@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix dc: . @prefix skos: . vivo:departmentOrSchool "Science, Faculty of"@en, "Botany, Department of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Durnford, Dion Glenn"@en ; dcterms:issued "2009-04-16T22:39:24Z"@en, "1995"@en ; vivo:relatedDegree "Doctor of Philosophy - PhD"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description """The light-harvesting complexes (LHC) of the unicellular marine chromophyte, Heterosigma carterae, were fractionated by sucrose-density gradient centrifugation, following digitonin solubilization, and by non-denaturing SDS-PAGE. The sucrose gradient allowed for the isolation of a major light-harvesting complex fraction, containing approximately 53% of the total chlorophyll, the majority of the chlorophyll c and a single polypeptide of 19.5 kDa. Up to 12 different polypeptides immunologically related to both the fucoxanthin-Chi a/c complexes (FCPs) and to the chlorophyll a + b-binding proteins (CABs) were detected in thylakoids and in the lower photosystem I (PS I) enriched fractions. Using a modification of the non-denaturing gel system of Allen and Staehelin (1991 Anal. Biochem. 194, 2 14-222) allowed the resolution of a number of large pigment-protein complexes which included several PS I and PS II fractions along with a predominant LHC fraction, an improvement over previously published methods. A Fcp eDNA from Heterosigina carterae has been cloned and sequenced. It encodes a 210 amino acid polypeptide that has similarity to other FCPs and to the CABs of terrestrial plants and green algae. Comparison of the FCP sequence to the recently determined 3- dimensional structure of the pea LHC II complex indicates that many of the key amino acids thought to participate in the binding of chlorophyll and in the formation of complex-stabilizing ionic interactions between hydrophobic regions of the protein are well conserved. In addition, the Fcp genes are part of a large multigene family with greater than 20 related members in Heterosigina. Phylogenetic analyses of the LHC protein sequences shows that the FCPs form a natural group separate from the iPCPs of the dinoflagellates. Though there are obvious similarities between the FCPs and the CABs, the relationships are very distant. Analyses of polypeptides in the red algae Aglaotharnn ion neglectuin and Porphyridium cruentum, in collaboration with Greg Wolfe and Beth Gantt, were the first to demonstrate that polypeptides immunologically related to the CABs and the FCPs are present within the Rhodophyceae. In addition, CAB/FCP-related LHCs have not been detected in a cyanobacterium (Nostoc) and a prochiorophyte (Prochlorothrix). This suggests the CAB/FCP related LHCs arose only once after the establishment of the chioroplast and provides some evidence that suggests chioroplasts evolved from a symbiotic cyanobacterium-like organism only once (monphyletic). The organization of the antennae in Heterosigma carterae is equally as complex as that in the terrestrial plants. This is indicated by the detection of at least 12 LHC-related polypeptides and the presence of a large multigene family encoding the FCPs. In addition, the immunological relatedness and the sequence conservation of the FCPs with the CABs indicates that the structure of the LHCs has been conserved throughout evolution and that these different antennae complexes share a common ancestor."""@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/7251?expand=metadata"@en ; dcterms:extent "4372063 bytes"@en ; dc:format "application/pdf"@en ; skos:note "AN ANALYSIS OF THE FUCOXANTHIN-CHLOROPHYLL PROTEINSAND THE GENES ENCODING THEM IN THE UNICELLULAR MARINERAPHIDOPHYTE, Heterosigma carterae:CHARACTERIZATION AND EVOLUTIONbyDION GLENN DURNFORDB.Sc. (Biology) Daihousie University 1989A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDEPARTMENT OF BOTANYWe accept this thesis as conformingto the required standardA..THE UNIVERSITY OF BRITISH COLUMBIAJune 1995©Dion Glenn Durnford, 1995In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis br scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)____________________________Department of r9NThe University of British ColumbiaVancouver, CanadaDate__________DE-G (2/88)ABSTRA CTThe light-harvesting complexes (LHC) of the unicellular marine chromophyte,Heterosigma carterae, were fractionated by sucrose-density gradient centrifugation, followingdigitonin solubilization, and by non-denaturing SDS-PAGE. The sucrose gradient allowed forthe isolation of a major light-harvesting complex fraction, containing approximately 53% ofthe total chlorophyll, the majority of the chlorophyll c and a single polypeptide of 19.5 kDa.Up to 12 different polypeptides immunologically related to both the fucoxanthin-Chi a/ccomplexes (FCPs) and to the chlorophyll a + b-binding proteins (CABs) were detected inthylakoids and in the lower photosystem I (PS I) enriched fractions. Using a modification ofthe non-denaturing gel system of Allen and Staehelin (1991 Anal. Biochem. 194, 2 14-222)allowed the resolution of a number of large pigment-protein complexes which included severalPS I and PS II fractions along with a predominant LHC fraction, an improvement overpreviously published methods.A Fcp eDNA from Heterosigina carterae has been cloned and sequenced. It encodes a210 amino acid polypeptide that has similarity to other FCPs and to the CABs of terrestrialplants and green algae. Comparison of the FCP sequence to the recently determined 3-dimensional structure of the pea LHC II complex indicates that many of the key amino acidsthought to participate in the binding of chlorophyll and in the formation of complex-stabilizingionic interactions between hydrophobic regions of the protein are well conserved. In addition,the Fcp genes are part of a large multigene family with greater than 20 related members inHeterosigina. Phylogenetic analyses of the LHC protein sequences shows that the FCPs form anatural group separate from the iPCPs of the dinoflagellates. Though there are obvioussimilarities between the FCPs and the CABs, the relationships are very distant.Analyses of polypeptides in the red algae Aglaotharnnion neglectuin and Porphyridiumcruentum, in collaboration with Greg Wolfe and Beth Gantt, were the first to demonstrate thatpolypeptides immunologically related to the CABs and the FCPs are present within theRhodophyceae. In addition, CAB/FCP-related LHCs have not been detected in a11cyanobacterium (Nostoc) and a prochiorophyte (Prochlorothrix). This suggests the CAB/FCPrelated LHCs arose only once after the establishment of the chioroplast and provides someevidence that suggests chioroplasts evolved from a symbiotic cyanobacterium-like organismonly once (monphyletic).The organization of the antennae in Heterosigma carterae is equally as complex as thatin the terrestrial plants. This is indicated by the detection of at least 12 LHC-relatedpolypeptides and the presence of a large multigene family encoding the FCPs. In addition, theimmunological relatedness and the sequence conservation of the FCPs with the CABs indicatesthat the structure of the LHCs has been conserved throughout evolution and that these differentantennae complexes share a common ancestor.111TABLE OF CONTENTSAbstract iiTable of contents ivList of figures viiiList of tables xList of abbreviations xiAcknowledgments xiiDedication xiiiCHAPTER 1 General introduction1.1 Heterosigma carterae: characteristics and taxonomy 21.2 Photosynthesis--an overview 41.3 Chioroplast characteristics and thylakoid ultrastructure of the major algal taxa. . .61.3.1 Terrestrial plants/green algae 61.3.2 Red algae 61.3.3 Glaucophytes 71.3.4 Euglenophytes 71.3.5 Dinoflagellates 81.3.6 Heterokonts/Haptophytes 81.3.7 Cryptophytes 91.4 Main photosynthetic components of the thylakoid membrane 131.4.1 Photosystem II 141.4.2 Cytochrome b6fcomplex and plastocyanin 161.4.3 Photosystem I 171.5 Light-harvesting antennal systems 191.5.1 Function of antennal complexes 19iv1.5.2 Intrinsic light-harvesting antennae 221.5.2.1 Bacterial LHCs 221.5.2.2 Eukaryotic LHCs 231.5.3 Extrinisic light-harvesting antennae 331.6 Concepts in chloroplast evolution 351.7 Methods in molecular phylogeny 381.8 Scope of this thesis 41CHAPTER 2 Characterization of the light-harvesting proteins from Heterosigma carterae2.1 Introduction 422.2 Materials and Methods 432.21 Heterosigma cultures 432.2.2 Heterosigma thylakoid fractionation 432.2.3 Denaturing SDS-PAGE and western blotting 442.2.4 Non-denaturing gel system 462.2.5 Spectroscopy and fluorescence measurements 462.3 Results 472.3.1 Fractionation of digitonin-solubilized membranes by sucrose gradientcentrifugation 472.3.2 Fractionation by non-denaturing PAGE 542.4 Discussion 59CHAPTER 3 An immunological characterization of LHC related-polypeptides in red algae3.1 Introduction 643.2 Materials and Methods3.2.1 Aglaothamnion cultures 65v3.2.2 Aglaotharnnion neglectuin thylakoid fractionation 663.2.3 SDS-polyacrylarnide gel electrophoresis 673.3 Results 683.4 Discussion 78CHAPTER 4 Characterization of the Fcp cDNAs of Heterosigma carterae4.1 Introduction 864.2 Materials and Methods4.2.1 Tryptic fragment sequencing 874.2.2 Heterosigma DNA and RNA isolation 874.2.3 cDNA library construction and screening 884.2.4 Rapid Amplification of cDNA Ends (RACE) 904.2.5 Southern blots 924.2.6 Northern blots 934.3 Results4.3.1 Identification and characterization of the Fcp 1 and Fcp2 cDNAs 934.3.2 Characterization of the Fcp gene family 1024.3.3 Characterization of the FCP protein sequence 1074.3.4 Analysis of the FCP transit sequence 1184.4 Discussion4.4.1 Fcp cDNA structure and multigene families 1204.4.2 Structural aspects of sequence comparisons 1234.4.3 The transit sequence and protein import 126CHAPTER 5 Phylogenetic analysis of the LHCs5.1 Introduction 129vi5.2 Materials and Methods5.2.1 Protein alignment 1305.2.2 Phylogenetic analysis 1315.2.3 Terms and concepts 1335.3 Results5.3.1 Assessment of phylogenetic signal 1355.3.2 Phylogeny of the tomato chlorophyll a + b-binding protein family 1435.3.3 Phylogeny of the chlorophyll a + b-binding protein family 1455.3.4 Phylogeny of the chlorophyll a + c-binding proteins 1535.3.5 Evolution of the LHC family of proteins 1555.4 Discussion5.4.1 CAB protein evolution 1595.4.2 Evolutionary relationships amongst the Cab gene family 1625.4.3 Evolution of the Chl a + b and Chi a + c gene families 1645.4.4 Evaluation of species relationships based on the LHC protein trees 1655.4.5 Light-harvesting protein evolution: pathways and evolutionary sources 169CHAPTER 6 Summary 173REFERENCES 176viiLIST OF FIGURESFigure 1.1 Heterosigma carterae photograph 3Figure 1.2 Thylakoid membrane components and electron transfer 5Figure 1.3 Structure of the main chiorophylls and carotenoids in plants and algae 21Figure 1.4 Schematic diagram of proposed endosymbioses leading to the chioroplasts 36Figure 2.1 Schematic diagram of the sucrose gradient fractionion of Heterosigma 48Figure 2.2 Spectral characteristics of the sucrose gradient fractions 49Figure 2.3 SDS-PAGE separation of sucrose gradient fractions 50Figure 2.4 Immunological analyses of the sucrose gradient fractions 53Figure 2.5 Mildly denaturing SDS-PAGE separation of Heterosigma thylakoids 54Figure 2.6 Analysis of pigment-protein complexes separated by SDS-PAGE 55Figure 2.7 Spectral analysis of pigment-protein complexes 1, lOa and 11 57Figure 3.1 Agloothamnion sucrose gradient fractionation 68Figure 3.2 Absorption spectrum of Aglaothamnion sucrose gradient fractions 69Figure 3.3 Analyses of LHC-related polypeptides in Aglaothainnion 71Figure 3.4 Immunological detection of Dl in Aglaotharnnion fractions 72Figure 3.5 Immunological analyses of thylakoids from diverse organisms 73Figure 3.6 Analyses of CAB and FCP-related polypeptides in Porphyridium and Nostoc....75Figure 3.7 Immunological analysis of LHC-related polypeptides in Prochlorothrix 77Figure 4.1 Rapid amplification of cDNA ends technique illustrated 91Figure 4.2 FCP tryptic fragment sequences 94Figure 4.3 Schematic representation of the Fcpl cDNA 96Figure 4.4 Nucleotide sequence of the Fcp 1 cDNA 97Figure 4.5 Alignment of nucleic acid sequences of the Fcp 1 and Fcp2 cDNAs 99Figure 4.6 Northern blot of Heterosigma total RNA 100Figure 4.7 Alignment of the nucleic acid sequences from the 3 end of the cDNA clones 103viiiFigure 4.8Figure 4.9Figure 4.10Figure 4.11Figure 4.12Figure 4.13Figure 4.14Figure 4.15Figure 4.16Figure 5.1Figure 5.2Figure 5.3Figure 5.4AFigure 5.4BFigure 5.5Figure 5.6Figure 5.7AFigure 5.7BSouthern blots of Heterosigma genomic DNA probed with the Fcpl and Fcp2cDNAs 104Southern blots of Heterosigma genomic DNA probed with the Fcpl eDNA 106Topological analysis of the Fcpl sequence 109Amino acid alignment of the first and third putative MSR of the FCP 110Amino acid alignment of the Chi a + c-binding proteins 111Amino acid alignment of the Chi a + c and Chi a + b-binding proteins 112Structural model of the pea LHC II 116Structural model of Heterosigma FCP 117Alignment of the known chromophyte transit sequences 119Amino acid alignment of sequences used in the phylogenetic analysis 137Random tree length distributions for the data sets used in the analyses 141Phylogenetic analyses of the tomato CAB proteins 144Distance matrix analysis of the LHCs from selected Chi a + b containingtaxa 146Parsimony analysis of the LHCs from selected Chi a + b containing taxa 147Phylogenetic analyses of the LHCs from terrestrial plants and green algae 150Phylogenetic analyses of the Chi a + c containing proteins 154Distance matrix analysis of the LHC I and LHC II proteins from the CAB, FCPand iPCP containing taxa 156Parsimony analysis of the LHC I and LHC II proteins from the CAB, FCP andiPCP containing taxa 158ixLIST OF TABLESTable 1.1Table 1.2Table 1.3Table 2.1Table 3.1Table 4.1ATable 4. lBTable 5.1Table 5.2Characterization of the major algal taxa 10Summary of the tomato CAB proteins 24Characteristics of chromophyte LHCs 31Summary of immunologic cross-reactivity of Heterosigma thylakoid proteins.. .53Some genes commonly used to infer phylogenetic relationships amongstphotosynthetic organisms 80Codon usage of the Heterosigma Fcp cDNA 101Third codon position data 101Amino acid characters used in phylogenetic analysis 131Species used in the phylogenetic analyses and their taxonomic affiliation 134xLIST OF ABBREVIATIONSAP Allophycocyaninbp base pairsCAB chlorophyll a + b-binding proteinChi chlorophyllFCP fucoxanthin-chiorophyll proteinkb kilobasekDa kiloDaltonLHC light-harvesting complexMSR membrane spanning regionPBS phycobilisomeiPCP intrinisic peridinin-chlorophyll-proteinsPCP soluble peridinin-chlorophyll proteinPC phycocyaninPE phycoerythrinPS I photosystem IPS II photosystem IISDS-PAGE sodium dodecyl sulfate-polyacrylamide gel electrophoresisxiA CKNOWLEDGMENTSThere are a few people I wish to acknowledge for help on various parts of thisproject. I thank Drs. Reudi Aebersold for tryptic fragment sequencing, Art Grossman and AnnEastman for antibodies, Tom Cavalier-Smith for help with the phylogenetic analysis and forallowing me access to his computer, and Carl Douglas for providing access to his laboratory andequipment. I wish to thank Dr. Ingo Damm for helpful comments, friendship, and for somepretty humorous times. Dingren Shen provided endless amounts of advice and his presence inthe lab (along with his command of ‘Leisure Suit Larry’) always kept things interesting. I alsowish to thank Dr. Hema Bandaranayake who, without hesitation, helped out with a numberthings directly and indirectly pertaining to this study. I also extend my gratitude to Dr.Laurence Baeza for assistance during her short stay in the lab. I wish I could have done more toassist her. Dr. Abdul Beihadri had helpful suggestions on PCR and was a very challengingchess partner. I thank those at the front lines of research who have provided advice and help onmethods and techniques. In particular, those past and present members of the Douglas lab whohave consistently provided cheerful advice on demand. I also thank Dr. Beverley Green fortaking the time to edit several versions of this dissertation along with other manuscripts. Inaddition, I thank her for helpful comments, encouragement and financial support throughout theduration of this study.I express my appreciation to Dr. Frank (Franz) Shaughnessy for many interestingdiscussions on statistics and ecology. In addition, I thank Frank, Christel Rasmussen and BevBenedict for being my field guides to the algae and the terrestrial plants (dry algae), though theycould use a few lessons in compass navigation. I particularly thank Bev Benedict who, in herown way, provided support and encouragement.xiiTo my parentsGordon and Sarah DurnfordxiiiCHAPTER]IntroductionThe goal of this study was primarily to characterize the light harvesting antennae from theRaphidophycean alga, Heterosigma carterae. This involved characterization of the fucoxanthinchlorophyll protein (FCP) complexes, cloning of the genes encoding them, and comparing themto the chlorophyll a + b-binding (CAB) family of proteins. The eukaryotic algae are diverse andhave been studied very little in regards to their photosynthetic structure and function. At the startof this study, the antennae from oniy three main groups of chromophytes had been examined inany detail: the brown algae, the diatoms and the dinoflagellates (discussed in section 1.5.2). Thestructural relatedness of the FCPs to the CABs had not been demonstrated until the first FCPgene from a diatom was sequenced (Grossman et al. 1990).Heterosigina carterae was chosen for this study for several reasons. First, the LHCs froma member of this class of algae had not been previously examined. Given the diversity amongstthe chrornophytes (see section 1.3 and Table 1.1), I thought it would be useful to examine adifferent representative from one of the other major algal taxa in order to assess the evolutionaryrelationships amongst the light-harvesting antennae. Second, like other chromophytes, thepresence of four membranes surrounding the chioroplast in Heterosigma (see section 1.3)indicated that nuclear encoded, chloroplast localized proteins probably have a differentmechanism for the targeting and translocation of these proteins into the chioroplast. Since achloroplast targeted-nuclear encoded gene had been determined from only a single chromophyte(the diatom FCP), I was interested in examining the transit sequence of the Heterosigma FCPsequence to get an idea as to possible mechanisms of chloroplast targeting. Third, the chloroplastgenome from Heterosigma carterae (published under the name of Olisthodiscus luteus) had been1well characterized and a few chioroplast genes had been sequenced by Rose-Anne Cattolico’s lab.In addition, easily maintainable, axenic cultures were available.1.1 Heterosigma carterae: characteristics and taxonomyHeterosigma carterae is a unicellular alga that is approximately 11-20 im long, 9-12 Imwide, has anywhere from 9-25 yellowish-brown chioroplasts and possesses two flagella (see Fig.1.1; Hulburt 1965; Hara and Chihara 1987). The chioroplasts have four surrounding membranes;two chioroplast envelope membranes and two membranes of the chioroplast E.R. (CER). Theouter CER membrane is not continuous with the nuclear envelope, as occurs with somechromophytes (Gibbs 1981). The chioroplasts are also distributed along the periphery of the cell.H. carterae possesses Chi a + c along with abundant amounts of fucoxanthin (74% of totalcarotenoid) followed by significant levels of diatoxanthin (14%) and 13-carotene (12%) (Riley andWilson 1967). H. carterae was originally grouped with the xanthophytes due to its yellowishcolour but ultrastructure analyses and other data indicates that it is a member of theRaphidophyceae (Hara et al. 1985). H. carterae is widely distributed in coastal marine habitats.It is an important component in red tides and has been implicated in fish kills, causing the loss ofmillions of dollars within the aquaculture industry of British Columbia (Taylor 1993).There is confusion in the literature regarding the taxa referred to as Olisthodiscus luteusCarter, Olisthodiscus carterae Hulburt (Hulburt 1965) and Heterosigma akashiwo Hada.Morphological and ultrastructural studies on Heterosigma akashiwo Hada and on various culturecollections identified as Olisthodiscus luteus Carter indicated that they were all very similar andcould be combined under the genus, Heterosigma (Hara and Chihara 1987). However, thesecollections were different from the originally described 0. luteus Carter culture (Hara et al. 1985)and appear to be the same as 0. carte rae, as described by Hulburt (1965). As this appears to bethe case, it has been suggested that the genus Heterosigina be retained and the previous speciesname (carterae) be used (Taylor 1992). A number of studies have been published on thechloroplast genome of H. carterae (e.g. Reith and Cattolico 1986) and some physiological work2on this alga has been done. However, most of this work has been published under the name 0.luteus due to a misidentification of the isolates in a number of culture collections (Taylor andHaigh 1993).Figure 1.1Photograph of Heterosigma carterae cells. In order to give a clearer view of the numberof chioroplasts, the prepared slide was allowed to dry partially to cause the cells to flatten.31.2 Photosynthesis—an overviewCyanobacteria, algae and terrestrial plants utilize a chlorophyll a based system for thelight reactions of photosynthesis. In these organisms there are two membrane integralphotochemical reaction centers: photosystem I (PS I) and photosystem II (PS II). Thesecomplexes, along with the membrane soluble plastoquinones (PQH-PQH2),the cytochrome b6fcomplex and plastocyanin (Pc) are responsible for the non-cyclic electron transfer. This electrontransport mechanism through the two photosystems is often referred to as the Z scheme and isillustrated in Figure 1.2 (Hill and Bendel 1960). The purpose of the process is for the productionof the reducing agent NADPH and for the generation of an electrochemical gradient through thenet transfer of protons from the stroma into the lumen. This gradient is then utilized for thegeneration of ATP via the thylakoid membrane bound ATP synthase (ATPase) (Fig. 1.2).NADPH and ATP are used for the fixation of carbon dioxide via the carbon reduction (Calvin)cycle, for the production of carbohydrate, and for many other cellular reactions. In addition,there is also a form of cyclic electron flow around PS I and the cytochrome bf complex mediatedby the ferredoxin NADP-reductase (FNR), as indicated by the dashed line in Figure 1.24Figure 1.2Thylakoid protein complexes involved in photosynthetic electron transport and ATPgeneration. The complexes include photosystem II, the cytochrome b6fcomplex, photosystem I,and the ATP synthase complex. Black filled areas represent the peripheral antennae complexesof PS I and PS II. Grey filled areas represent the core antennae of the reaction center (PS II) orthe chlorophyll binding reaction center complex of PS I.Photosystem IILumenCytochrome b6 f Photosystem I ATPase51.3 Chloroplast characteristics and thylakoid ultrastructure of the major algal taxaMany of the plastid and cytosolic features of the main algal groups are summarized inTable 1.1. This was included in order to emphasize the wide diversity amongst the algae and toprovide the necessary background that will allow one to assess the relationships between them.In this section I will briefly compare and contrast some of the key traits of the different algalgroups.1.3.1 Terrestrial plants/green algaeThe terrestrial plant and green algal chioroplast structure is well known. There are twomembranes of the chloroplast envelope surrounding a network of thylakoid membranes that forma unit with a single connecting lumenal space. The two chioroplast envelope membranes are notequivalent. The outer membrane is more permeable to low molecular mass substances and theinner chloroplast envelope membrane contains 5-10 times more intramembrane particles(Staehelin 1986). The thylakoids form both appressed (stacked, grana lamellae) andnonappressed (unstacked, stroma lamellae) regions. The grana lamellae can consist of a few ormany stacked thylakoids, depending on the conditions and the species. The external surfaces ofnonappressed thylakoids are directly exposed to the stroma. The thylakoid components arenonrandomly distributed between the appressed and nonappressed regions (discussed insection 1.4). The green algae and terrestrial plants are unique in that starch (amylose,amylopectin—an cd-4 glucan) is stored inside the chloroplasts.1.3.2 Red algaeThylakoids of the red algae, like those of the cyanobacteria, have a parallel arrangementof unstacked thylakoids. However, in some taxa the thylakoids are concentrically arranged, forma close association with the inner envelope of the chioroplast, and/or have one or several girdlethylakoid bands surrounding the internal parallel thylakoids (Staehelin 1986; Mörschel and Rhiel1987; Pueschel 1990). All red algae have the main antennae, phycobilisomes, attached to the6outer surfaces of the thylakoid membrane. Accessory chiorophylls, equivalent to Chi b or Chi c,are not present in the red algae. At one time, Chl d was thought to be an accessory chlorophyll inthe red algae, though it had only been observed in Gigartina (Gigartenalies) extracts and had notbeen shown to exist in vivo (Holt 1966). If Chi d is not an artifact of preparation, it is in a verylow concentration and will not make a significant contribution to the action spectrum (Gantt1990). Red algae store various forms of starch (cd,4-glucan) outside the chioroplast, in thecytoplasm.1.3.3 GlaucophytesThe glaucophytes are a group of unicellular algae that possess inclusions referred to ascyanelles, which are considered by some to be modified cyanobacteria functioning as plastids.However, the cyanelles have a reduced genome compared to free-living cyanobacteria and aredependent upon the host. In this regard, they resemble plastids. Most, but not all, contain arudimentary cell wall made up of peptidoglycan surrounding the cyanelle (for a review see Kiesand Kremer, 1990). This group possesses phycobilisomes, like the red algae, and the thylakoidsin the cyanelle are unstacked and usually concentrically arranged. The photosynthate reserve,starch, is stored in the cytoplasm.1.3.4 EuglenophytesEuglenophytes comprise a large group with both photosynthetic (approximately 1/3) andnonphotosynthetic representatives. The photosynthetic taxa contain Chi’s a + b and haveappressed thylakoid membrane regions in bands of three to many, though the thylakoids neverform grana like those in terrestrial plants (Gibbs 1970). Euglenophytes have a chioroplastenvelope with three membranes. The outer most membrane of the chioroplast envelope does notbind ribosomes (Gibbs 1978). Interestingly, the main xanthophylls in this group of algae—diadinoxanthin and diatoxanthin—are more typical of the chromophyte algae than the ChI a + bcontaining algae and terrestrial plants. Euglenophytes store paramylon (a 131-3 linked glucan) asa reserve in crystalline granules outside the chioroplast.71.3.5 DinoflageliatesThe dinoflagellates are considered chromophyte algae because of the presence of Chl c;however, there are also non-photosynthetic taxa. The primary xanthophyll, peridinin, gives theseorganisms the reddish-brown colour associated with red tides. Like the euglenophytes, some ofthe dinoflagellates have a total of three membranes surrounding the chioroplast, the outer (third)membrane lacking bound ribosomes on its cytoplasmic surface. The thylakoids are usuallyarranged in three appressed bands, resembling those of the euglenophytes. They differ from thethylakoids of other chromophytes by having a reduced diameter of the appressed regions and bythe lack of a surrounding thylakoid band (girdle thylakoid) (Staehelin 1986). Dinoflagellatesstore an a.1-4 glucan outside the chioroplast.1.3.6 Heterokonts/HaptophytesThe brown algae, diatoms, chrysophytes, xanthophytes, other heterokonts, and thehaptophytes possess a total of four membranes surrounding the chloroplast. The outer twomembranes are collectively referred to as the chloroplast ER (CER) due to the presence ofribosomes bound to the cytoplasmic side of the outermost membrane. The outer membrane ofthe CER tends to be continuous with the nuclear envelope when the number of chloroplasts arelow (1 or 2). Frequently, a girdle thylakoid of three appressed membranes surrounds the internalthylakoid membranes (Staehelin 1986), though this does not occur in the haptophytes (Gibbs1970). The internal thylakoid membranes usually form appressed regions of three bands. Theheterokonts and haptophytes typically store a f3 1-3 glucan outside the chioroplast, within thecytoplasm.81.3.7 CryptophytesThe cryptophytes are unique in that within the space between the CER and the chioroplastenvelope there is a putative vestigial nucleus (the nucleomorph). This is thought to be a remnantfrom a permanent endosymbiosis of a heterotrophic organism with a photosynthetic eukaryote(Ludwig and Gibbs 1987). The thylakoids are usually in pairs and appear thicker than otherthylakoid membranes (Gibbs 1970). Interestingly, the cryptophytes contain thephycobiliproteins, phycoerytherin or phycocyanin, within the thylakoid lumen (Spear-Bernsteinand Miller 1989). Cryptophytes store starch (an al-4 glucan) in the area between the CER andchloroplast envelope.9Table 1.1 Characteristics of the major algal groupsPlastid CharacteristicsTaxon Chis and main carotene2 thylakoid girdle # chipPBP1 xanth.2 iamellae lameliae membr’sCyanophyta a z (3 1 x napc,peapc,pecEiiiorophyta a, b, c * z J3 x naGlaucophyta 3 a, pc, ape z, c 13 1 x 2, pepRhEta z, n, xpc, pe, apcChiorophytaa)Chlorophyceae a, b n, z, 1, v CX, 13 3+ x 2b)Prasinophyceae a, b n, z, 1, ai (X, (3 3+ x 2Mg_2,4DDinophyta a, c2 p, d3 1 3 x 3(f) (2)Euglenophyta a, b d1, d2, n CX, (3 3+ x 3iChiorarachnida 4 a, b 1-3 x 4. nmtCiptopht a, c2, (cj) a2, c ]3T‘pe*, pc* nmtI)dLd3,n T 3II) f, d, d2 (x)Chryhyta a,cj,c — f, z, a 13Y T(c3) (d1,d2)T — 2(4)Bacillariophyta a,c],c2,c3 f,d1,2 1F *___anthophyta a,c],c Eli, d2, v2 13 T,c2 f 4(z, d1, d2)Eustigmatophyta a -— v, v2 13” T x 4Ttophyt a,c],c2,c3 7, d, d2 3’’ ** *_10Table 1.1 continued.mito.char.flagellarfeaturesstorageproductsTaxon type of number! features type! alternate namescristae type locationCyanophyta na x x na cyanobacteriablue green algaeProchiorophyta na x x naGlaucophyta F 2 mastig. ciJO GlaucophyceaeRhodophyta F x x a/0 red algaeChiorophyta a/Ia)Chlorophyceae F 2 (4) (M) (hairs) cuT green algaeb)Prasinophyceae F 2 (1,4, 8) scales micromonadsDinophyta T 2 hairs a/0 dinoflagellatesEuglenophyta D 2(4) hairs 13/0 EuglenoidsChiorarachnida T 1 hairs a/B green amoebaCryptophyta F T 2 mastig. a/B cryptomonadsRaphidophyta T 2 mastig. ?/0 chlorornonadsH golden-brown algaeChrysophyta T 2 mastig. /0H golden-brown algaei11E 2 mastig. - J3i0 * often gioupedH scales with ChrysophytesBaci11aiophyta 1 * mastig J37E * dToms —iit T T mastig. flenTPhaeophE 2 — mastig. J37D brown algaeEustigmatophyta T 1 (2) mastig. 7/0(H)T——7?:5Ha11Key to abbreviations/explanations for Table 1.1heading descriptionplastidfeatures:Chl’s/PBP -Chlorophyll’s and phycobiliproteins present. a, Chi a; b, Chi b; c1..3, Chi cj3’;pc, phycocyanin; pe, phycoerytherin; apc, allophycocyanin; pec,phycoerythrocyaninmain xanth. prominant xanthophylls present in order of relative abundance: a, anteroxanthin;a, alloxanthin; ci, Crocoxanthin; C2, f3-cryptoxanthin; d1, diadinoxanthin; d2,diatoxanthin; d3, dinoxanthin; f, fucoxanthin; 1, lutein; p, peridinin; vi,violaxanthin; n, neoxanthin; v2, vaucheriaxanthin; z, zeaxanthincarotene-main carotenoid present: o, 13,s-carotene; 13, f3,13-carotene; y, E,e-carotenethylakoids- number of appressed or loosely appressed thylakoid membranesgirdle -presence (‘v) or absence (x) of a surrounding thylakoid membranelamellae-number of membranes surrounding the chloroplast: pep, petidoglycan wallpresent-Type of stored carbohydrate: a, cij-4 glucan; 13, 131,-3 glucan: Location of storedreserve; I, inside chioroplast; 0, outside chioroplast; B, between chloroplastenvelope and chioroplast periplasmic membraneOther symbols used and noteso occasional occurrence/reports* located in thylakoid lument nucleomorph (nm) presentno ribosomes bound to outer membranef Glaucophyta contain a cyanelle andnot a “true” chloroplastna not applicable? unknownx absentJ presentReferences1 (Jeffery 1989)2 (Bjorland and Liaaen-Jensen 1989)(Kies and Kremer 1990)(Hibberd and Norris 1984)This table is a modification of tables found in (Sleigh 1989) and (Lee 1989).# chlpmembrMitochondrial Features:type of -F, flat cristae; T, tubular cristae; D, discoidal cristaecristeaFlagellar Features:number/ type -M, many; A, anisokont (flagella of unequal length); H, heterokont (organismwith a hairy and smooth flagellum); Ha, haptonemafeatures -hairs, fibrous flagellar hairs; mastig., mastigonemes or tripartite, tubular hairs.Storage Products:type/location121.4 Main photosynthetic components of the thylakoid membraneThe main complexes involved in the electron transport process are PS II, thecytochrome b6fcomplex (Cyt b6/), and PS I. An ATPase is also present within the thylakoidmembrane and it produces ATP using the proton gradient generated by the light reactions. Thesecomplexes are not equally distributed throughout the thylakoid membrane of the Chi a + bcontaining organisms. In many cases, the complexes are preferentially localized to eitherappressed or nonappressed thylakoid regions, referred to as lateral heterogeneity. In terrestrialplants, PS TI-LHC II complexes are primarily located in the appressed regions of the thylakoid(PS “a), though some PS II complexes are found in the unstacked sections and have a smallercross sectional absorbance (PS II). PS I is primarily found in the nonappressed (stroma) regionsalong with the ATP synthase complex. The Cyt bfcomplex, on the other hand, is equallydistributed in either region (Staehelin 1986). LHC II is found mainly in the appressed regions ofthe thylakoid, though it can be found in the nonappressed domains in a phosphorylationdependent manner. This reversible association of LHC II with PS II (appressed region) isbelieved to mediate the distribution of excitation energy between the two photosystems and to beimportant in the adaptation of the organism to varying light conditions (Allen 1992).In the chromophyte algae, lateral heterogeneity of the thylakoid membrane complexes,similar to that found in the terrestrial plants, does not occur. In the brown algae (Lichtlé et al.1992b), diatoms (Pyszniak and Gibbs 1992) and cryptomonads (Lichtld et al. 1992) the Chi a + cbinding proteins were demonstrated to be homogeneously distributed throughout the thylakoidmembrane, in both appressed and non-appressed regions, by immunolocalization. PS Icomplexes, however, were slightly enriched in the nonappressed regions of the chromophytethylakoids but this distribution did not resemble the almost exclusive localization of PS Tin thenonappressed regions of the terrestrial plant and green algal thylakoids. Moreover, the thylakoidmembranes of the chromophytes are typically associated in groups of three loosely appressedthylakoids; grana stacks similar to those found in terrestrial plants are not present. Thehomogeneous localization of PS I and the antennae in chromophytes is in agreement with a study13in diatoms that shows energy absorbed by the main LHC complex is distributed equally to PS Iand PS II (Owens 1986b). A role of light dependent phosphorylation in the regulation ofexcitation energy between the complexes has not been demonstrated. It is apparent that the modeof adaptation to light in the chromophytic algae is different from that in the terrestrial plants. Abrief discussion of the characteristics of each major complex in the thylakoid membrane is givenbelow.1.4.] Photosystem IIThe electron transfer reactions are initiated when an electron from a special chlorophyll a(P680), which is probably a dimer, is excited to the singlet state by light. The donation of anelectron from the excited P680 to pheophytin (Ph) results in a charge separation within thereaction center that is the driving force for the oxidation of water. This occurs via a manganesecluster (Mn) and is enhanced by the presence of an oxygen evolving complex (OEC). A series ofoxidation state changes in the manganese cluster, referred to as the water oxidation clock,eventually result in the oxidation of H2O (Rutherford 1989). A redox active tyrosine residue onDl mediates the transfer of electrons from the Mn cluster to P680+. This process results in thereduction of P680 and the liberation of 4 protons (Hj and oxygen within the lumen. P680 isthen able to undergo another round of light induced charge separation in the reaction center.From pheophytin, the electron is donated to the first bound plastoquinone, QA (on D2), then tothe second, QB (on Dl). An associated non-heme iron molecule may influence the stability of thequinones.The PS II complex of terrestrial plants consists of a reaction center and a core antennaesurrounded by more peripheral antennal complexes. The reaction center is a heterodimerconsisting of the Dl and D2 polypeptides, cytochrome b559 (Cytb559) and the psbl protein. BothDl and D2 are predicted to have five transmembrane helicies with a molecular mass of 32 and34 kDa, respectively. The Dl and D2 polypeptides are related to the M and L subunits of thepurple bacterial reaction center and, by analogy, are thought to associate in a similar fashion(Michel and Deisenhofer 1986; Trebst 1987). Since the determination of the 3-D structure of the14bacterial reaction center (Deisenhofer et al. 1985), it has been accepted that the core polypeptides(Dl + D2) bind the primary reactants involved in the initial light-induced charge separation, asdiscussed above. Evidence that Dl and D2 form the photosynthetic reaction center was providedthrough the isolation of a photochemically active core complex including these polypeptides andtwo hydrophobic polypeptides (4 and 9 kDa) of Cyt b559 (Nanba and Satoh 1987). The subunitsof the Cyt b559 complex probably assemble in a heterodimeric fashion and cooperatively bind theheme molecule via a conserved His residue in each (Tae and Cramer 1994). The function ofCyt b559 is not known at the present time and its presence is a major difference between PS II andthe bacterial reaction center. In addition, there are a number of other low molecular masspolypeptides associated with the PS II core reaction center (Hansson and Wydrzynski 1990).Also associated with the PS II complex are the core antennae CP43 and CP47. Theseantennae bind only Chl a and have fluorescence emission peaks at 685 and 695 nm, respectively.Though not in general agreement, it has recently been determined that CP43 binds 20 Chl a and 513-carotene while CP47 binds 2 1-22 Chl a and 4 13-carotene molecules (Alfonso et a!. 1994).These complexes are responsible for the coupling of the energy transfer from the peripheralantennae to the reaction center. In addition, lumen exposed regions of CP43 and CP47 may beimportant for the water splitting process as determined through inactivation and mutagenesisstudies (see Vermaas et al., 1993).In terrestrial plants and green algae there are three major extrinisic polypeptidesassociated with PS II on the lumenal side of the thylakoid membrane which influence theproperties of °2 evolution. They have apparent molecular masses of 33 (OEC1, psbO), 23(OEC2, psbP) and 17 kDa (OEC3, psbQ). These polypeptides are not directly involved inoxygen evolution though they probably have a structural or regulatory function. Incyanobacteria, OEC1 is not essential for 02 evolution though OEC1 deletion mutants are moresensitive to photoinhibition (Mayes et a!. 1991). In contrast, Chlamydomonas mutants with lowexpression of OEC1 lack 02 evolution capabilities and have low PS II stability (Mayfield et al.1987). Despite these differences, which may be related to the organization of the complexes inthe different organisms, it is generally agreed that OEC1 may help to stabilize the Mn cluster15(Ghanotakis and Yocum 1990; Vermaas et al. 1993). OEC2 and OEC3, in combination withOEC1, may be involved in concentrating specific ions near the catalytic site for enhanced wateroxidation properties. Interestingly, homologues of the OEC2 and OEC3 complexes are not foundin the cyanobacteria (Stewart et al. 1985). However, OEC1 is present and is immunologicallyrelated to the OEC 1 of terrestrial plants.1.4.2 Cytochrome b6fcomplex and plastocyaninAs the QB site on Dl becomes reduced twice (PQ—*PQH2), the plastoquinone diffusesfrom its binding site and becomes part of the plastoquinone pool’. An oxidized plastoquinonereplaces the reduced molecule in order to continue the electron transfer process. The transfer ofthe electrons from PS II to PS I is mediated by the thylakoid membrane intrinsic cytochrome b6fcomplex (Cyt b6f) and by the soluble plastocyanin (or cytochrome c553) protein. The Cyt b6fcomplex oxidizes plastoquinone and subsequently reduces plastocyanin in the non-cyclic electrontransfer pathway. It is also involved in the process of cyclic-electron flow around PS I. There isa net transfer of two protons from the chloroplast stroma to the thylakoid lumen for everyelectron donated to the Cyt b6fcomplex. The mechanism of plastoquinone oxidation is not fullyunderstood but is thought to occur via a ‘Q-cycle’ where oxidation of the plastoquinone occurs ina two step process (Malkin 1992). One electron is transferred to the Rieske iron-sulfur center —>cytochromef— plastocyanin pathway. The second electron is passed to the two b cytochromesin succession which, after two turns of the cycle, reduces a quinone. The net result after twocycles is the transfer of four protons into the lumen, the transfer of two electrons to plastocyanin,the oxidation of two plastoquinones and the reduction of a single plastoquinone (Malkin 1992).The purpose of the electron transfer process is in the generation of a proton gradient that is usedin the formation of ATP via a thylakoid membrane ATPase (Fig. 1.2).The Cyt bfcomplex is composed of four main polypeptides in both spinach andcyanobacteria. These include cytochromef (34 kDa), the 23 kDa polypeptide of cytochrome b6,the 20 kDa polypeptide of the Rieske FeS-protein and a 17 kDa subunit (subunit IV) (Hauska1986). Cytochromef(CytJ) has a membrane spanning region, anchoring the subunit to the16thylakoid membrane, and a lumen exposed portion that provides a heme binding site. It is a basicdomain on Cytf that is thought to interact with a acidic domain on plastocyanin to allow electrontransfer to occur between the two components (Cramer et a!. 1994). As well, electron transfer toplastocyanin is thought to be mediated by a Tyr residue near the heme molecule of Cytf.Plastocyanin is a small copper-binding protein (10 kDa) that is located in the thylakoid lumenand functions as an electron carrier between the Cyt b6fcomplex and PS I, where it reducesP700+. Plastocyanin has been found in all terrestrial plants and is present in many algae.However, in some algae, cytochrome c553 can replace plastocyanin.1.4.3 Photosystern IA second light induced charge separation occurs at PS I and starts with the photo-oxidation of P700, which is probably a chlorophyll a dimer (Golbeck and Bryant 1991). Theelectron released in this process follows a linear array of acceptors/donors starting with A0,which is a Chl a molecule. The electron is then donated to A1 (vitamin K1) and to F (a 4Fe-4Scluster). The exact order of electron transfer from Fx to FA (the second 4Fe-4S cluster) or to FB(the third 4Fe-4S cluster) (reviewed by Golbeck and Bryant 1991) has not been determinedconclusively. The presence of three 4Fe-4S clusters arranged in a triangular fashion has beenconfirmed by the 3-D structure of PS I at a resolution of 6A (Krauss et al. 1993). The terminalelectron acceptor is the soluble iron-sulfur protein, ferredoxin (Fd, Fig. 1.2). This protein ispresent in the stroma of the chioroplast and interacts with the soluble FerredoxinNADP+Reductase (FNR) that reduces NADP to NADPH.The primary reactants P700, A0, A1 and F are located on the core complex of PS I(CP1), which is composed of a heterodimer with polypeptides of approximately 83 (psaA) and82 kDa (psaB) (Golbeck and Bryant 1991; Krauss et al. 1993). However, CP I usually migratesat 60-70 kDa on SDS-polyacrylamide gels, probably due to the hydrophobic nature of thecomplex (Green 1988).The two iron-sulfur centers, FA and FB, are bound to PS I subunit VII (psaC) which isclosely associated with CP1 on the stromal side of the membrane. This 9 kDa protein is highly17conserved between cyanobacteria and terrestrial plants. There are many other smaller, non-pigmented polypeptides, in the size range of 4-22 kDa (PsaD—N in plants), associated with PS I.The functions of many of these subunits have not been elucidated. However, the stroma locatedsubunits, PsaD (22 kDa) and PsaE (10 kDa), have been shown to be closely associated with eachother and with PsaC by cross-linking and reconstitution studies (Golbeck 1992). They arethought to be important in the ‘docking’ of ferredoxin and possibly for the binding and orientationof the PsaC subunit. Furthermore, PsaE has been suggested to be important for cyclic electronflow around PS Tin cyanobacteria (Yu et al. 1993). PsaF, located in the thylakoid lumen, hasbeen crosslinked to plastocyanin in plants and is thought to be a plastocyanin docking protein(see Globeck 1992). However, inactivation of the psaF gene in a cyanobacterium didn’t alter theability to grow photoautotrophically and is apparently dispensable (Chitnis et al. 1991).The total molecular mass of PS I is estimated to be 340 kDa. It is thought to bind around100 chiorophylls. Almost half of the chlorophylls have been located around the transmembraneregions of the cyanobacterial PsaA and PsaB polypeptides in the 3-D structure (Krauss et al.1993). Tn terrestrial plants, the PS I core and LHC I together bind approximately 200 chlorophyllmolecules.Overall, there is remarkable conservation of the core reaction center complexes and otherthylakoid complexes directly involved in the electron transfer process. In fact, all of the majorcomplexes discussed above are quite highly conserved amongst the plants, eukaryotic algae andthe cyanobacteria. This is in contrast to a considerable amount of variation in the antennacomplexes which includes differences in size, chlorophyll and carotenoid content, and in theirstructural organization around the reaction centers.1.5 Light-harvesting antenna systems1.5.1 Function of antenna complexesThe light-harvesting antennae, located on the periphery of the reaction centers, increasethe absorptive cross-section of the photosystems thereby increasing the probability of absorbing18the available solar radiation. The antennae are protein complexes that specifically bind pigments(chiorophylls and carotenoids) in a manner that determines the position and orientation of thechromophores; this allows for efficient capture and transfer of the excitation energy. The proteinenvironment greatly influences the absorptive properties of the non-covalently boundchromophores.Plants and algae typically use solar radiation in the visible range (350-700 nm) and areable to absorb these low energy photons due to the conjugated bond system of the chlorophyllmolecule. The delocalization of electrons throughout this conjugated system lowers the energydifference between the ground and excited state allowing for the absorption of the photons in thevisible range. Chlorophyll a is present in all oxygenic photosynthetic organisms and is the onlychlorophyll type within the core complex of either PS I or PS II. Though Chl a is associated withall integral membrane (intrinisic) antennae, there is significant variation in the type of accessorychiorophylls and/or carotenoids that are also bound to the complex.The two main accessory chiorophylls in oxygenic organisms are chlorophyll b, (interrestrial plants, green algae, euglenophytes and the prochlorophytes—see Table 1.1) andchlorophyll c (cl-c3) (in the chromophytic algae). However, Chl c-like pigments have beenfound in the ancient green alga, Mantoniella and in some prochlorophytes (Table 1.1). Structuresof the different chiorophylls are shown in Figure 1.3. The accessory chlorophylls increase thelight capturing ability of the organism by broadening the absorption profile of the antennae.Chlorophylls absorb primarily in the blue (Soret band) and red (a-band) regions with lowabsorption in the green. Chl b, with respect to Chl a, is red shifted in the Soret band and blueshifted in the a-band, illustrating how the combination of the two chlorophylls results in abroadened absorption spectrum. Chl c absorption is similar to that of Chi b except absorption inthe a-band is lower (relative to the Soret band) and blue shifted.Carotenoids are important for both photoprotection and for light harvesting. In terrestrialplants the photoprotective role is of primary importance as the carotenoids do not make asignificant contribution to the absorption spectrum. Carotenoids act as photoprotectors byquenching chlorophyll triplet states, that can result in the production of highly reactive singlet19oxygen, or they can quench singlet oxygen states directly (Rau 1988). In addition, oxygenatedcarotenoids (xanthophylls) may also take part in the xanthophyll cycle which is involved indissipation of excess light energy.In the chromophytes carotenoids are important for light-harvesting, in addition to theirphotoprotective functions. In these cases, the carotenoids are abundant and make significantcontributions to the absorption properties of the cells. Carotenoids that play a dominant role inlight capture typically absorb in the 480-560 nm range, significantly broadening the absorptioncapabilities in a region where chlorophylls have poor absorption. The structures of some of themain carotenoids are shown in Figure 1.3 and will be discussed in the sections that follow.Increasing the antennal size surrounding the reaction centre necessitates an efficientprocess of excitation energy transfer inward towards the reaction center. When a chlorophyllabsorbs a photon, an electron in the chlorophyll is knocked from a ground state orbital to a higherenergy, excited state orbital. It is possible that the excitation may not be localized to the orbitalsof a single pigment but may be delocalized over several pigments (a delocalized exciton) whichmay contribute to the migration of excitation energy (Sauer 1986). Energy transfer from excitedchlorophyll complexes may also occur by inductive transfer (Förster transfer) which is effectiveover longer distances. In this mechanism, an excited donor pigment relaxes to the ground stateafter transferring the excitation energy to a neighboring acceptor, which is then excited (Sauer1986). In either energy transfer process, a separation of a positive and negative charge betweendonor and acceptor pigments (electron transfer) does not occur. This only occurs at the reactioncenter chlorophyll of PS I and PS II.The antennae are closely associated with the reaction center complexes in the thylakoidmembrane. The following discussion has separated the different LHCs into two main categoriesdepending on whether they are integral membrane complexes (intrinisic) or whether they areexternally bound to the thylakoid membrane (extrinisic). I will concentrate on the LHCs from theoxygen evolving organisms, both prokaryotic and eukaryotic. I will not cover the LHCs of theanoxygenic photosynthetic bacteria.20C IdHO OHaCH,OHOHH°°OCOC HhiHO< HOJOH- —-----------------HOCOMe H Me HFigure 1.3Structure of the main chlorophylls (a,b) and carotenoids (c-j) in the algae. (a)Chiorophylls Cl + c2. In c1 R is —C2H5;in c2 R is —CH=CH2.