required for assembly or core function of the Rhodobacter sphaeroides complex.Biochim. Biophys. Acta, 1710(1), 34–46.[72] Wen, J., Harada, J., Buyle, K., Yuan, K., Tamiaki, H., Oh-oka, H., Loomis, R.A.,and Blankenship, R.E. (2010) Characterization of an FMO variant of Chlorobaculumtepidum carrying bacteriochlorophyll a esterified by geranylgeraniol. Biochemistry,49(26), 5455–5463.[73] Williams, J.C. and Allen, J.P. (2009) Directed modification of reaction centers frompurple bacteria. In The Purple Phototrophic Bacteria, pp. 337–353. Springer.[74] Willows, R.D. and Kriegel, A.M. (2009) Biosynthesis of bacteriochlorophylls in pur-ple bacteria. In The Purple Phototrophic Bacteria, pp. 57–79. Springer.[75] Winn, M.D., Ballard, C.C., Cowtan, K.D., Dodson, E.J., Emsley, P., Evans,P.R., Keegan, R.M., Krissinel, E.B., Leslie, A.G.W., McCoy, A., McNicholas, S.J.,55BibliographyMurshudov, G.N., Pannu, N.S., Potterton, E.A., Powell, H.R., Read, R.J., Va-gin, A., and Wilson, K.S. (2011) Overview of the CCP4 suite and current de-velopments. Acta Crystallogr., Sect. D: Biol. Crystallogr., 67(4), 235–242. doi:10.1107/S0907444910045749.[76] Woodbury, N.W. and Allen, J.P. (1995) The pathway, kinetics and thermodynamicsof electron transfer in wild type and mutant reaction centers of purple nonsulfurbacteria. In Anoxygenic Photosynthetic Bacteria, pp. 527–557. Springer.[77] Wraight, C.A. and Gunner, M.R. (2009) The acceptor quinones of purple photosyn-thetic bacteria structure and spectroscopy. In The Purple Phototrophic Bacteria,pp. 379–405. Springer.[78] Yeates, T., Komiya, H., Chirino, A., Rees, D., Allen, J., and Feher, G. (1988)Structure of the reaction center from Rhodobacter sphaeroides R-26 and 2.4.1:protein-cofactor (bacteriochlorophyll, bacteriopheophytin, and carotenoid) interac-tions. Proc. Natl. Acad. Sci. U.S.A., 85(21), 7993–7997.56


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items