(b) Chiorophylls a + b. Inchlorophyll b the —CH3 on ring II is replaced by —CHO. Carotenoids are as follows: (c)13-carotene, (d) lutein, (e) siphonaxanthin, (f) vaucheriaxanthin, (g) fucoxanthin, (h)diadinoxanthin, (i) peridinin, and (j) alloxanthin.bMe,211.5.2 Membrane-intrinsic light-harvesting antennae1.5.2.1 Prokaryotic light harvesting complexesThe prochiorophytes are oxygen evolving prokaryotes that lack phycobisilomes andcontain Chi a + b. This similarity to the terrestrial plant chloroplasts lead to suggestions that theprochlorophytes were the ancestors of chioroplasts containing Chis a + b (Lewin, 1975). Thefirst identified prochlorophyte, Prochloron was not free-living and was found endosymbioticallyassociated with didemnid ascidians. The Chi a + b antenna from Prochloron is 34 kDa, has aChl a/b ratio of around 2.4, and can be phosphorylated (Schuster et al. 1984; Hiller and Larkum1985). Interestingly, Prochloron has recently been shown to contain significant amounts of aChl c-like pigment, in addition to Chl’s a and b (Larkum et al. 1994). The first discovered free-living prochlorophyte, Prochlorothrix hollandica, was a fresh water dwelling species. InProchlorothrix, there are several Chi a + b antennae containing proteins in the 32 to 38 kDarange having a Chi a/b ratio of 4 (Bullerjahn et a!. 1987; van der Staay 1992; van der Staay andStaehelin 1994). The LHC polypeptides from the above two prochiorophytes areimmunologically related to each other (Bullerjahn et al. 1990) but neither shows relatedness tothe CAB proteins of terrestrial plants, determined by a lack of cross reactivity with CAB directedantibodies (Hiller and Larkum 1985; Bullerjahn et al. 1990).Interestingly, the Prochlorothrix LHC reacts with an antibody directed against a Chi abinding, iron-stress induced protein from cyanobacteria (isiA) (Bullerjahn et al. 1987).Preliminary protein sequence of the LHCs from Prochlorococcus (LaRoche and Partensky,unpubl.), Prochioron (Hiller and Larkum, unpubi.) and Prochlorothrix (van der Staay and Green,unpubi.) indicates that the prochiorophyte LHC genes are related to the isiA gene product and tothe Chi a core antenna, CP43. This confirms a lack of relatedness to LHC II of the terrestrialplants and green algae.221.5.2.2 Eukaryotic light harvesting complexes (LHC)The main intrinisic eukaryotic LHCs are a family of functionally analogous complexesthat are evolutionarily related; discussed more thoroughly in Chapter 5. As the different pigmentcompositions of these complexes probably reflect significant evolutionary divergence, I willconsider each major complex separately. The main intrinisic LHCs are the Chlorophyll a + bbinding proteins (CABs), the fucoxanthin-chlorophyll proteins (FCPs), and the intrinisicperidinin-chlorophyll a + c-proteins (iPCPs) of dinoflagellates. As well, the antennae of otherchromophytes containing abundant xanthophylls other than fucoxanthin, will be reviewed.The chlorophyll a + b-binding proteins (CABs)The chlorophyll a + b-binding proteins are found in all terrestrial plants (vascular plants,ferns, mosses). They are also found in the Chiorophyta (Green algae) and in the photosyntheticmembers of the Euglenophyta. The green amoeba, Chlorarachnion, also contains Chl a + b(Hibberd and Norris 1984), but nothing is known about the LHCs of this organism. The bestcharacterized plants in terms of chlorophyll protein complexes are the angiosperms, particularlytomato, spinach and barley. Some work has also been done on the green alga Chlarnydomonas.The chlorophyll protein complexes and the different mildly-denaturing gel systems used toisolate them have been extensively reviewed (Green 1988; Thornber et al. 1991; Jansson 1994).The literature has been confused with a number of different labeling systems though there hasbeen a recent revision of the gene nomenclature (Jansson et al. 1992). I will adhere to this systemwhen referring to the Cab genes and will use the system of Green et al. (1991) when referring tothe protein complex. One exception is in the designation of CP29 type I and II which I will referto as CP26 and CP29, respectively (Bassi et al. 1990). This will be done due to the lack of closerelatedness between the two proteins and to avoid confusion. A table reviewing the CABnomenclature is shown in Table 1.2.23Table 1.2 Summary of the tomato CAB proteinsgene complex size amino function Chi a/b gene # of Chr.(kDa) acids location ratio copy # introns locationLhcbl LHCII 27 265 8 0 3&2type ILhcb2 LHC II 26 265 major 1.3 2 1 7 &12type II PS IILhcb3 LHC II 25 265 antenna 3 2 12type IIILhcb4* CP29 28 287 core ? ? ?(CP29-I) PS II 2.6Lhcb5 CP26 26 286 antenna l.9 1 5 ?(CP29-II)Lhcb6 CP24 24 210 minor PS II 2 1 7antennapsbS CP22 22 276 minor 3-4 1 3(22 kDa) antenna?Lhcal LHC I 22 246 LHCI-730 2 3 5type ILhca2 LHC I 23 270 LHC 1-680 1 4 10type II 2.2-3.5Lhca3 LHC I 25 273 LHC 1-680 1 2 10type IIILhca4 LHC I 21 250-251 LHC 1-730 2 2 3 & 6type IVArabidopsis gene; § immature polypeptide; Chr.=chromosomeTable 1.1 is a modification of Green at al. (1991) and Jansson (1994).The crystal structure of the pea LHC II complex has been determined at a 3.4 Aresolution. It contains three transmembrane ci-helicies and the complex is stabilized by ionicbonds between buried residues within the first and third transmembrane ct-helicies. The complexcontains 12 chlorophylls (tentatively identified as 7 Chl a and 5 Chl b) and two lutein moleculesat the center of the complex for photoprotection (Ktihlbrandt et al. 1994). The other xanthophyllsassociated with the complex were not localized though there is expected to be a neoxanthin and aviolaxanthin molecule also associated with LHC II (Peter and Thornber 1991).LHC II is the most predominant antenna in the terrestrial plants and green algae and formstrimers that are specifically associated with PS II (Kühlbrandt and Wang 1991). In terrestrialplants LHC II consists of three polypeptides with an unequal stoichiometry. Of these, the 28 kDaLHC II type I protein (Lhcb 1) is most abundant followed by the 27 kDa polypeptide (LHC IItype II; Lhcb2). These are encoded by a multigene family consisting of anywhere from 3 to 16members for Lhcbl and 1 to 4 for members for Lhcb2, depending on the plant (Green et al. 1991;24Jansson 1994). The 25 kDa LHC II type III complex (Lhcb3) is the next most abundant LHC IIcomponent. There are reports of different subcomplexes of the main LHC II complex with theLHC II type I and type II (Lhcbl and Lhcb2) proteins forming a more peripheral, mobile LHC IIantenna (Larsson et al. 1987; Larsson et al. 1987b; Peter and Thornber 1991; Jansson 1994). Thiswould be distal to a LHC II antenna complex with primarily LHC II type I (Larsson et al. 1987;Larsson et al. 1987b) and possibly LHC II type III polypeptides (Peter and Thornber 1991). Thissecond subcomplex would be more closely associated with the core complex of PS II.Presumably the excitation of PS TI is partially controlled by the reversible detachment of thedistal LHC II subcomplex (Larsson et al. 1987; Larsson et al. 1987b; Peter and Thornber 1991).Changes in light intensity, redox state of the plastoquinone pool or temperature can result in thephosphorylation of LHC II (Allen 1992). This leads to migration of the phosphorylated LHC IIfrom PS II in the appressed regions to PS Tin the nonappressed parts of the thylakoid membrane(Anderson and Andersson 1988).Plants grown under intermittent-light conditions (limits Chl b production) preferentiallyaccumulate LHC IT type III (Lhcb3) over the other LHC TI polypeptides indicating the productionof these complexes is differentially regulated (White and Green 1988; Mawson et al. 1994).Because of this, it has also been suggested that LHC TI type ITT is more proximally located to thereaction center and may function as a linker for the bulk LHC TI antennae containing types I andIT (Harrison and Melis 1992; Mawson et al. 1994). Tn agreement with this, regreeningexperiments with intermittent-light grown barley have shown that elevated levels of LHC IItype III accumulate early in the continuous light phase, relative to type I and II (Dreyfuss andThornber 1994).The minor antennal complexes associated with PS II include CP29 (Lhcb4), CP26(Lhcb5), CP24 (Lhcb6) and the 22 kDa polypeptide (psbS). They account for 6% of the totalPS TI associated Chl (Peter and Thornber 1991). CP29 (aka CP29 type IT) and CP26 (aka CP29type I) are often not recognized as distinct complexes since they comigrate in some gel systems.The Chl a/b ratio of CP29 and CP26 are different and estimated to be 2.6 and 1.9, respectively(Bassi et al. 1993). As these complexes (CP29 and CP26) were not as readily dissociated from25PS II with detergents as compared to LHC II, they were considered more tightly associated withthe reaction center (Barbato et al. 1989; Camm and Green 1989; Peter and Thornber 1991). Inaddition, CP29/CP26 remain present in the thylakoids of a barley Chi b-deficient mutant thatotherwise fails to accumulate LHC II (White and Green 1987b), indicating there is differentialstability of some of the LHC II complexes in the absence of Chl b. Both CP29 and CP26 arerelated to LHC II and to each other (Pichersky et al. 1991; Morishige and Thornber 1992), thoughthe relationship is distant (Chapter 5).The third minor chlorophyll-protein complex (CP24, Lhcb6) has a low Chi a/b ratio ( 1),a molecular mass of 24 kDa (Dunahay and Staehelin 1986) and is quite divergent from eitherLHC I or LHC II polypeptides (Jansson 1994). CP24 is not tightly associated with the PS IIcomplex and has been proposed to connect the LHC II complex to the PS II reaction center(Barbato et al. 1989). The final minor complex, CP 22 (with a single 22 kDa apoprotein) isassociated with the core complex of PS II and has recently been shown to bind chlorophyll (Funket al. 1994). Sequence of the gene encoding this protein (psbS) revealed limited similarity to theCABs and showed four putative membrane spanning regions (Kim et al. 1992).Evidence for an antennal complex specifically associated with PS I (LHC I) was firstshown in pea PS I preparations (Mullet et al. 1980). Since then it has been isolated and morethoroughly investigated in a number of terrestrial plant taxa (Haworth et al. 1983; Lam et al.1984; Lam et al. 1984b). The presence of a Chl a + b containing PS I associated LHC (CP 0) hasalso been demonstrated in the green alga Chlamydomonas (Ish-Shalom and Ohad 1983). Twomain subcomplexes of LHC I, termed LHC I 680 and LHC I 730, have been isolated and found tohave Chi a/b ratios in the range of 2.2-3.5 (Lam et al. 1984; Bassi and Simpson 1987). They areso named due to their 77 K fluorescence emission at 680 and 730 nm, respectively. In barley,LHC I 680 has been shown to consist of two separate polypeptides encoded by the Lhca2(LHC I type II) and Lhca3 (LHC I type III) genes which have apparent molecular masses of 23and 25 kDa, respectively. LHC I 730 consists of two polypeptides, with apparent molecularmasses of 22 and 21 kDa, and N-terminal sequencing has demonstrated that these polypeptidescorrespond to LHC I type I (Lhcal) and LHC I type IV (Lhca4), respectively (Knoetzel et al.261992). The instability of PS I and the easy detachment of LHC I 730 in a LHC I 680 depletedbarley mutant indicates that LHC I 680 is involved in binding LHC I 730 to the reaction centercore (Knoetzel and Simpson 1991).The CAB proteins ofgreen algaeThe pigment-protein complexes from Chi a + b containing algae (other thanChlamydornonas) have been studied very little. The LHC II sequences from Chlamydomonas,Dunaliella and the terrestrial plants are all highly conserved; these evolutionary relationships willbe more thoroughly considered in Chapter 5. The organization of the inner LHC II antennaeshould be quite similar between the green algae and the terrestrial plants as the same complexeshave been identified in each. However, there are indications of novel regulatory mechanisms inthe green alga, Dunaliella, which involve modifications in the Chi a/b ratio in changing lightintensities (Sukenik et al. 1987). Further work may unveil significant differences in the terrestrialplant-green algal CAB organization and regulation. Nonetheless, there are some green algal taxathat have a pigment composition significantly different from those of Chlamydomonas andDunaliella; these will be discussed below.In the green alga, Codium sp. (Siphonales), the CAB proteins contain the unusualcarotenoid siphonaxanthin (instead of lutein) which increases the absorbance in the 500-550 nmrange. The LHC II component has a lower Chl a/b ratio than the terrestrial plant LHC 11(0.7)and contains four polypeptides from 27-35.5 kDa (Anderson 1985). Two of these polypeptides(34 and 35.5 kDa) have sizes significantly different from the terrestrial plant CABs.Furthermore, a PS I specific antenna was isolated from this alga. It contained siphonaxanthin,had a Chl a/b ratio of 1.7 and a polypeptide composition resembling the LHC I polypeptides ofterrestrial plants (19-25 kDa) (Chu and Anderson 1985). The Codium CAB proteins areimmunologically related to those of terrestrial plants, despite the pigment differences betweenthem (Anderson et al. 1987).The Prasinophyte algae (Micromonadophyceae) have unique LHC antennae that containChl a, Chl b, and a Chl c-like pigment (Mg 2,4-divinyl pheoporphryin a monomethyl ester)27(Jeffery 1989). The carotenoid prasinoxanthin was also associated with the antenna complex.The LHC complexes from the prasinophyte, Mantoniella squamata, contain at least twopolypeptides of 20.5 and 22 kDa (Fawley et al. 1986b) and are arranged into larger oligomericcomplexes of 80 kDa (possibly trimers) (Rhiel et al. 1993). The presence of a Chl c-like pigmentand the smaller size suggested a closer relationship to the Chl a + c-binding proteins; however,sequencing of the a gene encoding this LHC has confirmed its relatedness (though distant) to theCABs (Rhiel and Mörschel 1993). There is some evidence that a unique PS I associated antennaemay not exist in Mantoniella and that the same protein complex excites both photosystems(Schmitt et al. 1993). The significance of this in terms of regulation and distribution of thecomplexes remains to be seen.Euglena gracilis also has Chi a + b-binding antennae that have sizes in the 26-28 kDarange (Cunningham and Schiff 1986). The predominant LHC has been reported to have amolecular ratio of 12 Chl a: 6 Chl b: 4 diadinoxanthin: 1 neoxanthin (Cunningham and Schiff1986b). Interestingly, the xanthophyll diadinoxanthin is more commonly found in thechromophytes rather than in Chl a + b-containing organisms. Nevertheless, sequencing of genesencoding LHC II and LHC I proteins from Euglena has confirmed their relatedness to the CABs(Houlné and Schantz 1988; Muchhal and Schwartzbach 1992). In spite of the sequencesimilarities, the LHCs from Euglena are uniquely translated into large polyprotein precursorsfrom unusually large mRNAs.The fucoxanthin-chlorophyll proteins (FCPs)Fucoxanthin is an oxygenic, allenic-xanthophyll (Fig. 1.3) that absorbs in the 450-550 nmrange. Its presence in some algae extends the absorption range of the LHC into the green regionof the spectrum. This would be particularly useful as the coastal ocean waters are usually limitedin the blue wavelengths of the spectrum and are more light limited than the terrestrial plants(Larkum and Barrett 1983). However, there is no correlation between light quality (due toattenuation of specific wavelengths of light at different depths) and the nature of the lightharvesting system used by the algae growing there (Saffo 1987).28Fucoxanthin-chiorophyll proteins occur in the diatoms, chrysophytes, phaeophytes,haptophytes, and some members of the raphidophytes (including Heterosigma). The exact molarratio of the pigments associated with the FCP are not precisely known. However, analyses ofcomplexes isolated using milder detergents suggests that there are 13 Chi a: 3 Chl c:10 fucoxanthin: 1 violaxanthin for a brown alga FCP (Katoh et al. 1989) and approximately12.5 Chl a: 5 Chl c: 24 fucoxanthin for a diatom FCP (Friedman and Alberte 1984). It is clearthat xanthophylls are much more abundant in these chlorophyll-proteins than in those of theterrestrial plants. Usually, no attempt is made to distinguish between the different Chl c forms(Cl, c2, c3), though in one study a Chl c2/cl ratio of 3.1 was observed for a Phaeodactylum FCP(Owens and Wold 1986). Both fucoxanthin and Chi c transfer excitation energy to Chi a,indicating their function in the harvesting of light energy (Duval et al. 1983).Typically, one or two polypeptides with a molecular mass of 15-2 1 kDa have beenidentified in FCP-containing fractions (see Table 1.3). Earlier studies reported the isolation of aChl a + c pigment-protein complex without fucoxanthin (Barrett and Anderson 1980; Alberte eta!. 1981; Peyriere et al. 1984; Owens and Wold 1986; Hsu and Lee 1987; Boczar and Prezelin1989). As the pigment content in these complexes was variable and the polypeptides had thesame apparent size as the FCPs, the nature of these complexes is uncertain. However, theChl a + c complexes were isolated using either the detergent triton X-100 or SDS and there is thepossibility that they are the result of pigment loss from the main FCP complex (Hiller et a!.1991). In this study, a specific Chi a + C-containing complex lacking fucoxanthin was not found.Structural relatedness of the FCPs to the CABs has been suggested on the basis ofimmunological cross-reactions of these polypeptides to antibodies directed towards one of thetwo LHC types (Caron et al. 1988; Passaquet et a!. 1991; Plumley et al. 1993). In thisdissertation, antisera specific for the two groups of LHCs were used to investigate the relatednessof the Heterosigma FCPs to the CABs and other FCP complexes. The FCPs, like the CABs, arealso nuclear encoded, translated on cytoplasmic ribosomes, and processed into the maturepolypeptide (Fawley and Grossman 1986). Their relatedness to the CABs was first confirmed bythe sequencing of a cDNA encoding the FCP from Phaeodactylum (Grossman et al. 1990). This29protein possessed four hydrophobic regions, three of which are present in the mature protein andmay potentially form membrane spanning regions. Since then, cDNAs encoding FCPs fromanother diatom (Odontella-gb X81054) a brown alga (Apt et al. 1994) and a haptophyte(LaRoche et al. 1994) have been sequenced, in addition to the Heterosigma Fcp cDNA.Other chromophyte antennaeOther taxa where xanthophylls, other than fucoxanthin, play a significant role in lightabsorption include the cryptophytes, the xanthophytes and the eustigmatophytes. Thecryptophytes possess a Chi a + c2 containing antenna complex which is abundant in thexanthophyll alloxanthin. The molar ratio of these pigments in an antenna fraction fromCryptomonas rufescens has been estimated to be 10 Chi a: 2Chl C: 3.4 alloxanthin (Lichtlé et al.1987). The polypeptides of these antennae are in the 18-24 kDa size range, similar to the FCPs,and Chi a/c ratios of 1.4 (Chroomonas), 1.7 (Cryptomonas maculata) and 4.8 (Cryptomonasrufescens) have been recorded (Ingram and Hiller 1983; Lichtlé et al. 1987; Rhiel et al. 1987). Asmall peptide sequence fragment from a cryptomonad LHC (Sidler et al. 1988) showedsimilarities to the diatom and brown algal FCP sequences.A 23 kDa LHC from the xanthophyte Pleurochioris meiringensis contains Chl a, Chl cand three abundant xanthophylls: diadinoxanthin, vaucheriaxanthin, and heteroxanthin (Wilhelmet al. 1988). These pigments have been reported to exists in the following molar ratio:1000 Chl a, 224 Chl c, 148 heteroxanthin, 264 diadinoxanthin, and 129 vaucheriaxanthin. A PS Iassociated antenna with a fluorescence emission different from the main LHC has also beenisolated from this alga (Büchel and Wilhelm 1993).In the eustigmatophytes there is a lack of Chi c while the xanthophyll, violaxanthin, isabundant and plays a significant light-harvesting role. The main LHC fraction from anothereustigmatophyte, Monodus subterraneus, was enriched in a 23 kDa polypeptide. The pigmentratios of this complex were estimated to be 10 Chi a: 2.8 violaxanthin: 1.3 vaucheriaxanthin ester(Arsalane et al. 1992). The 23 kDa polypeptide is immunologically related to the main FCP froma brown alga indicating a structural relatedness to this group of LHCs (Arsalane et al. 1992).30PaviovaIsochrysisChi Cl + C2fucoxanthinChi C 45diadinoxanthinvaucheriaxanthinheteroxanthinEustigmatophyta Nannochioropsis violaxanthinvaucheriaxanthin2.7 17.5, 18.5 52.3+5.0 15 63.0 18, 19, 19.52.7 16.4, 16.91.8-3.2 15-202.8 16, 181.5-1.7 =18nd 16.5, 17, 18, 1220.521 1318,20,24 14noChiC 26 17,25DinophytaChroornonas Chi C2alloxanthinCryptomonas1(Barrett and Anderson 1980)2(Caron et al. 1988)3(Berkaloff et a!. 1990)4(Katoh et al. 1989)5(Friedrnan and Alberte 1984)6(Gugliemelli 1984)7(Fawley and Grossman 1986)8(Owens and Wold 1986)9(Caron and Brown 1987)10(Hsu and Lee 1987)11(Brown 1988)‘2(Fawley et al. 1987)13(Hiller et al. 1988)14(LaRoche et al. 1994)15(Wilhelm et al. 1988)16(Büchel and Wilhelm 1993)17(Livne et al. 1992)18(Arsalane et al. 1992)‘9(Ingram and Hiller 1983)20(Rhiel et al. 1987)21(Boczar and Prezelin 1987)22(Knoetzel and Rensing 1990)23(Hiller et al. 1993)4(Iglesias-Prieto et al. 1993)25(Chrystal and Larkum 1987)Table 1.3 Summary of characteristics from chromophyte LHCsgroup organism accessory pigment Chl a/C polypeptides Refsratio kDaPhaeophyta(brown algae)B acilliarophyta(Diatoms)HaptophytaChi Cl + C2fucoxanthinChi Cl + C2fucoxanthinACrocarpia4 taxaFUCUSLaminariaDictyotaPhaeodactylumCylindrotheca5 taxa2nd5.63.34.3nd20-2321205493347891011Xanthophyta PleuroChioris4.722-23 1516MondusVisheriaesterCryptophytaGonyaulax Chi C2peridininAmphidinium3 taxa23 1820? 181.4 20,24 191.7 18, 19, 22? 201-4 17-24 2119? 221.7 19 23nd 19-20 2431Preliminary peptide sequence information from a LHC from Nannochioropsis (eustigmatophyte)(Livne et al. 1992) indicates that these polypeptides are related to the FCPs.The intrinisic peridinin-chiorophyll proteins (iPCPs)The dinoflagellate LHCs bind Chi a, Chi c2 (not c) and the xanthophyll, peridinin.Peridinin is brick red in colour and has aC37-skeleton (instead of C40 as with other xanthophylls)and an oxidized in-chain methyl group (see Fig. 1 .3i). There are two main light-harvestingcomplexes in the dinoflagellates: an intrinisic peridinin-chiorophyll a + c complex (iPCP) and awater-soluble PCP (sPCP). The latter LHC will be discussed in the next section. There is someevidence for the occurrence of a Chi a + c2 complex with a high proportion of Chl c2 (Chl a/cratio 0.3) and devoid of peridinin (Boczar et al. 1980; Boczar and Prezelin 1987). However,the status of this complex is uncertain due to the use of SDS in its isolation, variations in pigmentcontent and lack of its detection in other studies (Hiller et a!. 1991; Hiller et a!. 1993; IglesiasPrieto et al. 1993). The iPCP is the main light-harvesting antenna in the dinoflagellate,Symbiodinium, with a Chl a, Chl c, peridinin ratio of 1:1:2 and comprising 45, 75 and 70% ofthe total cellular levels of these pigments, respectively (Iglesias-Prieto eta!. 1993). InAmphidinium, the Chl a/c ratio is a little higher (1.7) though a similar ratio of Chi a/peridininoccurs (Hiller et al. 1993). In all studies, the polypeptides for the iPCPs are 19-20 kDa in size.Immunological cross-reactivity of FCP polypeptides with iPCP directed antibodies and theprotein sequence of the iPCP of Amphidinium demonstrates similarity to the FCPs, suggestive ofa common evolutionary origin (Hiller et a!. 1993; Hiller, unpubl. data-gb Z47563). Theindividual iPCP polypeptides from Aniphidiniuin are encoded by a putative polyprotein (Hiller,unpubl. data). This would be similar to coding arrangement of the Euglena LHC proteins(Houlnd and Schantz 1988). Significantly, photosynthetic members of the euglenophytes and thedinoflage!lates have a total of three membranes surrounding the chloroplast.321.5.3 The extrinisic light-harvesting antennaeThe extrinisic light-harvesting antennae are soluble complexes that are easily detachedfrom the membrane. This type of complex includes the soluble peridinin-chlorophyll proteincomplexes of the dinoflagellates, as well as the phycobilisomes of both the red algae andcyanobacteria.Soluble Peridinin- Chlorophyll Complex (sPCP)Soluble peridinin-chlorophyll a complexes were the first photosynthetic proteinscharacterized from the dinoflagellates. Early studies indicated that the complexes wereextrinisically associated with the thylakoid membrane, bound significant amounts of the totalperidinin and transferred excitation energy efficiently from peridinin to Chl a (Prézelin and Haxo1976; Song et al. 1976). The sPCPs also had a molar ratio of Chi a to peridinin of 1:4 and werebound to a monomeric complex of 31-35 kDa or to an apparent 15 kDa homodimer, dependingon the species (Govind et al. 1990). Recent studies on the sPCPs have estimated that they bind 5and 15% of the total cellular Chi a and peridinin, respectively (Iglesias-Prieto et al. 1993), thoughmuch higher estimates have been reported. This suggests that, though the sPCPs makesignificant contributions to the absorption of light, the iPCPs are the main antennal complexes.Recently the sequence of a gene encoding a sPCP from Symbiodinium was determined. Itencoded a 35 kDa polypeptide with a internal duplication, suggesting it arose from the fusion ofgenes encoding the 15 kDa sPCP form (Norris and Miller 1994). No similarity to any other LHCwas found. Interestingly, the resemblance of part of the sPCP transit sequence to the transitsequence of a thylakoid lumen localized polypeptide of terrestrial plants has lead to thesuggestion that the sPCPs may be localized within this compartment (Norris and Miller 1994).PhycobilisoinesThe phycobilisomes are large antennal complexes consisting of many chromophorebinding proteins which are responsible for light absorption in the 450-655 nm range, the33wavelengths where there is poor absorption by chlorophyll. The primary phycobiliproteinsmaking up the phycobilisome are phycoerytherin (PE) (Amax 560 nm), Phycocyanin (PC) (Amax620 nm) and allophycocyanin (AP) (Amax 650 nm). Each phycobiliprotein has two differentsubunits (o and 13) with a molecular mass in the 17-22 kDa range that form dimers. The c and 13polypeptides of each phycobiliprotein share a degree of sequence similarity and are evolutionarilyhomologous (Zuber 1986). Each subunit binds 1-4 open chain tetrapyrrole chromophores(phycobilins) that are attached by a thioether bond to a cysteinyl residue on the protein. Thespectral characteristics of the phycobiliprotein are partly influenced by the protein environment ofthe chromophore (Glazer 1989), as is the case with other pigment-protein complexes.Each phycobiliprotein dimer (ct,13) is arranged into cyclic trimers which form the buildingblocks of the phycobilisome. Two phycobiliprotein cyclic trimers ({a,13}3) are assembled intohexameric protein aggregates ({ a,j3 }6) of PE and PC, which make up the rod like structures ofthe phycobilisomes. These are bound to a usually triangular shaped core assemblage of AP. Thephycobiliprotein aggregates are assembled and held together with colourless linker polypeptides(Mörschel and Rhiel 1987). The exact composition of PBS rods depends upon the organism,growth and light conditions and has been previously reviewed (Mörschel and Rhiel 1987; Glazer1989; Gantt 1990; Grossman et al. 1993). Typically, the PBSs are organized in two ways: (1) ina hemi-discoidal fashion with a central core of AP from which the PE and PC rods radiate, or (2)in a hemi-ellipsoidal arrangement which is similar to the above organization except it is twice asthick (Gantt 1981). Both types of pycobilisomes are found in the cyanobacteria and red algae.The cryptomonads also contain phycobilins (phycoerythrin or phycocyanin) that arelocated within the thylakoid lumen (Spear-Bernstein and Miller 1989). This phycobiliproteindoes not form phycobilisomes but does specifically transfer excitation energy to PS II (Lichtlé etal. 1980).341.6 Concepts in chioroplast evolutionUnder the assumption that the endosymbiotic origin of organelles is generally accepted(Gray and Doolittle 1982; Taylor 1987), there are two main hypotheses as to the origin of thedifferent chioroplast types described in section 1.3. The idea of an endosymbiotic origin of thechioroplast was first proposed by Mereschowsky (1905) where he suggested that the differentplastid types were the result of endosymbioses of different cyanobacteria-like organisms withnonphotosynthetic phagotrophic protists. These ideas were revised more recently when it wassuggested that the diversity of the plastid types were the direct result of numerous endosymbioticevents with different prokaryotes already divergent in pigment biosynthesis, antennal systemsand biochemical pathways (Sagan 1967; Raven 1970; Whatley and Whatley 1981); this isreferred to as the polyphyletic view of chloroplast evolution. An alternative hypothesis is thatthere was only one primary endosymbiotic event between a cyanobacterium-like organism and aphagotrophic, nonphotosynthetic eukaryotic host. Subsequent divergence of this organism leadto the different plastid types observed in the plants and algae today. This view is referred to asthe monophyletic origin of plastids (Cavalier-Smith 1982; Taylor 1987)I am making a distinction between what I call the primary and secondary endosymbiosesleading to the chioroplasts (see Fig. 1.4). The term primary endosymbiosis refers to theestablishment of a chloroplast from a prokaryotic source. This generally is thought to haveresulted in the generation of an alga with a double membrane around the chloroplast, namely thered and green algae (Fig. 4.1, top). Debate tends to center around whether the chloroplast fromthese two organisms share a common ancestor (monophyletic) or have separate origins(polyphyletic). With the term secondary endosymbiosis, I am referring to events involved in theestablishment of chioroplasts directly from photosynthetic, eukaryotic sources (Fig. 4.1, bottom).This is generally thought to have occurred in the algae where there are more than two membranessurrounding the chloroplast (chromophytes, cryptophytes, Chiorarachnion, euglenophytes andthe dinoflagellates), as reviewed in section 1.3 (see McFadden and Gilson, 1995).35Primary endosymbiosis0PhotosyntheticprokaryotePhagotrophiceukaryote[0®1Photosyntheticeu karyote[0®1Photosyntheticeu karyotePhagotrophiceukaryotePhotosyntheticeukaryoteFigure 1.4Schematic diagram illustrating the proposed acquisition of chioroplasts (c) througha primary or a secondary endosymbiosis. The nuclei of the different organisms areindicated (ni or n2). The nucleus of the photosynthetic eukaryote in a secondaryendosymbiosis (ni) is thought to have given rise to the nucleomorph (if maintained).Secondary endosymbiosis36Traditional classification of the different algal groups was heavily based on their pigmentcontent. This resulted in the separation of three main taxonomic groups: the Rhodophyta (redalgae) with Chi a and phycobilisomes, the Chlorophyta (green algae) with Chl’s a and b, and theChromophyta (coloured algae) with Chi a, c and significant amounts of xanthophylls such asfucoxanthin, vaucheriaxanthin and peridinin. The accessory pigments and the proteins bindingthem have been assigned considerable weight in the speculations as to the number of prokaryotesinvolved in primary endosymbiotic events. The discovery of a chlorophyll a + b containingprokaryote (prochlorophyte) (Lewin 1975) with thylakoid stacking similar to green algae andterrestrial plants was interpreted as evidence for a polyphyletic chloroplast origin. Presumably,the green algae would have acquired a chloroplast from a prochlorophyte (or ‘green’ prokaryote)(Raven 1970). Moreover, the red algae were thought to have acquired a chloroplast from acyanobacterium based on the presence of phycobilisomes in both these organisms. Along thesame line of reasoning, the chromophytes were thought to have separately acquired a chloroplastfrom a ‘yellow’ prokaryote (Raven 1970) with chlorophylls a and c (Whatley and Whatley 1981)or from the anaerobic photoheterotrophic eubacterium Heliobacterium chiorum (Margulis andObar 1985), based on apparent pigment similarities. These comparisons were the foundationupon which the arguments for a polyphyletic view of chloroplast evolution were built.Though it is generally accepted that the red algal chloroplast evolved from acyanobacterium-like organism, the evolution of the green algal and chromophyte chloroplastfrom Chl a + b and Chl a + c-containing prokaryotes, respectively, is more controversial. Theproposed evolution of the chromophyte plastid from Heliobacteriuin was not widely accepted andhas subsequently been disproved by SSU rRNA analysis (Witt and Stackebrandt 1988). Theevolution of the green algal chloroplast from a prochlorophyte is still debated and will bediscussed in Chapter 3.Gene organization and phylogenetic analyses of chloroplast encoded gene sequencesprovided some evidence for a close relationship between the red and chromophyte chloroplasts.Chloroplast gene organization and gene localization of the ATPase subunits (Kowallik 1993),Rubisco large and small subunit (rbcL/S) operon (Douglas and Durnford 1989), psbA (Dl)37sequences (Winhauer et al. 1991) and other gene clusters (Reith and Munholland 1993;Douglas 1994b) have been primarily used for examining these evolutionary relationships. Inaddition, phylogenetic analysis of the psbA, rbcLIS , tufA (elongation factor Tu), (Morden et al.1992), and plastid 16s rRNA (Douglas and Turner 1991b) sequences also supports therelationship between the red and chromophyte plastids. This would indicate that thechromophyte plastid may have evolved from an association of a red algal-like ancestor with aphagotropic eukaryotic host, leading to the chromophyte lineage.The only clear evidence suggesting that a secondary endosymbiosis occurred was withrecent work on Cryptomonas (Douglas et al. 1991) and Chiorarachnion (McFadden et a!. 1994;McFadden and Gilson 1995). Like the chromophytes, both of these organisms have fourmembranes around the chioroplast. However, the cryptophytes and chiorarachniophytes have amembrane bound, nucleic acid containing organelle (called a nucleomorph) located in the spacebetween the outer two CER membranes and the inner two membranes of the chioroplast envelope(the periplastidal space) (Greenwood et al. 1977; Gillott and Gibbs 1980; Hibberd and Norris1984). It is generally thought that this organelle is the vestigial nucleus of the eukaryoticendosymbiont (Ludwig and Gibbs 1987). In the cryptomonads, the presence of phycobiliproteins(phycoerytherin or phycocyanin) and the storage of starch in the periplastidal space (the formercytoplasm of the putative endosymbiont) lead to the suggestion that the endosymbiont was a redalgal-like ancestor (Gillott and Gibbs 1980). This was supported by an analysis of the nuclearSSU rRNA sequence from the nucleomorph and the nucleus of Cryptomonas . Phylogeneticanalysis showed a loose affiliation of the Cryptomonas nucleomorph sequence with the red algaewhile the cryptomonad nuclear sequence was more distant in the tree (Douglas et al. 1991). Thenucleomorph rRNA transcripts were also localized in the nucleomorph and the periplastidal space(McFadden et al. 1994a). In a similar strategy, the nucleomorph rRNA sequence fromChiorarachnion was localized in the periplastidal space and was shown to be distinct from thenuclear rRNA sequence (McFadden et al. 1994b). The endosymbiont leading to the chioroplastof Chiorarachnion has been hypothesized to be a green alga (Hibberd and Norris 1984);however, phylogenetic studies were inconclusive (McFadden et al. 1994b).381.7 Methods used in molecular phylogenyThere are only a few reports of phylogenetic analyses of the Cab genes. Demmin et al.(1989) examined the relationships of the LHC II Cab gene family within the angiosperms. Fromtheir maximum likelihood analysis of Lhcbl (LHC II type I) and Lhcb2 (LHC II type II)sequences, they found that the angiosperm taxa grouped within their traditional taxonomicfamilies. Their trees suggested that the Lhcbl and Lhcb2 divergence occurred prior to themonocot/dicot separation. Matsuoka (1990) suggested the Lhcbl/2 divergence occurred prior tothe angiosperm and gymnosperm separation. Jansson (1994) also examined the evolutionaryrelatedness of the CAB proteins in a recent review and found there was a close association ofCP29 (Lhcb4) and LHC I type I (Lhcal). These trees also showed a separation between LHC I(Lhca) and LHC II (Lhcb). Within the LHC II assemblage, the green algal sequences formed aseparate branch from the Lhcbl-3 (LHC II types I-ITT) genes of the terrestrial plants.Unfortunately, little information was given on the alignment, characters utilized or specificsabout the method used in the analysis. As no indication of reliability was given, it is not possibleto judge which relationships may be significant. Recently, an analysis of a FCP from thehaptophyte, Jsochrysis galbana, suggested it was more related to a tomato LHC I sequence thanto a LHC IT sequence; therefore, the FCPs were suggested to have been derived from a LHC I-like sequence, before the appearance of the predominant LHC II antennae (LaRoche et al. 1994).One goal of this study is to examine the evolution of the Cab and Fcp gene families andrelate these gene relationships back to the function of the protein complex (Chapter 5). As well, Ihope to get a better idea of the Cab and Fcp gene relationships in order to determine when theymay have separated in relation to the functional separation of LHC I and LHC IT. Two methodswere used for the determination of phylogenetic relationships amongst the CABs and FCPs:maximum parsimony and distance matrix. Both methods are available in the PHYLIP computerpackage (Felsenstein 1992). Maximum parsimony is a character method based on the principlethat the evolutionary pathway requiring the fewest “steps” is the most likely. Parsimony attempts39to generate a tree which can be explained with the smallest number of mutational events (shortestlength). The characters used in parsimony can be of a molecular or morphological nature. In thisstudy, a character is a specific amino acid (e.g. Ala, Ser, etc.) at a particular position (e.g. the 49thposition of an alignment). Parsimony evaluates each character site separately on all possibleunrooted trees and records the number of changes required to explain the character distribution.The optimal tree is selected by adding the number of changes over each character site, for allpossible trees, and choosing the one with the fewest steps (see reviews by Swofford and Olsen,1991; Hillis et al., 1993 and Stewart, 1993). With large datasets, not all possible treecombinations can be tested as the number of possible trees increases exponentially with anincrease in the number of taxa. For example, with 13 taxa or more there will be over 13 billionpossible trees, making the calculations impractical (Hillis et al. 1993).To overcome this, PHYLIP uses a heuristic algorithm which starts with the first two (orthree) taxa and creates a tree. The remaining taxa are then added individually in a stepwisefashion to all possible positions on the tree. Each of the trees is evaluated at each step and theshortest tree is kept. This process continues until all the taxa have been added to the tree. Inorder to improve the chances of finding the most optimal tree, the branches undergo a series ofglobal rearrangements whereby groups (subtrees) are removed and added back to the tree in allpossible positions. After each addition, the length of the tree is again assessed and the shorterone is retained. This process continues until no further improvements in the tree topology havebeen recorded (Felsenstein 1992). The evaluation of the trees is based on the chosen optimalitycriteria. This refers to the method by which the evolutionary change of the characters is assessedor weighted (Swofford and Olsen 1990). In this study, an amino acid change is weightedaccording to the number of mutations required to explain the substitution, which is based on thegenetic code. This point is explained further in the methods section of Chapter 5.Distance methods are based on pairwise comparisons between taxa, the end result being asingle value which reflects the dissimilarity (or distance) between the two sequences. Distancesare calculated between all possible pairs of taxa and referred to as the distance matrix. With thedistance methods, characters are not considered individually between all taxa as they are in the40character based methods (parsimony). No ancestral state is implied or required as no distinctionis made between derived or ancestral character states (Sober 1988). The distance matrixcalculation can be based on three modes of amino acid substitution in the PROTDIST program,which is included in the PHYLIP package (Felsenstein 1992). In this study, distances werecalculated using the PAM (accepted point mutation) 100 matrix of Dayhoff (1978) which weightsamino acid changes on the basis of their probable occurrence. This is an empirically determinedmatrix calculated through the comparison of cytochrome c sequences from different species.This distance matrix is then used to infer a tree by any of several methods. This study routinelyused the Neighbor-Joining method (Saitou and Nei 1987) for tree construction as described inChapter 5.1.8 Scope of this thesisThis dissertation is primarily concerned with the characterization of the FCP complexesand the genes that encode them. I am also interested in using this information in an analysis ofevolutionary relatedness between the different light harvesting antennae. The FCPs areseparated and immunologically analyzed for structural relatedness to both the CABs and theFCPs in Chapter 2. The immunological analysis of the LHCs in the red alga Aglaothamnionneglectum is presented in Chapter 3. In addition, work done on the red alga, Porphyridiumcruentum, in collaboration with Beth Gantt and Greg Wolfe at the University of Maryland, ispresented in this Chapter. Characterization of the Heterosigma Fcp sequence and an analysis ofthe size and complexity of the nuclear encoded multigene family is presented in Chapter 4.Finally, Chapter 5 involves an analysis of the evolutionary relationships amongst the CABs andFCPs which combines many of the ideas from the previous chapters.41CHAPTER 2Characterization of the light-harvesting proteins from Heterosigma carterae2.1 IntroductionAs mentioned in the general introduction, there have not been any investigations into thechlorophyll-protein complexes of any raphidophycean alga. Even amongst the otherchromophyte algae, there have been relatively few studies characterizing the fucoxanthinchlorophyll proteins as compared to the CABs of terrestrial plants. As I am interested in thesimilarities and differences amongst the diverse antennal systems of the algae, I decided to workon a representative of the raphidophytes, Heterosigina carterae.In the initial part of this study I isolated and characterized the different chlorophyll-proteins contributing to the antennae of this organism. This study used two different methods forthe fractionation of the thylakoid membrane and subsequent characterization of its components,particularly the FCPs. One method was the fractionation of digitonin solubilized thylakoids on asucrose gradient, a method that has been used on chromophyte algae with success in the past(Hiller et al. 1991). The other method involved fractionation of solubilized thylakoids by apartially denaturing SDS-polyacrylamide gel electrophoresis (PAGE). This method has beeninvaluable in the characterization of the terrestrial plant CABs but has, so far, had limitedapplications for the non-green algal LHCs as the complexes are much more unstable under theseconditions.422.2 Materials and Methods2.2.] Heterosigma culturesAn axenic culture of Heterosigina carterae was maintained in an artificial sea watermixture as previously described (Cattolico et al. 1976). The media was prepared by adding saltsto distilled water at the following concentrations : 0.35 M NaC1, 0.02 M MgSO4, 0.021MMgC12, 7.8 mM CaC12, 7.5 mM KNO3, 0.37 mM KH2PO4and 0.37 mM NaHCO3. To this1 M Trizma base pH=7.6 (Sigma) was added to a final concentration of 1.9 mM. Stock A(50 mM Na2EDTA and 9 mlvi FeC13)was added at 0.76 mill followed by stock B (0.29 mMZnC12, 9.7 mMH3B0, 0.12 mM CoC12, 0.24 mM CuC12, 2.5 mM MnC12 and 0.03 mM(NH4)6M07024,again at 0.76 mill. A 0.1 iglml vitamin B12 solution was then added at0.38 mill. The media was made up to the appropriate volume and an additional 42 mill ofdistilled water was added to account for evaporation during autoclaving. Media was autociavedat 12 1°C for 30 minutes. Cells were grown in 1200 ml of media in three litre fernbach flasks.These were grown with continuous agitation on an orbital shaker at 75 rpm. The light was kepton a 12:12 hour light/dark cycle to induce synchronicity (Cattolico et al. 1976). Light levelswere maintained at 60 JiE/m2/min and the temperature was constant at 18°C throughout the lightand dark cycles. Cultures were routinely tested for contamination when inoculated by adding0.5 ml of culture to 5 ml of nutrient marine media (2.0 g nutrient broth and 1.25 g yeast extractper 250 ml artificial sea water). Cell counts were done using a standard hemacytometer. Cellcounts were made after they were killed by adding 0.5% formaldehyde (50 i1/l ml culture)2.2.2 Heterosigma thylakoidfractionationLate log phase cells were harvested at 400 x g for 12 minutes, resuspended in cold0.33 M sorbitol, 1 mM MgC12, 50 mM HEPES pH 7.6 and protease inhibitors (1 mM43phenylmethyl sulfonyl fluoride, 5 mM E-amino-n-caproic acid, 1 mM benzamidine-HC1,1 mg/mI leupeptin). Protease inhibitors were routinely used in solutions, being added from stocksolutions prior to use. Cells were lysed under 4000 kPa (600 psi) nitrogen in a Yeda Press (YedaResearch and Development Co. Ltd. Rehovot, Israel) to release the chloroplasts. Thechloroplasts were separated by differential centrifugation in a swinging bucket rotor at 6500 x gfor 12 minutes at 4°C. Chioroplasts were washed three times in cold 0.1 M NaC1, 5 mM MgCl,20 miM Tricine pH 8.0 (including protease inhibitors) yielding a washed thylakoid fraction.Thylakoids to be used for non-denaturing gel electrophoresis were made up to 10% glycerolprior to quick freezing in liquid nitrogen and storage at -80°C (Allen and Staehelin 1991).Thylakoids used for sucrose gradient fractionation were used fresh and solubilized withdigitonin at a detergent to chlorophyll ratio of 100:1, on ice for four hours with a constant gentlestirring. After centrifugation at 40 000 x g for 30 minutes, the supernatant was loaded onto a0.3 M-1.2 M linear sucrose gradient on top of 1.3 M and 1.6 M sucrose cushions. Sucrosesolutions were made up in 10 mM Tricine pH 8.0 containing 0.05% (w/v) digitonin. Sampleswere centrifuged for 24 hours at 250 000 x g in a swinging bucket rotor at 4°C. Fractions 2 and3 (Fig. 2.1) were precipitated at 40000 x g following dialysis at 4°C in 0.1 M CaC12, 10mMMgC12, 10 mM Tricine pH 8.0 including protease inhibitors (as above). Due to the large amountof detergent at the top of the gradient, Fraction 1 (Fig. 2.1) was pelleted at 100 000 x g followingextended periods of dialysis with many changes of the dialysis buffer. Chlorophyllconcentrations were determined in 90% acetone using the equations of Jeffrey and Humphreyshown below (Jeffrey and Humphrey 1975).Chl a = 11.47 (A664) -0.4 (A630) Chl c+ c2 = 24.36 (A630) -3.73 (A664)2.2.3 Denaturing SDS-PAGE and Western BlottingFor denaturing gel electrophoresis, samples were solubilized in 2% SDS, 65 mlvi TrisHC1 pH 6.8, 50 mlvi dithiothreitol, 10% glycerol and heat denatured for 1 minute at 100°C.44Thylakoid and sucrose gradient fractions were loaded on the basis of chlorophyll. Gel slicesfrom non-denaturing PAGE were incubated in 2X sample buffer (4% SDS, 132 mM Tris-HC1pH 6.8, 0.1 M dithiothreitol, 20% glycerol) at room temperature for 2 hours then heated to 80°Cfor 20 minutes. Polypeptides were separated on 12-16% or 7.5-15% SDS polyacrylamide gels(acrylamide: bis-acrylamide 37.5:1) containing 0.05% SDS and 1.32 M Tris-HC1 pH 8.8, with a2cm stacking gel containing 5% acrylamide, 0.1 M Tris-HC1 pH 6.1 and 0.1% SDS. Gels wererun for 18 hours at 4°C with the Laemmli buffer system (Laemmli 1970). Coloured molecularmass standards (Amersham) were used to estimate molecular mass.Proteins were electrotransferred to nitrocellulose in 50 mlvi sodium acetate pH 7.0overnight at 200 mA and 4°C. As a guideline, 115th the amount of chlorophyll loaded on gels tobe stained was used for the same samples destined to be used for western blotting. Westernblotting was carried out as previously described (White and Green 1987). Western blots werereblotted after stripping the nitrocellulose membrane in 0.1 M glycine-HC1 pH 2.2, 20 mM Mg-acetate, 50 mM KC1 (Legocki and Verma 1981) followed by reblocking in 3% Hipure liquidgelatin (Norland Products Inc. New Brunswick, N.J.) in phosphate buffered saline (1.37 M NaCl,27 mM KC1, 81 mM Na2HPO4,15 mM KH2PO4,pH 7.4). Proteins were Coomassie stained for2 hours using 0.1% Coomassie brilliant blue-R, 50% methanol and 7% glacial acetic acid.Gels were destained in several changes of a 20% methanol, 7% glacial acetic acid. Polyclonalantibodies used in this study were anti-CP1a (c-CP1a), specific for barley CP I plus LHC I(White and Green 1987) and anti-FCP (c-FCP), specific for Phaeodactylum tricornutumfucoxanthin-chiorophyll a + c (FCP) protein complex (Fawley and Grossman 1986), which wasprovided by Dr. Art Grossman. Other antibodies include the ct-PsaD antibody specific for a PS Iassociated subunit (also called PS I subunit #2) (Bengis and Nelson 1975) and ci-D1, specific forthe PsbA polypeptide of PS II, provided by L. McIntosh.452.2.4 Non-Denaturing Gel SystemThylakoids were solubilized with a mixture of 0.9% octyiglucoside, 0.9% decylmaltoside and 0.2% lithium dodecyl sulphate (in 2 mM Tris-maleate pH 8.0, 10% glycerol andwith protease inhibitors) and resolved on a non-denaturing gel system according to Allen andStaehelin (Allen and Staehelin 1991) except that a 7% acrylamide gel was used with anacrylamide to bisacrylamide ratio of 150:1. A stacking gel was not used because it resulted indegradation of the pigment-protein complexes. Samples were solubilized on ice for 30 minutesat an anionic detergent to chlorophyll ratio of 30:1 with occasional mixing, then centrifuged inan microfuge for 20 minutes at 4°C. Samples were electrophoresed at 10 mA for 1.5 to 2.5hours at 4°C. Estimations of molecular mass were done using non-denatured, high molecularmass markers (Pharmacia). Gel bands were excised and electrophoresed on a denaturinggradient gel as described above. Samples to be used for fluorescence data were excised from thegel and quick frozen in liquid nitrogen prior to storage at -80°C.2.2.5 Spectroscopy and Fluorescence measurementsAbsorption spectra were recorded on a Cary 210 Spectrophotometer at room temperature.P700 content was measured from the sucrose gradient fractions directly by monitoring therecovery of absorption at 700 nm after photo-oxidation by saturating red light with 1.7 mMascorbate and 0.075 mM methylviologen present in the reaction mixture (Marsho and Kok1972).Fluorescence emission spectra were recorded with a Perkin Elmer LS5O fluorometer withthe 77°K low temperature attachment and red sensitive photomultiplier. Excitation wavelengthwas 440 nm and the excitation and emission slit widths were adjusted to 10 nm and 5 nm,respectively. Spectra shown are an average of three scans. A 530 nm cut off filter helped toremove Rayleigh scatter in the 620 nm range. Gel slices from the non-denaturing gel system46were fitted in the cuvette with 60% glycerol and frozen in liquid nitrogen prior to measuring.Emission spectra were corrected for the drop in photomultiplier sensitivity in the 600-800 nmrange using an averaged correction factor provided by Perkin Elmer. Excitation spectra wererecorded from similarly prepared samples at 77 K. Emissions from the excitation spectra weredetected at 680 nm in all samples. The excitation and emission slit widths were 2.5 nm and10 nm, respectively. Scan rate was 300 nmlminute and the spectra shown are an average of twoscans.2.3 Results2.3.] Fractionation ofdigitonin-solubilized membranes by sucrose gradient centrifugationThylakoid membranes solubilized with digitonin were resolved into three major fractionson a sucrose gradient (Fig. 2.1). The top dark brown fraction (fraction 1) was rich in fucoxanthinand chlorophyll c as demonstrated by a broad shoulder from 488-540 nm and a prominentshoulder at 460 nm, respectively (Fig. 2.2A). Fraction 1 was removed from the 21% sucroselevel and contained approximately 53% of the total chlorophyll. It also showed visible redfluorescence upon excitation with long wavelength UV light, indicating the detachment of thelight-harvesting complex from the reaction center. A Chi a emission maximum of 675 nm, witha secondary peak at 732 nm, was recorded (Fig. 2.2B, solid line). A small peak at 637.5 nm isprobably the result of the partial uncoupling of Chl c fluorescence which is preventing completetransfer of excitation energy from Chi c to Chl a. Residual P700 activity was detected infraction 1, giving a Chi aIP700 ratio around 1200.47SucroseConcentration____Appearance Chi a I c Total Chi (%) Abs. Maxima0.2M-nfl1 Brown 4 53 672, 440, 460*,488*2 Green-brown 14 25 677, 437,460*,: } 3 light-brown 10 11 675, 437,460*,1.6 MFigure 2.1Schematic representation of sucrose gradient fractionation of digitonin solubilizedthylakoids, with Chi a / c ratios, percentage of total chlorophyll in each fraction andabsorption maxima data for the three major fractions. Asterisk in absorption dataindicates a shoulder.Fractions 2 (30% sucrose level) and 3 (34% sucrose level) contained 25 and 11% of thetotal chlorophyll, respectively (Fig. 2.1). Both contained significant amounts of carotenoid asthere is a prominent absorption at 496 nm in both fractions (Fig. 2.2A). An absorbance shoulderat 460 nm also indicates the presence of Chl c though the Chl a/c ratios were 14 and 10 forfraction 2 and 3, respectively. This indicated significant amounts of antennae were stillassociated with the complexes. Both fractions (2 and 3) were enriched in PS I with Chl a / P700ratios of 340 and 420. Both fractions 2 and 3 had emission maxima at 687 nm and significantshoulders at 717 nm, though fraction 2 had the greatest fluorescence in the 717 nm region48Aci)C-)Ccci0C))-Dci)>cclci)ci)0Cccl.00Cl).0ci)>Cclci)600 650 700 750 800Wavelength (nm) Wavelength (nm)Figure 2.2Spectral characteristics of the sucrose gradient fractions. (A) Room temperatureabsoption and (B) fluorescence emission spectra. Spectra are fraction 1 (.—), fraction 2 (---),fraction 3 (). A room temperature absorption spectrum (C) and fluorescence emissionspectrum (D) of thylakoid membranes is also shown. Emission spectra were excited with437 nm light at 77°K.C600 650 700 750wavelength (nm) Wavelength (nm)8004946-.3014.3 -—28.0—19.5—18.58.0—-— 7.5Figure 2.3Polypeptides of digitonin sucrose gradient fractions separated by denaturing 12-16%gradient SDS-PAGE. (A) Stained with Coomassie blue. Thy.: Whole thylakoids; 1:Fraction 1; 2: Fraction 2; 3: Fraction 3. (B) Western blot immunoprobed with the o-FCPantibody. Lanes labeled as in panel A. MW, molecular mass standards in kDa.Arrowheads on left of Panel B denote faint bands at 28.0, 17.0 and 17.5 kDa in wholethylakoids (Thy.) only.(Fig. 2.2B). Fluorescence in the 717 nm area is attributed to fluorescence from PS I, indicatingthat fractions 2 and 3 are enriched in PS I. The strong emission at 687 nm is probably from theantennae complexes associated with this fraction. In contrast, thylakoids had a fluorescenceemission maximum at 691 with a broad shoulder towards 740 nm (Fig. 2.2D). RoomMWAThy. 3 2 1 Thy.B1 2 3 MW21.5 —50temperature absorbance peaks at 440 and 674 nm, along with prominent shoulders at 460, 492and 534 nm, were also characteristic of the thylakoids (Fig. 2.2C). Thylakoids solubilized withdigitonin have a blue shifted fluorescence emission maximum that is more susceptible tochanges in the assay buffer conditions (not shown). Because of this, fluorescence emissionmaxima have to be interpreted with caution.SDS-PAGE (Fig. 2.3A) showed that there was a single polypeptide in Fraction 1. Itcross-reacted with an antibody specific for the FCP from the diatom Phaeodacrylum tricornutum(Fig. 2.3B). The purity of this fraction was ideal for obtaining tryptic fragments from the FCP,as will be discussed in Chapter 4. Though one polypeptide was usually found in this fraction,some extractions removed smaller amounts of the other three main FCP polypeptides, that areobvious in the thylakoid fraction (Fig. 2.3B).Fractions 2 and 3 showed a similar polypeptide pattern with a number of bands of 16-22 kDa, a sharp band at 37 kDa and diffuse bands in the 49-55 kDa range (Fig. 2.3A). The fourpolypeptides estimated as 20.5, 19.5, 18.5 and 18.0 kDa cross-reacted strongly with the o-FCPantiserum (Fig. 2.3B). Two minor polypeptides at about 17.5 kDa and 16.5 kDa were faintlyimmunostained (lower arrowheads). The FCP antibody also detected a polypeptide with anapparent molecular mass of 28 kDa, found only in the thylakoid fraction (upper arrowhead, seealso Fig. 2.4A). Note that in this Figure and subsequent figures, the apparent molecular massesdetermined by SDS-PAGE are used as labels to identify distinguishable polypeptides, and arenot meant to imply accurate molecular mass determinations.Using an antibody specific for barley CP 1 a (PS I core complex plus its correspondinglight-harvesting polypeptides (White and Green 1987)), a different subset of cross-reactingpolypeptides was found (Fig. 2.4). The c-CP1a antibody detected five major bands atapproximately 16, 17.5, 18.5, 19 and 21.5 kDa in fractions 2 (Fig. 2.4C, left) and 3, while anadditional four bands of about 18, 20, 22, 28 kDa could be resolved in thylakoids (Fig. 2.4A,left). To identify which of the polypeptides reacting with the c&CP1a antiserum were alsoimmunostained with the a-FCP antiserum, the immunoblots shown in the left panels of Figure512.4A-C were stripped and reblotted with the o-FCP (shown in Fig. 2.4A-C, right panels). Thebands of 20.5, 19.5, 18.5, and 18.0 that were prominent in thylakoids immunostained oniy withcx-FCP (Fig. 2.3) are heavily stained in Figure 2.4A (right panel) and are clearly distinguishedfrom the bands labeled 21.5 and 22.0 above them and the 16.0 band below which only crossreacts with the cx-CP1a antiserum. Similar results were obtained with fraction 2 when it wasreblotted with o-FCP (Fig. 2.4C, right panel). Note that the major light-harvesting polypeptidein fraction 1, which was immunostained with the cx*FCP, was not detected using the x-CP1a(Fig. 2.4B). Only four polypeptides at 28, 18.5, 18 and 17.5 kDa appeared to react with bothantisera. Results of the immunoblotting with the two antisera are summarized in Table 2.1.These results show that there are up to 12 polypeptides in the FCP/CAB family in Heterosigina,a larger number than previously reported for any chromophyte alga.52Western blot of sucrose gradient fractions immunoprobed with c-CPla (left panels) thenstripped, blocked and immunoprobed with c-FCP (right panels) on the same blot. (A)Whole thylakoids (B) Fraction 1 from sucrose gradient (C) Fraction 2 from sucrosegradient. Approximate molecular masses (kDa) are used as labels to distinguishindividual bands.Table 2.1 Summary of cross-reactivity with the o-CPla and o-FCP antiseraMolecular Mass (kDa)antiserum 28.0 22.0 21.5 20.5 20.0 19.5 19.0 18.5x-CP1a ± + + - + - + +*x-FCP* An apparent single band at 1-18.5 kDa may be a doubletBOP1 aCP1a +FOPMWCCP1aCP1a +FOPMW—28.0-- /20.5_z’9•5—18.517.0AMW OPla OPlaFOP28.0—22.019.017.5//’ ‘16.0”Figure 2.4MW21.5\\p20.5____190\\--;-—19.5 18.5— --18.518017.517.516.0”‘‘17.0+ - - + - + - + + + +18.0 17.5 17.0 16.0+* +-+C,-,3-,2.3.2 Fractionation by Non-Denaturing PAGEPigment-protein complexes were isolated from Heterosigma thylakoids solubilized in0.9% octylglucoside, 0.9% decyl maltoside and 0.2% lithium dodecyl sulphate, and separated bymeans of the non-denaturing gel system of Allen and Staehelin (Allen and Staehelin 1991) (Fig.2.5). With this system, eleven pigment-protein complexes were resolved (Fig. 2.5A). Bands 1and 10 were resolved into two Complexes (ib, lOb) with longer periods of electrophoresis(Fig. 2.5B), but this resulted in some degradation of the central pigment-protein complexes. Thefirst ten pigment-protein complexes were green and lacked noticeable fucoxanthin whileComplex 11 was a brown fraction making up approximately 40% of the total protein.Figure 2.5Unstained 7% polyacrylamide gel separating pigment-protein complexes of thylakoidssolubilized with 0.9% octylgiucoside, 0.9% decyl maltoside, 0.2% lithium dodecylsulfate (Allen and Staehelin 1991). (A) Electrophoresed for 1.5 hours at 10 mA at 4°C.(B) Similar gel electrophoresed for 2.5 hours. Apparent molecular masses are given inkDa.A B54AMW M 1 2 3 4 5 6 7 8 9 10 11 Thy--46IMW 1 2 3 4[53_ nw’ri5 6 7 8 9 lOalObliicpIFigure 2.6Analysis of isolated pigment-protein complexes. (A) 7.5-15% denaturing SDS-PAGE ofpigment-protein Complexes 1-11 separated from non-denaturing PAGE. (B) Westernblot analysis of the same pigment-protein complexes. Antibody used in sequentialimmunoprobing and area of cross reactivity indicated on the right; Molecular masses inkDa on the left. The bottom panel (CP I) is a western blot of the same gel fractions withthe cL-CP1 (PS I core complex) antiserum.3014.3BThy 1 2 3 4 5 6 7 8 9 lOa lOb 11 Thy322114 psaD55Denaturing SDS-PAGE of Complexes 1-3 (Fig. 2.6A) showed a broad stained band ofabout 53 kDa, a 37 kDa band, and a number of sharper bands in the molecular mass range (10-21 kDa) typical of the non-pigmented subunits of PS I (Golbeck and Bryant 1991). Animmunoblot of samples from a similar gel was sequentially probed with several antibodies todetermine the composition of the various bands in Figure 2.5. Immunodetection of the PS I-D(psaD gene product, Fig. 2.6B), and CP I polypeptides (Fig. 2.6, bottom panel) indicated thatpigment-protein Complexes 1-3 were PS I complexes. PS I core polypeptides were also detectedin Complex 4 but were lacking in the other fractions. When blots were probed with the c-FCPantibody, there appeared to be a number of antennal polypeptides in the 17-20.5 kDa rangeassociated with the PS I core complex (Fig. 2.6B). The immunologically detected LHCs inComplexes 1-3 had the same molecular mass as those in Complex 11 and may be similar to theLHC polypeptides found in sucrose gradient fraction 2 and 3. However, unlike sucrose gradientfractions 2 and 3, there did not appear to be any fucoxanthin or Chl c associated with Complexes1-3 (Fig. 2.7). This may have been the result of the detergents used in the extraction and theelectrophoretic forces partially denaturing the complexes but allowing some LHCs to remainassociated with the PS I core.Complex 1 had the long wavelength absorption maximum at 677 nm typical of PS I(Fig. 2.7A) and a long wavelength chlorophyll a fluorescence emission maximum at 717 nm(Fig. 2.7B), similar to that attributed to the PS I specific light-harvesting complex of land plants(Haworth et al. 1983; Murata and Satoh 1986). It also had a second fluorescence emissionmaxima at 676 nm which may indicate uncoupled chlorophyll resulting from the detergenttreatment (Fig. 2.7B). Excitation spectra of Complex 1 showed a Chi a peak at 437 nm and asecond minor peak at 535 nm (Fig. 2.7C).Complexes 6- lOa appeared to be PS TI-related, as they had a number of polypeptidesmigrating as somewhat diffuse bands in the 30-50 kDa range. An antibody specific to Dl (PS IIreaction center polypeptide encoded by psbA gene) (Fig. 2.6B) showed that it was found in allthese fractions but was absent from Complexes 1-5. Complexes 5-7 had long wavelength560)0Cci)0(I)0)1_0DUa)>ccici)Bci)0Cci)0C’)ci)0DLL0)>ccici)400 450 500 550Wavelength (nm)600 650 700 750Wavelength (nm)600800Figure 2.7Spectral characteristics of pigment-protein Complexes 1, lOa, and 11 in gel slices fromthe non-denaturing SDS-PAGE. (A) Room temperature absorption spectrum. Spectra areoffset for blarity. (B) Fluorescence emission spectra at 77 K of pigment-proteincomplexes. Spectra are Complex 1 (—); Complex lOa( ); Complex 11 ( ). (C)Fluorescence excitation spectra of the same three complexes and thylakoids at 77 K.Emission was detected at 680 nm.ci)0Ccci-Q0C,)-Qa)>ccici)400 500 600 700Wavelength (nm)C57maxima at 674 and Complexes 8-lOb at 671 nm (data not shown; except lOa, Fig. 2.7).Complex lOa had a fluorescence emission maximum at 686 nm (Fig. 2.7) which is typical ofPS II core complexes (Murata and Satoh 1986; Brown 1988). An excitation spectrum ofComplex lOa has a main Chl a peak at 437 nm and minor shoulder at approximately 470 and502 nm, indicating minor amounts of Chl c and a carotenoid (probably fucoxanthin). Thesepeaks are probably from minor amounts of FCPs detected in this fraction (Fig. 2.6B). Variabilityin the resolution of pigment-protein Complexes 4-9 was observed between different preparationsof thylakoids, possibly the result of different cell culture densities which could have altered lightconditions.Complex lOb, detectable after longer periods of electrophoresis, was enriched in the28 kDa polypeptide that cross reacted with the cx-FCP (Fig. 2.6B). It had absorbance peakstypically associated with Chi a, Chl c, and fucoxanthin but since it also contained some of thelower molecular mass light-harvesting polypeptides, I cannot say whether the 28 kDapolypeptide binds these pigments or not. This polypeptide may be analogous to the 31 kDapigment-protein complex containing only Chi a characterized from another chromophyte,Ochromonas (Gibbs and Biggins 1991).Complex 11 contained the majority of the fucoxanthin and had a low Chl a/Chl c ratio asestimated from the peaks at 440 nm and 460 nm respectively (Fig. 2.7A). It appeared to besimilar to fraction 1 from the sucrose gradient, but was not as pure. Rather than a singlepolypeptide, it had the three major light-harvesting polypeptides at 20.5, 19.5 and 18.5 kDa thatcross reacted with the diatom o-FCP antibody. Moreover, polypeptides in the 2 1-33 kDa rangewere also weakly stained with Coomassie blue in this fraction (Fig. 2.6A). One-third the amountof Complex 11 was loaded onto the gel in Figure 2.6 in order to prevent overloading; therefore,comparison of absolute protein amounts can not be made. Complex 11 had a fluorescenceemission maximum at 681 nm at 77°K (Fig. 2.7) which is comparable to the LHC II of landplants (Murata and Satoh 1986). The excitation spectrum of Complex 11 has peaks at 437 and460 nm from Chl a and Chi c, respectively. As well, a broad area from 480 to 540 nm results58from excitation energy transfer of xanthophylls (fucoxanthin) to Chi a (Fig. 2.7C). Thisindicates that the accessory pigments are still coupled to Chl a. Although some Dl isimmunodetected in Complexes lOb and 11, it would be premature to conclude that the 28 kDapolypeptide or any of the FCPs are preferentially associated with PS II, as it is impossible to ruleout comigration of individual polypeptides in this region of the gel.An orange free pigment zone migrated just ahead of Complex 11. Absorbance spectraindicated that it contained carotenoids and a small amount of chlorophyll (data not shown). Nopolypeptides were detected following Coomassie staining.2.4 DiscussionSucrose gradient separation following digitonin solubilization has been successfully usedin the isolation of light-harvesting complexes from a number of algae (Berkaloff et al. 1990;Hiller et al. 1991; Arsalane et al. 1992). This fractionation technique usually results in a light-harvesting antennal fraction at the top of the gradient and a few additional denser pigment-protein complexes. I have found that the raphidophycean alga, Heterosigma, like otherchromophyte algae, has a predominant fucoxanthin-chiorophyll a/c pigment-protein complexreleased by digitonin solubilization. This complex has a single polypeptide with an apparentmolecular mass of 19.5 kDa and spectral characteristics comparable to the predominant LHCfrom other chromophytes (Hiller et al. 1991). It appears to be the more abundant of fourpredominant FCPs in the thylakoids as determined immunologically with the ct-FCP antiserum.Since it is easily dissociated from the core complexes, this suggests it may be peripherallylocated and possibly analogous to land plant LHC II. In addition, the 19.5 kDa polypeptide ispreferentially removed in most digitonin extractions, suggesting it is even more distal to theother predominant FCPs. In contrast, the FCP fraction obtained by non-denaturing SDS-PAGE59(Complex 11) appeared to contain all the major FCP polypeptides. The detergent used for thesolubilization of thylakoids in the non-denaturing gel system were obviously more penetrating’and this resulted in a more vigorous extraction of the FCPs.The denaturing SDS-PAGE system used allowed for the resolution of up to 12 separatecross-reacting LHC related polypeptides in Heterosigina. They were in the same size range (15-22 kDa) as those reported from other chromophytes (Hiller et al. 1991). Most published workhas reported one to four light-harvesting polypeptides (Hiller et al. 1991), although as many assix polypeptides from four chromophyte species have been reported to cross react with anantibody raised to Chiatnydoinonas (Chlorophyceae) LHC (Plumley et al. 1993). Differences inthe number of polypeptides detected may partly be due to the different electrophoretic systemsused to resolve the complexes and the nature of the antiserum. In order to rule out the possibilitythat some of the immuno-reactive bands were the result of proteolytic cleavage of largerpolypeptides, Heterosigina thylakoids were isolated and incubated at 37°C, in the presence orabsence of protease inhibitors, with no difference in the number of LHC bands detected. Wholecells solubilized directly in 2X SDS sample buffer also showed the same pattern as thylakoids(data not shown). The FCP antenna family in Heterosigma may therefore be as complex as theCAB antenna family in land plants (Green et al. 1991; Green et al. 1992).It is also interesting to note that these light-harvesting polypeptides were detected usingLHC specific antibodies from a different class in the Chromophyta and antibodies fromterrestrial plants. Other studies using antibodies specific for land plant and chiorophyte LHCsshow cross-reactivity with various members of the Chromophyta (Caron et al. 1988; Passaquet etal. 1991; Plumley et al. 1993); others show cross-reactivity within the Chromophyta (Fawley etal. 1987). These results indicate the presence of commonly conserved antigenic determinantsassociated with all light-harvesting polypeptides, suggestive of a common evolutionary origin.This structural similarity is confirmed by sequences of the FCP genes from the diatom,Phaeodactylum tricornutum (Grossman et al. 1990). Protein sequences of tryptic fragmentsfrom a Cryptomonad (Sidler et al. 1988), the dinoflagellate Amphidinium (Hiller et al. 1993) and60from Heterosigma (Green et al. 1992, Chapter 4) also demonstrate the apparent structuralsimilarity between the Chi a + b binding proteins and the Chl a + c binding proteins.Solubilized chromophyte thylakoids have previously been fractionated by non-denaturingPAGE, especially in sodium deoxycholate (Caron and Brown 1987; Brown 1988) or Deriphat160 gel systems (Boczar et al. 1980; Peyriere et al. 1984; Boczar and Prezelin 1989; Knoetzeland Rensing 1990). I was unable to obtain satisfactory results with either of these gel systems orwith the non-denaturing gel systems used successfully with land plants (Camm and Green 1989;Thornber et al. 1991). This indicates the labile nature of these complexes as compared to theLHCs of terrestrial plants; a problem that is impeding characterizations of the number and typesof chlorophyll-proteins in the chromophytes. However, a modification of the non-denaturing gelsystem devised by Allen and Staehelin (Allen and Staehelin 1991) proved to be successful in theseparation of Heterosigma pigment-protein complexes, allowing the preservation of a number oflarge complexes with apparent molecular masses of over 200 kDa. This system represents animprovement over other electrophoretic separation techniques for chromophyte algal pigment-protein complexes, being able to separate several PS I fractions, a number of PS II fractions, anda dominant LHC fraction.An important feature of the non-denaturing electrophoretic separation technique is theability to isolate light-harvesting antennal proteins still associated with the core complexes.Complexes 1-3 were PS I fractions and appeared identical except that the slowest migratingappeared to have larger amounts of associated light-harvesting polypeptides. Complexes 2 and 3lacked the majority of the LHC and tended to retain high levels of the lower cross reactingpolypeptide (17 kDa) suggesting it may be closely associated with the PS I core complex.Previous studies on isolated PS I complexes of chromophyte algae found a 715-720 nmfluorescence emission peak which is usually assumed to be due to PS I reaction center inassociation with its light-harvesting antenna (Brown 1988; Berkaloff et al. 1990). As well, aPS I specific antenna with different fluorescence emission characteristics has been identified inthe xanthophyte alga, Pleurochioris (BLichel and Wilhelm 1993). It appears that the presence of61a light harvesting complex associated with PS I is a common feature of the Chromophyta(Berkaloff et al. 1990), as is the case with green algae and land plants where a number of uniqueantennal proteins in the size range of 2 1-24 kDa are specifically associated with PS I (Mullet etal. 1980; Haworth et al. 1983; Lam et al. 1984b).Complexes 1-3 do contain associated LHC-related polypeptides though the total amountis low compared to the overwhelming occurrence of these polypeptides in Complex 11. Inaddition, these polypeptides do not appear to bind significant amounts of fucoxanthin or Chi c.As these polypeptides are of comparable sizes to the other LHCs in Complex 11 and inthylakoids, it would seem that these polypeptides are partially denatured and have lost theaccessory pigments without being removed from their association with the core complex. Thismay be a result of the detergents used, the forces exerted during electrophoresis, or acombination of both. This would not be too surprising as the LHCs are quite susceptible todegradation. Because of the possibility of a nonspecific interaction with the high molecular masscomplex and an inability to distinguish between LHCs associated with this PS I fraction andthose in the LHC fraction (Complex 11), I can not yet assign any specific LHC polypeptidesexclusively to PS Tin Heterosigma.The ability to resolve a number of PS I and PS II complexes is comparable to the resultsobtained with the green alga, Chiarnydomonas, (Allen and Staehelin 1991) though there aredifferences in the associations of the LHCs with these complexes. The number of LHCsresolved also appears to differ, illustrating the differences between the chromophytes, greenalgae and land plants. At the present time I am unable conclude, with any certainty, the nature ofthe complex organization in the thylakoid with regard to the localization of PS I and PS II orwhether the pigment-protein complexes separated on the non-denaturing gel system representdifferent environments within the thylakoid membrane.The recent immunocytochemical localization of FCP and PS I complexes within thethylakoids of members of the Chromophyta (Lichtlé et al. 1992; Pyszniak and Gibbs 1992;Lichtlé et al. 1992b) show that PS I and PS II are not as highly segregated within appressed and62non-appressed regions as they are in terrestrial plants. Though there was only a slight preferencefor PS Tin the ‘nonappressed regions of the chromophyte thylakoid, the FCPs werehomogeneously distributed throughout all parts of the membrane. In Heterosigma, there is anassociation of a large number of antennae polypeptides with the PS I enriched fractions from thesucrose gradient. It is unlikely that these are all specific PS I antennae as most of the prominentLHC polypeptides detected in thylakoids are present in these fractions. At the moment it is notclear which of these polypeptides may be preferentially associated with PS Ito the exclusion of aPS TI association. The consistent association of the FCPs with these lower fractions suggests thatthe separation of the main antennae is not as sharply defined as is the case with LHC I andLHC II in terrestrial plants. The homogeneous distribution of the FCPs in appressed andnonappressed thylakoid regions of chromophytes may explain their prevalent association withthe PS I enriched fractions of the sucrose gradient. This would agree well with work showingthat the excitation energy captured by the main antennae of a diatom was equally distributed toboth photosystems (Owens 1986b). However, this remains speculative at the moment since theassociation of the FCPs with PS I may be a result of contamination during the fractionationprocedure.63CHAPTER 3An immunological characterization of LHC related-polypeptides in red algae3.1 IntroductionThis Chapter is concerned with the immunological characterization of LHC proteins fromtwo red algae. In this Chapter I will first describe the immunological analyses I did withAglaothamnion. This will be followed by the immunological work done with Porphyridium incollaboration with Beth Gantt’s group at the University of Maryland. Greg Wolfe, in the lab ofBeth Gantt, was the first to demonstrate that a PS I fraction from the red alga Porphyridiumcontained the core complex and an array of smaller polypeptides in the 11-24 kDa range, typicalof the PS I polypeptide distribution of chlorophytes and terrestrial plants (Wolfe et al. 1992;Wolfe et al. 1994b). In his work, PS I and PS II fractions from Porphyridiuin were isolated. Thespectral characteristics and the immunological detection of D2, CP43 and CP47 were used toidentify the PS II fraction. The discovery of a putative LHC I complex in Porphyridium lead toa collaboration with our lab in order to examine the immunological relatedness of these red algalchlorophyll-proteins to other antennae. This collaboration showed that these polypeptides wereindeed structurally related to the CABs and FCPs of the terrestrial plants and diatoms,respectively (Wolfe et al. 1994). In addition, I examined a more distantly related red alga,Aglaotharnnion neglectum, to look for similar immunologically related polypeptides todetermine whether the CAB-related antennae were a general occurrence amongst the red algaeor, alternatively, if they were unique to the primitive red algal class represented byPorphyridiurn.Light-harvesting polypeptides resembling the CABs of the Chi a + b-containing64organisms have not been previously discovered in the red algae. Many of the earlier attempts atisolating pigment-protein complexes involved solubilization of the thylakoids with SDSfollowed by polyacrylamide gel electrophoresis. This frequently resulted in the release of aconsiderable amount of free chlorophyll, though a PS I fraction usually remained intact (Hillerand Larkum 1981; Redlinger and Gantt 1983). More recently, a PS I fraction was isolated fromthe unicellular red alga, Cyanidium caldarium, and was found to contain a high molecular massband and four smaller polypeptides in the 13-18 kDa range (Yurina et al. 1991). In a separatestudy, the thylakoid membrane composition of Porphyridiurn was analyzed by fractionation ofdetergent solubilized thylakoid membranes on a sucrose gradient. This resulted in the separationof PS I and PS II fractions (Marquardt and Ried 1992). Neither of the above two studiesreported the occurrence of LHC polypeptides associated with either PS I or PS II. The latterstudy used a partially-denaturing gel system to examine the thylakoid composition which wouldnot be expected to resolve the LHC polypeptides. In addition, the lack of LHC detection ineither of these studies may have been due to the degradation of the complexes as a result of themethods used to fractionate the thylakoids (Wolfe et al. 1992).3.2 Materials and Methods3.2.] Aglaothamnion culturesAglaotharnnion neglectum Feldmann-Mazoyer is a filamentous red alga (classRhodophyceae) originally collected off the shores of Hawaii and belonging to the subclassFlorideophycidae—order Ceramiales (in the family Ceramiaceae). An axenic culture ofAglaothamnion was provided by Dr. Kirk Apt to whom I am grateful. These cultures weremaintained in the same artificial sea water medium described in Chapter 2. No difference wasseen when Aglaothamnion was cultured in Provasoli’s enriched sea water media as described byMagruder (1984), so the artificial media was used. These algae were kept on a 16 hour light: 865hour dark cycle at 24°C and 30 jJE/m2/min, with constant bubbling of air through 0.2 jimmillipore filters to maintain an axenic culture. The cultures were kept in six litre flasks with fourlitres of media. Bubbling air at 4000 cm3/min with an aquarium pump was sufficient to keep thecultures agitated. In order to subculture the algae and to maintain an even growth of the tissue,the algae were fragmented in a sterilized blender for 10-15 seconds before inoculating freshcultures.3.2.2 Aglaothamnion neglectum thylakoidfractionationAglaothamnion was harvested by pouring the culture through four layers of cheese cloth.Thylakoids were prepared using a modification of a method used for Porphyridiurn (Wolfe et al.1992). The material was quick frozen in liquid nitrogen and ground to a fine powder in a prefrozen mortar. This powder was resuspended in cold 50 mM NaPO4pH 7.0, with the proteaseinhibitors (1 mM phenylmethyl sulfonyl fluoride, 5 mM E-amino-n-caproic acid, 1 mlvibenzamidine-HC1, 1 mg/mI leupeptin). The extract was then put through a pre-chilled Frenchpress at 1300 psi. This was repeated three times and the effluent was kept on ice between runs.The extract was centrifuged at 1000 x g for 10 minutes at 4°C to remove unbroken cells andcellular debris. The supernatant was then centrifuged at 28 000 rpm in a SW28 rotor for 45minutes at 4°C. The green pellet was resuspended in cold 10 mM NaPO4pH 7.6, with proteaseinhibitors (PIs) included in the buffer. The fraction was then centrifuged through a sucrose stepgradient (0.5 M/ 0.8 M/ 1.6 M sucrose steps in 10 mM NaPO4pH 7.0, plus PIs) at 27 000 rpm ina SW28 rotor for three hours at 4°C. The pellet was removed and resuspended in cold 10 mMNaPO4, 150mM NaC1, plus PIs. If necessary, the sample was then quick frozen in liquidnitrogen and stored at -80°C for use at a later time.Thylakoids were resuspended in 4 ml ice-cold 1 M NaBr and centrifuged at 15 000 rpmin a SS34 rotor for 10 minutes. The pellet was resuspended in 1 ml cold 50 mM HEPES pH 7.4,10 mM NaCl, 10% glycerol, and 0.75 M sucrose (+PIs). Thylakoids were solubilized in3-Dodecylmaltoside (13-DM) at a detergent:chlorophyll a ratio of 24:1 for 2 hours at 4°C, in the66dark. Solubilized thylakoids were diluted 1:1 with 50 mM HEPES pH 7.4 (+ PIs) and loadedonto a linear 15-30% (w/v) sucrose gradient (in 50 mM HEPES pH 7.4, 10 mM NaC1,10% glycerol, 0.04% (wlv) 13-DM and PIs). The gradient was centrifuged in a SW41 rotor at36 000 rpm for 20 hours. Pigmented bands were dialyzed against 2 x 2 liters of ice-cold 50 mMHEPES pH 7.4, 10 mM NaCl plus PIs. Samples were stored at -80°C until required.Chlorophyll concentrations were determined in N-N-dimethyl formamide using the followingequation: ChI =(A664) 10.65 (a=83.9 mM1cm)(Moran 1982).3.2.3 SDS-polyacrylamide gel electrophoresisAglaothainnion thylakoid samples were denatured as described in Chapter 2. Proteinswere separated on 14% acrylamide gels (acrylamide: bis-acrylamide 37.5:1) containing 0.8 MTris-HC1 pH 8.8 and 0.1% SDS. A 2 cm stacking gel containing 5% acrylamide, 0.1 M Tris-HC1pH 6.1, and 0.1% SDS was used. Gels were typically run at 18 mA for 17 hours at 4°C.Electrophoresis buffer used was the standard SDS-Tris-glycine buffer described in Chapter 2.Staining, electrotransfer to nitrocellulose and western blotting of the polypeptides were done asdescribed in Chapter 2 (section 2.2.3). The o-D1 (from pea) and o-OEC3 (from spinach) were agift from Ann Eastman. The c-OEC2 antiserum was donated by Dr. Ekramadoullah. Thespinach o-PsaD and cx-PsaF antisera (from spinach) were from Bengis and Nelson (1975).67percent chlorophyll absorbance maximaI 14% 417, 435, 482*, 672 nmII 86% 421,437,493,678nmFigure 3.1.Schematic diagram of linear 15-30% Sucrose gradient used to fractionated f3-dodecylmaltoside solubilized thylakoids of A. neglectum. Resolved chlorophyll-binding complexes areindicated as fraction I or II. The percentage of chlorophyll and the absorption maxima of eachfraction are given. An (*) indicates the presence of an absorbance shoulder.3.3 ResultsFractionation of the thylakoids of Aglaothamnion neglectuin on a 15-30% linear sucrosegradient resolved two green bands: band I and band II (fig. 3.1). The top fraction (I) was lightgreen and contained only 14% of the total chlorophyll present on the gradient. The darker greenbottom fraction (II) contained the majority of the chlorophyll (86%). A very small reddish-brown pellet present at the bottom of the gradient contained very little or no chlorophyll. Roomtemperature absorption spectra of the two fractions were taken (Fig. 3.2). Fraction I hadabsorbance maxima at 417 nm, 435 nm and with a long-wavelength absorbance maximum of672 nm. A broad shoulder at 482 nm was also present. Fraction II had an absorption maxima inthe 493 nm range and a long-wavelength form of chlorophyll a at 678 nm. Soret peaks at437 nm and 421 nm were also observed. The distinct absorbance properties in the 480-550 nmregion between the two fractions indicates differences in carotenoid distribution.68C000U)Wavelength (nm)Figure 3.2Room temperature absorption spectrum of sucrose gradient fractions I (top spectra) and II(bottom spectra). Spectra are offset for clarity.In Figure 3.3 (panel B), the polypeptide composition of fraction II and thylakoids from A.neglectum are shown along with a thylakoid preparation from Heterosigma and a PS I fractionfrom spinach. Analyses were limited to fraction II and to thylakoids of A. neglectum because thepresence of excessive amounts of detergent prevented proper resolution of the polypeptideconstituents in fraction I. Polypeptides in the size range of 18-22 kDa, 30-34 kDa, and 40100 kDa were especially abundant in the thylakoid lane (fig. 3.3 B, lane 6). Fraction II had adistinct band at 66 kDa and diffuse polypeptides in the 18-22, 30-34 and 40-46 kDa regions (fig.3.3 B, lane 5). The LHCs of Heterosigma carterae in the 15-22 kDa range were the mostabundant thylakoid proteins (fig. 3.3 B, lane 7; Chapter 2). In the spinach PS I preparation therewere a number of polypeptides in the 14-69 kDa range but the 2 1-27 kDa antennae were themost prevalent (fig. 3.2 B, lane 8).400 500 600 70069A western blot with the same fractions was immunostained with the x-CP1a antiserum(fig. 3.3 A). This antibody (described in Chapter 2) cross-reacted with four polypeptides in thespinach PS I enriched fraction, the two lower polypeptides belonging to LHC I (25 and 22 kDa)and the two upper polypeptides (26.5 and 27 kDa) to contaminating LHC II. This antiserumrecognizes epitopes from both LHC I and LHC II polypeptides (White and Green 1987). The H.carterae thylakoid fraction cross-reacts with up to six polypeptides in the size range of 17-23 kDa (fig. 3.3 A, lane 3). The pattern of cross-reacting polypeptides appears a little differentfrom similar immunoblots in Chapter 2 because a different gel system was used to resolve them.Of particular interest was the detection of four cross-reacting polypeptides (19.0, 18.5,18.0 and 17.5 kDa) in the A. neglectum thylakoid fraction (fig. 3.3 A, lane 2). In addition, theremay be a fifth immunoreactive polypeptide with a molecular mass of 19.5 kDa, as the 19.0-19.5 kDa band appears to be a doublet. Fraction II contains two cross-reacting polypeptides witha size of 19.0 and 19.5 kDa. The other cross reacting polypeptides were probably removedduring fractionation. It is interesting to note that these polypeptides are significantly smallerthan the corresponding polypeptides in terrestrial plants and green algae (21-24 kDa). In termsof size, they more closely resemble the FCPs from the chrornophytes (16-2 1 kDa). The c&CP1aantibody also cross-reacts with the core complex of PS I, as seen in the 66 kDa area of all fourfractions (fig. 3.3 B).Using an antibody specific to the pea Dl protein, the 32 kDa core complex polypeptideof PS II (Dl) was detected in both fraction II and in the thylakoid fraction from Aglaothamnion(fig. 3.4, lanes 1 and 2). The Dl polypeptide was also detected in the Heterosigina thylakoidfraction (lane 3) but was absent from the spinach PS I preparation (lane 4). Though fraction IIcontains both PS I and PS II specific polypeptides, it appears to be enriched in PS I complexes.Immunological analyses have demonstrated that Dl and CP47 (PS II core antenna) arestructurally conserved in Synechocystis (cyanobacterium), Prochlorothrix, Heterosigma, andAglaothamnion (data not shown). The immunological similarities of these polypeptides in thered alga, Porphyridium have also been demonstrated (Marquardt and Ried 1992). The putative70Aco \\çç\\ ?s\\—0B =9746946Figure 3.31 2 3 43021.514.3Composition and immunological analysis of polypeptides in Aglaothamnion thylakoidsand fraction II. A) Western blot immunostained with the o-CPla antiserum (described in text).Samples are; 1, Aglaothamnion fraction II; 2, Aglaothamnion thylakoids; 3, Heterosigmacarterae thylakoids; 4, spinach PS I fraction. B) Gel stained with coomassie blue. Samples are;5, Aglaothamnion fraction II; 6, Aglaothamnion thylakoids; 7, Heterosigma carterae thylakoids;8, spinach PS I preparation; 9, markers. Size of markers in kDa are indicated on the right.5 6 7 8 971A.n. A.n. H.c. Sp. MF II Thy. Thy. PSI1 2x-D1— 3OKDaFigure 3.4Immunological detection of the Dl protein in Aglaothamnion sucrose gradient fraction II(lane 1), Aglaothamnion thylakoids (lane 2), Heterosigma thylakoids (lane 3) and in a spinachPSI fraction (lane 4). The molecular mass marker (30 kDa) is in lane 5. An anti-Di antiserum(a-Dl) derived from spinach is used.ferredoxin docking and PsaC stabilizing protein, PsaD (PS I subunit II, psaD), was also presentand immunologically reactive to a spinach PsaD specific antibody in Synechocystis,Heterosigma, and Aglaothannion(fig. 3.5, psaD). However, there were considerable variationsin size (14-22 kDa) and the reactions were weak in the Aglaothamnion and Heterosigmathylakoid lanes (fig. 3.5, psaD). In contrast, only Heterosigma and Aglaothamnion thylakoidsshowed a cross-reaction with an antibody directed to the putative plastocyanin docking protein,PsaF (PS I subunit III), of spinach (fig. 3.5, psaF). A polypeptide immunologically related toPsaF was not detected in Synechocystis, though it may have been removed during isolation sinceit is extrinisically associated with the membrane.3 45721 2 34OEC1—301 2 34—30OEC21 2 34—— 21psaD—141 2 34—21psaF—14Figure 3.5Immunological analyses of thylakoids from Heterosigma (lane 1), Aglaothamnion(lane 2), Synechocystis (lane 3), and spinach (lane 4). Western blots were immunostained withthe anti-OEC1 (oxygen evolving complex 1), the anti-OEC2 (oxygen evolving complex 2), theanti-psaD (PS I subunit II) and anti-psaF (PS I subunit III) antisera as indicated. Molecular massmarkers are indicated on the right.73Western blots were done to examine the presence or absence of polypeptides making upthe oxygen evolving complex (fig. 3.5). This was attempted immunologically using spinachderived antibodies specific to the OEC1 (33 kDa) and OEC2 (23 kDa) polypeptides. Since theOEC2 polypeptide is not present in cyanobacteria (Stewart et al. 1985), it was thought that anexamination of the red algae for the presence or absence of this complex would provide someuseful phylogenetic information. The OEC1 polypeptide was only very weakly detected inSynechocystis and Heterosigma and this antibody did not react significantly with anything in theAglaothamnion thylakoid fraction (fig. 3.5, OEC1). In addition, only the spinach thylakoidfraction cross-reacted with the OEC2 antibody, even though four times the normal amount ofchlorophyll was loaded on the gel (fig. 3.5, OEC2). Unfortunately, I am unable to conclude ifAglaothamnion or Heterosigma have a protein homologous to the OEC2 polypeptide as the lackof immunological cross-reaction could be due to sequence divergence and not complete absence.Greg Wolfe et al. (1992, 1994) at the University of Maryland analyzed the pigment-protein complexes of the unicellular red alga, Forphyridiuin cruentum. They were the first todemonstrate the presence of a putative chlorophyll-binding complex analogous to LHC I ofterrestrial plants. In a collaboration with this group, we were able to analyze immunologicallythe structural relatedness of these polypeptides to those from the CAB and FCP family. Thisalga is a member of the class Bangiophycidae (order Porphyridiales), which is considered to beprimitive with respect to the Florideophycidae, of which Aglaothamnion is a member.Figure 3.6 shows the results of immunoblotting two identical gels with different antisera; the cxjCP1a polyclonal antibody and the o-FCP antibody (described in Chapter 2). At least fivepolypeptides in the P. cruentum PS I fraction cross-reacted with the a-CPla antibody; they hadapparent molecular masses of 19.5, 20, 22, 23 and 23.5 kDa (Fig. 3.6, panel B, lane 3). Thereare two additional polypeptides cross-reacting in the thylakoid lane at 19 and 20.5 kDa (Fig. 3.6,panel B, lane 4). The spinach PS I fraction showed a number of cross-reacting polypeptides(22-29 kDa) from LHC I and contaminating LHC II and CP29 (Fig. 3.6, lane 2, panel B) as thesample was overloaded. A major cross-reacting band at 65 kDa in all four lanes is the corecomplex of PS I. It is significant that thylakoid proteins from the cyanobacterium, Nostoc, lack74Figure 3.612341234Immunological analysis of polypeptides from: 1, Nostoc thylakoids; 2, spinach PS Ifraction; 3, Porphyridium cruentum PS I fraction; 4, Porphyridium cruentum thylakoid fraction.Molecular mass markers are on the right. Antisera used included the a-FCP antiserum specificfor a diatom FCP complex (panel A) and the o-CPla antiserum (panel B).75cross-reactivity with the c-CP1a antiserum in the size range typical for a LHC (Fig. 3.6, lane 1,panel B).A second antibody, c-FCP, was used to assess relatedness of the low molecular masspolypeptides of Porphyridium to the FCPs of the chromophytic algae (Fig. 3.6, panel A). Thea-FCP antibody cross-reacted with five polypeptides from the P. cruenturn thylakoid fraction(Fig. 3.6, panel A, lane 4). Interestingly, the 19 and 19.5 kDa bands appear lessimmunologically reactive to this antiserum compared to their reaction with the c-CP1aantiserum. The 20 and 23.5 kDa polypeptides did not appear to cross-react at all with the a-FCPantiserum. These observations indicated a degree of structural divergence between some of thepolypeptides. At least three polypeptides (19.5, 22, and 23 kDa) in the P. cruentum PS I fractioncross-reacted with the o-FCP antiserum (Fig. 3.6, panel A, lane 3) . Again, the 20.5 kDapolypeptide in the PS I complex was only weakly detected and must have been removed duringpurification.These results show that the complex pattern of polypeptides immunologically related tothe LHCs in the unicellular red alga, Porphyridium, is comparable to that in the filamentous redalga, Aglaothamnion, except for differences in the apparent size of the polypeptides. This maybe due to the use of a different SDS-gel system. No polypeptides in the 18-23 kDa range weredetected in the thylakoids of the cyanobacterium, Nostoc, with the ct-FCP antiserum (Fig. 3.6,panel A, lane 1). This supports the idea that the cyanobacteria do not possess intrinisic antennaerelated to those of algae or terrestrial plants. However, a large molecular mass polypeptide(69 kDa), of unknown identity, was weakly immunostained with the x-FCP antiserum.Thylakoid and PS I fractions from the Chl a + b-containing prochlorophyte,Prochlorothrix hollandica (a gift from Georg van der Staay), were immunoblotted with thebarley o-CP1a antiserum (Fig. 3.7, lanes 1 and 2) to search for evidence of any CAB relatedLHCs. Neither of these fractions showed any immunologically reactive polypeptides in the sizerange of a typical LHC (18-34 kDa). However, the CAB polypeptides were detected in thespinach thylakoid control, as expected (fig. 3.7, lane 3). In both the thylakoid and the PS Ifractions of Prochlorothrix, the core complex of PS I was detected at 65 kDa.76Figure 3.712 3M66453121Immunological analysis of Prochlorothrix hollandica PS 1(1) and thylakoid (2) fractions.A spinach control (thylakoid fraction) is also shown (3). Molecular mass standards (M) areshown on the right. The blot was probed with the o-CPla antiserum.773.4 DiscussionAnalyses with the unicellular alga, P. cruentum, demonstrated that there are severalpolypeptides that are immunologically related to both the CABs and FCPs (Wolfe et al. 1994).Furthermore, some of these are specifically associated with a purified PS I complex (Wolfe et al.1992; Wolfe et al. 1994). This is significant as it is the first demonstration of such membraneintrinisic light-harvesting antennae within the Rhodophyta. Despite recent analyses of red algalthylakoid membrane polypeptide composition, CAB-related LHCs have not been previouslyreported. The LHCs in Porphyridiurn also cross-react to different extents with the CAB andFCP-specific antisera. This indicates that though the different polypeptides share commonimmunologically reactive epitopes, there is structural variability between these antennal proteins;this may be related to different functions. These observations suggest closer evolutionaryrelationships between the red, green and chromophytic plastids than previously thought. Recentprotein sequencing data from these polypeptides have confirmed their relatedness to the LHC Iproteins of terrestrial plants (Beth Gantt, unpubl. data).When I examined a second red alga, Aglaothamnion neglectum, up to five polypeptidesin the thylakoid membrane fraction were antigenically related to the Chl a + b-binding proteinsof terrestrial plants. Fewer cross-reacting polypetides were found in the sucrose gradientfraction, probably a result of selective loss during the detergent solubilization procedure. It hasbeen immunologically determined that fraction II from the sucrose gradient containspolypeptides from both PS I and PS II; therefore, it is not possible to conclude whether they areassociated specifically with PS I or PS II. By analogy with Porphyridium, I would expect anassociation of these LHCs with PS I in A. neglectum. However, this does not exclude thepossibility that CAB/FCP-related polypeptides are associated with PS II. In both red algae, therewere fewer detectable LHCs in the detergent solubilized fractions as compared to the thylakoidfractions. This potential for degradation may be one reason these polypetides were notpreviously reported.It is significant that CAB/FCP-related polypeptides were found in representatives of both78the Porphyridiales (Porphyridiuin) and the Ceramiales (Aglaothamnion), as they represent verydiverse lineages separated by large evolutionary distances (Garbary and Gabrielson 1990). Thissuggests that these LHC-related polypeptides are not limited to a specific taxon; likely being acommon feature of the red algae. I suspect that similar polypeptides will be detected in other redalgal orders.Plastids are thought to have evolved from endosymbiotic cyanobacteria and/orprochlorophytes; therefore, it was significant to find a lack of CAB/FCP-related polypeptides inNostoc and Prochlorothrix. This is particularly significant with the prochlorophytes as it hasbeen demonstrated that they possess Chi a + b + c-binding antennae. The lack of immunologicalrelatedness of these antennae has been demonstrated by others (Hiller and Larkum 1985;Bullerjahn et al. 1990). In addition, preliminary sequence information from the prochiorophytes,Prochioron (Hiller & Larkum, unpubl.), Prochiorococcus (LaRoche & Partensky, unpubl.) andProchlorothrix (van der Staay & Green, unpubl.) shows that the 34 kDa antennae are related tothe isiA gene product and to the inner Chl a antenna, CP43. There were no similarities to theCABs or the FCPs. This proves that the prochiorophyte LHCs evolved independently of theCABs and FCPs.The presence of CAB/FCP-related polypeptides in the red algae and their absence in thecyanobacteria and prochlorophytes is significant in terms of both the evolution of the light-harvesting antennal proteins and the evolution of the chioroplast. The significance with regard tothe evolution of the light-harvesting proteins as an extended gene family will be dealt with inChapter 5. Much of the remaining discussion will deal with the implications of theseobservations for the different hypotheses of chloroplast evolution (discussed in section 1.6),especially with regards to the red and green algae. In analyzing chloroplast evolution, bothchioroplast and nuclear characters can be used (see table 3.1). Plastid characters, includingplastid encoded genes and plastid localized gene products encoded in the nucleus (such as theCABs and FCPs), were probably present in the original endosymbiont and comparisons willnormally reflect the ancestry of the chloroplast. Cytoplasmic/nuclear characters will generallyreflect the ancestry of the original phagotrophic host. Comparing evidence of both types is79necessary to develop a better understanding of the possible evolutionary pathways leading to thechloroplast.Many of the plastid encoded characters point toward a monophyletic scheme ofchloroplast evolution. These include the plastid 1 6s rRNA, the psbA, tufA and atpB genes (seetable 3.1) which have been reviewed by Morden et al. (1992). Recent phylogenetic analysis ofthe plastid encoded atpB gene (H-ATPase, f3-subunit) suggests that there is a deep divisionbetween the green and non-green algal chioroplast lineage’s, though they cluster together withinthe cyanobacterial line, indicating a single origin for the chloroplast (Kowallik 1993; Douglasand Murphy 1994).Table 3.1 Some genes commonly used to infer phylogenetic relationships amongstphotosynthetic organismsgene protein /function locationpsbA Di-core PS II reaction centre protein ChioroplastatpB 13-subunit of chloroplast ATPase ChioroplastrbcL rubisco large subunit ChloroplastChloroplastrbcS rubisco small subunit /nucleartufA elongation factor Tu Chioroplast1 6s rRNA small subunit rRNA Chloroplastglyceraldehyde-3-phosphateGapAJB dehydrogenase nucleus1 8s rRNA small subunit rRNA nucleusHowever, there is evidence that is consistent with a polyphyletic interpretation ofchloroplast evolution. A seven amino acid deletion in the C-terminus of Dl (psbA gene) inProchlorothrix, the green algae and in the terrestrial plants but not in the Dl protein ofcyanobacteria, red algae and the chromophytes is suggestive of a polyphyletic origin of thechloroplast (Morden and Golden 1989; Golden et al. 1993). This emphasizes a character sharedby a prokaryotic! chioroplast pair containing Chl a and b to the exclusion of another prokaryotic!80chioroplast pair containing PBSs or Chi a + c —satisfying the criterion for a polyphyletic origin.However, it has recently been reported that the prochiorophyte, Prochioron dideinni, lacks thisdeletion, indicating that it is not a common occurrence amongst the prochlorophytes (Lockhart etal. 1993).It is reasonable that insertion/deletion (indel) events would be considered goodphylogenetic indicators as two independent deletion events in an identical area seems unlikely.However, I argue that the significance of this particular indel may have been over estimated.Besides having representative sequences from relatively few taxa, there are two knownvariations in the Dl indel location; one in the green alga, Chlamydomonas reinhardtii (Ericksonet al. 1984), and the other in a euglenophyte, Euglena gracilis (Karabin et al. 1984). In theseexamples, there is an eight and 16 amino acid deletion at the C-terminal end of Dl, respectively.The fact that these algae possess additional or separate deletions different from the otherChl a + b-containing organisms suggests that there are lower constraints on the 9-16 amino acidsat the C-terminal end. Significantly, this region of Dl is post-translationally removed duringprocessing into the mature form. In addition, the C-terminal region is neither required forprotease binding nor for recognition (Taguchi et al. 1993). Phylogenetic analysis of the wholepsbA gene is consistent with the monophyletic view of chloroplast evolution with Prochlorothrixclustering with the cyanobacteria, separate from the eukaryotic algal groups (Morden et al.1992). All things considered, it seems that the deleted region has an ambiguous evolutionaryhistory of questionable importance. Furthermore, based on the accumulation of molecularevidence from plastid/cyanobacterial characters, a direct evolutionary relationship between theprochlorophytes and the green algal chioroplasts are now considered less likely (Turner et al.1989; Morden et al. 1992; Golden et al. 1993; Swift and Palenik 1993).Plastid characters that do offer strong evidence for a polyphyletic origin of thechloroplast are the plastid encoded rbcL and rbcS (rubisco large and small subunit) data whichclearly show the Chl a + b-containing eukaryotes in one lineage and the red algae andchromophytes in the other. In these studies, the cyanobacteria are closely related to theChi a + b-containing lineage, while the most similar prokaryotic ancestor to the red algal/81chromophyte line is the f3-purple bacterium, Alcaligenes eutrophus (Douglas et al. 1990).However, due to the lack of phycobilisomes, Chi a, and oxygenic photosynthesis, a 3-purplebacterium is not a logical choice for a hypothetical chioroplast ancestor. Several hypotheseshave been put forth to explain the relatedness between the 3-purple bacterial and chromophyterbcLIS sequences: (1) lateral gene transfer of the rbcLIS operon from a purple bacterium to analga leading to the red algal/chromophyte lineage (Boczar et a!. 1989b); (2) the possibility thattwo organisms provided genes to the phagotrophic host (Assali et al. 1990); (3) since somepurple bacteria possess two rbcL genes, it has been suggested that the original eukaryoticphotoautotroph had two copies of the rbcLIS operon with the green algal and chromophytelineages each retaining a different copy; (4) finally, transfer of rbcL from an oL-purple bacterium,the proposed mitochondrial ancestor, to the chloroplast ancestor has also been suggested (Martinet al. 1992). The significance of the rbcL data will remain an issue that will need to be resolvedthrough the analysis of other data.The evidence supporting or refuting either hypothesis of chloroplast evolution (asdiscussed in section 1.6 and above) has been plagued by conflicting and ambiguous evidence.Deciding between alternative hypotheses of chioroplast evolution can be done by applying thefollowing two criteria, as emphasized by Reith and Munholland (1993): first, the polyphyleticview of chloroplast evolution would be supported when characters (morphological or molecular)are found in one chloroplast type that are shared with a putative prokaryotic ancestor, to theexclusion of another chloroplast type-prokaryote pair. Second, if chloroplast evolution is to bedeemed monophyletic then the different chloroplast types should all share a certain trait orcharacter, which is more related to each other than either is to a prokaryotic (cyanobacterial)ancestor (Reith and Munholland 1993). This trait or character could then be interpreted ashaving been derived within a single lineage, following the primary endosymbiotic event that leadto the first chloroplast containing eukaryote.The immunological relatedness of the red algal polypeptides to the CABs and FCPs issignificant because it provides evidence that links organisms possessing the three major antennalsystems: the Chi a + b-containing antennae, the Chi a + c-containing antennae, and the82phycobilisomes. Previously, these characters formed the basis of an algal taxonomic system thatseparated the Chl a -- b containing green algae (Chiorophyta), the Chi a + c containingchromophytic algae (Chromophyta), and the PBS containing red algae (Rhodophyta) into majordivisions. This finding also demonstrates that the thylakoid membrane intrinisic LHCs and thesoluble phycobilisome types of antennal systems can be present in the same organism.The lack of detectable, immunologically related LHCs from the cyanobacterium, Nostoc,and the prochlorophyte, Prochlorothrix, suggests that the intrinisic light-harvesting antennaeevolved after the primary endosymbiotic event that gave rise to a photosynthetic eukaryote. Thepresence of CAB/FCP related LHCs in all eukaryotic algae, to the exclusion of the putativechloroplast ancestors (cyanobacteria and prochiorophytes), is consistent with the monophyleticview of chloroplast evolution because it seems unlikely that structurally similar proteins wouldevolve independently in the red, green and chromophyte algae from prokaryotic precursors.These data also support the idea that both the PBS and the CAB/FCP antennal systems existed inthe first photosynthetic eukaryote from which the green algae and red algae diverged.Divergence of the intrinisic LHCs with the gain of Chl c (in the chromophytes andprasinophytes) and the loss of PBSs (in the chiorophytes and chromophytes) could explain thedifferences in antennal systems presently observed (Cavalier-Smith 1982).There can be alternative interpretations or hypotheses invoked to explain the LHCdistributions between the prokaryotes and eukaryotes. One alternative is that the cyanobacteriaand prochlorophytes both contained CAB/FCP-related antennae but they were selectively lostover time while being retained in the immediate chloroplast ancestor (Bryant 1992). Such ascenario does not necessarily exclude a monophyletic view of chloroplast evolution.Alternatively, the different LHCs may have evolved independently following separatechloroplast acquisitions (polyphyletic) leading to the major algal divisions. However, thisalternative requires the acceptance of additional evolutionary steps and seems less likely.Nevertheless, the independent evolution of LHC proteins from related cyanobacterial precursorsis possible and could explain the overall low degree of similarity between the CABs and FCPs.Some possible molecular sources from which the LHC antennae may have evolved are discussed83in section 5.4.6 (Chapter 5).The acceptance of a monophyletic chioroplast origin requires the assumption that the redand green algae are related and have diverged from a common ancestor. Such a relationship wasproposed by Cavalier-Smith (1981, 1987) who grouped the red and green algae together in thekindom Plantae. In order to demonstrate a evolutionary relationship between the red and greenalgae it will be necessary to show similarities between nuclear encoded characters in addition tothe chloroplast characters.More recently, the nuclear encoded (but chloroplast localized) glyceraldehyde-3-phosphate dehydrogenases (GapAIB) from two different red algae have been sequenced andphylogenetic analyses suggests that the red and green algae form a monophyletic group with acyanobacterial ancestor (Zhou and Ragan 1993; Liaud et al. 1994). These studies also indicatethat the green algal and red algal lineages separated very early in evolution. Moreover, thetransit peptide from the GapAfB precursors in red algae resembles the chioroplast targetingtransit peptides of terrestrial plants (Zhou and Ragan 1993; Liaud et al. 1994). In vitro transportstudies with the transit peptide of the nuclear encoded y-subunit of phycoerytherin inAglaothamnion has shown that it can direct rubisco (RbcS) into pea chloroplasts (Apt et al.1993), suggesting the red and green algae have a similar chioroplast import mechanism. Thismay be more easily explained by a monophyletic origin of the red and green algal chloroplastsdue to the perceived difficulty in acquiring import capabilities (Cavalier-Smith 1982; Cavalier-Smith et al. 1994b).A better idea of the relationships between the red and green algae can be achievedthrough the analysis of nuclear encoded characters from the two groups. This can giveinformation on the nature of the eukaryotic ‘host for the chloroplast. The most widely utilizednuclear character to date has been the small subunit rRNA (SSU, 18s) (Bhattacharya et al. 1990;Douglas et al. 1991; Hendricks et al. 1991; Maier et al. 1991; Cavalier-Smith et al. 1994b) . Thenuclear encoded large subunit rRNA (LSU, 23s) has been used infrequently (Perasso et al. 1989)due to its larger size and the smaller dataset. Most of these studies did not support a commonorigin of the red and green algae. A recent maximum likelihood analysis of nuclear encoded84SSU rRNA sequences, however, showed that the red and green algae grouped together on thesame branch as would be expected for a monophyletic origin (Cavalier-Smith et al. 1994b). Thisgrouping does not occur when a distance matrix method is used, which was the case in all thestudies mentioned above (Cavalier-Smith, pers. comm.; Cavalier-Smith et al. 1994). Comparingthe studies mentioned above, the relationships between the red algae and other groups were notconsistent. This was probably due to the inclusion of different taxa, the number of charactersincluded in the analysis and the limitations of the method used. Many of these studies also didnot give an estimate of the tree node reliability, so it is hard to determine how consistent thedeeper branches of these trees are.Overall, there is accumulating evidence suggesting a monophyletic origin of thechloroplast, though the question is far from settled. If true, this would mean that the red andgreen algae share a common ancestor and that the acquisition of a chioroplast from a prokaryoticancestor occurred only once. Divergences over a long period of time would have resulted in thedifferences between the two groups today. The presence of CAB and FCP-related LHCs in thered algae provides some evidence for the monophyletic origin of the different chloroplast types,though other alternative evolutionary pathways can be envisaged. At the moment, there isinsufficient evidence to clearly decide whether chloroplasts with two surrounding membraneshave arisen through one or multiple endosymbiotic events. The analysis of more characters froma diverse array of red, green and chromophytic algae along with cyanobacteria should help todetermine relationships amongst these groups and help to resolve conflicting interpretations ofsome data.85CHAPTER 4Characterization of Fcp cDNAs from Heterosigma carterae4.1 IntroductionThis section examines cDNAs encoding the FCP family of proteins in Heterosigma. Atthe beginning of this project only a single chromophyte algal FCP sequence had beencharacterized (Grossman et al. 1990). As the chromophytes consist of many distinct phyla, Ithought this research would begin to fill a gap in the literature and aid in our understanding ofthe relationships between the different antennae and their diversity. As well, it would be animportant piece of information for assessing the evolutionary relationships amongst the FCPsand the CABs of the terrestrial plants and green algae. In this chapter I will first describe thecloning and sequencing of a number of cDNAs encoding FCPs. I will then examine thecomplexity of the Fcp gene family and concentrate on the relationships of the Heterosigmasequence to other FCPs and the CABs. A cDNA library was constructed and screened with anucleic acid probe in order to clone cDNAs encoding the FCPs. This will complement theprotein characterization work in Chapter 2. In addition, I was interested in the processesinvolved in the targeting and translocation of proteins into the chloroplast. As these organismspossess two additional membranes around the chloroplast, an examination of the leadersequences of nuclear encoded, chioroplast localized precursors should provide some clues on thenature of the transport mechanism in the chromophytes.864.2 Materials and Methods4.2.1 Tryptic fragment sequencingMaintenance of Heterosigma cultures and preparation of thylakoids have been previouslydescribed in Chapter 2. Sixty jig of chlorophyll (four lanes at 15 jig/lane) from sucrose gradientfraction one (Fl, Fig. 4.2; chapter 2) was separated on a 12-17% gradient gel (Chapter 2) andtransferred to nitrocellulose (Biorad) in 50 mM sodium acetate pH 7.0 at 200 mA, for 1 8hours at4°C. The transferred band was stained with amido black (0.1% amido black, 45% H20, 45%methanol, and 10% acetic acid) and destained in 50% methanol/l0% acetic acid. The proteinband was cut out from the membrane, digested on the support with trypsin (Aebersold et al.1987) and separated by narrow-bore reverse phase HPLC on a Waters peptide analyzer equippedwith a Vydac C-4 column. Individual peptides were collected manually and sequenced usingstandard pulsed-liquid phase or solid-phase sequencing procedures (Aebersold et al 1990).4.2.2 Heterosigma DNA and RNA isolationGenomic DNA was isolated from 4 liters of late log phase cells. Cells were harvested bycentrifugation at 1500 x g for 10 minutes. The pellet was resuspended in 2 ml extraction buffer(0.05 M Tris-HC1 pH 8.0 and 0.05 M EDTA pH 8.0) that was added while vortexing gently.16 ml of lysis buffer (2% sarkosyl, 10mM EDTA pH 8.0, and 20 mlvi Tris-HC1 pH 7.6) wasthen added and gently mixed. Proteinase K (8 mg) was added and swirled at 37°C for 1 hour tobreak down proteins. Four ml of 5 M NaC1 was added and mixed. This mixture was extractedwith an equal volume of Tris equilibrated phenol (pH 7.6) for 30 minutes at room temperature.Following centrifugation at 9000 x g for 20 minutes, the supernatant was extracted with an87equal volume of chloroform:butanol (4:1), centrifuged as above and repeated. The DNA wasethanol precipitated with an equal volume of 95% ethanol and spooled out with a glass rod. TheDNA was resuspended in T.E. pH 8.0 and re-extracted twice with phenol:chloroform, as above.The supernatant was ethanol precipitated and resuspended in a total of 11 ml T.E. pH 8.0. DNAwas purified by CsC1 density gradient ultracentrifugation, using standard protocols (Sambrooketal. 1989).Total RNA was isolated from 4 liters of log phase cells of Heterosigina using aguanidinium thiocyanate extraction method (Chromczynski and Sacchi 1987). Cells wereharvested at 1500 x g for 10 minutes and resuspended in guanidinium solution (4 Mguanidinium thiocyanate, 25 mM sodium citrate pH 7, 0.5% sarcosyl, and 0.1% 13-mercaptoethanol) at 1 ml per 0.1 g cells, followed by hand homogenization in a 5 ml PotterElvehjem homogenizer. The following solutions were then added to the homogenized extract:1/lOth volume of 2 M sodium acetate pH 4, 1 volume water saturated phenol, and 115th volumechloroform:isoamyl alcohol (49:1). The sample was mixed following each addition. Thesample was vigorously vortexed for 10 seconds, cooled on ice for 15 minutes and centrifuged at10 000 x g for 20 minutes. An equal volume of isopropanol was added to the supernatant inorder to precipitate the nucleic acids. After centrifugation at 10 000 x g for 20 minutes thepellet was resuspended in 1.5 ml guanidinium solution and again precipitated with an equalvolume of isopropanol. Precipitated samples were resuspended in T.E. and stored at -80°C. Thepoly A mRNA used in the construction of the cDNA library was isolated from total RNA usingan oligo (dT) cellulose (Pharmacia) column (Sambrook et al. 1989).4.2.3 cDNA library construction and screeningThe cDNA library was constructed from 5 ig of poly A mRNA. The poiy A mRNAwas considered to be of high quality since in vitro translation using a wheat germ system(Promega) yielded many proteins in the 14-70 kDa range. The library was constructed using the88lambda ZAP II kit from Stratagene according to the manufacturers instructions. The synthesis ofcDNA was initiated with a poly-T primer with an Xho I adapter and after the second strandsynthesis, the ends were blunt ended and EcoR I linkers were ligated to the cDNA. Directionalcloning was achieved by ligating the cDNAs into an Xho 1/EcoR I cut 2ZAP vector.Screening of the amplified library was done using a nucleic acid probe created through apolymerase chain reaction. The amino acid sequence information from the tryptic fragments(Fig. 4.2) was used to create degenerate primers for amplification of a Fcp-specific geneproduct. The degenerate primers were based on tryptic fragments Ti (TVEIK) and T3(YDLAGDQ) as shown in Figure 4.2. This would give an amplified fragment of approximately100 bp if the alignments with the diatom FCP were correct. The primers included a nonspecificadapter sequence with an Xba I or Pst I restriction site in the 5’ region (indicated in bold). Thesequences of the primers used were as follows: 5’-GCTCTAGA AC(GICIT/A) T(TIGICIA)GA(AJG) AT(T/C/A) AA(A/G) CA(T/C) GG-3 for P1 and 5’-CGCTGCAG TG(AIG)TC(AIGICIT) CC(AIG/CIT) GC(AIGICIT) A(G/A)(G/A) TCA (A/G)A -3’ for P2. The 36 cyclePCR reaction profile was as follows: cycle 1—4, 95°C-2 mm., 52°C-i .5 mm., and 72°C-15 sec.;cycle 5—24, 95°C-i mm., 52°C-i .5 mm., and 72°C-30 sec.; cycle 25—>36, 95°C-i mm., 52°C-1 mm.; and 72°C-i mm.. The reaction included 2.0mM MgC12, 100 pmol of each primer (P1and P2), 50 mM KC1, 10 mM Tris-HC1 pH 8.3, 0.2 mM dNTPs and 0.5 mg Heterosigmacarterae genomic DNA as the target sequence. The reaction was done under mineral oil with0.5 units of Taq polymerase (AmpliTaq, Perkin-Elmer-Cetus). The reaction mixture wasextracted with phenol-chloroform (1:1), ethanol precipitated and run on a 4% Nusieve agarosegel (Mandel Sci.) along side Xi74-Hae ifi markers (NEBL). A 118 bp fragment was isolatedusing Whatman DE-81 paper and directly labeled withct-32P-dCTP (3000 mCi/mmolAmersham) using a random primer labeling kit (BRL), according to the manufacturersinstructions. The cDNA library was screened at high plaque densities (10 000 pfu/plate) for thefirst round of screening. Further rounds of screening were done at low densities so thatindividual plaques were well separated. Plaque lifts, denaturation, fixing and hybridization were89done as descibed in the Stratagene cDNA kit instruction manual and according to Sambrook etal. (1989).Prospective positive clones went through three to four rounds of screening until plaqueswere homogeneous. The recombinant pBluescript phagemid (with insert) was excised from thek-ZAP vector using the ExAssist/SOLR system (Stratagene). Miniprep plasmid DNA wasprepared using an alkaline lysis method (Sambrook et al. 1989) and was often sequenced directlyafter RNAse treatment, two phenol-chloroform extractions and ethanol preciptation. Large scaleplasmid preparations were also done using an alkaline lysis method followed by plasmidprecipitation with polyethylene glycol (Sambrook et al. 1989). Double stranded sequencing with35S-dATP was done using the dideoxy chain terminating method with a T7 polymerase(Pharmacia). Standard sequencing gel receipes and running conditions were used (Sambrook etal. 1989). Sequencing gels were lifted directly off plates onto Whatman 3MM paper, withoutfixing, and dried for 45 minutes at 80°C under vacuum.4.2.4 Rapid Amplification of cDNA Ends (RACE)As the isolated cDNA clones were truncated near the N-terminus, it was necessary to usealternative methods to clone this region. The 5’ ends of the truncated cDNA clones weredetermined using the rapid amplification of cDNA ends (RACE) technique (as illustrated inFig. 4.1) using a modification of previously described methods (Frohman 1990; lain et al. 1992;Schuster et al. 1992). Reverse transcription of 1 tg of poly-A mRNA (heated at 70°C, 10 mmand quenched on ice) was done using 200 units superscript II reverse transcriptase (BRL),200 IIM dNTPs, 20 units RNAse inhibitor (RNasin, Promega), 10 mM DTT, and 10 pmol of agene specific primer (gspl) located near the 5’ end of the truncated cDNA clone[5’AATGAAGCCGATGGTCT3’j. The reverse transcription reaction was incubated at 42°C for60 minutes followed by a 50°C incubation for 15 minutes. The cDNA was treated withRNAse H (Pharmacia) at 42°C for 15 minutes to remove the RNA template. The reverse90gsp 1mRNA—A reverse transcribecDNAA add poiy A tail withterminal transferaseaaaaaaaaAgsp2-adpaaaaaaaattttttttdt-adpA amplifyFigure 4.1 Rapid amplification of cDNA ends (RACE) technique,illustrated. The gene specific primers are gspl and gsp2.The adapter portion of the primer, adp, is indicated.Mechanism is explained in the text.transcription primer and dNTPs were removed by ultra-filtration through ultrafree-MCcentrifugation filters (30,000 NMWL, Millipore) (Jam et al. 1992). Following a washing step inthe ultra-filtration unit, the samples were concentrated under vacuum to approximately 10 pi. Apoly-A tail was then added to the extended cDNA by using terminal deoxy transferase(Pharmacia) in a standard PCR buffer (10 mM Tris-Hcl pH8.3, 50 mMKCI, 25 mM MgCI2)with2 mM dATP (Schuster et al. 1992).Amplification of the cDNA was done using a poly T primer with a 5’ adapter (indicatedin bold) [5’ GACTCGAGTCGACATCGTTTTTTTTTTTTTTTTT 3’] and a second, nested91gene specific primer (gsp2) closer to the 5’ end of the truncated clone, also with a 5’ adapter[GACTCGAGTCGACATCGAGCAGGCAGAGCAGACA3’]. The amplification reaction wascarried out as before (section 4.2.3) but with 2.5 mM MgCI2 and 10 pmol of each primer. Thereaction was denatured at 95°C for 5 mm. and brought down to 80°C when 2 units of Taqpolymerase were added and the reaction layered with oil. The reaction then continued with thefollowing profile: cycle 1, 55°C-5 mm., 72°C-20 mm.; cycle 2—>37, 95°C-45 sec., 55°C-i mm.,72°C-2 mm.; cycle 38, 95°C-45 sec., 55°C-i mm., 72°C-lO mm. The amplified product wasisolated, as previously described, and cloned in a pBluescript II vector (Stratagene) that was cutwith EcoR V and tailed with ddTTP (Holton and Graham 1991).4.2.5 Southern blotsFor a Southern blot, 4 jig of genomic DNA was digested with an appropriate restrictionenzyme using the conditions recommended by the manufacturer. In order to ensure completedigestion, 10 units of enzyme were added at one hour intervals for approximately three hours,mixing gently at each addition to avoid mechanical shearing. DNA was run on large (20 X28 cm) 0.8% agarose gels in a Tris-Borate-EDTA electrophoresis buffer (Sambrook et al. 1989)with 100 jig/mi ethidium bromide. Gels were transferred onto Hybond-N nylon membrane(Amersham) by capillary action in 20X SSC, as described in Sambrook et al. (1989). Nucleicacids were fixed onto the membrane by baking for 2 hours at 80°C. Membranes wereprehybridized in Church buffer (0.25 M sodium phosphate, 7% SDS, 1 mM EDTA) for 1 hourat 65°C (Church and Gilbert 1984). Hybridization buffer was changed prior to adding denaturedprobe and then incubated for 16 hours. Membranes were typically washed at low stringency(1X SSC/0.i% SDS at 65°C) three times (15 minutes each) or at a higher stringency wash(0.1X SSC/ 0.1%SDS at 65°C), as required.924.2.6 Northern blotsRNA was quantitated spectrophotometrically, and an appropriate amount precipitatedand resuspended in a RNA loading buffer. Samples were heated at 70°C for 10 minutes andquenched on ice for 5 minutes prior to loading the gel. A 1.5% agarose-formaldehyde gel madewith a lx MOPS buffer (20 mM MOPS pH7, 5 mM sodium acetate, 1 mlvi EDTA) and 6% (v/v)formaldehyde (2.2 M) was used. The gel was run in a 1X MOPS buffer with 2.2 Mformaldehyde (Douglas et al. 1990). RNA was transferred to Hybond-N, fixed and hybridizedas in section 4.2.5.4.3 Results4.3.1 Identification and characterization of the Fcpl and Fcp2 cDNAsA cDNA expression library was constructed from Heterosigina mRNA. Because of this,the initial several rounds of screening were attempted with both the cx-FCP and o-CP 1 a antisera.However, this method was not successful and was abandoned. An alternative was to screen thelibrary with a nucleic acid probe. A heterologous Fcp gene probe from a diatom did nothybridize to Heterosigma genomic DNA; therefore, I had to obtain protein sequence informationdirectly from a Heterosigma FCP in order to generate primers for the amplification of ahomologous nucleic acid probe.The sequences of seven tryptic fragments were obtained from the 19.5 kDa polypeptidein fraction 1 (Fl) of the sucrose gradient (Fig. 4.2, Fl). This polypeptide was easily purifiedfrom the other three dominant FCPs normally present in the thylakoids (Fig. 4.2, Chapter 2).The 19.5 kDa polypeptide was immunologically related to the diatom FCP sequence (Fig. 4.2,93Fl). A total of 60 amino acid positions were obtained from the sequencing of seven trypticfragments. Only two of these tryptic fragments (Ti and T2) could be unambiguously alignedwith the Phaeodactylum FCP sequence and both were within the first putative membranespanning region. The sequence information from tryptic fragments Ti and T2 was used to createdegenerate primers for the amplification of a Fcp specific probe; this was used to screen thecDNA library.30 —21.5 —Figure 4.2Thy Fl Tryptic FragmentsT1- T V E I KT2-. XSMLAYLG?FLVTFAGT3- LPGAYDLAGDQFSSLP1’T4- S/LPGAYDLAGDQFT5- E E L/A E/A G/A GX E AT6- IINSLLGSPVEFT7- IINSLLGSPVE/DFNAGFHTryptic fragment sequences determined from the abundant protein in fraction 1 (Fl) ofthe sucrose gradient (Chapter 2). The left panel shows a western blot of a thylakoid fraction(Thy) and fraction 1 (Fl) immunostained with the ct-FCP antiserum. Molecular mass standardsin kDa are indicated on the left. Sequences from the tryptic fragments Ti -T7 are indicated onthe right. Ambiguous amino acids are indicated by a question mark.94More than 90 000 recombinant cDNA clones were screened with the PCR amplifiedprobe resulting in the detection of 3 clones. However, all were truncated upstream of the firstputative membrane spanning region. A re-screening with one of the truncated cDNA clonesdetected an additional 30 positive clones of which 15 were isolated; most were smaller than theoriginal. The RACE technique was used to amplify the 5’ end of the Fcp cDNA directly frompolyA+ rnRNA using a reverse transcription reaction followed by PCR amplification (see Fig.4.1, methods). A number of nearly identical fragments differing in only a few nucleotides,mainly within the third codon position, were obtained from the amplification reaction. TheRACE product that was identical to the truncated cDNA clone in the 200 bp overlapping regionwas used to generate the full length sequence. The sequencing strategy and the proportion of thefull length sequence determined from the RACE product and the cDNA clone are indicated inFigure 4.3.The cDNA encoding the FCP (Fcp 1) is 858 base pairs (bp) long with an open readingframe of 625 nucleotides (Fig. 4.4). This gives an immature polypeptide with 210 amino acidresidues which is typical for the FCPs (Grossman et al. 1990; Apt et al. 1994). The proposedtransit peptide cleavage site is at serine 36 (Fig 4.4), by analogy to the Phaeodactylumprocessing site (Bhaya and Grossman 1991). With the cleavage site at Ser 36, the mature proteinwould have a calculated molecular mass of 18.9 kDa. This is very close to the estimated size of19.5 kDa based on the SDS polyacrylamide gel (Fig. 4.2; Chapter 2). A typical polyadenylationsignal at the 3’ end of the cDNAs (AATAAA) was not present in any of the cDNA clones.In line with the proposed nomenclature for the Cab genes (Jansson et al. 1992), I use theterm Fcp 1 to refer to a specific cDNA/gene type (type 1) that includes all the nearly identicalmembers. As I have only a single complete cDNA of Fcp type 1, this term refers directly to thesequence shown in Figure 4.4.The tryptic fragment sequences were identical to the inferred protein sequence in allpositions except for two from fragment T2. These residues are indicated by square brackets [unFigure 4.4. The first conflict has a Tyr [Y] instead of the inferred Ile 82 (I). However, the latter95ORFnon-codingBstX BstElI regionBsmA Styl BssHII PvuIl Apal-100 200 300 400 500 600 700 800___________RACE_ProductcDNA cloneFigure 4.3Schematic representation of the Fcp 1 cDNA and the sequencing strategy used. Shading representsthe FCP open reading frame (ORF). The sections of the Fcp sequence determined from the RACEproduct and the cDNA clone are indicated. Key restriction sites are given and the arrows indicatethe sequencing done on both strands.96•+l/1 •28/1ONucleotide sequence of the Fcp 1 cDNA from Heterosigma. The nucleotide and amino acidpositions are given above the first nucleotide of the codon (•). Tryptic fragment sequences areshown in bold below the appropriate amino acid. Ambiguous amino acids are indicated by aquestion mark (?). Amino acids not matching the derived gene sequence are indicated by squarebrackets El. A putative processing site is indicated by an open triangle (A.aaa atg tct ctc aag ctc gcc acc ctc gct gcc gcc ctc atg ggt gcc tcc gcc ttc gtgmet ser leu lys leu ala thr leu ala ala ala leu met gly ala ser ala phe val•58/20 •88/30gcc ccc aac aag atg ggc gtg gcc atg agc ttc gag aacala pro asn lys met gly val ala met ser phe glu asnzX.118/40gag atc ggc gtc cag gcc cct ctg ctt ctg gat gag gccglu ile gly val gln ala pro leu leu leu asp glu ala•178/60gac cag gag cgg ttc gac cgc ctc ggc cgc atc tcc atgasp gln glu arg phe asp arg leu gly arg ile ser metX S M.238/80ctt gcc atc ctt ggc cac ctg gtg cct ggc gct tac gacleu ala ile leu gly his leu val pro gly ala tyr aspL A [Y] L G F? L V P G A Y D.298/100ctg gct ggc gat cag ttc tcc agc ctg tct gct ctg cctleu ala gly asp gin phe ser ser leu ser ala leu proL A G D Q F S S•358/120gct gct ggt gtg gcc cag acc atc ctg ggt ttt gct cagala ala gly val ala gln thr ile leu gly phe ala gln•4l8/140atc aag gag gag ctg gag gct gac gcc ggg tgg gat gacile lys glu glu leu glu ala asp ala gly trp asp aspE E L? E? G? G?.478/160gag aag aag gac tcc aag cgc gca gcc gcc cag atg ggtglu lys lys asp ser lys arg ala ala ala gin met gly•538/180atc ctt gcc ttg atg gtc cac gag atc atc aac tct ctgile leu ala leu met val his glu ile ile asn ser leuI I N S L•598/200ctg ggc tcc cct gtg gac ttc aac ttt ttt cct tct ccaleu gly ser pro val asp phe asnL G S P VE/D F Naaa att ttt gac ttt ctc ttg accggg ttg ctg tgc tgg gtt agt tcgtct ttt gtt ggt atc ttg act tgattt gaa cta tag tac cct tct ttcFigure 4.4aag tcc tcc tcc gcc ctg aaglys ser ser ser ala leu lys•l48/50ggc ttc tgg gac cct ctg ggcgly phe trp asp pro leu gly•208/70cgt acc gtc gag atc aag cacarg thr val glu ile lys hisT V El K•268/90acc acc gct ggt gtg cgt ctgthr thr ala gly val arg leuT [F]A G? L•328/llOctg ccc acc ggc ctg aag gctleu pro thr gly leu lys alaL PT.388/130ggc ttc att ggt ctg att gaggly phe ile gly leu ile glu.448/150tgc gag gcc cgc atg gac gctcys glu ala arg met asp alaX E A.508/170att gag ctg aac aac ggc cgtile glu leu asn asn gly arg•568/190cag ctg gac aac aac cct tacgin leu asp asn asn pro tyr•628/210gct ggc ttc taa aca gtg tttala gly phe OCHA G F H?ttc tga ctt ttc tgg gcc cag gct ctt tct agt gttcag ttt gta gcc tga tgc gtg ttt cat cat ata gacctt ctt ttg ctt gac tgc tca caa tct ttc ata cac.85897is consistent with the other FCP sequences as there is an Tie and a Val at an homologous positionin the FCPs from both Macrocystis and Phaeodactylum, respectively. The other conflictinvolves a Phe [Fj in the sequence of T2 but a Thr 89 (T) in the inferred protein. At homologouspositions in the other FCP sequences, there is either a glutamate (E), a glutamine (Q) or anarginine (R) residue. These conflicts could be the result of sequencing ambiguities.Alternatively, since the FCPs are encoded by several nearly identical multigene family members,the differences may reflect true sequence polymorphisms.A second cDNA clone from the first round of screening was significantly divergent fromthe first cDNA (Fcpl). An alignment of this second clone (Fcp2) with Fcpl is shown inFigure 4.5. Because Fcp2 is not full length, the alignment starts at amino acid position 73. Thetwo clones are 75% identical at the nucleotide level and 70% identical at the amino acid level.When the chemical similarities of the amino acids are considered, Fcp 1 and Fcp2 areapproximately 82% similar. Most of the differences between the two clones occur in theconnectors linking the putative membrane spanning regions (MSR). Several changes in thesecond MSR were also observed, though most were conservative substitutions. On the otherhand, few amino acid changes are observed in the first and third MSRs. Because this sequenceshows significant differences from the main Fcp 1 it may represent a different Fcp type;therefore, it is referred to as Fcp2 (Fcp type 2). However, such a designation requires furthercharacterization.A wide band with an average size of 0.98kb is detected using the Fcpl clone to probetotal RNA (Fig. 4.6); this is consistent with the size of the cDNA clone. The wide hybridizationsignal indicates that there is probably a collection of related transcripts with small sizevariations. This is consistent with the variable sizes of the 3’ non-coding region in different FcpcDNA clones (not shown). The strong hybridization signal obtained in a short period of time(8 hrs) indicates the Fcp genes are abundantly expressed.An analysis of the codon usage is given in table 4. 1A. In the third position of the codonthere is an overall preference for pyrimidines (C + T), representing over 67% of the codons98•217/73 .247/83K H G R I S M L A I L G H L V T T A G Vaag cac ggc cgc atc tcc atg ctt gcc atc ct ggc cac ctg gtg acc acc gct ggt gtgaag cat ggc cgc atc tct atg ctt gcc atc ctt ggc cac atc ctc acc act gcc ggt gctK H G R I S M L A I L G H I L T T A G A.457/153 •487/163gac gct gcc ggg tgg gat gac gag aag aag gac tcc aag cgc gca att gag ctg aac aacgat gag cag ggc tgg gat gag gcc aag aag gac tcc aag cgc gcc att gag ctg aac aacD E Q G W D E A K K D S K R A I E L NN•517/173 .547/183G R A A Q M G I L A L N V H S Qggc cgt gcc gcc cag atg ggt atc ctt gcc ttg atg gtc cac gag cagggc cgt gct gct cag atg ggt atc ctc gct ctg atg gtg cac gag accG R A A 0 M C I L A L M V H E Tgtgtttttttttccttctccaaaaatttttgactttctct>>tcctttatcctgtaacataattttaaatgaccccatagag4tFigure 4.5L D N Nctg gac aac aacatc aac aac gatI N N DComparison of the Fcpl sequence (fig.4.4) with the Fcp2 partial sequence. Fcpl aminoacid and nucleotide sequences (top two lines) are displayed starting at amino acid 73, accordingto the labeling in figure 4.4. The partial Fcp2 nucleotide and amino acid sequences are in thebottom two lines of each row. Nucleotide/amino acid positions are indicated above theappropriate residues. Lines above and below selected residues indicate the potential membranespanning regions. Amino acids and nucleotides in bold indicate areas differing between the twosequences. The # at the end of Fcp2 indicates the start of the poly A-tail. The>> symbol at theend of Fcpl indicates the presence of additional sequence that is not shown..277 /93 •307/103R L P G A Y D L A G D Q F S S L P T C Lcgt ctg cct ggc gct tac gac ctg gct ggc gat cag ttc tcc agc ctg ccc acc ggc ctgcgc tgg cct ggt gcc gtg gat ctg tcc ggc aag aca tac gcc gag atc cct gct ggt atcR W P C A V C L S C K T Y A E I P A G I•337/113 .367/123K A L S A L P A A G V A Q T I G F I G Laag gct ctg tct gct ctg cct gct gct ggt gtg gcc cag acc atc ggc ttc att ggt ctgaag gcc ctt ggt gcc ctt cct ttt gct ggc gtc tgc cag att gtg gcc ttc att ggt ctgK A L G A L P F A G V C 0 I V A F T 0•397/133I E L C F Aatt gag ctg ggt ttt gctatt gag ctt ggt ttc tccI E L C F S•427/143D A A G WQ I K E E L E A D C E A R Ncag atc aag gag gag ctg gag gct gac tgc gag gcc cgc atgaag tgc cag gat gac gtg gca gcg ttc tgc gag ggc aag atgK C Q D D V A A F C E G K ND D E K K C S K R A I S L N N•577/193 •607/203P Y I I N S L L G S P V D F N A G F Zcct tac atc atc aac tct ctg ctg ggc tcc cct gtg gac ttc aac gct ggc ttc taa acacct tat gtg atc aac tct ctc ctc ggt gcc cct gtg gac ttc aac gcc ggc ttc taa tttP Y V I N S L L G A P V D F N A C F Z99lOpg 2Opg— 4.5— 2.4— 0.24Figure 4.6Northern blot of Heterosigma total RNA probed with the Fcpl cDNA probe. 10 and20 .tg of total RNA were loaded. Molecular mass markers (kb) are shown.100Table 4.1A Codon usage of the Heterosigma Fcp 1 cDNA210 codons MM: 22314 DaltonTTT phe FTTC phe FTTA leu LTTG leu LCTT leu LCTC leu LCTA leu LCTG leu LTCT ser STCC ser STCA ser STCG ser S4 CCTproP5 CCCproP- CCAproP18 CCGproP3 TATtyrY8 TACtyrY- TAAOCHZ- TAGAMBZ6 CAThisH2 CAChisH- CAAg1nQ- CAGg1nQ3413 ACTthrT9 ACCthrT- ACAthrT8 ACGthrT- AGTserS8 AGCserS- AGAargR10 AGGargRGTT val VGTC val VGTA val VGTG val VGCT alaAGCC ala AGCA ala AGCG ala A12 GAT asp D17 GACaspD1 GAAg1uE- GAGg1uE6131Table 4. lB Third Codon Position Datanucleotide group total (%) G + C%Pyrimidines 67C 71T 2980Purines 33G 97A 318112- TGTcysC2 TGCcysC1 TGAOPAZ- TGGtrpW- CGTargR3 CGCargR- CGAargR7 CGGargRATT ile IATC ile IATA ile IATG met M- AATasnN6 AACasnN- AAA1ysK- AAGIysK3623 GGTg1yG10 GGCg1yG- GGAg1yG13 GGGg1yG101(Table 4. 1B). The pyrimidine bias is particularly evident in the alanine and glycine codonusage. Cytosine accounts for 71% of the codons in the third position when a pyrimidine occurs.Likewise, purines (G + A) occur only 33% of the time in the third codon position, though thereis an extreme bias (97%) for guanine over adenine. Overall, there is a strong bias for guanineand cytosine in the third position, occurring 80% of the time. Comparatively, the G + C contentof the entire Fcp 1 cDNA is 63%, though it is only approximately 40% in the non-coding region.The codon usage in the partial Fcp2 clone does not appear to be significantly different from thatin the Fcp 1 clone.4.3.2 Characterization of the Fcp gene familyFifteen truncated cDNA clones of various lengths were isolated and sequenced. Six werederived from unique genes as determined by differences in the untranslated region at the 3 endof the cDNA (Figure 4.7). Of these six, five were nearly identical in the coding region and weretherefore considered to be the same type (Fcp1*15). Preliminary sequencing evidence indicatesthere is at least one additional unique Fcp cDNA clone (Fcpl*6). The asterisk followed by anumber indicates a different cDNA representative of the same gene type based on sequencecomparison. However, as these cDNAs are not full length, the designation of these cDNAs asbeing nearly identical to the full length Fcpl cDNA clone requires further analysis. The valuesto the right of Figure 4.7 indicate the number of identical copies isolated for each of the eDNAclones. Three copies of Fcp 1 (Fcp 1 * 1 in Fig. 4.7) were isolated. In addition, six clones with anuntranslated region identical to Fcpl*3 were isolated and two equivalent cDNA clones ofanother type were observed (Fcpl*5, Fig. 4.7). The isolation of these different cDNAs indicatesthat the major FCP protein is encoded by a multigene family as in the terrestrial plants (Green etal. 1991) and diatoms (Grossman et al. 1990).Hybridization of the Fcp2 probe to genomic DNA detected 4-7 bands, depending on therestriction enzyme used (Fig. 4.8, lane 2). When the Fcpl cDNA was used as a hybridization102No.copiesdetectedFcpl*13Fcpl*26Fcpl*32Fcpl*41Fcpl*51Fcp21SequencealignmentoftruncatedcDNAclonesstarting18bpbeforethestopcodon(bold).Codingregionisshowninuppercase,whilethenoncodingregionisinlowercase.Thenumber ofidenticalcDNAclonesisolatedforeachisgiven.A(#)denotesthesiteofthepoiyA-tail.A(>>)indicatesonlyaportionofthe3’endhasbeenshown.codingregionEFNAGFZ•.GAcTTcAAcccTGGcTTcTAAacagtgttttttttttccttctccaaaaatttttgactttctcttgaccttctgacttttctgggcc>>•..GAcTTcAAcocTGGcTTcTAAatcaagtctctaattgtaacattagtgtgctgtttttggcctcgtttcattttggagctggctgaca>>•GAcTTcAAcGccGoTTTcTAAatacaatctagttggtctggattgcgatcccgcctctgggttggcgttgaggatgccatagtggact>>•GAGTTcAAcGccGGcTTcTAgcgatgttcctgttctcatgacttgcaaattttccaactggaatggactttcttgtaacctttcggg>>•..GAcTTczAcGcTGGcTTcTAAgcagtgttttttcttctctttcccaaaattgacccccctgaccttctgacttttctgggcttgggcc>>•.•GAcTTcAAcGccGGcTTcTAtttcctttatcctgtaacataattttaaatgaccccatagag#Figure4.71.35—1.1—Figure 4.8Southern blots of Heterosigma genomic DNA probed with the Fcpl cDNA (1) or theFcp2 cDNA (2). Genomic DNA was digested with Sac I, Dra I or both Sac I and Dra I, asindicated. Molecular mass markers (kb) are indicated on the left. Open arrows indicate twoDNA fragments that appear to hybridize to both Fcp cDNA probes.12 129.4—6.4—2.3—2.0—12DralSad SadD ra I104probe there were approximately 11 to 17 fragments detected (lane 1). These fragments ranged insize from ito 18 kb. However, not all the hybridization signals were of the same intensity.This could be due to multiple gene copies on the fragment and/or to the close migration of morethan one DNA segment. At this washing stringency (iX SSC, 65°C) the Fcpl and Fcp2 probesappear to hybridize to fragments of the same size such as those indicated by the open triangles inFigure 4.8. This could occur as the result of the two genes being linked on the same fragment.Alternatively, hybridization to genes of intermediate sequence divergence could account forfragments of weaker intensity and for some of the shared hybridizing bands in both lanes. Inaddition, bands with strong hybridization signals may have multiple copies of one gene type onthat fragment. Overall, there are multiple copies of both Fcp gene types (Fcpl and Fcp2), theFcp 1 type having more members. This indicated the existence of a very large Fcp multigenefamily in Heterosigma.A series of genomic Southern blots were done in order to investigate the complexity ofthis multigene family more thoroughly. Figure 4.9 shows a single blot that was sequentiallywashed at increasingly stringent conditions, as labeled. The probe was a Sty IJBssH II fragmentfrom Fcpl (see Fig 4.3), which included sequences from the middle portion of the putativeMSR1 to the middle of MSR3. At the low stringency wash (1X SSC, 65°C) there are nineprominently hybridizing bands in the 3.6-18 kb size range and a few smaller hybridizing bands(1.4-3.3 kb) in the Sac I digest. The Dra 1(D) digest shows over 20 hybridizing bands, as doesthe Sac L’Dra I double digest. At the next level of stringency (0.iX SSC, 0.18 M Na) there waslittle obvious change in the number of hybridizing bands or in the signal ratio between them inthe Sac I digest. In the Dra I digest one hybridizing band was lost (4.1 kb) and the signalintensity of five other bands decreased relative to the others (12, 8, 5.5, 5.4, and 3.6 kb). In thedouble digest, three hybridizing bands were removed (8, 5.8 and 2.2 kb) and a fourth had areduced signal (3.8 kb). The hybridization conditions described above were both below thetheoretical Tm of the probe, which was 93°C at 0.18 M Na (1X SSC) and 77°C at 0.0 18 M Na(0.iX SSC).1056.44.423—2.0—1 35—1.1Figure 4.9SSID D SS/D De0.01X SSC65°CSouthern blots of Heterosigma genomic DNA probed with a32P-labeled Sty L[BssH IIFcp 1 cDNA fragment. The blot was successively washed under increasingly stringentconditions as indicated on the bottom of each panel. Molecular size markers (kb) are indicatedon the left of the first panel. The restriction enzymes used were as follows: 5, Sac I; S/D,Sac JIDra I; D, Dra I.94—SS/D D1XSSC 0.1XSSC65°C 65°C106The final washing condition (0.0 1X SSC, 65°C) exceeded the theoretical Tm of the probe(6 1°C at 0.00 18 M Na). After this wash, the probe was selectively removed from a number offragments. Three fragments with a strong signal (6.6, 5.5, and 3.6 kb) remained in the Sac Idigest after the 0.O1X SSC wash, along with one weakly hybridizing fragment (4.5 kb). In theDral lane, 8 prominent hybridizing bands remained with a size of 4.5, 3.8, 3.0, 2.5, 1.9, 1.8, 1.3,and 1.2 kb. Six main bands were detectable in the Sac IJDra I digest (4.5, 3, 2.5, 1.9, 1.3, and1.2 kb). Though it is difficult to get an accurate estimate of gene copy number using genomicSoutherns, there appear to be 6-8 copies of the Fcp 1 gene. This is in good agreement with thenumber of characterized cDNA clones.There seem to be as many as 20 or more related Fcp genes present on the nucleargenome of Heterosigma, as seen in the number of hybridizing fragments at the lower stringencywashes. These may represent different Fcp gene types similar to the gene types in the Cabfamily. However, the presence of additional hybridizing bands due to the existence ofpseudogenes cannot be ruled out. Though care was taken to achieve complete digestion and thepattern was repeatable, some of the weakly hybridizing bands may represent intermediates of anincomplete digestion. Neither Sac I or Dra I cut within the full length Fcpl cDNA clone,however, internal restriction sites in other genes could cause an over-representation of detectablebands.4.3.3 Characterization of the FCP protein sequenceThe FCP protein is predicted to have three membrane spanning regions using the Kyteand Doolittle scale of amino acid hydrophobicity (Kyte and Doolittle 1982) with a window sizeof 19 (K-D plot, Fig. 4. bA). The hydrophobic regions of the mature protein detected in thehydropathy plot include residues 73-96, 106-141, 173-192, and the C-terminus (200-210) (aslabeled in Fig. 4.4, 4.10, 4.12). A fourth hydrophobic region is also detected in the aminoterminus of the protein (residues 1-28), which is entirely within the predicted transit sequence107(TS) domain. The membrane spanning regions (MSR), as labeled in Figure 4. bA, werepredicted by comparison to the K-D plot and by comparing the FCP sequence with the CABs(Fig. 4.13). The regions that span the thylakoid membrane in the pea LHC TI structure(Kühlbrandt et al. 1994) are conserved in the FCPs; therefore, a similar topology was thought tobe a reasonable assumption (see Fig. 4.15 for model). A feature of the prediction is the presenceof distinct hydrophilic domains at the start of the putative membrane spanning regions 1 and 3(Fig. 4.10, 4.15). These areas may protrude from the bilayer in a continuation of the ct—helicalstructure, as appears to be the case with the pea LHC II (KUhlbrandt and Wang 1991;Kühlbrandt et al. 1994). Figure 4. lOB shows the distribution of acidic and basic amino acidswithin the protein by vertical bars. Acidic amino acids (top panel (A) of Fig. 4. lOB) whichinclude glutamic acid (full bar) and aspartic acid (intermediate bar). The positions of the basicamino acids (bottom panel (B) of Fig. 4. lOB) are also shown and include arginine (full bar),lysine (intermediate bar), and histidine (short bar). There is an enrichment for acidic and basicamino acids in the regions preceding, and at the start of, the predicted MSRs which are locatedon the stromal side of the membrane. Stroma exposed areas are indicated by shading in Figure4.10. The areas exposed to the thylakoid lumen are indicated by the hatched regions (Fig. 4.10).Few acidic and basic residues are present in the predicted lumen exposed sections of the FCP(Fig. 4. lOB). There are a few occurrences of acidic and basic residues within the membraneexposed portion of the protein, some of which are probably involved in the binding ofchlorophyll (discussed later).The first and third putative membrane spanning regions of the FCP are related as is thecase for the CABs; first recognized in a tomato Cab gene by Hoffman et al. (1987). Thisinternal similarity between the two regions is suggestive of a gene duplication. These regions(MSR1 and MSR3) are approximately 49% similar to one another (Fig. 4.11), comparable to thedegree of relatedness between MSR1 and MSR3 of the CAB proteins. With the CABs,similarity between MSR1 and MSR3 extends into the domains immediately preceding the startof the putative transmembrane c-helix. This is not the case with the FCPs.108A100 200B IIIII,,JIIJI 11)11A iiit AB iIIiI1ii1!.iiIiiiI!i ii.IqiII ,t;::” BTopological analysis of the Heterosigma Fcp 1 full length sequence. A) Kyte-Doolittlehydropathy plot done using a sliding window of 19 amino acids. Hydrophobic areas areassigned a positive value and hydrophilic ones are negative The transit sequence (TS) and theputative membrane spanning regions (MSR1-3) are labeled and roughly correspond to the clearareas. B) an acidic-basic map of the FCP protein. The position of the acidic amino acids, areindicated in the top panel (A) by a full vertical bar (E-glutamic acid) and an intermediate verticalbar (D-aspartic acid). The bottom panel (B) shows the position of the basic amino acids with afull bar (R-arginine), an intermediate bar (K-lysine), and a short bar (H-histidine). In all threepanels, shading represents regions of the protein exposed to the stroma of the chioroplast.Hatches indicate areas that are exposed to the thylakoid lumen.TS MSR1 MSR2 MSR3-2-1.5—1-0.500.511.52Figure 4.10100 200109MSR1 61 DQERFDRLRTVEIKKHGRISMLAILGHLVTTAGVRLP 95MSR3 159 DDEKKDSKRAIEL NNGRA1QMGILALMVHEQLDNNP 193Figure 4.11Alignment of the first and third putative membrane spanning regions of theHeterosigma FCP. Similar amino acids are in bold. Numbers indicate the positions of thefirst and last amino acids shown, relative to the entire protein.The alignments in figures 4.12 and 4.13 were done with MACAW (Schuler et al. 1990)and adjusted by hand to maximize similarity. The Chi a + c-binding proteins (FCPs/iPCPs) arevery well conserved between the different algal taxa. The Heterosigma FCP sequence is roughly77% similar to the other known algal FCP sequences. The boxed regions in Figure 4.12represent similarities between the FCPs and the iPCPs. Shaded regions are amino acidsconserved only in the FCPs, excluding the Isochrysis sequence. The greatest similarity betweenthe Chl a + c-binding proteins is within the putative membrane spanning regions; particularlyMSR1 and MSR2 (Fig. 4.12). The N-terminal portion preceding MSR1 is also very highlyconserved, though the analogous region in front of MSR3 is not. This is contrary to therelationships between the stromal exposed regions in front of MSR 1 and MSR3 in the CABs(see Fig. 4.13).Amino acid similarities between the CABs, FCPs and iPCPs are indicated by the boxedregions in Figure 4.13. The Heterosigma FCP sequence is approximately 40% similar to boththe tomato LHC I and LHC II sequences, when the unambiguously aligned positions arecompared (see Fig. 5.1, chapter 5). For comparison, the LHC I and LHC II sequences fromtomato are approximately 57% similar using the same amino acid positions in the calculation.The Heterosigina FCP sequence shows the greatest similarity to the green algal and terrestrialplant CAB sequences, almost exclusively within the predicted membrane spanning regions ofthe mature protein, primarily within the first and third MSRs. However, the sequenceconservation between the CABs and the FCPs does extend into the stromal side of the first MSR110iPcp-AcFcp-HcFcpl-MpFcp-LsFcp-OsFcpl-PtFcp-IgFcp-P1MSR3,.WZWZZZWZZZZZZZZWZZi160 170 180 190 200SK1DLER [ 4MIIG!4F ET-GSA1t)S jKIELNNGR A IIAL 7IELD-NNPMS1’QAS KPELNNGR Jt4-NKPAMS$TEASjKBNGRA fl illTF*rKLS 11 $L . ttV43-GSI PIVGEMTFS1KKLQ C1J1ELNQGR tL2LI4 jVHE-VSI LPPSDEFRL WGPYWGDATFFigure 4.12Amino acid alignment of the Chi a+c-binding proteins from chromophytic algae. Taxainclude: Ac, Amphidinium carterae; He, Heterosigma carterae; Mp, Macrocystis pyrfera;Ls, Laminaria saccharina; Os, Odontella sinensis; Pt, Phaeodactylum tricornutum; 1g.Isochrysis galbana and P1, Paviova lutherii. Hatched boxes indicate putative membranespanning regions (MSR 1-3). Boxed regions indicate sequence similarities between theFCP and iPCP sequences. Shaded regions indicate FCP similar regions, except Ig and P1.Amino acids are numbered with reference to the Heterosigma Fcpl sequence as in fig. 4.4.iPep-AcFcp-HcFcpl-MpFcp-LsFcp-OsFcpl-PtFcp-IgFcp-P1ipcp-AcFcp-HcFcpl -MpFcp-LsFcp-OsFcpl-PtFcp-IgFcp-P1v/I1 10 20 30 40 50 60IESELGV IPTGFWDPLGL AKMKKNSLKLAT FAAALM- -GA SAFVAPNKNG VAKSSSALKM ENEZGV IDPG WEI4I4DMKSAVMAVA CAAAPGFRGP SAFNGAAL SAKACSANKM E2IGA {GpPGI 1A442EI*DPt.GI LDtIFMKL--- AIAALLAGSA AAF-APAQ-- SGKASTALNM E iLG NMKF- - - AVFAFLLASA AAF-APAQQ- SARTSVATNM GA 1tGt VADtMMTL--- ASLPSTAIAG LASAAPKVQ- -PRMAANDEF -GL G GDPAGLKGDLEVYA ILGFYDPLGL LDNEEYEMSR170 80 90 100 110R IKHG RI TPEL KFP G -YLSPSMG L1EDIPIL IWTTAG t1 -AYDL-AG DSSLPjtQQN- LP G -MLSNSAN LFtADMPRLRVàVKHG RI r4I’QQN- -NLSNSARIRYVEVKRG RIAIW 1VTRNG IDYAGNSF DPNGW I PPtVYJHG R QEAG P -DIDYS-G TESIP F{GQEK-- --PLFSGDNG PAIEQIPRMVE LMSR2-wzw/zw//zzz//zI120 130 140 150siz] IJjY cQDQSEG SAGEAGDFGF KVLTSGI A IGEX I FAQIKE ELEAD--CEA RNSdXAGI I I VMKNVE GS---FPGF TLG----GNPDIG I AFV I VMKDVT G-EGEFPGF R NGASd IAFI E VMKDIT G--GEFVGDF R NNYQI.PYWLW I IG-f...RIQKGW AKVNPETGKA DSALREGYEPEV1J Q AFV LEL V-KDVT G-EGDFPGDFDAAGWD.FGASW---- --LDFGW1-. -LDFGWP—--- -GDLGFDPLGL A200iPcp-AcFcp-HcFcpl-MpFcp-LsFcp-OsFcpl-PtFcp- IgFcp-P1YGDWANFTAS PLRYIINSLLGSP VDFNAGFYVINDLVGAS YTFN111Figure 4.13Amino acid alignment of select Chl a+b and Chl a+c binding proteins from terrestrialplants, green algae, and chromophytes. Taxa include: Le, Lycopersicon esculentum; Cr,Chlainydomonas reinhardtii, Egi, Euglena gracilis; Ms, Mantoniella squamata; Ac,Amphidinium carterae; Hc. Heterosigma carterae; Mp, Macrocystis pyrifera and Pt,Phaeodaciylum tricornutum. Hatched boxes indicate putative membrane spanning regions(MSR 1-3). Boxes indicate similarities between Fcps, iPcps and Cabs. Dashes (-)represent gaps introduced to increase similarity. Proposed Chi a (A) and b (, ) bindingresidues of the pea LHCII are indicated. Amino acids are numbered with reference tothe Lhcbl-Le sequence, consistent with the numbering used in figure 4.14.11210 20 30 40 50Lhcal-LeLhcbl-LeLhcb-CrLhcb-EglLhcb-MsiPcp-AcFcp-HcFcpl-MpFcpl-PtMSR1-////-60 70 80LFYESI *VP IIVPLGLARELE\\I CPNLGL GDVFPLLAR1RlETEII {APNI4rL PLLAKL*tREAIEt1I 4ZNLGL 7\\1t11PtELLAGS AE I G GTG90 100GN- -WVKAQE WAAIPGGQATNGV-KFGEAV WFKAGSGT-KFGEAV WFKAGNGVPFGEGAV WYKGIP WFTAGTLCTPFPGLPGLPGLPGSEGGLDYLGNSEGGLDYLGNSADGLNYLGNP-GAVAPLAPSMGLKYEDIP-AGDQFSSLPSANLSFADMPS-GTSFESIPI I ILhcal-Le PSFLSSTKSK FAAAMPVSVG ATNSNSRFSM SADWMPGQPRLhcbl-Le GNGRITNRKA VARSA PSSSPWYGPI) RVKYLGPFSGLhcb-Cr LQVTCKATGK KTAAKARAPK SSGVEFYGPN RAKWLGPYSELhcb-Egl MLATSGRKA KAAPKS DNLSQWYGPD RARWLGPLLhcb-Ns N ACIASSFVGS VAALKATKVQ AKSVSTVVKAiPcp-AcFcp-Hc MSLKL--AT FA-AALMGASFcpl-Mp MKSAVNAVAC AAAPGFRGPSFcpl-Pt MKF AVFAFLLASA- -- PSYLIDES-PSYLTNATPAYLTEV-PSYLIDIYPEFGTFESEL-- -FENEI---FESEI-- -FENEIAFVAPNKMGVAFNGAALTTSAAFAPAQ-QS333‘3‘33 APGDFG3 FPGDYC3 FPGDYC3 LPGDYC- YPG--CT AAPTGFJ QAPLGP QAPLGQQPLGAKSSSALFiSACSAMFJISARTSVATNMA3EVPA--NtrA IPE--TWLtTA IPE--TWP’TA 43SPT--TGESP IIPF GSCAfrDPI IG14DQERfrJDPLI GLL4ArDQER13DQEK110YLGQPVPW- -SQIFAQIFAQIFDDCTAVADKFYLSPAYDLMLSNDIDYMSR2120GIPSLVHAQIPSLVHAQIPSLIHAQEGSGY-PSFNGLGAISTGLKALSNGVAALS INGFAALS130? TILAIEFI -iw vfvfr -J IrSA’IILM-I LTFLSTIIM -1 jAvLAIEv\\4rV -IGWIILY?P tGVAII3F ‘P PAGLAIFAF IP GIAI1kF IGCCFCCCYPGL YPGL YPGDTL----YPGVGDV---SPGFGFKVLT-- -CEARMGDFTLG-GNPGDFR---NNY160GA--FDPLGYGS--FDPLGLES--FDPLGLGP--FDPLGLGR- -FDSLGLDAAGWD-- - -FGASWD-- --LDFGWD---Lhcal-LeLhcbl-LeLhcb-CrLhcb-EglLhcb-MsiPcp-AcFcp-HcFcpl-MpFcpl-PtLhcal-LeLhcbl-LeLhcb-CrLhcb-EglLhcb-MsiPcp-AcFcp-HcFcpl-MpFcpl-Pt140 150FfQR-S NEKDSEKKKIEYRIA GGPLGEVVDPLIEYRVN GGPAGEGLDPTEYRYG GGVGDFGRELUEkYRTG LSDSPFEELTEj2SQDQ SEGSAGEAGDL1IE1LGFAQ IKEELEAD- -FILE1LAVMK NVEGS- --FPFAVMK DITG- -GEFVQQSAYPGTGPQAIjT-GKGPQAIIvT-GKGPQPL4P-KAGPQALAT-QEGPQDc4T-GSAYHE-NNPYHEIL-NKPYHEJ3-VSIL200MSR3170SKDPAKFEELAEDPFAELADDPLDjrFAELANDLAELAESG-LEELSKDEF43LERK-- -LEKKDSKAMSE14I’QASKTFSKKLQK210 220 230LLLADHLADVDDHLANVETFHLADIAJWQAHVADGDWANFTASPIILLGSPVVIDLVGASYP LiIPKGIFPNFVPGKTKFTPSAGFAASGFGFYPWHNNIGDVIPVNNNAWAFAPTVNNAPAFAPSANIFFTSPVHANVLTNALRDFNAGFTFN113(amino acid residues 44-5 1, Fig. 4.13). This primarily includes the WDPLGL motif of the FCPsto the (F/Y)DPLGL motif of the LHC I CABs or the (W/F)DTAGL motif of the LHC II CABs.In this case, the PL of the (F/Y)DPLGL motif is a signature sequence of the LHC I CABswhich, interestingly, also occurs in the FCPs. In the CAB sequences, there is an additionalFDPLGL motif in front of the third predicted MSR (residues 16 1-166; Fig.4.13), which is highlyconserved and is part of a local 2-fold symmetry between the MSR1 and MSR3 (Kühlbrandt etal. 1994). In the FCP/iPCP sequences, there are no obviously conserved areas in front of thethird putative MSR. However, the Isochrysis FCP sequence is an exception to this statement asit does have a FDPLGL motif in front of the third MSR, more closely resembling a CAB protein(Fig. 4.12).Figures 4.14 and 4.15 are designed to demonstrate the location of the conserved aminoacids with respect to the proposed structure of the LHC. A model of the pea sequence is shownin Figure 4.14 along with the identified chlorophyll molecules (represented by porphyrin rings)and the approximate location of the thylakoid membrane (hatched regions), though the latter hasnot been accurately determined. This diagram is labeled to show the amino acids conservedamongst all LHC types (CABs, FCPs, and iPCPs). The residues conserved between all knownCAB, FCP and iPCP sequences are indicated with black circles; most of these are within MSR1and MSR3. Significantly, many of the conserved residues are within the area of close contactbetween the two transmembrane c-helicies (MSR1 and MSR3). The area of close contactincludes residues Ser 69 to Ala 76 and Gly 184 to Met 191 (Fig. 4.14), as defined by Kühlbrandtet al. (1994). Other conserved residues are thought to function as Chl a ligands in pea, whichinclude the following residues: Glu 65, His 68, Gly 78 (backbone carbonyl), Glu 139, Glu 180,Asn 183 and Gin 197 (His in FCPs) (Fig. 4.13 and 4.14). All ligands thought to bind Chl a inpea (six in all) are conserved in the Chl a + c-binding proteins; presumably they would also beinvolved in Chl a binding.A putative model of the Heterosigma FCP sequence is illustrated in Figure 4.15. Thetopology was predicted by aligning the FCP sequence with the CAB LHC II sequences and then114modeled by analogy to the pea LHC II 3-D structure. It also shows the location of the conserved(chemically similar) amino acids when the FCPs and the iPCPs are compared. The FCP/iPCPsequences are mainly conserved within MSR1 and MSR3. The similarity within the secondputative MSR is significant but lower than when the first and third MSRs are compared. Thereis a lack of conserved residues in the lumen exposed portions of the protein and in the connectorbetween MSR2 and MSR3. Residues thought to bind Chl a, by analogy to the pea structure, arewell conserved (solid triangles, Fig. 4.13). The degree of sequence conservation in themembrane spanning regions is significant enough to suggest that the FCPs may have a similarstructural topology, though the aqueous exposed areas are not conserved and may be structurallydistinct.115Figure 4.14Amino acid comparisons based on the structural models for the pea LHC II complex.Porphyrin rings represent the approximate location of chlorophyll molecules determined fromthe pea LHC II structure. The approximate location of the thylakoid membrane in each Figure isindicated by the hatched regions. The pea LHC II sequence is compared to both the CABs andthe FCPs. The key to the conserved (similar) amino acids is given in the Figure. The Figure is amodification of Figure 4 from Kühlbrandt et al. (1994).3O©®®®© NMSR1 MSR3pea LHCII model150®Lumen120 1104(D)®®®®®®®®®® C Conserved in all CABs and FCP/PCPs220 © Conserved in all CABsQ Conserved in most CABs116Heterosigma FCP model® ®120®(j) ) D190 ©dOO• Conserved nail FCPsand PCPs‘-‘ ‘o Q Conserved in all (6/7), except 1g.Q Conserved in all FCPs. except 1g.Figure 4.15Model proposed for the Heterosigma FCP protein. Porphyrin rings represent theapproximate location of putative chlorophyll molecules by analogy to the pea LHC II structure.The location of the chlorophylls in the FCP model are purely speculative and are included onlyto illustrate the conservation of the residues binding these particular molecules in the peastructure. The approximate location of the thylakoid membrane is indicated by the hatchedregions. The Heterosigma FCP sequence is compared to the FCPs/iPCPs. The key to theconserved residues is indicated in the Figure.MSR1 MSR3(B) (A)50Thylakoidmembrane(C)Lumen1174.3.4 Analysis of the FCP transit sequenceThe N-terminus of the Heterosigma FCP was blocked; therefore, the precise cleavagesite of the transit peptide is not known. However, comparison to the Phaeodaciylum FCPcleavage site (Bhaya and Grossman 1991) would suggest that it occurs after Met 35 (Fig. 4.16),which is within a well conserved region. A comparison of the Heterosigma N-terminal region tothat of other chromophyte transit sequences is shown in Figure 4.16.Most of the chromophyte transit sequences have a basic amino acid within the first fourresidues (Fig. 4.16). This is followed immediately by a hydrophobic region (15-17 amino acidslong) emphasized by the black section of Figure 4.16. The CAB transit peptides lack both astrongly hydrophobic region and a positively charged amino acid at the N-terminus. A prolineresidue followed by a positively charged amino acid usually occurs at or near the end of thehydrophobic section (Fig 4.16). This proline may function as a helix breaker should thehydrophobic region form an cr-helix. After the hydrophobic region there is an increase in thenumber of hydrophilic residues, such as basic (Arg, Lys) and hydroxylated (Ser, Thr) aminoacids. Few, if any, acidic residues (Glu, Asp) are present within the targeting sequences. Abasic N-terminus, a hydrophobic region followed by a more polar section are all characteristicsof a signal sequence (von Heijne 1990). However, these traits are not present in the IsochrysisFCP sequence (LaRoche et al. 1994). Though there is a hydrophobic region (Fig 4.16), a basicamino acid in the first four residues is not present. This sequence also has a number ofhydroxylated amino acids at the beginning of the transit peptide (Fig 4.16), unlike the otherchromophyte sequences.Putative processing sites for the signal sequence-like regions are indicated in Figure 4.16(open triangles) and are based on the (-3,-1)-rules for the prediction of signal sequence cleavagesites (von Heijne 1986). There are other processing sites that can be predicted using the (-3,-i)-rules though only one is shown. The section immediately following the hypothetical processing118site in the FCPs, and prior to the start of the mature polypeptide, is enriched in hydroxylated andbasic amino acids (Fig 4.16). This amino acid composition is similar to the transit peptide ofchioroplast localized proteins of terrestrial plants. Presumably, the remaining region would thentarget the polypeptide to the chloroplast envelope once it crossed the ER-like membrane. Such abipartite transit sequence in the chromophytes has been suggested by Bhaya and Grossman(1991) and expanded on by Pancic and Strotmann (1993). In addition, the Chroomonasphycoerythrin transit sequence has two hydrophobic domains within it (Fig. 4.16). The firstregion is similar to a signal sequence and the second hydrophobic region resembles the transitsequence of a thylakoid lumen localized protein (Hiller et al. 1990).VVASSSALKM FTSAKSSANKM:VATNN AFSKASTALNM AFF.PAVGGATSNVFSESSSPAHRNRRTIVM DGVQPMA.ANDE FASNDMGEVV____________GAPAGAAN•AGAAAAVTPFigure 4.16Analysis of the N-terminal transit peptide from some chromophyte nuclear encoded sequences.The sequences are from the FCPs of; Hc, Heterosigma carterae; Mp, Macrocystis pyrifera;Pt, Phaeodactylum tricornutum; Os, Odontella sinensis; and Ig, Isochrysis galbana. Othersequences are the the Odontella sinensis- y-atpase (Os-a) and the Chroomonas (cryptophyte)phycoerythrin a-subunit (C-pe). Areas enclosed in black indicate hydrophobic regions. Shadedregions emphasize hydroxylated amino acids, serine and threonine. Basic amino acids (Arg, R;Lys, K) are underlined. Transit peptide cleavage sites are indicated by solid triangles (y). Opentriangles (V) represent possible signal peptide cleavage sites based on the -1 ,-3 rule.119HcMpPt IvOsOs-aIgC -peGAEPFNGAALLATFAAALMGASAFVAPAVMAIACAAAPGI4_AVIASLIAGAZAFAPALAIAALLAGSAAAFAPA.AIAGLASAAPLASLAVIGSAAAYVpNNIv]NMFZ4.4 Discussion4.4.1 Fcp cDNA structure and multigene familiesThe cloned Fcp cDNA (Fcp 1) encodes one of the predominant polypeptides present inthe thylakoids of Heterosigma, as confirmed by protein sequencing. The Fcp genes are nuclearencoded by a multigene family and the northern blot indicates they are highly expressed.Southern blot analyses of genomic DNA and cDNA sequencing indicate that there are at least 6-8 nearly identical copies of the Fcp 1 gene in Heterosigma. In addition, there appear to be over20 closely related Fcp gene sequences. This is comparable to the complexity observed in thetomato Cab gene family (table 1.2) where there are eight LHC II type I genes and greater than13 genes encoding LHC II proteins in total (Green et al. 1991).Similarly large LHC gene families have been reported in many other terrestrial planttaxa, such as petunia, where there are an estimated 16 Lhcbl genes (Dunsmuir et al. 1983).Multi-Lhcbl gene families are also present in moss (Long et al. 1989), green algae (Imbault etal. 1988; LaRoche et al. 1990) and in Euglena gracilis (Muchhal and Schwartzbach 1992). Inthe diatom Phaeodactylum there are at least six related members of the Fcp gene family (Bhayaand Grossman 1993). A similar number of Fcp cDNAs have been cloned and sequenced fromthe brown alga, Macrocystis (Apt et al. 1994). Moreover, the presence of multiple copies of themain LHC appears to be a general occurrence.The gene copy number of the primary LHC II (Lhcb 1) is quite variable between differentterrestrial plant taxa. One mechanism for the generation of multiple gene copies is throughunequal cross-over events which usually result in the creation of tandem repeats (Langridge1991). When this occurs, subsequent unequal cross-over events can lead to increases ordecreases in the size of the gene family. A high degree of similarity between the duplicatedmembers of the Fcp or Cab gene families may be the result of either a recent gene duplication120event, if there is similarity in the surrounding non-coding sequences, or through concertedevolution (Tanksley and Pichersky 1988; Bhaya and Grossman 1993).Large numbers of duplicated genes are usually assumed to be required for the generationof sufficient mRNA transcripts for the production of abundant proteins (Li 1983). As the FCPsare the most abundant proteins in the thylakoid membrane, this explanation seems reasonable.Alternatively, the presence of multiple gene copies may be less related to abundant proteinproduction but more to the lack of negative selection against multiple copies on the genome. Inthis scenario, only a minimum number of Cab genes may be required and selected for.I have named the more divergent Fcp sequence, Fcp2, to emphasize its distinction fromthe Fcpl sequences. The categorizing of the Fcp2 sequence as being a unique gene ‘type fromthe Fcp 1 sequences requires more detailed examination of the different FCP complexes and theirfunctions. However, as the immunological analysis in Chapter 2 suggested, there is quite anintricate antennal system in Heterosigma. This would suggest that many other Fcp gene ‘types’exist. If the sequence divergence between the various Cab gene types is any indication, it isunlikely that Southern blots using either the Fcp 1 or Fcp2 cDNA probes, under the previouslydescribed hybridization conditions, have detected the full extent of the Fcp gene family inHeterosigma. If distinct LHC I and LHC II antennae exist in Heterosigma, as is likely, then thesequence divergence between them may be too large to be easily detect by hybridization underthe conditions used in this study. The differential cross-reactions of the FCPs with the twodifferent antisera would also suggest this (Chapter 2). Bhaya and Grossman (1993) identifiedseveral Fcp genes that were present on two different genomic clones, which may be on the samechromosome. These sequences range in amino acid similarity from 86-99%. It would beinteresting to examine the functional differences between these clones, if any. The Fcp genefamily in Phaeodactylum does not appear to be as large as that in Heterosigma, based onreported Southern hybridizations (Bhaya and Grossman 1993).The Fcp genes of Heterosigma are highly expressed as indicated by the stronghybridization signal on the northern blot. The cDNA preferentially utilizes 28 of the possible121codons in a manner similar to the Phaeodactylurn Fcp (Grossman et al. 1990). The trends seenin the analysis of highly expressed yeast genes are generally consistent with the Fcp codon usagepatterns (Bennetzen and Hall 1982). The codon usage in the Heterosigma Fcp cDNA shows avery strong bias for G + C in the third codon position (80%, Table 4.1 B) and this trend isevident in the other Fcp cDNAs. However, the bias is not as prominent for the diatom Fcpgenes (Phaeodaciylum, 64%; Odontella, 72%) or the Isochrysis Fcp sequence (67%). TheMacrocystis Fcp cDNA shows extreme bias for G + C in the third position with approximately91% occurrence for the five Fcp genes. The total G + C content for all codon positions of theMacrocystis Fcp cDNA is, however, similar to the that of Heterosigma. The analysis of Cabgenes in angiosperms has shown a bias for G + C occurrence in the third codon of gymnosperms(Jansson and Gustafsson 1990) and monocots (Brinkmann et al. 1987) but not dicots.A high degree of codon bias for a particular gene has been correlated with the relativeconcentrations of particular tRNA molecules within the cell (Ikemura 1982). Codon bias is alsomore pronounced in abundantly expressed proteins in bacteria (Sharp and Li 1986) and yeast(Bennetzen and Hall 1982). Together, the coincidence of tRNA poois with non-random codonpreference in abundantly expressed proteins would be consistent with an increase in translationalefficiency. For highly expressed genes, the use of rare codons may result in a decreased rate orpremature termination of translation (Robinson et al. 1984) and be selected against. However,abundance of a particular protein did not appear to be correlated to the degree of codon bias interrestrial plants (monocots) (Campbell and Gown 1990) so the universality of such a proposalremains uncertain. Alternatively, the differences in G + C bias may be more related to theoccurrence of different G/C ratios within certain regions of the eukaryotic genorne, calledisochores (for example, see Sharp 1991), as occurs in mammalian nuclear DNA.The gene encoding the P-type ATPase from Heterosigma akashiwo (which is equivalentto H. carterae, see section 1.1) has just recently been published (Wada et al. 1994). It alsoshows codon usage bias, with 73% of the third codon positions being G or C. However, there isno bias for pyrimidines in the third codon position, as is the case with the Heterosigma Fcpl122cDNA. In addition, the degree of preference for the predominant codons in the ATPase gene isalso not as distinct as in the Fcp] cDNA. The significance of the difference in bias between thetwo genes from Heterosigma is not known.4.4.2 Structural aspects of sequence comparisonThe similarity of the FCPs to the CABs was confirmed when the sequence of the firstFcp cDNA was determined (Grossman et al. 1990). In addition, there were three putativemembrane spanning regions within the mature protein. As membrane spanning regions typicallyform ct-helicies, the prediction of three transmembrane domains in the FCPs was consistent withthe amount of measured o-helica1 structure (Hilier et al. 1987). When the Heterosigma FCP iscompared to other FCPs or to the CABs, most of the conserved amino acids are within themembrane spanning, hydrophobic regions. This is particularly obvious within the first and thirdmembrane spanning regions where there are probably considerable selective pressures againstchanges due to their importance in chlorophyll and carotenoid binding. The four amino acidsthat form the ionic bonding pairs between MSR 1 and MSR 3 in the pea LHC II structure areconserved in the Chi a + c-binding proteins (E 71—R 174; R 76—E169, Fig. 4.15). In additionto binding chlorophyll, these residues are thought to be the main protein stabilizing force withinthe lipid bilayer (Kühlbrandt et al. 1994). Moreover, within the area of close contact betweenthese two c-helical membrane spanning regions—on the sides that face one another—there are anumber of conserved smaller residues, probably selected for close packing (Green andKühlbrandt 1995). In addition, one of the two Met residues (Met 73, Fig. 4.14), involved in thesterospecific binding of the internally located carotenoids in pea, is conserved in the Chi a + cbinding proteins. However, the position analogous to Met 188 in the pea structure is notconserved in the FCPs (or in all CABs) and its role may be fulfilled by a conserved Gin residue.Overall, this indicates that the structure and topology of the FCPs and the iPCPs are very similarto the CABs.123However, there are some noticeable differences between the two main LHC types (CABsand FCPs). First, compared to the CABs, the FCPs are shorter in the connectors joining thetransmembrane regions, in the C-terminus and in the N-terminus, which accounts for theirsmaller size on polyacrylamide gels. The shorter connecting regions may be an indication thatthe FCPs helicies are more tightly packed together than those of the CABs (Green andKühlbrandt 1995). Second, the two-fold symmetry that exists between the surface exposedregions just before the first and third MSRs in pea appears to be lacking in the FCPs since theseareas are not conserved. What effect this may have on the organization of the FCPs is notknown. Despite these differences, the overall structural relatedness clearly indicates that thedifferent LHC types (CAB/FCPs etc.) are evolutionarily related and share a common ancestor.This also confirms the immunological work in chapter 2, and elsewhere (Caron et al. 1988;Passaquet et al. 1991; Hiller et al. 1993; Plumley et al. 1993; Durnford and Green 1994), thatindicated the CABs and the FCPs are structurally related and share common antigenic epitopes.Since the 3-D structure of the pea LHC II has been determined (Kühlbrandt et al. 1994),it is possible to use this information in locating putative pigment binding residues in other LHCs.Overall, the conservation of six of the seven putative Chi a-binding residues in the FCPs wouldsuggest that these residues also function in the binding of chlorophyll. Presumably theseresidues would bind Chi a, but the possibility of some of these residues ligating Chi c or nochlorophyll at all can not be dismissed. The ligand of Chi a7 in pea is not known and the regionunder this Chl in the CABs (WFKAG; Fig. 4.14) is not conserved in the FCPs (Green andKühlbrandt 1995). As this Chl is not bound by a side chain group, it is not possible to accuratelyassess whether it may be present in the FCPs.Three Chi b ligands in the pea LHC II structure have been tentatively identified(Kühlbrandt et al, 1994). The alignment in Figure 4.13 suggests that two of these putative Chi bligands, within MSR2 in pea (Gin 131 and Giu 139, indicated by open triangles), are conservedin the FCPs and iPCPs. However, a gap was inserted in the CAB sequence to maximizesimilarity to the other LHCs. In an os-helical structure, these residues would be on opposite124sides; therefore, it is unlikely that both would bind chlorophyll in the FCPs/iPCPs (Green andKühlbrandt 1995).The first of these conserved amino acids (E134, Fig. 4.12, 4.15) probably does not bindchlorophyll in the FCPs since most lack an Arg residue at position 137. In the pea LHC II, anArg in a homologous position interacts with the conserved Glu to bind Chi b5 (Kühlbrandt et al.1994). However, the presence of an Arg three positions away from the conserved Glu in thesecond MSR of the Isochrysis FCP, suggests that this residue may be a chlorophyll ligand in thiscase (see Fig. 4.12). Nonetheless, this Glu is very conserved in all FCPs and iPCPs and, if notbinding chlorophyll, probably has another important function (Green and KUhlbrandt 1995).Another amino acid with the potential to bind chlorophyll in the Heterosigma FCP isposition 123 (V123 Fig. 4.12,4.15). This could occur via a backbone carbonyl group because itwould not have a H-bonding partner in the ct-helix due to a conserved proline residue at position119 in the FCPs (Fig 4.12, 4.15). Interestingly, this residue would be on the same side as theconserved Glu in the second MSR if an ct-helix is formed (E 134, Fig. 4.12, 4.15). As well, theconserved Gin (Q125, Fig. 4.15) could provide a chlorophyll ligand, but not in addition to theformer two because they would be on different faces of the ct-helix. The FCPs probably lack achlorophyll binding residue near the C-terminus, as occurs in the pea LHC II (position 211, Fig.4.13), because this region is not conserved and it is absent in the Phaeodactylum FCP sequence.An accurate determination of the Chl a/c ratio of the FCPs has not been done althoughestimates in the range of 1.4-5.6 have been reported (table 1.3). With an average Chi a/c ratioaround 3 and assuming at least six bound Chl a molecules (since six of the residues thought tobind Chi a in the pea LHC II are conserved in the FCPs), there would be approximately two Chic molecules per FCP polypeptide. In addition, anywhere from 5 to 12 fucoxanthin moleculesmay be present based on calculated molar ratios of Chl a, c and fucoxanthin (Friedman andAlberte 1984; Katoh et al. 1989). Although there are a few putative chlorophyll ligands in thesecond MSR of the FCPs, it is not possible to conclude if they would bind accessorychlorophylls (Chi c) as they are suspected of doing in the CABs (Chi b).125Other domains besides the membrane spanning regions may participate in thechlorophyll or carotenoid binding. The conserved domain in front of MSR1 in both the FCPsand CABs suggests that this region may be important for this purpose. In the pea LHC IIstructure, this region shields Chl a4 and lutein2 from the aqueous environment and may beinvolved in binding one of the centrally located lutein molecules (Kühlbrandt et al. 1994; Greenand Kühlbrandt 1995). Presumably this region has an analogous function in the FCPs and may,in addition, provide a site for the binding of fucoxanthin. Similarly, in pea LHC lithe positionimmediately preceding MSR3 is positioned above Chl al and luteini and is thought to beimportant in their binding (Kühlbrandt et al. 1994; Green and Kühlbrandt 1995). Although thereis no obvious sequence conservation amongst the FCPs in this domain, it still may be importantfor both chlorophyll and carotenoid binding. It is likely that the polar groups at each end offucoxanthin and the other carotenoids form hydrogen bonds to polar groups within the MSRconnector regions. Similar interactions with the carboxyl group on Chl c, due to the lack of aphytol tail, could help bind or stabilize this molecule. Though not well conserved in the FCPs,the other domain linking MSR1 to MSR2 and the C-terminal section may have a role in eitherchlorophyll or carotenoid binding, as suspected for the pea LHC II structure (Kühlbrandt et al.1994).4.4.3 The transit sequence and protein importThe N-terminal region of the FCP protein from Heterosigma resembles a signal sequencethat targets proteins to the ER in eukaryotes. The targeting of nuclear encoded, plastid localizedproteins in chromophyte algae was first hypothesized to be mediated by a eukaryotic-like signalsequence by Sarah Gibbs (1979). The presence of a signal sequence-like transit peptide inplastid localized precursors was confirmed by Grossman et al. (1990) when the first nuclearencoded Fcp gene from a chromophyte was sequenced. The presequence of the HeterosigmaFCP can be separated into several regions: a positively charged amino terminus (residues 1-4), a126hydrophobic section (residues 5-16), followed by a more polar region preceding the cleavagesite. Though primary sequence conservation is not usually apparent, these are generalcharacteristics of all signal peptides (von Heijne 1990).The presence of a signal sequence would correlate well with the ultrastructure of thechromophyte plastid which has two additional membranes surrounding it (Gibbs 1970). Theoutermost membrane has ribosomes bound to the outer surface, like ER, and is commonly calledthe chioroplast ER (CER). The presence of a signal sequence offers an explanation for hownuclear encoded precursors may cross these two additional membranes. This was followed upby Bhaya and Grossman (1991) where cotranslational transport and processing of the FCPprecursor was observed in an in vitro microsomal (ER) membrane system. In addition,terrestrial plant chloroplasts were not able to import the FCP precursor, illustrating the differencein transit sequence specificity. These observations lead to the proposal that the synthesis ofchioroplast precursors occurred on the CER bound ribosomes and were cotranslationallytransported through the membrane (Bhaya and Grossman 1991).The presence of a putative bipartite transit sequence in the Odontella y-ATPase subunitsuggests that after the removal of the signal sequence, the remaining peptide directs the proteinacross the two membranes of the chloroplast envelope (Pancic and Strotmann 1993). Thepresequence of this protein is much larger than those of the FCP sequences shown in Figure4.16. If the FCP presequence is bipartite, then the resulting transit peptide after the cleavage ofthe putative signal sequence would share some characteristics with the chloroplast transitpeptides of plants (Bhaya and Grossman 1991; Apt et al. 1994). This similarity includes a highproportion of hydroxylated and basic amino acids. However, with the FCPs, this putativesecond leader sequence is predicted to be quite small compared to a typical plant transit peptide.It would be interesting to test various portions of the last half of the Heterosigma FCPpresequence for transport competence into the pea in vitro chloroplast system.In addition, it has been suggested that the transport of chloroplast precursors, followingtranslocation through the outer membrane, occurs via membranous vesicles. These vesicles have127been observed between the two sets of surrounding chioroplast membranes in the chromophytesand are referred to as the periplastidal reticulum (Gibbs 1979). The periplastidal reticulum washypothesized to pinch off from the CER and fuse with the outer chioroplast envelope, apparentlydepositing protein precursors into the lumen of the chioroplast envelope. If so, the bipartitesequence may still be important in directing the precursor across the inner chioroplast envelope.An interesting exception is with the haptophyte FCP presequence. Though ahydrophobic region is present, the N-terminal region does not contain a basic amino acid whichis a standard part of a signal sequence. This region is relatively rich in hydroxylated amino acidsas compared to the other FCPs. It has been suggested that an unformylated initiation Metresidue can compensate for the lack of a basic amino acid and that there is not an absoluterequirement for a positive charge (von Heijne 1990). Otherwise, the Isochrysis FCPpresequence is unlike that of other chromophytes. The significance of these differences, in termsof the transport mechanism, needs to be investigated.128CHAPTER 5A phylogenetic analysis of the LHCs5.1 IntroductionIn this study I was interested in investigating the evolution of the two main LHC genefamilies: the Cabs and the Fcps. This analysis involves three lines of investigation. (1) I wasinterested in analyzing the relationships amongst the different gene members of the Cab family.Since most of the genes from the tomato Cab gene family have been cloned, sequenced and wellcharacterized (Green et al. 1991), it provided an opportunity to examine the evolutionaryrelationships amongst them and relate these trends to the proposed functions of the differentprotein complexes. (2) I wanted to examine the relationships between the known chromophyteFcp and iPcp gene sequences; however, as there are few known Fcp sequences, this analysis willbe limited. (3) I also wanted to examine the evolutionary relatedness of the FCPs to the CABs.Analysis of the relationships between the different Cab gene types to the Fcp genes may provideinformation as to when the divergence of the chromophyte antennae occurred. In addition, theusefulness of the LHCs in assessing organismal phylogeny will be discussed.1295.2 Methods5.2.1 Protein AlignmentPhylogenetic analyses were carried out on the inferred protein sequences from a numberof different light harvesting antenna! proteins, from a diverse range of organisms. An originalalignment was done using MACAW 2.02 (Schuler et al. 1990) with the 250 PAM matrix(Dayhoff et al. 1978) scale. Refinements to this alignment were made by hand using the GeneticData Environment (GDE) software on a SUN workstation. The process of generating anysequence alignment carries with it the assumption of positional homology between the residuesbeing compared. Highly divergent areas of the protein, which were impossible to alignunambiguously, were omitted from the analysis. This was done to avoid the comparison of non-homologous regions of the protein which would violate the first assumption, positionalhomology, and could significantly alter the resulting tree topology (see Schlegel, 1991).Furthermore, regions of ambiguous similarity and multiple insertion/deletion events (indels)probably represent areas of ‘multiple hits’ and contain little, if any, phylogenetic signal (Swoffordand Olsen 1990). Gaps were introduced in the alignments to maximize similarity. Those thatwere larger than one amino acid were treated as a single deletion event in the analysis.The exact residues used in each analysis are indicated in Table 5.1 by residue numberaccording to the labeling in Figure 5.1. The number of residues varies depending on the degreeof sequence conservation between the taxa selected for the analysis. Amino acids were analyzed,rather than nucleotides, for a three reasons: (1) the third nucleotide of the codon tends to becomerandomized over large phylogenetic distances and is expected to have little phy!ogenetic signal,(2) the protein sequence is somewhat less sensitive to biases in the G + C content, and (3) itallowed the inclusion of LHC sequences that were determined only at the amino acid level(Swofford and Olsen 1990).130Table 5.1 Amino acid characters used in the phylogenetic analysesdatasetTomato CABs (Fig. 5.3)Total CABs (Fig. 5.4)Green algal CABs (Fig. 5.5)FCP analysis (Fig. 5.6)Total LHC analysis (CABs/FCPs) (Fig. 5.7)*as labeled in Figure 5.15.2.2 Phylogenetic analysischaracters*48-71; 79-111;160-192; 219-223; 225-232, 241-286as above48-71; 79-111;160-192; 219-231, 241-30148-73; 79-111;140-188; 191-198; 241-27548-71: 79-111:159-194: 241-281Phylogenetic analyses were done with PHYLIP version 3.5 (Felsenstein 1992) using bothparsimony (PROTPARS) and distance matrix (PROTDIST) algorithms. Both algorithms wereused to determine if putative taxon relationships were consistent between each tree constructionmethod. However, consistency between methods is not necessarily a good indication that thederived tree topology is the correct one (Felsenstein 1992). Comparing trees generated using thedistance method should help in assessing which taxa may be rapidly evolving, based on thelength of the branches in the tree. This should give some idea as to the likelihood of artifactsoccurring in the parsimony tree since parsimony tends to fail under circumstances when the rateof change between different taxa is large (Schlegel 1991). In such cases, there is a tendency togroup faster changing sequences together; a case referred to as long branches attracting(Felsenstein 1978). As well, distance methods tend to be sensitive to the number of taxa in theanalysis, which can alter tree topology (Schiegel 1991).Parsimony trees were done with the jumble option in effect, which randomizes the orderin which each taxon is added to the analysis. This was repeated ten times for each tree since the131input order of the taxa can influence the final tree topology (Felsenstein 1992). Changes inamino acids were assigned a value conesponding to the number of minimal mutational stepsrequired to cause such a change, based on the genetic code. Mutations resulting in a synonymousamino acid change are not included in the calculation. It is assumed that they occur much morefrequently than non-synonymous amino acid changes and are phylogenetically unimportant(Felsenstein 1992). With the exception of omitting ambiguously aligned regions, which assigns aweight of zero to these areas, no external weighting was used. If more than one mostparsimonious tree was found then a consensus tree was shown. Consensus trees include branchtopologies that occur most frequently; branch relationships appearing in more than 50% of thetrees of equal length (equally parsimonious) are definitely included (Felsenstein 1978).Distance matrix analyses utilized the Dayhoff accepted point mutation (PAM) matrix(Dayhoff et al. 1978) in the calculation of distances between the different taxa. This is anempirical matrix which assigns a probability for the conversion of one amino acid to another.Gaps are treated as unknown amino acids and dropped from the calculation (Felsenstein 1992).From the distance matrix, the tree topology was determined using the neighbor-joining method(Saitou and Nei 1987). The neighbor-joining method determines tree topology by starting with astar-like tree and successively clustering taxa into neighbors. Neighbors are taxa that areconnected by a single interior node and are clustered on the basis of a calculated minimumdistance between them. The joining of neighbors continues until only one possible unrooted treeexists. The distance values between each taxon were randomly inputted using the jumble option.An estimate of branch stability was assessed by a bootstrap analysis with 100 replicates.Bootstrapping is a process that randomly resamples the character sites (eg. amino acids at aspecific position) until a dataset the same size as the original is obtained; the number of taxa inthe analysis does not change. This is repeated any number of times, as determined by the user.Depending on the size of the dataset and the number of taxa, bootstrap trials can take aconsiderable amount of computer time. Because of this, trials are routinely limited to 100replicates. The phylogenetic analysis is then done on each of the replicated datasets and the132results are expressed as the number of times, out of 100, a particular node on the resulting treewas supported (Felsenstein 1985). The bootstrap values are typically displayed directly on thebranches of the tree they correspond to. It is an estimate of whether the tree is likely to change ifthe number of characters in the analysis had been larger. Bootstrap values should be consideredas an indication of the consistency of the individual characters in determining a specific treetopology and not a probability that the relationships depicted are the true phylogeny. It is alsoimportant to realize that bootstrap analysis is not able to detect errors in phylogeny that are theresult of systematic errors. These are errors that result when the evolutionary processes violatethe assumptions of the phylogenetic method used (Swofford and Olsen 1990).5.2.3 Terms and conceptsA phylogenetic tree is a type of dendrogram, ie., a branching diagram depictinghypothesized genealogical relationships of the taxa. The taxa are displayed at terminal nodes andthe branches joining them are connected at internal nodes. The branching of the tree normallyoccurs in a bifurcating fashion. The trees in either method are constructed based on a specificcriterion. This refers to the manner in which the evolutionary characters are weighted andassessed. Characters generally refer to a measurable observation of any trait pertaining to anorganism or gene (taxon). In this study, a character refers to an amino acid residue located at aspecific position within the protein. The identity of a character at a specific position (whether itis Ala, Leu, or Ser, etc.) is called the character state. A monophyletic group refers to two or moretaxa that share the same ancestral taxon. Alternatively, groups are considered polyphyletic whenthe taxa within them are derived from two or more distinct ancestral genes or species. Paralogoussequences (eg. Lhcal and Lhcbl) do not share the same evolutionary history as duplication anddivergence of the gene occurred prior to the divergence of the organism (ie., a speciation event)(Schlegel, 1991). Genes of the same type (e.g. all Lhca4 sequences) are considered orthologous133because they diverged from a common ancestral sequence and should parallel the evolution of theorganism (Schlegel, 1991).The taxonomic positions of the organisms, from which the LHC sequences used in thisanalysis originated, are given in Table 5.2. The reader should also refer to Table 1.1, in thegeneral introduction (Chapter 1), for further characteristics of the algal groups in question.Table 5.2 Species and genes used in the phylogenetic analysesOrganism Taxomonic position/gene ReferenceLycopersicon esculentum Angiosperm: dicot(tomato) Lhcbl Pichersky et al., 1985Lhcb2 Pichersky et al., 1987Lhcb3 Schwartz et al., 1991aLhcb5 Pichersky st al., 1991Lhcb6a/b Schwartz and Pichersky, 1990Lhcal Hoffman et al., 1987Lhca2 Pichersky at al., 1987Lhca3 Pichersky et al., 1989Lhca4 Schwartzetal., 1991bpsbS Wallbraun et al., 1994Spinacia oleracea (spinach) Angiosperm: dicotLhcb6 Spangfort et al., 1991psbS Kim et al., 1992Arabidopsis thaliana Angiosperm: dicotLhcal Jensen et al., 1992Lhca2 Zhang et al., 1992Lhca3 Wang et al., 1993Lhca4 Zhangetal., 1991Lhcb4 Green and Pichersky, 1993Lemna gibba (duckweed) Angiosperm: monocotLhcb2 Karlin-Neumann et al., 1985Gossypium hirsutum (cotton) Angiosperm: dicotLhcb2 Sagliocci et al., 1992Oryza sativa (rice) Angiosperm: monocotLhcbl Matsouka, 1990Pinus sylvestris (Scots pine) GymnospermLhcbl Jansson and Gustafsson, 1990Lhcb2 Jansson and Gustafsson, 1990Lhcb5 Jansson and Gustafsson, 1992Lhcal Jansson and Gustafsson, 1991Lhca2 Jansson and Gustafsson, 1991Lhca3 Jansson and Gustafsson, 1991Lhca4 Jansson and Gustafsson, 1992Ginkgo biloba GymnospermLhcbl Chinn and Silverthorn, 1993Polystichum munitum Pteridophyta (fern) Pichersky et al., 1990Physcomitrella patens Bryophyta (moss) Long et al., 1989134Chlainydomonas reinhardtii ChiorophytaLhcb Imbault et al., 1988Lhca Hwang and Herrin, 1993Chlamydomonas moewusii ChiorophytaLhcb Larouche et al., 1991Chiamydomonas stellata ChiorophytaLhcb Wolfe et al.,1993-unpubl.Lhca (2C.stellata) pir S33466, S31393Chiamydomonas eugametos Chiorophyta Gagné and Guertin, 1992Dunaliella tertiolecta ChlorophytaLhcb LaRoche et al., 1990Dunaliella sauna ChiorophytaLhcb Longetal., 1989Mantoniella squamata Chlorophyta: Rheil and Mörschel, 1993PrasinophyceaeEuglena gracilis Euglenophyta Muchhal and Schwartzbach,Lhcb 1-4 1992Lhca (35 & 38) Houlné and Schantz, 1988Phaeodactylum tn cornutum Chromophyta: Diatom Grossman et al., 1990Bhaya and Grossman, 1993Odontella sinensis Chromophyta: Diatom Thelen & Pancic -gb 81054Macrocystis pyrfera Chromophyta: Brown alga Apt et al., 1994Heterosigina carterae Chromophyta: Durnford, D. this studyRaphidophytaIsochrysis galbana Haptophyta LaRoche et al., 1994Pavlova lutherii Haptophyta Hiller, R. unpublished*Amphidinium carterae Dinophyta: Dinoflagellate Hiller et al., 1993 ** Sequence determined at the protein level. However, recently the nucleic acid sequence of tnegene encoding the dinoflagellate iPCP protein has been determined (gb Z47562, Z47563).5.3 Results5.3.1 Assessment ofphylogentic signalThe datasets used in each analysis include the sequences shown in each Figure (Fig. 5.3-5.7) and the characters outlined in Table 5.1. The phylogenetic signal of the dataset was assessedby calculating the skewness of tree-length distributions from 10,000 randomly generated treesusing the ‘random tree” option in the computer program PAUP (Phylogenetic Analysis UsingParsimony) (Swofford 1991). The random tree option randomly selects a tree topology and135calculates the length of that tree based on the dataset. It repeats this process 10 000 times andplots the tree lengths against the frequency of their occurrence. Random sequences producenearly symmetrical distributions from a parsimony analysis of all possible tree lengths whilethose that contain a phylogenetic signal have a skewed distribution (Hillis et al. 1993). Thecharacters in the FCP sequence dataset (used in Fig. 5.6) were manually randomized and therandom tree distribution calculated (Fig. 5.2A) to compare it to the random tree distributionproduced with the non-randomized sequences (Fig. 5.2D). As expected, the distribution wasnormal as compared to the very skewed distribution of the non-randomized dataset (compareFig. 5.2A & B). This indicates that this is a useful method of assessing the potential phylogeneticsignal in the datasets.The tomato CAB (Fig. 5.2B) and FCP (Fig. 5.2D) datasets had a stongly skewed randomtree distribution. The green algal dataset also showed a strongly skewed distribution, though to alesser extent (Fig. 5.2C). This indicates that these datasets have not diverged beyond the point ofbeing a potentially useful phylogenetic indicator. The last two datasets (Fig. 5.2E and F) giveonly a moderately skewed distribution. Though this is a conservative estimate, it indicates thatthe dataset may be approaching a limit of change where convergent or back mutations havebecome as frequent as divergent mutations (Meyer et al. 1986) and is approaching its limit as aphylogenetic indicator. However, PAUP does not assign weights to amino acid changes, nordoes it reflect the genetic code, as PHYLIP does. Without weights, PAUP essentially calculatesthe similarity based on amino acid identity. This underestimates the true relatedness of thesequences and the random tree distributions should be considered a lower limit estimate ofphylogenetic signal. In addition, since PAUP does not use a character weighting scheme, therandom tree lengths in Figure 5.2 can not be compared directly to the length of the trees I willshow.136Figure 5.1Amino acid sequence alignment of selected sequences used in the phylogenetic analysis.Numbers on top of the alignment indicate character positions and are not an indication of theprotein size. Regions of some proteins have been omitted to save space, and have been indicatedby the presence of a *• The sequences are arranged by gene/protein type followed by the speciesit is from. Gene types are: Lhca and b, Light harvesting complexes associated with PS I or PS II,respectively, according to the nomenclature of Jansson et al. (1992); iPCP, intrinisic peridininchlorophyll protein; FCP, fucoxanthin-chiorophyll protein. The taxa are: Ac, Amphidiniumcarterae; At, Arabidopsis thaliana; C.eugam., Chiamydoinonas eugametos; Cr, Chiamydomonasreinhardtii; Cs, Chiamydomonas stellata; Ds, Dunaliella sauna; Eg, Euglena gracilis; Gb,Ginkgo biloba; Hc, Heterosigma carterae; Ig, Isochrysis galbana; Le, Lycopersicon esculentum(tomato); Lg, Lemna gibba; Mp, Macrocystis pyrifera, Ms, Mantoniella squalnata; Os, Odentellasinensis; P1, Pavlova lutherii (partial sequence); Pm, Polystichuni muniturn; Pp, Physcomitrellapatens; Ps, Pinus sylvestris; Pt, Phaeodactyluin tricornutum; So, Spinacia oleracea. Gaps insequence are indicated by a -.137MSR111121314151IIIIIIII6171819110111100Lhcb6a-LeV--PP---KRFVVAAAAVAPKKSWIPAVXSGGNLVDPEWL00SLPGDFGFDPLGLGKDPAFLKWYREAELINGRWAMAAVLGIFVGQAWSGIP---psbS-So*KStJAPKCVEKPKLK-VEDGLFGTSG01GFTKENELFVGRVANIGFAASLLGEGITGKGIL-Lhcb5-Le*JtJ(7ppKAKAAAVSPADDELAKWYGPDRRIFLPEGLLDRSEI-PEYLNGEVPGDYGYDPFGLSKKPEDFAEYQAYELIHARWAMLGAAGFIIPEAFNKFGA--Lhcb4-At*GSGRFTAVFGFGKKX-AAPKKSAKKPV’ITDRPLWYPGAISPOWLOGSLVGDYGFDPFGLGKPAEYLQFDI*GIQRPRECELIHGRWAATLGALSVEWLTGVT---Lhcal-Le*LSSTKSKFAA)MPVYVGATNFMSRFSMSADWMPGQPRPSYL00SAPGD9’GFDSLGLGEVPANLERYKESELIHCRWAML.AVPGIIVPEALGLGN----Lhca2-Le*pPLRVSKYSPTPTARSATVCVAADPDRPLWFPGSTPPPWLDGSLPGDFGFDPLGLASDPESLRWNQQAELVHCRWANIGAAGIFIPELLTKIG---Lhcal-Le*LSSGFJREVSFRPSTSSSYNS1ICVEAK(GQWLPGLASPDYLDGSLPGDNGFDPLGLVEDPENLKWPIQAELVNGRWAMLGVAG1LPEVFTSIG---Lhca3-Le*IGSRISQSVT!RKASFVVRAASTPPVKQGANRQLWFASKQSLSYLDGSLPGDFGFDPLGL-SDPEGTGGFI-EPKWLAYGEVINGRFAMLGAAGAIAPEILGKAGLIPLhcbl-Le*GNGRIKAVAI(SAPSSSPWYGPDRVKYLGPFSGES---PSYLPGEFPGDYGWDTAGLSADPETFAXNRELEVIHCRWANLGALGCVFPELLAflNGV---Lhcb2-Le*GEGRIRTVKSAPQSIWYGEDRPKYLGPFSEQT----PSYLTGEFPGDYGWDTAGLSADPETFARNRELEVIRCRWAMLGALGCVFPEILSKNGV--Lhcb3-Le*--NANPLRDWAMGS-ARFTMSNDLWYGPDRVXYLGPFSAQT---PSYLNGEFPGDYGWDTAGLSADPEAFAXNRALEVINGRWAMLGALGCIFPEVLEKWVKV- -Lhcb-GbGEGRITJ(TASKKV---VASSGSPWYGPDRVKYLGPFSGEA----PSYLTGEFPGDYGWDTAGLSADPETFA)DRELEVIHSRWANLGALGCVFPELLSRNGV--LhCb2-Lq*S00RF5RWKAVPQSIWYGADRPKFLGPFSEQT----PSYLTGEFPGDYGWDTAGLSADPETFA1RELEVIHSRWAMLGALGCIFPELLSKNGV--Lhcb-PpTVSKSAGSD’PIWYGADRPKFLGPFSGET----PSYLNGEFAGDYGWDTAGLSSD?EPPATIII), Lhcal—4(LHC I types I—IV), Lhcb4 (CP29), Lhcb5 (CP26), Lhcb6alb (CP24), psbS (the 22 kDa protein).Further explanations and characterizations of the complexes are given in Table 1.2. Bootstrapvalues calculated from 100 replicates are given at the appropriate nodes.ALhcalLhca3Lhca2Lhca4psbS0.10psbSLhca4Lhca3Lhca2Lhcb6bLhcb6aLhcb4ArabidOPSiSLhcalLhcb5Lhcb3144Changing the outgroup from PsbS (CP22) to Lhcb6 (CP24) did not significantly alter therelationships presented in Figure 5.3. The relationship of the Lhcb4/6 branch, with respect to themain Lhca branch (Lhca2-4), could be altered depending on the alignment used and number ofcharacters selected. However, no significant change in tree topology was observed when theamino acid changes were weighted according to their chemical similarity. Little change was alsoobserved when a tree was constructed with the same distance data but using a different algorithm(eg. Fitch). Despite this, rates of divergence are relatively large in the Lhca portion of the tree asindicated by the long branch lengths in Figure 5.3A. Bootstrap values are also low, suggestingthe characters used are inconsistent in their phylogeny prediction.5.3.3 Phylogeny of the Chlorophyll a + b-binding protein family:This analysis shows the relationships between the diverse CAB proteins from differenttaxa. In this case, I intentionally included LHC I and LHC II sequences to assess theirrelatedness to one another and to make inferences about the evolution of this gene family. Thefirst analysis used 146 amino acid positions (characters), defined in Figure 5.1. Figure 5.4B is aconsensus of four most parsimonious trees with a length of 1469.There is an apparent separation of the LHC II proteins (Lhcbl, 2, 3 and 5) and the LHC Iproteins (Lhcal-4) in the distance (Fig. 5.4A) and the parsimony trees (Fig. 5.4B). The mainexception to this is the presence of Lhcb4 and Lhcb6 within the LHC I branch, which was thecase in the previous analysis (Fig. 5.3). The Euglena Lhca sequences are also outside this maindivision. However, the separation of LHC I and LHC II is not well supported by the bootstrapvalues in either tree. Other branch variation, indicated by low bootstrap values, was within theestablished branches of the tree and did not alter the separation of the Lhca and Lhcb lineages (ie.within the Lhcbl-2 branches).1450.10C P24/C P29LHCI-IIFigure 5.4ADistance matrix analysis of the LHC proteins from select Chlorophyll a + b containingtaxa, including both PS I associated (Lhca) and PS II associated (Lhcb) antennal complexes. Thetree was constructed from the distance matrix using the neighbor-joining method described in theresults section. Sequences are labeled according to their gene names and the identification of thecorresponding complexes are indicated in the figure legend of Figure 5.3 and in table 1.2. Referto Table 5.2 for references and full species names. Bootstrap values calculated from 100replicates are given at the appropriate nodes.__________n—Spinach-psbSL Tomato-psbS412LhcbEugle99 Lhcb-MantoniellalL Lhcb-Mantoniella2C.eugametosCP26Lhcb5-TomatoLhcb5-PineLhcb-EuglenalLhcb-D.tertidecta ILhcb-C.moewusii LHCII- Lhcb-D.salina Green AlgaeLhcb-C.reinhardtii— Lhcb3-Tomato ILHcII-IIILhcb-MossLhcb-FemLhcb-GinkgoLhcbl-Pine LHCII-ILhcbl -TomatoLhcbl -RiceLhcb2-Cotton LHCII-IILhcb2-TomatoLhcal -Tomato ILhcal-Arabidopsis LHC-I- Lhcal-Pine ILhca-C.reinhardtii ILhcb6b-Tomato I- Lhcb6a-Tomato 1Lhcb6-Spinach ILhcb4-Arabiciopsis i— Lhca2-Pine— Lhca2-Tomato I— Lhca2-ArabidopsisLhca4-TomatoLhca4-ArabidopsisLHCI-IVLhca4-PineLhca3-Tomato I— Lhca3-Arabidopsis I LHCI-IlILhca3-Pine ILHCILhca-Euglenal4914637— Lhcb-C.moewusii LHCII45 Lhcb-C.reinhardtii6 Lhcb-D.sali Green AlgaeLhcb-D.tertiolecta97 Lhcb2-TomatoLhcb2-Cotton62 27 24 Lhcbl-RiceLhcbl -Tomato60 1 35 Lhcbl -Pine LH8 61 Lhcbl-Ginkgo6 Lhcb-FernLhcb-moss29 Lhcb3-Tomato I LHCII-IIILhcb-Euglenal17 100 Lhcb5-Pine CP26Lhcb5-Tomato100 Lhcb-Mantoniellal31 Lhcb-Mantoniella2C.eugametosLhcal -PineLhcal -Arabidopsis LHC -Lhcal -TomatoLhca-C.reinhardtiiLhcb4-Arabidopsis I CP29Lhcb6a-TomatoLhcb6b-Tomato CP2420 Lhcb6-Spinach99 Lhca2-Tomato100 Lhca2-Pine LHCI-tI70 Lhca2-Arabidopsis69 Lhca3-ArabidopSiS100 Lhca3-Tomato LHCI-III76 Lhca3-Pne58 Lhca4-Pine100 Lhca4-Tomato LHCI-IVLhca4-ArabidopsisLhcb-Euglena2Lhca-EuglenalLhca-Eugena2psbS-SpinachpsbS-TomatoFigure 5.4BParsimony analysis of the LHC proteins from select Chlorophyll a + b containing taxa,including both PS I associated (Lhca) and PS II associated (Lhcb) antennal complexes. Analysiswas done using an identical dataset as in (A), as described in the results section. Refer to Table5.2 for references and full species names. Bootstrap values calculated from 100 replicates aregiven at the appropriate nodes.147The distance and parsimony trees were consistent in the clustering of the same CAB types(e.g. Lhca4-pine with Lhca4-tomato), and this was true for all known angiosperm andgymnosperm taxa. This suggests that the major LHC Cab gene lineages diverged before theangiosperm / gymnosperm separation. The association of the chiorophyte Lhcb sequences,however, does not follow this pattern. In both trees, they form a monophyletic branch separatefrom the Lhcbl, Lhcb2, and Lhcb3 sequences of the terrestrial plants. The separation of theLhcbl/2 sequences from the green algal Lhcb sequences is supported (bootstrap value 75 or 62,Fig. 5.4), though the separation of Lhcb3 is not as well supported in the parsimony tree(Fig. 5.4B). This suggests that Lhcbl-3 diverged after the chlorophyte and terrestrial plantlineages had separated.The fern Lhcb sequence clusters with the Lhcbl dade in the distance tree but before theLhcbl/2 separation in the parsimony tree. The bootstrap values coinciding with these branchesare also low. Because of this, the relationship is considered unstable and conclusions regardingthe divergence of the fern sequence, before or after the Lhcb 1/2 functional separation, areinconclusive (Fig. 5.4 A/B). Divergence of the moss Lhcb sequence before the Lhcbl/2divergence is moderately supported by the bootstrap analysis (Fig. 3, A/B), though this is underthe assumption that the moss sequence represents the main LHC II antennal protein and is not aparalogous gene. Some caution must be exercised in the analysis of the Lhcb relationshipsbetween the fern, moss, and chlorophyte CABs as a limited number of sequences have beencharacterized; these are often assumed to be homologous to the type 1 LHC II (Lhcbl) sequenceof the angiosperms but they could just as well be paralogous sequences.Both trees show that the Lhcb3 sequence has separated at an earlier point than the Lhcbland 2 gene types. Although this is a consistent relationship, it is not well supported (bootstrapvalue 67 or 61, Fig. 5.4AJB). This is suggestive of a divergence of this CAB type around, or justafter, the chiorophyte/terrestrial plant separation. Lhcb5 diverged early in the Cab geneevolution, before the separation of the chlorophytes. The precise position of Lhcb5, however, isnot well supported by bootstrap analysis. The close association of Lhca types 2, 3 and 4 are148supported in both trees (Fig. 5.4). The clustering of Lhcal, Lhcb4, and Lhcb6 is consistentbetween methods but is not substantiated with the bootstrap trials. The relative relationships ofthe Euglena (Euglenophyte) and Mantoniella (Prasinophyceae) LHCs were not well resolved onthese trees.A second analysis was done to resolve the relationships amongst the green algae andselected terrestrial plant sequences (Fig. 5.5, A/B). This analysis included 20 taxa and a largernumber of amino acids (170), making up approximately 75% of the mature protein (Fig. 5.2A).The outgroup was the Chiamydomonas eugalnetos light responsive gene that has similarity to theCAB proteins (Gagne and Guertin 1992). This was used as the outgroup because it is one of themost divergent green algal sequences. Figure 5.5 A and A’ are both unrooted trees and areidentical except that A’ is displayed in the radial format.The distance tree (Fig. 5.5, A) shows a topology that strongly supports the association ofthe green algal LHC II sequences with three out of four Euglena LHC II sequences. However,there is a fair degree of divergence between these sequences, as indicated by the length of thebranches. The Euglena2 sequence is more distantly related than the others but it is stillassociated with the LHC II branch. The Euglena sequences were isolated by heterologoushybridization with a plant LHC II probe (Muchhal and Schwartzbach 1992). Their identificationas LHC II sequences was by analogy to the terrestrial plant system. The LHC I sequences fromthe green algae and euglenoids tend to form a monophyletic assemblage separate from the LHC IIgroup but the branches are long and not strongly supported by the bootstrap analysis. TheEuglena-Lhca sequences were cloned by immunoscreening a cDNA library with a LHC IIpolyclonal antibody. However, sequence comparisons and immunoprecipitation with a barleyLHC I monoclonal antibody provided some evidence that they are LHC I sequences (Houlné andSchantz 1988). The identification of the C. reinhardtii Lhca sequence as a PS I associatedantenna was confirmed by N-terminal sequencing of a PS I associated polypeptide (p22) (Bassi etal. 1992; Hwang and Herrin 1993). The characterization of the Lhca gene from C. stellata hasnot been published and little information on its function or organization is known. However, it149Figure 5.5Phylogenetic analysis of the LHC proteins from select green algae and terrestrial plantCAB proteins, including both PS I associated (Lhca) and PS II associated (Lhcb) antennalcomplexes. The tomato sequence is a Lhcbl and the others are Lhcb types unless otherwiseindicated. (A) Distance matrix analysis using the neighbor-joining tree construction method.The inset (A) is identical to (A) except it is displayed in a radial format to facilitateinterpretation. (B) Parsimony analysis using an identical dataset as in (A), as described in theresults section. Refer to Table 5.2 for references and full species names. Bootstrap valuescalculated from 100 replicates are given at the appropriate nodes.150C slellala-LOcaMantoniella5Mantoniella2C.reinhardtii-LhcaTomato-LhcalArabidopsis-Lhcb4CstelIata-LhcaTomato-Lhca4Tomato-Lhca2Tomato-Lhca3Eug!enal-LhcaEuglona2-LhcaC.eugametoBTomato-Lhcblmoss-LhcbD.tertiolectaC.moewusllD.salinaC.reinhardAA’Arebidopss.1.hcb4TOmato4J,ealC. reinhard(ü-LhceC.ugareooD.salinaTomatomossC.stellataEugfena4Euglena2CieILet Eogeea40.100.10Mantoniella5Mantoniefla2Etiglena3EuglenalTomato-Lhca4C.steIata-LhcaTomato-LIica2Tomato-Lhca3Tomato-LhcalC.eugametosArab Id opsis-Lhcb4Creinhardtii-Lhca151does share a degree of similarity to the N-terminal sequence of a C. reinhardtii LHC I protein,p15.1 (Bassi et al. 1992).The Mantoniella Lhcb sequences form a branch at the base of the LHC II cluster in thedistance tree, as they did in Figure 5.4. However, a large divergence between the MantoniellaLHCs and the Lhca and Lhcb sequences are indicated by the length of the branches separatingthem. This is clearly evident in the radial tree (Fig. 5.5 A’). The position of the Mantoniellasequence is not supported in a bootstrap trial (28%) and this low value may reflect a fair amountof noise within the sequence. This noise may be the result of non-divergent mutations (parallel,non-divergent changes) occurring at a significant rate.Parsimony analysis of this dataset gave two trees with a length of 1356. The differencesbetween the two were within the green algal LHC II branch and no significant deviations fromthe consensus tree (Fig. 5.5 B) were observed. The overall topology of the parsimony treeconcurs with the distance tree. The main difference between the two methods is in the placementof the Mantoniella LHC branch. With parsimony, the Mantoniella sequence separates before thealgal LHC II and LHC I divergence. I am skeptical about this topology because of the longbranches connecting the Mantoniella and C. eugametos sequences (Fig. 5.5 A’). The presence ofthese very divergent sequences can result in treeing artifacts with a parsimony analysis(Felsenstein 1978), which is probably the case here. Though the distance tree is probably a moreaccurate reflection of the true gene phylogeny, I am unable to conclude whether the MantoniellaLHC is more closely related to either LHC I or LHC II type sequences.In both trees (Fig 5.5), the two Chiamydomonas Lhca sequences cluster with a specificterrestrial plant Lhca sequence and not with each other (Fig. 5.4 and 5.5). The bootstrap valuesare low for these branches so the exact association of the two sequences can not be concluded.Nevertheless, this offers preliminary evidence that the divergence of the LHC I sequencesoccurred early in the evolutionary history; before the green algal/terrestrial plant separation.1525.3.4 Phylogeny of the chlorophyll a + c-binding proteinsThis analysis examines the relationships of the chlorophyll a + c binding proteins (FCPs,iPCPs) from several groups of non-green algae. The dataset consists of 9 sequences (from sixtaxa) and includes 153 amino acids (90% of the mature protein) in the analysis, indicated inTable 5.1. The protein sequences are from two diatoms (Phaeodactylurn and Odontella), a brownalga (Macrocystis), a raphidophyte (Heterosigma) and a haptophyte (Isochrysis). All of thesealgae contain fucoxanthin, Chi a and Chl c. The other protein sequence is from thedinoflagellate, Amphidinium, which contains the predominant carotenoid peridinin, instead offucoxanthin. The Amphidinium iPCP sequence was determined directly by amino acidsequencing, while the FCP protein sequences were inferred from the gene sequence.Both the distance (Fig. 5.6 A) and parsimony (Fig. 5.6 B) trees showed a clustering of theFCP sequences separate from the iPCP (intrinisic peridinin-chlorophyll protein) of thedinoflagellate, Amphidinium. This is with the exception of the Isochrysis sequence which wasdistant from the other sequences, even though it has a fucoxanthin binding LHC (FCP). Thedistance analysis did not resolve the relationships amongst the FCP sequences of the brown alga,diatoms or the raphidophyte.The parsimony analysis (Fig. 5.6 B) gave only one most parsimonious tree with a lengthof 490. The pennate (Phaeodactylum ) and centric (Odontella) diatoms form a monophyleticgroup in agreement with traditional morphological characters (Round and Crawford 1990),though the length of the branches separating them (Fig. 5.6 A) indicate an early divergence. TheHeterosigma sequence is also grouped on this branch, which is separated from the brown algaldade. These relationships are supported by high bootstrap values and the branch is consideredrobust. The Isochrysis sequence is very divergent and is at the base of the tree. Interestingly, theHeterosigma sequence groups with the brown algae when the Amphidinium sequence is removedfrom both the parsimony and distance analyses (not shown).153ABCFigure 5.6HeterosigmaOdontellaPhaeodactylum2Phaeodactyum 1Macrocystis2MacrocystislPhaeodactylum2Phaeodactyum 1PaviovaOdontellaPhylogenetic analysis of the FCP proteins from various Chi a + c containing organisms.(A) Distance matrix analysis using the neighbor-joining tree construction method. (B) Parsimonyanalysis using an identical dataset as in (A) and described in the results section. (C) Parsimonyanalysis using a limited number of characters (87) in order to incorporate the partial sequenceinformation from Paviova lutherii (.). Refer to Table 5.2 for references and full species names.Bootstrap values calculated from 100 replicates are given at the appropriate nodes.Isochrysis0.10Amphidinium2AmphidiniumiMacrocysfis2MacrocystisiIsochrysisOdontellaHeterosigmaAmphidinium2AmphidiniumisochrysisPhaeodactylum2PhaeodactylumlMacrocystisiMacrocystis2HeterosigmaAmphidinium2Amphidiniumi154The very distant placement of the Isochrysis sequence was odd in light of the partialsequence information from Paviova, another haptophyte (Roger Hiller, unpubi.). I ran aparsimony tree with the partial Paviova sequence included in the alignment (Fig. 5.6 C •).Because of its inclusion, only 87 amino acids (49% of mature protein) were used in the analysis(Fig. 5.1). There are some obvious differences in topology between Figure 5.6 B and 5.6 C. Themain point is the close affinity of the Paviova partial sequence to the main FCP cluster, which isseparate from the Isochrysis sequence. Though these are preliminary data, it seems unusual thattwo taxonomically related species show such a remarkable difference. The most obviousexplanation is that the Isochrysis sequence is paralogous and not the predominant FCP in theorganism.5.3.5 Evolution of the LHCfainily ofproteinsThis analysis looks at the evolutionary relationships between the CABs, FCPs and theiPCPs; the main intrinisic light-harvesting antennae. A dataset of 39 taxa and 135 amino acids(61-79% of mature polypeptides) was constructed. The exact characters used in the analysis aregiven in Table 5.1. The taxa included were from the main LHC proteins of the non-green algae(FCP/iPCPs), the green algae and the terrestrial plants (CABs). In addition, representatives fromother LHC types, such as LHC I (Lhcal-4), CP24 (Lhcb6), CP26 (Lhcb5) and CP29 (Lhcb4),were included.With this dataset, the interpretation can vary depending on whether parsimony or distancemethods are used. In the distance tree there are three major groups (Fig. 5.7 A, A): (1) thechlorophyll a + c proteins (FCP, iPCP) that group with the C. eugalnetos (CAB) sequence, (2) theLHC II (Lhcb) sequences, and (3) the LHC I (Lhca) sequences that group with Lhcb6 (CP24)and Lhcb4 (CP29). The Euglena LHC I (Lhca-Euglenal/2) sequences form the earliest branchin the tree. In this tree it appears that Chl a + b sequences from LHC I and II are more closelyrelated to each other than either is to the Chl a + c lineage. This is clearly evident in the radial155Figure 5.7ADistance matrix analysis of all LHC proteins from select Chi a + b-containing and Chla + c-containing taxa, including both PS I associated (Lhca) and PS II associated (Lhcb) antenna!complexes. The tree was constructed from the distance matrix using the neighbor-joining methodas described in the results. Inset (A’) is a radial display of tree (A) and is otherwise identical.Refer to Table 5.2 for full species names and references. Bootstrap values calculated from 100replicates are given at the appropriate nodes.1569— psbS-SpinachL psbS-Tomato—Pcp-ArnphidiniumlPcp-Amphidinium2Fcpl -MacrocystisFcp2-MacrocystisFcp-OdonteliaFcpl -Phaeodactylum— Fcp1-PhaeocactyIumFcp-Heterosigma169Lhcb-EuglenalI — Lhcb-D.tertiolectajy Lhcb-D.salinajj L lhcb-C.moewusii44jLhcb-C.reinhardtiiI P— Lhcb-C.stellata61 r- Lhcb3-Tomato23 U r- Lhcb-MossLhcb-Fern47-1i1- Lhcb-GinkgoLhcbl -Tomato{j- Lhcb2-Tomato74L Lhcb2-LemnaLhca3-TomatoLhca2-TomatoLhca4-Tomato—Lticb6a-TomatoLhcb6b-Tomato-Lhcb4-ArabidopsisLhcal -TomatoLhca-C. reinhardtii93E Lhca-Euglenalh— Lhca-Euglena20.10Fcp-IsochrysisC.eugametosLhcb-Euglena217 Lhcb5-TomatoAPepFcp651 Lhcb-MantoniellalLhcb-Mantoniella21520GreenAlgaeTerrestrialP Ia ntsLHCIILHCICP24!CP29Macrocystis1 AmptiidinitsoehrysisC.ougamatos MantoniellapsbS-SpinachsbS -TomatoLhca-EuglenLhca3Lhca20.10157BLhcbl -TomatoGinkgoLhcb-MossLhcb-FemLhcb2-TomatoLhcb2-LemnaLhcb3-TomatoLhcb-C.stellataLhcb-C.reinhardtiiLhcb-D.salinaLhcb-C.moewusiiLhcb-D.tertiolectaLhcb-Euglenal100 Fcpl -Macrocystis47 Fcp2-Macrocystis29 Fcpl -Heterosigma32 Fcpl-Phaeodactylum88100Fcp-Odontefla56 100 Pcp-Amphidiniuml69 Fcp-lsochrysis21 c.eugametos100Lhcb6a-TomatoL__ Lhcb6b-Tomato— Lhcb4-ArabidopsisLhca-C.reinhardtiiLhcal -TomatoLhca2-Tomatoi— Lhca3-TomatoLhca4-Tomato1 Lhcb-Euglena2Parsimony analysis of all LHC proteins from select Chi a + b-containing and Chl a + ccontaining taxa including both PS I associated (Lhca) and PS II associated (Lhcb) antennalcomplexes. Analysis was done using an identical dataset as in (A), as described in the resultssection. The tomato PsbS protein is used as the outgroup. Refer to Table 5.2 for references andfull species names. Bootstrap values calculated from 100 replicates are given at the appropriatenodes.io—---- psbS-Tomato— psbS-SpinachLhcb-MantoniellalL Lhcb-Mantoniella2Lhcb5-TomatoLHCIIGreenAlgae161313Figure 5.7BFcpPcpFcp60 LHCIC P24/C P29158tree (Fig. 5.7 A’). Because of the low bootstrap values on the main branches leading to thesebranches, however, interpretations have to be made with caution since the resolution in thisregion of the tree is poor.With parsimony, six equally parsimonious trees were found with a length of 1792. Theconsensus of these six trees is shown in Figure 5.7 B. Most of the variations were within theinternal nodes of the main branches and were considered insignificant. This analysis produced atree topology where the Chi a + c sequences and the LHC I (Lhca) lineage formed amonophyletic group; this relationship was not supported by bootstrap replicates. As the branchlengths are very different between the FCPs and the CABs, I am skeptical of this association andsuspect it may be the result of treeing artifacts that are common to parsimony; particularly incases where there are unequal rates between the lineages. The low bootstrap values separatingthe two lineages (16%) makes any firm conclusions regarding the relationships of the CABs andFCPs unsupported. Relationships within the major lineages are generally the same as withdistance analysis.1595.4 Discussion5.4.1 CAB protein evolutionThe trees suggest that the two main LHC II types (Lhcb 1-2) form a monophyletic dade inagreement with the analysis of Jansson (1994). The Lhcbl and Lhcb2 polypetides make up themain LHC II complex and are generally 90% identical in the mature polypeptide (excluding theN-terminal region) so their close association on the tree is not surprising. Together, Lhcbl andLhcb2 make up the main peripheral antennae of PS II. Its principal function is the capture oflight followed by the transfer of the excitation energy to the core reaction center of PS II. It hasalso been implicated in the mediation of thylakoid appression, the regulation of energydistribution between the photosystems and in photoprotection, all of which have been previouslyreviewed (Anderson and Andersson 1988; Bassi et al. 1990; Jansson 1994). LHC II is thought toexist as a trimeric unit (Kühlbrandt and Wang 1991) composed of Lhcb 1 and Lhcb2polypeptides at a ratio of approximately 2:1 (Jansson 1994). Both Lhcbl and Lhcb2 havephosphorylatable threonine residues in the amino terminus of the mature polypeptide that arethought to be responsible for the state transition observed in thylakoids (Mullet 1983). The closeevolutionary relatedness of these two complexes is reflected in their functional similarities.However, the Lhcb2 complex is enriched in a peripheral subpopulation of LHC II, has differentphosphorylation kinetics and appears later in development (Larsson et al. 1987). Thesedifferences are thought to be important in the adaptation to different light intensities. It is likelythat the appearance of the Lhcbl I Lhcb2 lineages was due to a fine tuning of the light adaptationresponse.The more distantly related Lhcb3 (LHC II type III) sequence is a minor antennalcomponent that is about 80% identical to the Lhcbl and Lhcb2 complexes. This protein has ashorter N-terminus and is often found in the LHC II complex (Green et al. 1992b). Lhcb3 mayfunction as a linker between the bulk trimeric LHC II complex and the PS II core. This is160suggested because of its close association with the core complex of PS II in barley Chi b-lessmutants (Harrison and Melis 1992). The more proximal location of Lhcb3 to PS II, with respectto Lhcbl/2, would accommodate the hypothesis of an earlier evolutionary history of this protein,as depicted in the tomato CAB tree (Fig. 5.3). The addition of the peripherally located Lhcbl andLhcb2 complexes could occur without much change in the preexisting inner antennaorganization.The Lhcb5 (CP26) sequence is consistently found at the base of the Lhcb dade,suggesting it has evolved prior to the peripheral LHC II complex. CP26 remains associated withthe PS II core when the main peripheral LHC II antennae are removed (Camrn and Green 1989)and is considered an inner antenna. It is likely that the peripheral antennal proteins of LHC IIarose from the divergence of a gene encoding a inner antennal, CP26-like, protein. Duplicationand divergence of the Lhcb5 gene would have to be accompanied by a change in size andchlorophyll binding capabilities. The Chl a/b ratio of LHC II is 1.3 compared to approximately3.3 for CP26 (van Amerongen et al. 1994; Table 1.2). This suggests that changes in the amountsor relative proportion of Chl a and b bound by each polypeptide has changed in the course ofevolution (Green and Kühlbrandt 1995). Evolution of Lhcbl/2 at a later point would also suggestan increased level of regulation by phosphorylation as CP26 is not reversibly phosphorylated.An interesting finding is the close association of Lhcb6 (CP24) and Lhcb4 (CP29) in thetrees, separate from Lhcb5 (CP26). This is somewhat surprising as the CP29 and CP26 pigment-protein complexes often copurify (Green 1988) and both are associated with the core complex ofPS II (Camm and Green 1989). Their presence in both grana (cc) and stromal (13) localized PS IIcenters (Allen and Staehelin 1992) suggests that they are part of the basic PS II unit. Because ofthe biochemical similarities, the complexes are also known as CP29 type I (CP26) and CP29type II (Cp29) (Pichersky et al. 1991). The distant relationship between Lhcb4 and Lhcb5 hasbeen suggested through direct sequence comparison (Morishige and Thornber 1992; Green andPichersky 1993) and by phylogeny construction (Jansson 1994). Lhcb6 and Lhcb4 consistentlygroup together, though the long branches joining them (Fig. 5.3 A) suggest they diverged from161one another long ago. The minor CP24 complex binds little chlorophyll and, like CP29, isenriched in PS II fractions (Dunahay and Staehelin 1986).The Lhcal sequence (LHC I type I) is closely related to the PS II inner core antennae,CP29 and CP24. Although consistent between methods, this relationship is only moderatelysupported by bootstrap replicates in the distance trees (Fig. 5.3/4 A). A clustering of the CP29(Lhcb4) and LHC I (Lhcal) sequences was also supported in a recent dendrogram (Jansson1994). PS I antennae, Lhca2-4, appear to be more related to one another than either is to Lhcalor Lhcb6/4. However, all the Lhca sequences are quite divergent from one another as indicatedby the long branch lengths separating them on the distance trees.The relationship amongst the different LHC I antennae are unstable and not stronglysupported by bootstrap replicates in most cases. Parsimony analysis consistently places Lhca2and Lhca3 as the closest relatives within the LHC I lineage; this is supported by bootstrapreplicates in Figure 5.3B. Distance methods, however, consistently place Lhca2 and Lhca4 as theclosest relatives, although this is not strongly supported by bootstrap trials in any of the treesshown. The latter relationship has been observed in another dendrogram that is based on adistance method (Jansson 1994). However, no indication of reliability or significance of thisbranching order was given. It is tempting to conclude a direct evolutionary relationship betweenLhca2 and Lhca3 as they are the polypeptides making up the LHC 1-680 complex. In tomato, theLhca2 and Lhca3 genes are linked and have been mapped to chromosome 10, suggesting a closeraffinity due to a more recent duplication event (Pichersky et al. 1989). On the other hand, thegenes encoding Lhcal and Lhca4, which make up the LHC 1-730 pigment-protein complex, areon different chromosomes in tomato (Pichersky et al. 1987; Schwartz et al. 1991) and despite theapparent functional association, there does not seem to be any direct evolutionary relationshipbetween the two complexes.5.4.2 Evolutionary relationships amongst the Cab gene family162The analysis of the Cab gene family from a diverse array of organisms reinforces the ideathat two particular CAB types (eg. Lhcbl) from different organisms are more similar to oneanother than either is to different CAB types (eg. Lhca4) from the same organism. Comparisonof Lhcb 1 and Lhcb2 sequences previously revealed that the same CAB type encodes a nearlyidentical polypeptide with similar numbers of introns (Chitnis and Thornber 1988). Thedistinction between Lhcb 1 and Lhcb2 was also made by detecting signature amino acids (Janssonand Gustafsson 1990), which are conserved residues specific to a particular CAB type. Inaddition, a phylogenetic analysis of angiosperm Lhcbl and Lhcb2 sequences (Demmin et al.1989; Matsuoka 1990) also revealed this trend.Sequencing of a fern (Polystichum munitum) (Pichersky et al. 1990) and moss Lhcb(Physcomitrella patens) (Long et al. 1989) provided useful markers in estimates of the mainLHC II divergence. The analysis of the fern sequence was unable to clearly resolve thebranching order. However, the moss Lhcb sequence is consistently at a branch before the fernand Lhcb2 separations, indicating that the functional separation of the Lhcbl and Lhcb2complexes occurred after the bryophyte lineage separated. This is a tentative conclusion untilmore representatives from bryophytes are sequenced, in order to rule out the possibility ofmisleading tree topology as a result of comparing paralogous genes. Overall, the duplication andseparation of the Lhcb2 (LHC II type II) sequences probably occurred after the bryophyte lineageseparated and at about the same time as the lineage leading to the pteridophytes (ferns) diverged.This is in agreement with earlier predictions by Pichersky et al. (1990). It is apparent that thedivergence of all the major CAB types occurred early. This definitely occurred before theangiospermlgymnosperm separation, as the same CAB types from pine consistently grouped withthe same CAB types from the angiosperms. It is not possible to draw further conclusions as tothe earliest separation of most the CABs because there are too few sequences known from theother taxonomically distinct groups such as the Chiorophyta and the Euglenophyta.There is a distinct separation of the green algal LHC II sequences from the terrestrial plantLhcb 1-3 sequences suggesting that the minor Lhcb3 antenna appeared just after the green algal163lineage diverged. This may indicate a fundamental difference in the regulation (Sukenik et al.1987) and organization of the peripheral LHC II antennae in green algae. The lack of Lhcbl andLhcb2 type antennal proteins in green algae suggests the Lhcbl/2 duplication and functionaldivergence occurred in the lineage leading to the terrestrial plants after the green algal lineageshad separated.The identification of an LHC I sequence from the green alga Chiamydoinonas reinhardtiiwas confirmed by peptide sequencing of a PS I associated protein (Hwang and Herrin 1993).This analysis confirms its identity as a Lhcal type, as this sequence tends to cluster with theangiosperm Lhcal sequences; however, the relationship is not supported (bootstrap value 30 or54, Fig. 5.5). In this case, the Chiamydomonas sequence is more closely related to tomato Lhcalthan to the Lhca sequence from Chiamydomonas stellata. Should this relationship hold as moreLHC I sequences from green algae are determined, it would suggest that the LHC I sequenceshad diverged before the chlorophyte/land plant separation. This brings up the interestingsuggestion that the different LHC I complexes evolved prior to the establishment of the differenthomologous LHC II complexes (Hwang and Herrin 1993), which will be discussed in thefollowing sections.The Euglena sequences for LHC I and LHC II are very divergent, as compared to theirgreen algal and terrestrial plant counter parts, indicating that this alga separated from the greenalgal lineage very early. Of the four complete LHC II proteins, three are more related to oneanother (Euglena sequences 1, 3 and 4) and form a branch just before the green algal! land plantLhcbl/2 sequences. The respective LHC II and LHC I proteins from the green algae and Euglenagroup together, although the branch lengths in the LHC I lineage are very long on the distancetrees. This suggests that LHC I and LHC II polypeptides had functionally separated prior to theappearance of the euglenoid chloroplast. It also shows a close direct link between the green algaland Euglena chloroplasts, previously hypothesized on the basis of the presence of chlorophyll aandb.1645.4.3 Evolution of the Chi a + b and Chi a + c gene families:The relative position of the Chi a + c-binding proteins in relation to the Chi a + b-bindingproteins of LHC I and LHC II is not clear from an examination of either the parsimony ordistance trees. The distance trees (Fig. 5.7 A, A’) suggest that LHC I sequences are more closelyrelated to the CAB LHC II sequences whereas the parsimony tree (Fig. 5.7 B) supports the ideathat the FCPs are more related to the CAB LHC I sequences; neither tree is supported bybootstrap replicates. I am inclined to believe that the distance tree is a more accurate reflection ofthe true relationships because the dataset has a fair bit of noise and the branches between sometaxa are very long. Under these conditions, the parsimony tree may not be reliable (Stewart1993). However, these proteins have diverged to such an extent that the resolution of suchdistant events may not be possible.Of importance is the time of divergence of LHC I and LHC II CAB complexes and theseparation of the lineage leading to the Chi a + c-binding proteins. It would be interesting toknow if the FCP complex had evolved from ancestral PS II associated or PS I associatedantennae. It has been suggested that the FCPs evolved from a LHC I or CP24-like ancestor andthat the LHC II complexes of higher plants evolved after the divergence of the chromophytes andthe chlorophytes (LaRoche et al. 1994). I agree that the major peripheral LHC II CAB sequencesevolved after the divergence of the chlorophytes and chromophytes. However, there isinsufficient (convincing) evidence, at the moment, to suggest a closer relationship of the FCPs tothe Lhca-CABs. In fact, the distance trees indicate that the CAB LHC II and LHC I sequencesare more closely related to one another than either is to the FCPs. This would suggest that theFCPs diverged from the ancestral LHC before their was a separation of the LHC I and LHC IItype genes.It is not possible to make a firm conclusion regarding the ancestry of the FCPs except tosay that they diverged from the CABs very early. The long branches (large divergences) betweenthese taxa, the possibility of comparing paralogous genes, and the limited information on other165members of the FCP family make any conclusions tentative. A clearer picture of the FCP/CABrelationships will develop when LHC I sequences from the chromophytes are characterized.5.4.4 Evaluation of species relationships based on the LHC protein trees:The usefulness of the nuclear encoded CAB proteins as an index of phylogeneticrelationships is limited and will depend on several factors: the distance between the organismsstudied, the evolutionary questions asked, and the availability of the appropriate sequenceinformation. A significant obstacle in the utilization of CAB proteins for phylogeny is their smallsize. The relatively small number of useful characters raises questions about the reliability orsignificance of the topology observed. This “uncertainty” is often reflected by low bootstrapvalues. One also has to be cautious in the construction and interpretation of phylogenies based ona single gene/protein, which could lead to erroneous trees (Cao et al. 1994). As most of theconserved residues are thought to be membrane spanning, there is probably a functionalconstraint on these regions for being hydrophobic. If so, the more distant relationships may morereadily reach a point where non-divergent (homoplasmic) mutations occur at a significant rate,which can mask true evolutionary relationships (Meyer et a!. 1986). Another problem in usingthe CAB proteins for phylogeny, which is a common concern with nuclear encoded proteins, isthe possibility of comparing paralogous genes. Since the CABs and FCPs are encoded by amultigene family, care must be taken to assure that the sequences used have shared the sameevolutionary pathway. This is not always possible to judge due to the fragmentary nature of thesequence information. In most cases, there is only one LHC polypeptide identified from aparticular organism and is assumed that it is analogous to Lhcb 1 of the terrestrial plants.However, this assumption is dubious because there is insufficient characterization of the LHCsfrom anything other than select terrestrial plants and Chiamydornonas. With the Chi a + ccontaining organisms, the cloned sequences typically encode the most abundant LHC in theorganism. However, there is little direct evidence for a preferential association with either166photosystem. Without structural/functional information it is difficult to judge which sequencesmay be orthlogous or paralogous. Nonetheless, the observed gene relationships give someimportant clues as to taxon relationships in combination with morphological, biochemical andother molecular sequence studies; some of these cases are mentioned below.This study, and others (Muchhal and Schwartzbach 1992; Jansson 1994), clearly showthat the light harvesting proteins from Euglena are homologous to those of the green algae andland plants. The presence of three membranes around the chloroplast and the possession ofChls a + b suggests that the euglenoid chloroplast was acquired secondarily; evolving from asymbiotic green algae (Gibbs 1978). Phylogenetic analyses of psbA (Dl), rbcL/S (Rubisco largeand small subunit), tufA (Morden et al. 1992), chloroplast 5S rRNA (Somerville et al. 1992), andpsaB (PS I core complex) (Assali and Loiseaux-de Goër 1992) also provides evidence for a closerelationship between the green algal and euglenoid chioroplasts.Euglena contains the xanthophylls diadinoxanthin and diatoxanthin and stores a 13-1,3-glucan, paramylon, in the cytoplasm. These characteristics more closely resemble thechromophytes rather than the green algae. As well, an analysis of the chloroplast encoded SSUrRNA points toward a closer affinity of the Euglena chioroplast with those of the chromophytes(Douglas and Turner 199lb; Giovannoni et al. 1993). However, there may be problemsinvolving biased base composition in these studies that may have caused this association (seediscussion by Lockhart et al. 1994).A couple of statements can be made concerning the branch topology for the Mantoniellasequences (Fig. 5.4, 5.5). First, the early branching of the Mantoniella sequences is in agreementwith the early divergence of the Prasinophyceae from the green algal lineage based onmorphological characteristics, such as the synthesis of a Chl c-like pigment (Mg-2,4 D) and thepresence of scales on the cell body and flagella (Melkonian 1990), and from rRNA phylogeneticanalysis (Steinkötter et al. 1994). Second, based on the tree topology the acquisition of achloroplast (via a secondary endosymbiosis with a green alga), or genes, by a phagotrophic hostleading to the euglenophytes would have occurred after the separation of the prasinophytes from167the other green algal lineages. An alternative interpretation that could explain the early branchingof Mantoniella is that the LHC evolved from a paralogous member of the LHC family that wasdifferent from the gene leading to the chlorophyte LHC II gene lineage.The relationships between the fucoxanthin-containing algae are not well resolved andthere are too few complete sequences to make it interesting. However, the fucoxanthincontaining chromophytes form a distinct group separate from the peridinin-containingdinoflagellate, Amphidinium. This is in agreement with the traditional view of a distantrelationship between the dinoflagellates and the other chromophytes. This was based onmorphological characters, such as differences in the xanthophyll content, presence of a uniquesoluble LHC complex, the presence of only three membranes around the chioroplast, the apparentlack of histones, and persistently condensed chromosomes (Taylor, 1990). In addition,Phylogenetic analysis of nuclear rRNA consistently shows a deep divergence between thedinoflagellates and the other chromophytes (Bhattacharya et al. 1990; Hendricks et al. 1991;Cavalier-Smith et al. 1994b). Though the iPCPs are definitely FCP-related, whether this arose asthe result of a divergence from the chromophyte line or from an independent evolution of thechloroplast, can not be resolved with this data (but see Cavalier-Smith, 1994).The positions of the both haptophytes depicted in the FCP I iPCP trees (Fig. 5.6) are notconsistent with standard taxonomic position of this group. The first is the early branching ofIsochrysis before the dinoflagellate, and the other chromophytes. This would not be an accuratereflection of the organismal relationships as a number of studies indicate that the haptophytesform a sister group to the heterokontloomycete lineage (Andersen 1991; Bhattacharya et al. 1992;Cavalier-Smith 1994; Medlin et al. 1994; Cavalier-Smith et al. 1994b). A likely explanation forthe odd position of Isochrysis is that the FCP sequence is a paralogous gene, resulting in anerroneous tree topology. The sequence was isolated by immuno-screening a cDNA library with aFCP specific antibody that could have detected a product paralogous to the other sequences sincemost of the antennal proteins are immunologically related. Furthermore, there is evidence thatthere is only a single copy of this Jsochrysis Fcp gene (LaRoche et al. 1994), though in terrestrial168plants and all other known chromophytes, the main antennal protein is encoded by a multigenefamily (Green et al., 1991; Bhaya et al., 1993; Apt et al., 1994; Chapter 4). Furthercharacterization of the haptophyte family of antennae will have to be done before this can beresolved.The second unusual relationship is the very close association of Paviova with the diatomswhen the tree is constructed with the available Pavlova FCP protein sequence data (87characters). The Pavlova polypeptide binds both ChI’s a, c and fucoxanthin, and is biochemicallyand immunologically quite similar to other FCPs (Fawley et al. 1987; Hiller et al. 1988). itscurrent tree position is not expected. However, being a haptophyte, I would have expected it toform a deep branch at the base of the FCP lineage if it is a true reflection of organismalphylogeny. Nevertheless, the dataset is small and the species relationships may change when thecomplete sequence is determined.5.4.5 Light-harvesting protein evolution: pathways and evolutionary sourcesThe FCP and CAB proteins are clearly homologous and were derived from a singleancestral gene, though the trees show there was probably an early separation of the FCP and CABlineages. I propose that the antennal proteins associated with PS I were some of the first proteinsthat acquired the function of light-harvesting, probably from one of a photoprotective role. Itwould have been from this complex that the CABs and FCPs diverged at separate times. This issuggested for a few reasons: first, the LHC I proteins in the terrestrial plants originated from veryearly duplications as indicated by the large divergence between them, as compared to the smallerdivergence between the LHC II antennal proteins. Second, the different LHC I proteins from thegreen algae seem to have a greater affinity for specific LHC I types of the terrestrial plants, ratherthan to each other. Furthermore, the green algal LHC II proteins are clearly separated from thoseof the terrestrial plants. This suggest that the LHC I genes had diverged into the different typesbefore the separation of the green algae and land plants and prior to the duplication and169divergence of the LHC IT-related genes (Hwang and Herrin 1993). Third, the presence of aCAB/FCP related LHC I complex in the red algae (Chapter 3; Wolfe et al. 1994) along withPBSs, also suggests that the LHC I antennal proteins originated prior to the membrane intrinsicLHC II protein complex. Though none have been reported, it remains to be seen whether there isan intrinisic PS II associated antennae in red algae. As well, the sequence of the red algal LHC Iantennae needs to be determined to get a better idea of possible relationships.If the presence of a suitable LHC I associated antennae complex had been established,then the loss of PBS, due to light or nutrient stresses, may have necessitated the adaptation of anLHC I-related antennae to associate with PS II. This seems plausible as some representatives ofthe PS II associated inner antennae of land plants (CP24 and CP29) are evolutionarily closer tothe LHC I proteins than to LHC II proteins. The fact that some inner antennae of PS II areLHC I-related suggests that as additional complexes were recruited, they were added on andbecame more peripherally located. In addition, the LHC 1-680 (Lhca2 & 3) (Knoetzel et al.1992) and the LHC II complexes have similar fluorescence emission maxima (680 nm) making afunctional transition from a PS I association to a PS II association seem plausible. Alternatively,a separate LHC I-related complex may have been stressed induced for a photoprotective role intimes of PBS degradation.Since the chioroplast is generally thought to have evolved from a cyanobacterium, theCABIFCP related LHCs must have become the main antennae after the establishment of thechloroplast. In order to explain the present day pigmentation of the algae, there must have beenat least two independent losses of phycobilisomes from the ancestral organisms; once leading tothe green algae and at least one other time leading to the chromophyte plastid. Moreover, therewould have to a gain in the ability to synthesize Chl b and Chi c in the green algae andchromophytes, respectively.It is probable that the phycobilisomes were replaced by a LHC I-related complex that wasinduced for photoprotection during times of stress. The presence of an inducible LHC I-relatedsystem could act to protect the photosystem in the event of a loss or reduction of the PBS due to170either high light or nutrient deprivation (Bryant 1992). The PBSs are efficient but metabolicallyexpensive, requiring about ten times more amino acids per chromophore (Bryant 1992), so aninitial loss of this antenna due to a nutrient deficiency (Grossman et al. 1993) or other stressrelated events seems reasonable. In time and with sufficient modifications, this system couldeventually replace the PBS system of the cyanobacterial or red algal endosymbiont.Such a senario seems more likely since there are a number of LHC related complexes thatare induced during either a light or nutrient stress. These include the early light-inducibleproteins (ELIP5) in the terrestrial plants (Grimm et al. 1989; Adamska and Kloppstech 1994), thecarotenoid biosynthesis-related (Cbr) proteins in Dunaliella (Levy et al. 1992) and the high light-inducible proteins (HLIPs) in cyanobacteria (Dolganov et al. 1994). These proteins havesequence similarities to the CAB proteins in the hydrophobic domains, primarily in the first andthird transmembrane regions. Though a number of putative chlorophyll a ligands are wellconserved in the ELIPS and HLIPS (Green and Kühlbrandt 1995), it has not been conclusivelydetermined whether they bind any chlorophyll or carotenoids.Though unrelated to the CAB proteins, there is an iron stress inducible protein (isiA) incyanobacteria that is homologous to CP43 (psbC) (Laudenbach and Straus 1988) and bindschlorophyll (Burnap et al. 1993). It has been postulated to act as a chlorophyll reserve (Burnap etal. 1993) or as a antennal replacement in the absence of PBS (Pakrasi et al. 1985). The similarityof the isiA gene to the LHCs of the prochlorophytes (Hiller & Larkum; LaRouche & Partensky;van der Staay & Green; unpubi.) demonstrates the potential of stress induced proteins in thecreation of novel antennal systems.There are two potential cyanobacterial molecular sources from which the eukaryoticLHCs may have evolved after the evolution of the chioroplast; the psbS gene and the HLIPs.Though the psbS gene product has been immunologically detected in a cyanobacterium (Nilssonet al. 1990), two other groups have failed to do so (Kim et al. 1994; Vermaas, pers. comm.). Atthe moment it has only been cloned from tomato and spinach so its presence in cyanobacteria isuncertain. Nonetheless, the psbS protein binds chlorophyll and is predicted to span the171membrane four times (Kim et al. 1992; Funk et al. 1994). If present in the cyanobacteria, a C-terminal deletion could give a LHC precursor with three transmembrane helicies (Green andPichersky 1994).The HLIPs are another potential source for the evolution of the LHCs as they are knownto occur in cyanobacteria (Dolganov et al. 1994) and they have homologues in red algae (Reith,unpubi.; gb X62578) and the Glaucophyta (Stirewalt and Bryant, unpubl.) (see Green andKühlbrandt 1995). The HLIPs are only 72 amino acids long, yet there is sequence similarity tothe first or third membrane spanning region of the ELIPs and CABs. It has been proposed thatthese proteins function as homodimers and were the evolutionary source of the eukaryotic LHCs.One can envisage a series of gene duplications and fusions that could give a LHC that spannedthe membrane three times and adopted the role of light-harvesting (Dolganov et al. 1994; Greenand Kühlbrandt 1995).With the available sequence information it is not possible to make any firm conclusionsregarding the ancestry of the FCPs. This will have to wait until more diverse FCP familymembers have been identified and sequenced. The sequence of the red algal LHC proteins willalso be an important piece of information. I would predict that they will be more closely relatedto the FCPs rather than to the CABs.172CHAPTER 6SummaryThis dissertation demonstrates that the light-harvesting antennae of Heterosigma carteraeform an intricate system comparable to the complexity of the LHC antennae seen in the terrestrialplants. Heterosigma possesses up to 12 differently migrating polypeptides that cross react todifferent extents with CAB and FCP specific antisera. There are four prominent LHCs inHeterosigma with apparent molecular masses of 18-21 kDa. The gene encoding one of theseantennae, the 19.5 kDa polypeptide, was cloned and sequenced. Based on Southern hybridizationand eDNA sequencing, there are approximately 6-8 gene copies encoding this polypeptide.Overall, there are probably over 20 related genes encoding the FCPs in Heterosigma.The FCPs are structurally related to the CABs as determined by the immunological cross-reaction data and through direct sequence comparisons. This implies they are evolutionarilyrelated and had evolved from the same ancestral gene. Both the CABs and the FCPs have threeputative membrane spanning regions. In the pea LHC II two of the three membrane spanningregions (MSR1 and 3) interact to bind carotenoids and chlorophyll. The corresponding regions inthe FCPs are very conserved. Some of the highly conserved amino acids in the FCPs are thoughtto bind chlorophyll and/or are important in helix-helix interactions that can help to stabilize thecomplex. These striking similarities indicate that the FCPs and the CABs are structurally quitesimilar.The LHCs from the diverse algal taxa that utilize fucoxanthin as an accessory chlorophyll(the FCPs) are related and form a natural monophyletic group. The intrinisic peridininchlorophyll protein complexes (iPCPs) from the dinoflagellates are definitely more closelyrelated to the FCPs than they are to the CABs and they form a sister group to the FCP dade. Itseems that the acquisition of xanthophylls for a primary role in light harvesting, rather than solely173a photoprotective one, occurred early in the evolution of these LHCs. The use of differentprimary xanthophylls in the distinct antennae complexes (fucoxanthin, peridinin,vaucherioxanthin, etc.) was then the result of divergence following the separation of the mainalgal taxa. This was probably related to the light-environment the algal group experiences in themarine habitat.The presence of CAB/FCP-related LHCs in the red algae provides a link between theantennal systems of the three major groups of photosynthetic organisms. In addition, the lack ofsuch immunologically related LHCs in the cyanobacteria and prochiorophytes suggests that themembrane intrinisic LHCs originated following the endosymbiotic origin of the chloroplast thatgave rise to the first true photosynthetic eukaryote. Such a scenario implies a monophyleticorigin for the chloroplast as it is unlikely that related proteins could evolve independently indifferent lineages.This work could continue in a number of directions. It would be interesting to furthercharacterize the chlorophyll protein complexes in Heterosigma in addition to the sequencing ofthe genes encoding them. This would give an indication of the gene family complexity and howmuch diversity exists between the different members. This will allow the determination of thenature of the divergence that is responsible for the differential immunological cross-reactivityseen in Chapter 2. This information would be useful if antennae gene characterization studiesfrom representatives of other major algal taxa were also being done. This comparison would bevery useful in determining if specific gene types are conserved between the diverse algal groupssimilar to the conservation of the Cab gene types between the angiosperms and gymnosperms.Should specific gene types exist it would make an interesting gene evolution study; it wouldallow one to determine at what point certain gene duplications had occurred in relation to thephylogeny of the algae being compared. This would also complement the work currently beingdone with the Cab gene family and may help to determine which specific Cab gene types aremore closely related to the Fcp genes.The analyses of the gene family complexity have to be done in conjunction with structuraland functional studies on the pigment-protein complexes. Such functional/structural studies174include whether the complex is specifically associated with either PS I or PS II, more accurateestimates of the pigment content of the different antennae complexes and the regulatory role ofthe divergent FCPs in the adaptation to different light regimes. These studies will allow for abetter understanding of how the antennal systems of divergent organisms are organized and givean idea of the similarities and differences in the ways they adapt to harvest light. The correlationof the function of a specific complex to the estimated time of appearance will allow for theassessment of the evolution of the photosynthetic antennal systems in different organisms.175REFERENCESAebersold, R., J. Leavitt, L.E. Hood and S.B.H. Kent (1987). Internal amino acid sequenceanalysis of proteins separated by one- or two- dimensional gel electrophoresis after in situprotease digestion on nitrocellulose. Proc. Nail. Acad. Sci. USA 184: 6970-6974.Adamska, I. and K. Kloppstech (1994). Low temperature increases the abundance of early light-inducible transcript under stress conditions. J. Biol. Chem. 269 (48): 30221-30226.Aebersold, R., G.W. Pipes, R.E.H. Wettenhall, H. Nilco and L.E. Hood (1990). 187: 56-65.Alberte, R.S., A.L. Friedman, D.L. Gustafson, M.S. Rudnick and H. Lyman (1981). Light-harvesting systems of brown algae and diatoms. Isolation and characterization of thechlorophyll a/c and chlorophyll a/fucoxanthin pigment-protein complexes. Biochim. Biophys.Acta 635: 304-3 16.Alfonso, M., G. Montoya, R. Cases, R. Rodriguez and R. Picorel (1994). Core antennacomplexes, CP43 and CP47, of higher plant photosystem II. Spectral properties, pigmentstoichiometry, and amino acid composition. Biochemistry 33: 10494-10500.Allen, J.F. (1992). Protein phosphorylation in regulation of photosynthesis. Biochim. Biophys.Acta 1098: 275-335.Allen, K.D. and L.A. Staehelin (1991). Resolution of 16 to 20 Chlorophyll-Protein ComplexesUsing a Low Ionic Strength Native Green Gel System. Anal. Biochem. 194: 2 14-222.Allen, K.D. and L.A. Staehelin (1992). Biochemical characterization of photosystem II antennapolypeptides in grana and stroma membranes of spinach. Plant Physiol. 100: 15 17-1526.Andersen, R.A. (1991). The cytoskeleton of chromophyte algae. Protoplasma 164: 143-159.Anderson, J.M. (1985). Chlorophyll-protein complexes of a marine green alga, Codium species(Siphonales). Biochim. Biophys. Acta 806: 145-153.Anderson, J.M. and B. Andersson (1988). The dynamic photosynthetic membrane andregulation of solar energy conversion. Trends Biochem. Sci. 13: 35 1-355.Anderson, J.M., P.K. Evans and D.J. Goodchild (1987). Immunological cross-reactivity betweenthe light-harvesting chlorophyll a/b-proteins of a marine green alga and spinach. Physiol.Plantarum. 70: 597-602.Apt, K.E., D. Bhaya and A.R. Grossman (1994). Characterization of genes encoding the light-harvesting proteins in diatoms: biogenesis of the fucoxanthin chlorophyll a/c proteincomplex. J Applied Phycol. 6: 225-230.Apt, K.E., N.E. Hoffman and A.R. Grossman (1993). The y-subunit of R-phycoerythrin and itspossible mode of transport into the plastid of red algae. J. Biol. Chem. 268(22): 16208-16215.Arsalane, W., B. Rousseau and J.C. Thomas (1992). Isolation and Characterization of NativePigment-Protein Complexes From Two Eustigmatophyceae. J. Phycol. 28: 32-36.Assali, N.E. and S. Loiseaux-de Goër (1992). Sequence and phylogeny of the psaB gene ofPylaiella littoralis (Phaeophyta). J. Phycol. 28: 209-213.176Assali, N.E., R. Mache and L.-d. Goër (1990). Evidence for a composite phylogenetic origin ofthe plastid genome of the brown alga Pylaiella littoralis (L.) Kjellum. Plant Mol. Biol.15(307-315):Awramik, S.M. (1992). The oldest records of photosynthesis. Photosyn. Res. 33: 75-89.Barbato, R., F. Rigoni, M.T. Giardi and G.M. Giacometti (1989). The minor antenna complexesof an oxygen evolving photosystem II preparation: purification and stoichiometry. FEBS lett.251(1/2): 147-154.Barrett, J. and J.M. Anderson (1980). The P-700-chlorophyll a-protein complex and two majorlight-harvesting complexes of Acrocarpia paniculata and other brown seaweeds. Biochim.Biophys. Acta 590: 309-323.Bassi, R., B. Pineau, P. Dainese and J. Marquardt (1993). Carotenoid-binding proteins ofphotosystem II. Eur. J. Biochem. 212: 297-303.Bassi, R., F. Rigoni and G.M. Giacometti (1990). Chlorophyll binding proteins with antennafunction in higher plants and green algae. Photochem. Photobiol. 52(6): 1187-1206.Bassi, R. and D. Simpson (1987). Chlorophyll-protein complexes of barley photosystem I. Fur.J. Biochem. 163: 221-230.Bassi, R., S.Y. Soen, G. Frank, H. Zuber and J.D. Rochaix (1992). Characterization ofchlorophyll a/b proteins of photosystem I from Chlamydomonas reinhardtii. J. Biol. Chem.267(36): 25714-25712.Bengis, C. and N. Nelson (1975). Purification and properties of the Photosystem I reactioncenter from chioroplasts. J. Biol. Chem. 250: 2783-2788.Bennetzen, J.L. and B.D. Hall (1982). Codon selection in yeast. J. Biol. Chem. 257(6): 3026-303 1.Berkaloff, C., L. Caron and B. Rousseau (1990). Subunit organization of PSI particles frombrown algae and diatoms: polypeptide and pigment analysis. Photosyn. Res. 23: 181-193.Bhattacharya, D., H.J. Elwood, L.J. Goff and M.L. 50gm (1990). Phylogeny of Gracilarialemanetformis (Rhodophyta) based on sequence analysis of its small subunit ribosomal RNAcoding region. J. Phycol. 26(181-186):Bhattacharya, D., L. Medlin, P.O. Wainwright, E.V. Ariztia, C. Bibeau, S.K. Stickel and M.L.Sogin (1992). Algae containing chiorophylls a + c are polyphyletic: molecular evolutionaryanalysis of the Chromophyta. Evolution 46(6): 1801-18 17.Bhaya, D. and A. Grossman (1991). Targeting proteins to diatom plastids involves transportthrough an endoplasmic reticulum. Mol. Gen. Genet. 229: 400-404.Bhaya, D. and A.R. Grossman (1993). Characterization of gene clusters encoding thefucoxanthin chlorophyll proteins of the diatom Phaeodactylum tricornutum. Nucl. Acid Res.2 1(19): 4458-4466.Bjorland, T. and S. Liaaen-Jensen (1989). Distribution patterns of carotenoids in relation tochromophyte phylogeny and systematics. The Chroinophyte Algae: Problems andPerspectives. Oxford, Clarendon Press. 37-60.177Blankenship, R.E. (1992). Origin and early evolution of photosynthesis. Photosyn, Res. 33: 91-lii.Boczar, B.A., T.P. Delaney and R.A. Cattolico (l989b). The gene for the ribulose-1,5-bisphosphate carboxylase small subunit protein of the marine chromophyte Olisthodiscusluteus is similar to that of a chemoautotrophic bacterium. Proc. Natl. Acad. Sci. USA 86:4996-4999.Boczar, B.A. and B.B. Prezelin (1987). Chlorophyll-protein complexes from the red-tidedinoflagellate, Gonyaulax polyedra Stein. Plant Physiol. 83: 805-812.Boczar, B.A. and B.B. Prezelin (1989). Organization and Comparison of Chlorophyll-ProteinComplexes from Two Fucoxanthin-Containing Algae: Nitzschia closterium(Bacillariophyceae) and Isochrysis galbana (Prymnesiophyceae). Plant Cell Physiol. 30(7):1047-1056.Boczar, B.A., B.B. Prezelin, J.P. Markwell and J.P. Thornber (1980). A chlorophyll ccontaining pigment-protein complex from the marine dinoflagellate, Glenodinium SP. FEBSlett. 120(2): 243-247.Brinkmann, H., P. Martinez, F. Quigley, W. Martin and R. Cerff (1987). Endosymbiotic originand codon bias of the nuclear gene for chloroplast glyceraldehyde-3-phosphatedehydrogenase from maize. J. Mol. Evol. 26: 320-328.Brown, J.S. (1988). Photosynthetic pigment organization in diatoms (Bacillariophyceae). J.Phycol. 24: 96-102.Bryant, D.A. (1992). Puzzles of chloroplast ancestry. Current Biology 2(5): 240-242.BUchel, C. and C. Wilhelm (1993). Isolation and characterization of a photosystem I-associatedantenna (LHCI) and a photosystem I-core complex from the chlorophyll c-containing algaPleurochloris meiringensis (Xanthophyceae). .1. Photochein. Photobiol. B: Biol. 20: 87-93.Büchel, C., C. Wilhelm, N. Hauswirth and A. Wild (1992). Evidence for a lateral heterogeneityby patch-work like areas enriched with photosystem I complexes in the three thylakoidlamellae of Pleurochloris meiringensis (Xanthophyceae). Crypt. Bot. 2: 375-386.Bullerjahn, G.S., T.C. Jensen, D.M. Sherman and L.A. Sherman (1990). Immunologicalcharacterization of the Prochlorothrix hollandica and Prochloron sp. chlorophyll a/b antennaproteins. FEMS Micobiol. letts. 67: 99-106.Bullerjahn, G.S., H.C.P. Matthijs, L.R. Mur and L.A. Sherman (1987). Chlorophyll-proteincomposition of the thylakoid membrane from Prochlorothrix hollandica, a prokaryotecontaining chlorophyll b. Eur. J. Biochem. 168: 295-300.Burnap, R.L., T. Troyan and L.A. Sherman (1993). The highly abundant chlorophyll-proteincomplex of iron-deficient Synechococcus sp. PCC7942 (CP43’) is encoded by the isiA gene.Plant Physiol. 103: 893-902.Campbell, W.H. and G. Gown (1990). Codon usage in higher plants, gren algae, andcyanobacteria. Plant Physiol. 92: 1-11.Camm, E.L. and B.R. Green (1989). The chlorophyll ab complex, CP29, is associated with thephotosystem II reaction centre core. Biochim. Biophys. Acta 974: 180-184.178Cao, Y., I. Adachi, A. Janke, S. Pääbo and M. Hasegawa (1994). Phylogenetic relationshipsamong Eutherian orders estimated from inferred sequences of mitochondrial proteins:instability of a tree based on a single gene. J. Mol. Evoi. 39: 5 19-527.Caron, L. and J. Brown (1987). Chiorophyll-carotenoid protein complexes from the diatom,Phaeodaclylum tricornutum: Spectropotometric, pigment and polypeptide analyses. PlantCell Physiol. 28(5): 775-785.Caron, L., R. Remy and C. Berkaloff (1988). Polypeptide composition of light-harvestingcomplexes from some brown algae and diatoms. FEBS lett. 229(1): 11-15.Cattolico, R.A., J.C. Boothroyd and S.P. Gibbs (1976). Synchronous growth and plastidreplication in the naturally wall-less alga Olisthodiscus luteus. Plant Physiol. 57: 497-503.Cavalier-Smith (1987). Glaucophyceae and the origin of plants. Evol. Trends Plants 2: 75-78.Cavalier-Smith, T. (1982). The origins of plastids. Biol. J. Linn. Soc. 17: 289-306.Cavalier-Smith, T. (1993). The origin, losses and gains of chioroplasts. Origins of Plastids. NewYork, Chapman and Hall. 29 1-348.Cavalier-Smith, T. (1994). Origin and relationships of haptophyta. The Haptophyte Algae.Oxford, Clarendon press. 4 13-435.Cavalier-Smith, T., M.T.E.P. Allsopp and E.E. Chao (1994b). Chimeric conundra: arenucleomorphs and chromists monophyletic or polyphyletic? Proc. Natl. Acad. Sci. USA 91:11368-11372.Chitnis, P.R., D. Purvis and N. Nelson (1991). Molecular cloning and targetted mutagenesis ofthe gene psaF encoding subunit III of photosystem I from the cyanobacteria Synechocystissp. PCC 6803. 1. Biol. Chem. 266(30): 20146-20151.Chitnis, P.R. and J.P. Thornber (1988). The major light-harvesting complex of photosystem II:aspects of its molecular and cell biology. Photosyn. Res. 16: 4 1-63.Chromczynski, P. and N. Sacchi (1987). Single-step method of RNA isolation by acidguanidinium tiocyanate-phenol-chloroform extraction. Anal. Biochem. 162: 156-159.Chrystal, J. and A.W.D. Larkum (1987). Pigment-protein complexes and light harvesting inEustigmatophyte alage. Progress in Photosynthesisi Research. Dordrecht, Martinus NijhoffPublishers. 189-192.Chu, Z.X. and J.M. Anderson (1985). Isolation and characterization of a siphonaxanthinchlorophyll a/b-protein complex of photosystem I from a Codiuin species (Siphonales).Biochirn. Biophys. Acta 806: 154-160.Church, G.M. and W. Gilbert (1984). Genomic sequencing. Proc. Natl. Acad. Sci. USA 81:1991- 1995.Cramer, W.A., S.E. Martinez, D. Huang, G.-S. Tae, R.M. Everly, J.B. Heymann, R.H. Cheng,T.S. Baker and J.L. Smith (1994). Structural aspects of the cytochrome b6f complex;structure of the lumen-side domain of cytochromef. I. Bioenerg. Biomembr. 26(1): 31-47.Cunningham, F.X. and J.A. Schiff (1986). Chlorophyll-protein complexes from Euglenagracilis and mutants defective in chlorophyll b: polypeptide composition. Plant Physiol. 80:23 1-238.179Cunningham, F.X. and J.A. Schiff (1986b). Chlorophyll-protein complexes from Euglenagracilis and mutants deficient in chlorophyll I: pigment composition. Plant Physiol. 80: 223-230.Dayhoff, M.O., R.M. Schwartz and B.C. Orcutt (1978). A model of evolutionary change inproteins. Atlas ofprotein sequence and structure. Washington, D.C., National BiomedicalResearch Foundation. 345-352.Deisenhofer, J., 0. Epp, K. Miki, R. Huber and H. Michel (1985). Structure of the proteinsubunits in the photosynthetic reaction centre of Rhodopseudomonas viridis at 3A resoluton.Nature 318: 618-624.Demrnin, D.S., Y.C. Stockinger and L.L. Walling (1989). Phylogenetic relationships betweenthe chlorophyll aJb binding protein (CAB) multigene family: an intra- and interspecies study.J. Mol. Evol. 29: 266-279.Dolganov, N.A.M., D. Bhaya and A.R. Grossman (1994). Cyanobacterial protein withsimilaroty to the chlorophyll afb binding proteins of higher plants: evolution and regulation.Proc. Nati. Acad. Sci. USA In Press:Douady, D., B. Rousseau and L. Caron (1994). Fucoxanthin-chlorophyll a/c light-harvestingcomplexes of Laminaria saccharina: partial amino acid sequences and arrangement inthylakoid membranes. Biochem. 33(11): 3165-3 170.Douglas, S.E. (1994b). Chioroplast origins and evolution. The molecular biology ofcyanobacteria. Boston, Kiuwer Academic Publishers.Douglas, S.E. and D.G. Durnford (1989). The small subunit of ribulose-1,5-bisphosphatecarboxylase is plastid-encoded in the chlorophyll c-containing alga Cryptomonas ‘2. PlantMol. Biol. 13: 13-20.Douglas, S.E., D.G. Durnford and C.W. Morden (1990). Nucleotide sequence of the gene for thelarge subunit of ribulose- 1 ,5-bisphosphate carboxylase/oxygenase from cryptomonas cP:Evidence supporting the polyphyletic origin of chioroplasts. J. Phycol. 26: 500-508.Douglas, S.E. and C.A. Murphy (1994). Structural, transcriptional, and phylogenetic analysis ofthe atpB gene cluster from the plastid of Cryptomonas 1 (Cryptophyceae). J. Phycol. 30:329-340.Douglas, S.E., C.A. Murphy, D.A. Spencer and M.W. Gray (1991). Cryptomonad algae areevolutionary chimaeras of two phylogenetically distinct unicellular eukaryotes. Nature 350:148- 15 1.Douglas, S .E. and S. Turner (1991 b). Molecular evidence for the origin of plastids from acyanobacterium-like ancestor. J. Mol. Evol. 33: 267-273.Dreyfuss, B.W. and J.P. Thornber (1994). Assembly of the light-harvesting complexes (LHCs)of photosystem II. Plant Physiol. 106: 829-839.Dunahay, T.G. and L.A. Staehelin (1986). Isolation and characterization of a new minorchlorophyll a/b-protein complex (CP24) from spinach. Plant Physiol. 80: 429-434.180Dunsmuir, P., S.M. Smith and J. Bedbrook (1983). The major chlorophyll a/b binding protein ofpetunia is composed of several polypeptides encoded by a number of distinct genes. J. Mol.Appl. Genet. 2: 285-300.Durnford, D.G. and B.R.G. Green (1994). Charcacterization of the light harvesting proteins ofthe chromophytic alga, Olisthodiscus luteus (Heterosigma carterae). Biochim. Biophys. Acta1184: 118-123.Duval, J.C., H. Jupin and C. Berkaloff (1983). Photosynthetic properties of plastids isolatedfrom macrophytic brown seaweeds. Physiol. Veg. 2 1(6): 1145-1157.Erickson, J.M., M. Rahire and J.D. Rochaix (1984). Chiamydomonas reinhardtii gene for the32,000 mol. wt. protein of photosystem II contains four large introns and is located entirelywithin the chloroplast inverted repeat. EMBO 3(12): 2753-2762.Fawley, M.W. and A.R. Grossman (1986). Polypeptides of a light-harvesting complex of thediatom Phaeodactylum tricornutum are synthesized in the cytoplasm of the cell asprecursors. Plant Physiol. 81: 149-155.Fawley, M.W., S.J. Morton, K.D. Stewart and K.R. Mattox (1987). Evidence For a CormnonEvolutionary Origin of Light-Harvesting Fucoxanthin Chlorophyll a/c-Protein Complexes ofPavlova gyrans (Prymnesiophyceae) and Phaeodactylum tricornutum (Bacillariophyceae). J.Phycol. 23: 377-38 1.Fawley, M.W., K.D. Stewart and K.R. Mattox (1986b). The novel light-harvesting pigment-protein complex of Mantoniella squamata (Chlorophyta): Phylogenetic implications. I. Mol.Evol. 23: 168-176.Felsenstein, J. (1978). Cases in which parsimony or compatibility methods will be positivelymisleading. Syst. Zool. 27: 401-410.Felsenstein, J. (1985). Confidence limits on phylogenies: an approach using the bootstrap.Evolution 39(4): 783-791.Felsenstein, J. (1992). PHYLIP (phylogeny inference package). Seattle, WA., University ofWashington.Friedman, A.L. and R.S. Alberte (1984). A diatom light-harvesting pigment-protein complex.purification and isolation. Plant Physiol. 76: 483-489.Frohman, M.A. (1990). RACE:rapid amplification of cDNA ends. PCR Protocols. San Diego,Academic Press. 28-38.Funk, C., W.P. Schröder, B.R. Green, G. Renger and B. Andersson (1994). The intrinisic 22kDa protein is a chlorophyll-binding subunit of photosystem II. FEBS Lett. 342: 26 1-266.Gagné, G. and M. Guertin (1992). The early genetic response to light in the green unicellularalga Chlamydomonas eugametos grown under light/dark cycles involves genes that representdirect responses to light and photosynthesis. Plant Mol. Biol. 18: 429-445.Gantt, E. (1981). Phycobilisomes. Annual review ofplant physiology. Palo Alto, Ca., AnnualReviews Inc. 327-347.Gantt, E. (1990). Pigmentation and photoacclimation. Biology of the red algae. New York,Cambridge University Press. 203-219.181Garbary, D.J. and P.W. Gabrielson (1990). Taxonomy and evolution. Biology of the Red Algae.Cambridge, Cambridge University Press. 477-498.Ghanotakis, D.F. and C.F. Yocum (1990). Photosystem II and the oxygen-evolving complex.Annu. Rev. Plant Physiol. Plant Mol. Biol. 41: 255-276.Gibbs, P.B. and J. Biggins (1991). In vivo and in vitro protein phosporylation studies onOchromonas danica, an alga with a chlorophyll a/c/fucoxanthin binding protein. PlantPhysiol. 97: 388-395.Gibbs, P.B. and J. Biggins (1991). Thylakoid organization in the Chromophytic algaOchromonas dancia. Isolation and characterization of a new pigment-protein complex.Plant Physiol. 97: 38 1-387.Gibbs, S .P. (1970). The comparative ultrastructure of the algal chloroplast. Ann. N.Y Acad. Sci.175: 454-473.Gibbs, S.P. (1978). The chloroplasts of Euglena may have evolved from a symbiotic greenalgae. Can. .J. Bot. 56: 2883-2889.Gibbs, S.P. (1979). The route of entry of cytoplasmically synthesized proteins of algaepossessing chloroplast ER. I. Cell Sci. 35: 253-266.Gibbs, S.P. (1981). The chloroplast endoplasmic reticulum: structure, function and evolutionarysignificance. Inter. Rev. Cytology 72: 49-99.Gillott, M.A. and S.P. Gibbs (1980). The crytomonad nucleomorph: its ultrastructure andevolutionary significance. J. Phycol. 16: 558-568.Giovannoni, S.J., N. Wood and V. Huss (1993). Molecular phylogeny of oxygenic cells andorganelles based on small-subunit ribosomal RNA sequences. Origins of Plastids. NewYork, Chapman and Hall. 159-168.Glazer, A.N. (1989). Light guides: directional energy transfer in photosynthetic antenna. J. Biol.Chem. 264(1): 1-4.Golbeck, J.H. (1992). Structure and function of photosystem I. Annu. Rev. Plant. Physiol. PlantMol. Biol. 43: 293-324.Golbeck, J.H. and D.A. Bryant (1991). Photosystem I. Current Topics in Bioenergetics. SanDiego, Academic Press. 83-177.Golden, S.S., C.W. Morden and K.L. Greer (1993). Comparison of sequences and organizationof photosynthesis genes among the prochlorophyte Prochlorthrix hollandica, cyanobacteria,and chioroplasts. Origins ofplastids. Symbiogenesis, prochiorophytes, and the origins ofchioroplasts. New York, Chapman and Hall. 141-158.Govind, N.S., S.J. Roman, R. Iglesias-Prieto, R.K. Trench, E.L. Triplett and B.B. Prézelin(1990). An analysis of the light-harvesting peridinin-chlorophyll a-proteins fromdinoflagellates by immunoblotting techniques. Proc. R. Soc. Lond. B 240: 187-195.Gray, M.W. and W.F. Doolittle (1982). Has the endosymbiont hypothesis been proven?Microbiol. Rev. 46(1): 1-42.Green, B.R. (1988). The chlorophyll-protein complexes of higher plant photosyntheticmembranes. Photosyn. Res. 15: 3-32.182Green, B.R., D.G. Durnford, R. Aebersold and E. Pichersky (1992). Evolution of structure andfunction in the chi a/b and chl a/c antenna protein family. Research in Photosynthesis. NewYork, Kiuwer Academic Publishers. 195-201.Green, B.R. and W. Kühlbrandt (1995). Sequence conservation of light-harvesting and stress-response proteins in relation to the three-dimensional molecular structure of LHCII.Photosyn. Res. In press:Green, B.R. and E. Pichersky (1993). Nucleotide sequence of an Arabidopsis thaliana Lhcb4gene. Plant Physiol. 103: 145 1-1452.Green, B.R. and E. Pichersky (1994). Hypothesis for the evolution of three-helix chl a/b and chia/c light-harvesting antenna proteins from two-helix and four-helix ancestors. Photosyn. Res.in press.Green, B.R., E. Pichersky and K. Kloppstech (1991). Chlorophyll a/b-binding proteins: anextended family. Trends Biochem. Sci. 16: 181-186.Green, B.R., D. Shen, R. Aebersold and E. Pichersky (1992b). Identification of the polypeptidesof the major light-harvesting complex of photosystem II (LHCII) with their genes in tomato.FEBS letts 305(1): 18-22.Greenwood, A.D., H.B. Griffiths and U.J. Santore (1977). Chloroplasts and cell compartmentsin Cryptophyceae. Br. Phycol. J. 12: 119.Grimm, B., E. Kruse and K. Kloppstech (1989). Transiently expressed early light-induciblethylakoid proteins share transmembrane domains with light-harvesting chlorophyll bindingproteins. Plant Mol. Biol. 13: 583-593.Grossman, A., A. Manodori and D. Snyder (1990). Light-harvesting proteins of diatoms: theirrelationship to the chlorophyll a/b binding proteins of higher plants and their mode oftransport into plastids. Mol. Gen. Genet. 224: 9 1-100.Grossman, A.R., M.R. Schaefer, G.G. Chiang and J.L. Collier (1993). The phycobilisome, alight-harvesting complex responsive to environmental conditions. Microbiol. Rev. 57(3):725-749.Gugliemelli, L.A. (1984). Isolation and characterization of the pigment-protein particles fromthe light-harvesting complex of Phaeodactyum tricornutum. Biochim. Biophys. Acta 766: 45-50.Hansson, O. and T. Wydrzynski (1990). Current perceptions of photosystem II. Photosyn. Res.23: 131-162.Hara, Y. and M. Chihara (1987). Morpology, ultrastructure and taxonomy of theRaphidophycean alga Heterosigma akashiwo. Bot. Mag. Tokyo 100: 151-163.Hara, Y., I. Inouye and M. Chihara (1985). Morphology and ultrastructure of Olisthodiscusluteus (Raphidophyceae) with special reference to the taxonomy. Bot. Mag. Tokyo 98: 251-262.Harrison, M.A. and A. Melis (1992). Organization and stability of polypeptides associated withthe chlorophyll a-b light-harvesting complex of photosystem-II. Plant Cell Physiol. 33(5):627-637.183Hauska, G. (1986). Composition and structure of cytochrome bc1 and b6fcomplexes.Photosynthesis III: photosynthetic membranes and light-harvesting systems. New York,Springer-Verlag. 496-507.Haworth, P., J.L. Watson and C.J. Arntzen (1983). The detection, isolation and characterizationof a light-harvesting complex which is specifically associated with photosystem I. Biochim.Biophys. Acta 724: 151-158.Hendricks, L., R. De Baere, Y. Van de Peer, P.J. Neefs, A. Goris and R. De Wachter (1991).The evolutionary Position of the rhodophyte Porphyra umbilicalis and the BasidiomyceteLeucosporidium scottii among other eukaryotes as deduced from the complete sequences ofsmall ribosomal subunit RNA. J. Mol. Evol. 32: 167-177.Hibberd, D.J. and R.E. Norris (1984). Cytology and ultrastructure of Chiorarachnion reptans(chiorarachniophyta divisio nova, Chlorarachniophyceae classis nova). J. phycol. 20: 310-330.Hill, R. and R. Bendel (1960). Function of the two cytochrome components in chloroplasts: aworking hypothesis. Nature 186: 136-137.Hiller, R.G., J.M. Anderson and A.W.D. Larkum (1991). The chlorophyll-protein-complexes ofalgae. Chiorophylls. Boca Raton, CRC Press. 529-547.Hiller, R.G., A.-M. Bardin and E. Nabedryk (1987). The secondary structure content of apigment-protein complex from the thylakoids of two chromophytic algae. Biochim. Biophys.Acta 894: 365-369.Hiller, R.G. and A.W.D. Larkum (1981). Chlorophyll-proteins of the red alga Griffithsiamonilis. Photosynthesis III. Structure and molecular organisation of the photosyntheticapparatus. Philadelphia, Balaban International Science Scervices. 387-396.Hiller, R.G. and A.W.D. Larkum (1985). The chlorophyll-protein complexes of Prochioron sp.(Prochlorophyta). Biochim. Biophys. Acta 806: 107-115.Hiller, R.G., A.W.D. Larkum and P.M. Wrench (1988). Chlorophyll proteins of theprymnesiophyte Paviova lutherii (Droop) comb. nov.: identification of the major light-harvesting complex. Biochim. Biophys. Acta 932: 223-231.Hiller, R.G., P.M. Wrench, A.P. Gooley, G. Shoebridge and J. Breton (1993). The majorintrinsic light-harvesting protein of Amphidinium: characterization and relation to other light-harvesting proteins. Photochem. and Photobiol. 57(1): 125-131.Hillis, D.M., M.W. Allard and M.M. Miyamoto (1993). Analysis of DNA sequence data:phylogenetic inference. Methods in Enzymolgy. New York, Academic Press. 456-487.Hoffman, N.E., E. Pichersky, V.S. Malik, C. Castresana, K. Ko, S.C. Darr and A.R. Cashmore(1987). A cDNA clone encoding a photosystem I protein with homolgy to photosystem IIchlorophyll alb-binding polypeptides. Proc. Natl. Acad. Sci. 84: 8844-8848.Holt, A.S. (1966). Recently characterized chlorophyll. The Chlorophylls. New York, AcademicPress. 111-118.Holton, T.A. and M.W. Graham (1991). A simple and efficient method for direct cloning ofPCR products using ddT-tailed vectors. Nucl. Acids Res. 19(5): 1156.184Houlné, G. and R. Schantz (1987). Molecular analysis of the transcripts encoding the light-harvesting chlorophyll aTh protein in Euglena gracilis: Unusual size of the mRNA. Curr.Genet. 12: 611-616.Houlné, G. and R. Schantz (1988). Characterization of cDNA sequences for LHCI apoproteinsin Euglena gracilis: the mRNA encodes a large precursor containing several consecutivedivergent polypeptides. Mol. Gen. Genet. 243: 479-486.Hsu, B.-D. and J.-Y. Lee (1987). Orientation of pigments and pigment-protein complexes in thediatom Cylindrothecafusiformis. A linear-dichroism study. Biochim. Biophys. Acta 893:572-577.Hulburt, E.M. (1965). Flagellates from brackish waters in the vicinity of Woods Hole,Massachusetts. J. Phycol. 1: 87-94.Hwang, S. and D.L. Herrin (1993). Characterization of a cDNA encoding the 20-KDaphotosystem I light-harvesting polypeptide of Chiamydomonas reinhardtii. Curr. Genet. 23:5 12-5 17.Iglesias-Prieto, R., N.S. Govind and R.K. Trench (1993). Isolation and characterization of threemembrane-bound chlorophyll-protein complexes from four dinoflagellate species. Phil.Trans. R. Soc. Lond. B 340: 381-392.Ikemura, T. (1982). Correlation between the abundance of yeast transfer RNAs and theoccurrence of the respective codons in protein genes. J. Mol. Biol. 158: 573-597.Imbault, P., C. Wittemer, U. Johanningmeier, J.D. Jacobs and S.H. Howell (1988). Structure ofthe Chiamydomonas reinhardtii cabli-! gene encoding a chlorophyl-alb-binding protein.Gene 73: 397-407.Ingram, K. and R.G. Hiller (1983). Isolation and characterization of a major chlorophyll a/c2light-harvesting protein from a Chroomonas species (Cryptophyceae). Biochim. BiophysActa 722: 310-319.Ish-Shalom, D. and I. Ohad (1983). Organization of chlorophyll-protein complexes ofphotosystem Tin Chiamydomonas reinhardii. Biochim. Biophys. Acta 722: 498-507.Jam, R., R.H. Gomer and J.J. Murtagh (1992). Increasing specificity from the PCR-RACEtechnique. BioTechniques 12(1): 58-59.Jansson, S. (1994). The light-harvesting chlorophyll a/b-binding proteins. Biochim. Biophys.Acta 1184: 1-19.Jansson, S. and P. Gustafsson (1990). Type I and type II genes for the chlorophyll a/b-bindingprotein in the gymnosperm Pinus sylvestris (Scots pine): cDNA cloning and sequenceanalysis. Plant Mol. Biol. 14: 287-296.Jansson, S. and P. Gustafsson (1991). Evolutionary conservation of the chlorophyll alb-bindingproteins:cDNA5 encoding type I, II and III LHC I polypeptides from the gymnosperm Scotspine. Mol. Gen. Genet. 229: 67-76.Jansson, S., E. Pichersky, R. Bassi, B.R. Green, M. Ikeuchi, A. Melis, D.J. Simpson, M.Spangfort, L.A. Staehelin and J.P. Thornber (1992). A nomenclsture for the genes encodingthe chlorophyll a/b-binding proteins of higher plants. Plant Mol. Biol. Reporter 10(3): 242-253.185Jeffery, S.W. (1989). Chlorophyll c pigments and their distribution in the chromophyte algae.The Chromophyte Algae: Problems and Perspectives. Oxford, Clarendon Press. 13-36.Jeffrey, S.W. and G.F. Humphrey (1975). New spectrophotometric equations for determiningchiorophylls a, b,c1and c2 in higher plants, algae and natural phytoplankton. Biochem.Physiol. Pflanzen 167: 191-194.Karabin, G.D., M. Farley and R.B. Hallick (1984). Chloroplast gene for Mr 32000 polypeptideof photosystem II in Euglena gracilis is interrupted by four introns with conserved boundarysequences. Nuci. Acid Res. 12(14): 5801-58 12.Katoh, T., M. Mimuro and S. Takaichi (1989). Light-harvesting particles isolated from a brownalga, Dicryota dichotoma. A supramolecular assembly of fucoxanthin-chiorophyll-proteincomplexes. Biochim. Biophys. Acta 976: 233-240.Kies, L. and B.P. Kremer (1990). Phylum Glaucocystophyta. Handbook ofProtoctista. Boston,Jones and Bartlett. 152-166.Kim, S., E. Pichersky and C.F. Yocum (1994). Topological studies of spinach 22 KDa proteinof photosystem II. Biochim. Biophys. Acta 1188: 339-348.Kim, S., P. Sandusky, N.R. Bowiby, R. Aebersold, B.R.G. Green, S. Vlahakis, C.F. Yocum andE. Pichersky (1992). Characterization of a spinach psbS cDNA encoding the 22 kDa proteinof photosystem II. FEBS 3 14(1): 67-71.Knoetzel, J. and L. Rensing (1990). Characterization of the photosynthetic apparatus from themarine dinoflagellate Gonyaulax polyedra I. Pigment and polypeptide composition of thepigment-protein complexes. J. Plant Physiol. 136: 27 1-279.Knoetzel, J. and D. Simpson (1991). Expression and organization of antenna proteins in thelight- and temperature-sensitive barley mutant chlorina-104.Planta 185: 111-123.Knoetzel, J., I. Svendsen and D.J. Simpson (1992). Identidication of the photosystem I antennapolypeptides in barley: isolation of three pigment-binding antenna complexes. Eur. I.Biochem. 206: 209-215.Kowallik, K.V. (1993). Origin and evolution of plastids from chlorophyll-a+c-containingalgae:suggested ancestral relationships to red and green algal plastids. Origins ofplastids.New York, Chapman and Hall. 223-263.Krauss, N., W. Hinrichs, I. Witt, P. Fromme, W. Pritzkow, Z. Dauter, C. Betzel, K.S. Wilson,H.T. Witt and W. Saenger (1993). Three-dimensional structure of system I of photosynthesisat 6A resolution. Nature 361: 326-331.Kühlbrandt, W. and D.N. Wang (1991). Three-dimensional structure of plant light-harvestingcomplex determined by electron crystallography. Nature 350: 130-134.Kühlbrandt, W., D.N. Wang and Y. Fujiyoshi (1994). Atomic model of plant light-harvestingcomplex by electron crystallography. Nature 367: 614-621.Kyte, J. and R.F. Doolittle (1982). A simple method for displaying the hydrophobic character ofa protein. J. Mol. Biol. 157: 105-132.Laemmli, U.K. (1970). Cleavage of structural proteins during the assembly of the head ofbacteriophabe T4. Nature 227: 680-685.186Lam, B., W. Ortiz and R. Malkin (1984). Chlorophyll a/b proteins of photosystem I. FEBSletters 168(1): 10-14.Lam, E., W. Ortiz, S. Mayfield and R. Malkin (1984b). Isolation and characterization of a light-harvesting chlorophyll a/b protein complex associated with photosystem I. Plant Physiol. 74:650-655.Langridge, 1. (1991). Molecular genetics and comparative evolution. Taunton, England,Research Studies Press LTD.Larkum, A.W.D. and I. Barrett (1983). Light-harvesting processes in algae. Advances inBotanical Research. New York, Academic Press. 1-2 19.Larkum, A.W.D., C. Scaramuzzi, G.C. Cox, R.G. Hiller and A.G. Turner (1994). Light-harvesting chlorophyll c-like pigment in Prochloron. Proc. Nati. Acad. Sci. USA 91: 679-683.LaRoche, J., J. Bennett and P.G. Falkowski (1990). Characterization of a cDNA encoding forthe 28.5-KDa LHCII apoprotein from the unicellular marine chlorophyte, Dunaliellatertiolecta. Gene 95: 165-171.LaRoche, J., D. Henry, K. Wyman, A. Sukenik and P. Falkowski (1994). Cloning andnucleotide sequence of a cDNA encoding a major fucoxanthin-, chlorophyll a/c-containingprotein from the chrysophyte Isochrysis galbana: implications for evolution of the cab genefamily. Plant Mol. Biol. 25: 355-368.Larouche, L., C. Tremblay, C. Simard and G. Bellemare (1991). Characterization of a cDNAencoding a PSII-associated chlorophyll a/b-binding protein (CAB) from Chiamydomonasmoewusii fitting into neither type I or type II. Curr. Genet. 19: 285-288.Larsson, U.K., J.M. Anderson and B. Andersson (1987). Variations in the relative content of theperipheral and inner light-harvesting chlorophyll a/b-protein complex (LHC II)subpopulations during thylakoid light adaption and development. Biochim. Biophys. Acta894: 69-75.Larsson, U.K., C. Sundby and B. Andersson (1987b). Characterization of two differentsubpopulations of spinach light-harvesting chlorophyll a/b-protein complex (LHCII):polypeptide composition, phosphorylation pattern and association with photosystem II.Biochim. Biophys. Acta 894: 59-68.Laudenbach, D.E. and N.A. Straus (1988). Characterization of a cyanobacterial iron stress-induced gene similar to psbC. J. Bacteriol. 170(11): 5018-5026.Lee, R.E. (1989). Phycology. New York, Cambridge University Press.Legocki, R.P. and D.P.S. Verma (1981). Multiple immunoreplica technique: screening forspecific proteins with a series of different antibodies using one polyacrylamide gel. Anal.Biochein. 111: 385-392.Levy, H., I. Gokhman and A. Zamir (1992). Regulation and light-harvesting complex IIassociation of a Dunaliella protein homologous to early light-induced proteins in higherplants. I. Biol. Chem. 267(26): 18831-18836.Lewin, R.A. (1975). Extraordinary pigment composition of a prokaryotic algae. Nature 256:735-737.187Li, W.-H. (1983). Evolution of duplicate genes and pseudogenes. Evolution of genes andproteins. Sunderland, Mass., Sinauer Associates INC. 14-37.Liaud, M.F., C. Valentine, W. Martin, F.Y. Bouget, B. Kloareg and R. Cerff (1994). Theevolutionary origin of red algae as deduced from the nuclear genes encoding cytosolic andchioroplast glyceraldehyde-3-phosphate dehydrogenases from C’hondus crispus. J. Mol. Evol.38: 319-327.Lichtlé, C., J.C. Duval and Y. Lemoin (1987). Comparative biochemical, functional andultrastructural studies of photosystem particles from a Cryptophyceae: Cryptomonasrufescens; isolation of an active phycoerytrin particle. Biochi,n. Biophys. Acta 894: 76-90.Lichtlé, C., H. Jupin and J.C. Duval (1980). Energy transfers from photosystem II tophotosystem Tin Cryptomonas rufescens (Cryptophyceae). Biochim. Biophys. Acta 591: 104-112.Lichtlé, C., R.M.L. McKay and S.P. Gibbs (1992). Immunogold localization of photosystem Iand photosystem II light-harvesting complexes in cryptomonad thylakoids. Biol. Cell 74:187- 194.Lichtlé, C., A. Spilar and S.P. Gibbs (1992b). Immunogold localization of light-harvesting andphotosystem I complexes in the thylakoids of Fucus serratus (Phaeophyceae). Protoplasma166: 99-106.Livne, A., D. Katcoff, Y.Z. Yacobi and A. Sukenik (1992). Pigment-protein complexes ofNannochioropsis sp. (Eustigmatophyceae): an alga lacking chiorophylls b and c. Research inPhotosynthesis. Boston, Kiuwer Academic Publishers. 203-206.Lockhart, P.J., D. Penny, M.D. Hendy and A.W.D. Larkum (1993). Is Prochlorothrixhollandica the best choice as a prokaryotic model for higher plant Chi a/b photosynthesis?Photosyn. Res. 37: 61-68.Lockhart, P.J., M.A. Steel, M.D. Hendy and D. Penny (1994). Recovering evolutionary treesunder a more realistic model of sequence evolution. Mol. Biol. Evol. 11(4): 605-612.Long, Z., S.U. Wang and N. Nelson (1989). Cloning and nucleotide sequence analysis of genescoding for the major chlorophyll-binding protein of the moss Physcomitrella patens and thehalotolerant alga Dunaliella sauna. Gene 76: 299-3 12.Ludwig, M. and S.P. Gibbs (1987). Are the nucleomorphs of cryptomonads and Chlorarachnionthe vestigial nuclei of eukaryotic endosymbionts? Ann. NYAcad. Sci. 501: 198-211.Magruder, W.H. (1984). Specialized appendages on spermatia from the red alga Aglaothamnionneglectum (Ceramiales,Ceramiaceae) specifically bind with trichogynes. J. Phycol. 20: 436-440.Maier, U.G., C.J.B. Hofmann, S. Eschbach, J. wolters and G.L. Igloi (1991). Demonstration ofnucleomorph-encoded eukaryotic small subunit ribosomal RNA in cryptomonads. Mol. Gen.Genet. 230: 155-160.Malkin, R. (1992). Cytochrome bc1 and b6fcomplexes of photosynthetic membranes. Photosyn.Res. 33: 121-136.Margulis, L. and R. Obar (1985). Heliobacterium and the origin of chrysoplasts. Biosystems 17:3 17-325.188Marquardt, J. and A. Ried (1992). Fractionation of thylakoid membranes from Porphyridiumpurpureum using the detergent N-lauryl-3-iminodipropionate. Planta 187: 372-380.Marsho, TN. and B. Kok (1972). Methods in Enzymology. New York, Academic Press. 515-522.Martin, W., C.C. Somerville and S. Loiseaux-de Goër (1992). Molecular phylogenies of plastidorigins and algal evolution. J. Mol. Evol. 35: 385-404.Matsuoka, M. (1990). Classification and characterization of cDNA that encodes the light-harvesting chiorphyll a/b binding protein of photosystem II of rice. Plant Cell Physiol. 3 1(4):5 19-526.Mawson, B.T., P.J. Morrissey, A. Gomez and A. Melis (1994). Thylakoid membranedevelopment and differentiation: Assembly of the chlorophyll a-b light-harvesting complexand evidence for the origin of Mr19, 17.5 and 13.4 kDa proteins. Plant Cell Physiol. 35(3):34 1-35 1.Mayes, S.R., K.M. Cook, S.J. Self, Z. Zhang and I. Barber (1991). Deletion of the geneencoding the photosystem II 33 kDa protein from Synechocystis sp. PCC 6803 does notinactivate water-splitting but increases vulnerability to photoinhibition. Biochim. Biophys.Acta 1060: 1-12.Mayfield, S.P., P. Bennoun and J.-D. Rochaix (1987). Expression of the nuclear encoded OEE1protein is required for oxygen evolution and stability of photosystem II particles inChiamydomonas reinhardtii. EMBO J. 6(2): 313-318.Mayfield, S.P., M. Schirmer-Rahire, G. Frank, H. Zuber and J.D. Rochaix (1989). Analysis ofthe genes of the OEE1 and OEE3 proteins of photosystem II complex from Chiamydomonasreinhardtii. Plant Mol. Biol. 12: 683-693.McFadden, G. and P. Gilson (1995). Something borrowed, something green: lateral transfer ofchloroplasts by secondary endosymbiosis. TREE 10(1): 12-17.McFadden, G.I., P.R. Gilson and S.E. Douglas (1994a). The photosynthetic endosymbiont incryptomonad cells produces both chloroplast and cytoplasmic-type ribosomes. J. Cell. Sci.107: 649-657.McFadden, G.I., P.R. Gilson and D.R.A. Hill (1994c). Goniomonas: rRNA sequences indicatethat this phagotrophic flagellate is a close relative of the host component of cryptomonads.Eur. J. Phycol. 29: 29-32.McFadden, G.I., P.R. Gilson, C.J.B. Hoffman, G.J. Adock and U.-G. Maier (1994b). Evidencethat an amoeba aquired a chioroplast by retaining part of an engulfed eukaryoti alga. Proc.Nati. Acad. Sci. USA 91: 3690-3694.McGrath, J.M., W.B. Terzaghi, P. Sridhar, A.R. Cashmore and B. Pichersky (1992). Sequenceof the fourth and fifth photosystem II type I chlorophyll a/b-binding protein genes ofArabidopsis thaliana and evidence for the presence of a full complement of the extendedCAB gene family. Plant Mol. Biol. 19: 725-733.Medlin, L.K., G.L.A. Barker, M. Baumann, P.K. Hayes and M. Lange (1994). Molecularbiology and systematics. The Haptophyte Algae. Oxford, Clarendon Press. 393-4 11.Melkonian, M. (1990). Phylum Chlorophyta: Class Prasinophyceae. Handbook of Protoctista.Boston, Jones and Bartlett. 600-607.189Meyer, T.E., M.A. Cusanovich and M.D. Kamen (1986). Evidence against use of bacterialamino acid sequence data for construction of all-inclusive phylogenetic trees. Proc. Natl.Acad. Sci. USA 83: 217-220.Michel, H. and J. Deisenhofer (1986). X-ray diffraction studies on a crystalline bacterialphotosynthetic reaction center: a progress report and conclusions on the structure ofphotosystem II reaction centers. Photosynthesis III: photosynthetic membranes and light-harvesting systems. New York, Springer-verlag. 371-381.Moran, R. (1982). Formulae for determination of chlorophyllous pigments extracted with N,-Ndimethylformamide. Plant Physiol. 69: 1376-1381.Morden, C.W., C.F. Delwiche, M. Kuhsel and J.D. Palmer (1992). Gene phylogenies and theendosymbiotic origin of plastids. Biosystems 28: 75-90.Morden, C.W. and S.S. Golden (1989). psbA genes indicate common ancestry ofprochlorophytes and chloroplasts. Nature 337: 382-385.Morishige, D.T. and P.1. Thornber (1992). Identification and analysis of a barley cDNA cloneencoding the 3 1-kilodalton LHC ha (CP29) apoprotein of the light-harvesting antennacomplex of photosystem II. Plant Physiol. 98: 23 8-245.Mörschel, E. and E. Rhiel (1987). Phycobilisomes and thylakoids: the light-harvesting system ofcyanobacteria and red algae. Membranous structures. New York, Academic Press. 209-254.Muchhal, U.S. and S.D. Schwartzbach (1992). Characterization of a Euglena gene encoding apolyprotein precursor to the light-harvesting chlorophyll a/b-binding protein of photosystemII. Plant Mol. Biol. 18: 287-299.Mullet, J.E. (1983). The amino acid sequence of the polypeptide segment which regulatesmembrane adhesion (grana stacking) in chloroplasts. J. Biol. Cheni. 258(16): 9941-9948.Mullet, J.E., J.J. Burke and C.J. Arntzen (1980). Chlorophyll proteins of photosystem I. PlantPhysiol. 65: 814-822.Murata, N. and K. Satoh (1986). Light-emission by plants and bacteria. San Diego, AcademicPress. 137-159.Murray, E.E., J. Lotzer and M. Eberle (1989). Codon usage in plant genes. Nucl. Acids Res.17(2): 477-498.Mustardy, L., F.X. Cunningham and E. Gantt (1990). Localization and quantitation ofchioroplast enzymes and light-harvesting components using immunocytochemical methods.Plant Physiol. 94: 334-340.Nanba, 0. and K. Satoh (1987). Isolation of a PS II reaction center consisting of D-1 and D-2polypeptides and cytochrome b-559. Proc. Natl. Acad. Sci. USA 84: 109-112.Nilsson, F., B. Andersson and C. Jansson (1990). Photosystem II characteristics of a constructedSynechocystis 6803 mutant lacking synthesis of the Dl polypeptide. Plant Mol. Biol. 14:105 1-1054.Norris, B.J. and D.J. Miller (1994). Nucleotide sequence of a cDNA clone encoding theprecursor of the peridinin-chiorophyll a-binding protein from the dinoflagellateSymbiodinium sp. Plant Mol. Biol. 24: 673-677.190Ortiz, W., E. Lam, M. Ghirardi and R. Malkin (1984). Antenna function of a chlorophyll a/bprotein complex of photosystem I. Biochim. Biophys. Acta 766: 505-509.Osafune, T., J.A. Schiff and E. Hase (1991). Stage-dependent localization of LHCP IIapoprotein in the golgi of synchronized cells of Euglena gracilis by immunogold electronmicroscopy. Exper. Cell Res. 193: 320-330.Owens, T.G. (1986). Photosystem II heterogeneity in the marine diatom Phaeodactylumtricornutum. Photochein. Photobiol. 43(5): 535-544.Owens, T.G. (1986b). Light-harvesting in the diatom Phaeodactylum tricornutum II.Distribution of excitation energy between the photosystems. Plant Physiol. 80: 739-746.Owens, T.G. and E.R. Wold (1986). Light-harvesting function in the diatom Phaeodactylumtricornutum: I. isolation and characterization of pigment-protein complexes. Plant Physiol.80: 732-738.Pakrasi, H.B., H.C. Riethman and L.A. Sherman (1985). Organization of pigment proteins in thephotosystem II complex of the cyanobacterium Anacystis nidulans R2. Proc. Natl. Acad. Sd.USA 82: 6903-6907.Pancic, P.G. and H. Strotmann (1993). Structure of the nuclear encoded y subunit of CF01ofthe diatom Odontella sinensis including its presequence. FEBS lett. 320(1): 6 1-66.Passaquet, C., J.C. Thomas, L. Caron, N. Hauswirth, F. Puel and C. Berkaloff (1991). Light-harvesting complexes of brown algae. Biochemical characterization and immunologicalrelationships. FEBS lett. 280(1): 2 1-26.Perasso, R., A. Baroin, L.H. Qu, J.P. Bachellerie and A. Adoutte (1989). Origin af algae. Nature339: 142-144.Peter, G.F. and J.P. Thornber (1991). Biochemical composition and organization of higher plantphotosystem II light-harvesting pigment-proteins. J. Biol. Chem. 266(25): 16745-16754.Peyriere, M., L. Caron and H. Jupin (1984). Pigment complexes and energy transfers in brownalgae. Photosynthetica 18(2): 184-191.Pichersky, E., T.G. Brock, D. Nguyen, N.E. Hoffman, B. Piechulla, S.D. Tanksley and B.R.Green (1989). A new member of the CAB gene family: structure, expression andchromosomal location of Cab-8, the tomato gene encoding the Type III chlorophyll a/b-binding polypeptide of photosystem I. Plant Mol. Biol. 12(257-270):Pichersky, E., N.E. Hoffman, R. Bernatzky, B. Piechulla, S.D. Tanksley and A.R. Cashmore(1987). Molecular characterization and genetic mapping of DNA sequences encoding theType I chlorophyll a/b-binding polypeptide of photosystem I in Lycopersicon esculentum(tomato). Plant Mol. Biol. 9: 205-216.Pichersky, E., D. Soltis and P. Soltis (1990). Defective chlorophyll a/b-binding protein genes inthe genome of a homosporous fern. Proc. Natl. Acad. Sci. USA 87: 195-199.Pichersky, E., R. Subramaniam, M.J. White, J. Reid, R. Aebersold and B.R. Green (1991).Chlorophyll a/b binding (CAB) polypeptides of CP29, the internal chlorophyll a/b complexof PSII: characterization of the tomato gene encoding the 26 KDa (type I) polypeptide, andevidence for a second CP29 polypeptide. Mol. Gen. Genet. 227: 277-284.191Plumley, F.G., T.A. Martinson, D.L. Herrin, M. Ikeuchi and G.W. Schmidt (1993). Structuralrelationships of the photosystem I and photosystem II chlorophyll a/b and a/c light-harvesting apoproteins of plants and algae. Photochem. and Photobiol. 57(1): 143-15 1.Prézelin, B.B. and F.T. Haxo (1976). Purification and characterization of peridinin-chlorophylla-proteins from the marine dinoflagellates Glenodinium sp. and Gonyaulaxpolyedra. Planta(Bert.) 128: 133-141.Pueschel, C.M. (1990). Cell structure. Biology of the red algae. New York, CambridgeUniversity Press. 7-4 1.Pyszniak, A.M. and S.P. Gibbs (1992). Immunocytochemical localization of photosystem I andthe fucoxanthin-chlorophyll alc light-harvesting complex in the diatom Phaeodactylumtricornutum. Protoplasma 166: 208-2 17.Randall, S., A.L. Friedman, D.L. Gustafson, M.S. Rudnick and H. Lyman (1981). Light-harvesting systems of brown algae and diatoms. Isolation and characterization of chlorophyllalc and chlorophyll alfucoxanthin pigment-protein complexes. Biochim. Biophys. Acta 635:304-316.Rau, W. (1988). Functions of carotenoids other than in photosynthesis. Plant Pigments. NewYork, Academic Press. 231-255.Raven, P.H. (1970). A multiple origin for plastids and mitochondria. Science 169: 641-646.Redlinger, T. and E. Gantt (1983). Photosynthetic membranes of Porphyridium cruentum. Ananalysis of chlorophyll-protein complexes and heme-binding proteins. Plant Physiol. 73: 36-40.Reith, M. and R.A. Cattolico (1986). Inverted repeat of Olisthodiscus luteus chloroplast DNAcontains genes for both subunits of ribulose- 1 ,5-bisphosphate carboxylase and the 32,000-dalton QB protein: Phylogenetic implications. Proc. Natl. Acad. Sci. USA 83: 8599-8603.Reith, M. and J. Munholland (1993). A high-resolution gene map of the chloroplast genome ofthe red alga Porhyra purpurea. Plant Cell 5: 465-475.Rhiel, E., W. Lange and E. Mörschel (1993). The unusual light-harvesting complex ofMantoniella squamata: supramolecular composition and assembly. Biochim. Biophys. Acta1143: 163-172.Rhiel, E. and E. Mörschel (1993). The atypical chlorophyll a/b/c light-harvesting complex ofMantoniella squamata: molecular cloning and sequence analysis. Mol. Gen. Genet. 240: 403-413.Rhiel, E., E. Morschel and W. Wehrmeyer (1987). Characterization of and structural analysis ofa chlorophyll a/c light-harvesting complex and of photosystem I particles isolated fromthylakoid membranes of Cryptomonas maculata (Cryptophyceae). Eur. I. Cell Biol. 43: 82-92.Riley, J.P. and T.R.S. Wilson (1967). The pigments of some marine phytoplankton species. J.Mar. Biol. Ass. U.K. 47: 35 1-362.Robinson, M., R. Lilley, S. Little, J.S. Emtage, G. Yarranton, P. Stephens, A. Millican, M.Eaton and G. Humphreys (1984). Codon usage can affect efficiency of translation of genes inEscherichia coli. Nucl. Acids Res. 12(17): 6663-667 1.192Round, F.E. and R.M. Crawford (1990). Phylum Bacillariophyta. Handbook ofProtoctista.Boston, Jones and Bartlett. 574-597.Rutherford, A.W. (1989). Photosystem II, the water-splitting enzyme. Trends Biochem. Sci. 14:227-232.Saffo, M.B. (1987). New light on seaweeds. Bioscience 37(9): 654-664.Sagan, L. (1967). On the origin of mitosing cells. J. Theoret. Biol. 14: 225-274.Saitou, N. and M. Nei (1987). The neighbor-joining method: a new method for reconstructingphylogenetic trees. Mol. Biol. Evol. 4(4): 406-425.Sambrook, J., E.F. Fritsch and T. Maniatis (1989). Molecular cloning, a laboratory manual.Cold Spring Harbour Laboratory Press.Sauer, K. (1986). Photosynthetic light reactions--physical aspects. Photosynthesis III:photosynthetic membranes and light-harvesting systems. New York, Springer-Verlag. 85-97.Schirmer, T., W. Bode, R. Huber, W. Sidler and H. Zuber (1985). J. Mol. Biol. 184: 257-277.Schlegel, M. (1991). Protist evolution and phylogeny as discerned from small subunit ribosomalRNA sequence comparisons. Europ. J. Protistol. 27: 207-219.Schmitt, A., A. Herold, C. Welte, A. Wild and C. Wilhelm (1993). The light-harvesting systemof the unicellular alga Mantoniella squamata (Prasinophyceae): evidence for the lack of aphotosystem I-specific antenna complex. Photochem. Photobiol. 57(1): 132-138.Schuler, G.D., S.F. Altschul and D.J. Lipman (1990). A workbench for multiple alignmentconstruction and analysis. Proteins: Structure, function, and Genetics in press:Schuster, D.M., G.W. Buchman and A. Rashtchian (1992). A simple and efficient method foramplification of cDNA ends using 5’ RACE. Focus 14: 4 1-47.Schuster, G., D.C. Owens, Y. Cohen and I. Ohad (1984). Thylakoid polypeptide compositionand light-independent phosphorylation of the chlorophyll a,b-protein in Prochioron, aprokaryote exhibiting oxygenic photosynthesis. Biochim. Biophys. Acta 767: 596-605.Schwartz, E. and E. Pichersky (1990). Sequence of two tomato nuclear genes encodingchlorophyll a/b-binding proteins of CP24, a PSII antenna component. Plant Mol. Biol. 15:157- 160.Schwartz, E., D. Shen, R. Aebersold, J.M. McGrath, E. Pichersky and B.R. Green (1991).Nucleotide sequence and chromosomal location of Cabi 1 and Cab 12, the genes for thefourth polypeptide of the photosystem I light-harvesting antenna (LHCI). FEBS lett. 280(2):229-234.Sharp, P.M. and W.-H. Li (1986). Codon usage in regulatory genes in Eschericia coli does notreflect selection for ‘rare’ codons. Nuci. Acids Res. 14(19): 7737-7749.Sidler, W., W. Wehrmeyer and H. Zuber (1988). Structural studies on chlorophyll a/c lightharvesting complex from the cryptomonad Cryptomonas maculata: Partial amino acidsequences. Experientia 44: A60.Sleigh, M. (1989). Protozoa and other protists. New York, Cambridge University Press.193Sober, E. (1988). Reconstructing the past: parsimony, evolution and inference. Cambridge, MA,The MIT Press.Somerville, C.C., S. Jouannic and S.L. Goer (1992). Sequence, proposed secondary structure,and phylogenetic analysis of the chioroplast 5S rRNA gene of the brown alga Pylaiellalittoralis (L.) Kjellm.Song, P.S., P. Koka, B.B. Prézelin and F.T. Haxo (1976). Molecular topology of thephotosynthetic light-harvesting pigment complex, peridinin-chlorophyll a-protein, frommarine dinoflagellates. Biochem. 15(20): 4422-4427.Spear-Bernstein, L. and K.R. Miller (1989). Unique location of the phycobiliprotein light-harvesting pigment in the Cryptophyceae. .J. Phycol. 25: 4 12-419.Staehelin, A. (1986). Chioroplasts structure and supramolecular organization of photosyntheticmembranes. Photosynthesis III. Photosynthetic membranes and light harvesting systems.New York, Springer-Verlag. 1-84.Steinkötter, J., D. Bhattacharya, I. Semmelroth, C. Bibeau and M. Melkonian (1994).Prasinophytes from independent lineages within the Chiorophyta: evidence from ribosomalRNA sequence comparisons. J. Phycol. 30: 340-345.Stewart, A., U. Ljungberg, H.-E. Akerlund and B. Andersson (1985). Studies on the polypeptidecomposition of the cyanobacterial oxygen-evolving complex. Biochim. Biophys. Acta 808:353-362.Stewart, C.B. (1993). The powers and pitfalls of parsimony. Nature 361: 603-607.Sukenik, A., K.D. Wyman, J. Bennett and P.G. Falkowski (1987). A novel mechanism forregulating the excitation of photosystem II in a green alga. Nature 327: 704-707.Swift, H. and B. Palenik (1993). Prochlorophyte evolution and the origin of chioroplasts:morphological and molecular evidence. Origins ofplastids. Syinbiogenesis, prochiorophytes,and the origins of chioroplasts. New York, Chapman and Hall. 123-139.Swofford, D.L. (1991). PAUP: Phylogenetic analysis using parsimony. Champaign Ii, IllinoisNatural History Survey.Swofford, D.L. and G.J. Olsen (1990). Phylogeny reconstruction. Molecular Systematics.Sunderland, Sinauer Assoc., Inc. 411-501.Tae, G.-S. and W.A. Cramer (1994). Topography of the heme prosthetic group of thecytochrome b-559 in the photosystem II reaction center. Biochemistry 33: 10060-10068.Taguchi, F., Y. Yamamoto, N. Inagaki and K. Satoh (1993). Recognition signal for the C-terminal processing protease of Dl precursor protein in the photosystem II reaction center.FEBS lett. 326(1-3): 227-23 1.Tanksley, S.D. and E. Pichersky (1988). Organization and evolution of sequences in the plantnuclear genome. Plant Evolutionary Biology. New York, Chapman and Hall. 55-83.Taylor, F.J.R. (1987). An overview of the status of evolutionary cell symbiosis theories. Ann.NYAcad. Sci. 503: 1-16.Taylor, F.J.R. (1990). Phylum dinoflagellata. Handbook of Protoctista. Boston, Jones andBartlett. 419-437.194Taylor, F.J.R. (1992). The taxonomy of harmful marine phytoplanton. Giornale BotanicoItaliano 126(2): 209-219.Taylor, F.J.R. (1993). Current problems with harmful phytoplankton blooms in BritishColumbia waters. Toxic Phytoplankton Blooms in the Sea. Elsevier Science Publishers. 699-703.Taylor, F.J.R. and R. Haigh (1993). The ecology of fish-killing blooms of the chioromonadflagellate Heterosignia in the Straight of Georgia and adjacent waters. Toxic PhytoplanktonBlooms in the Sea. Elsevier Science Publishers. 705-7 10.Thornber, J.P., D.T. Morishige, S. Anandan and G.F. Peter (1991). Chlorophyll-carotenoidproteins of higher plant thylakoids. Chiorophylls. Boca Ranton, CRC Press. 549-586.Trebst, A. (1987). The three-dimensional structure of the herbicide binding niche on the reactioncenter polypeptides of photosystem II. Z. Naturforsch. 42c: 742-750.Turner, S., T. Burger-Wiersma, S.J. Giovannoni, L.R. Mur and N.R. Pace (1989). Therelationship of a prochiorophyte Prochlorothrix hollandica to green chioroplasts. Nature337: 380-382.Valentin, K., R.A. Cattolico and K. Zetsche (1993). Phylogenetic origin of the plastids. Originsofplastids. New York, Chapman and Hall. 193-2 12.van Amerongen, H., B.M. van Bolhuis, S. Betts, R. Mei, R. van Grondelle, C.F. Yocum and J.P.Dekker (1994). Spectroscopic characterization of CP26, a chlorophyll a/b binding protein ofthe higher plant photosystem II complex. Biochim. Biophys. Acta 1188: 227-234.van der Staay, G.W.M. (1992). Functional localization and properties of the chlorophyll bbinding antennae in the prochiorophyte Prochlorothrix hollandica. Universiteit vanAmsterdam.van der Staay, G.W.M. and L.A. Staehelin (1994). Biochemical characterization of proteincomposition and protein phosporylation patterns in stacked and unstacked thylakoidmembranes of the prochiorophyte Prochlorothrix hollandica. J. Biol. Chem. 269(40): 24834-24844.Vermaas, W.F.J., S. Styring, W.P. SchrOder and B. Andersson (1993). Photosynthetic wateroxidation: the protein framework. Photosyn. Res. 38: 249-263.von Heijne, G. (1986). A new method for predicting signal sequence cleavage sites. Nuci. AcidsRes. 14(11): 4683-4690.von Heijne, G. (1990). The signal peptide. J. Membrane Biol. 115: 195-201.Wada, M., M. Shono, 0. Urayama, S. Satoh, H. Yukichi, Y. Ikawa and T. Fujii (1994).Molecular cloning of P-type ATPases on intracellular membranes of the marine algaHeterosigma akashiwo. Plant Mol. Biol. 26: 699-708.Walibraun, M., S. Kim, B.R. Green, B. Piechulla and E. Pichersky (1994). Nucleotide sequenceof a tomato psbS gene. Plant Physiol. 106: 1703-1704.Walne, P.L. and P.A. Kivic (1990). Phylum Euglenida. Handbook of Protoctista. Boston, Jonesand Bartlett. 270-287.195Whatley, J. and F.R. Whatley (1981). Chioroplast evolution. New Phytol. 87: 233-247.White, M.J. and B.R. Green (1987). Antibodies to the photosystem I chlorophyll a + b antennacross-react with polypeptides of CP29 and LHCII. Eur. J. Biochem. 163: 545-55 1.White, M.J. and B.R. Green (1987b). Polypeptides belonging to each of the three majorchlorophyll a + b protein complexes are present in a chlorohyll-b-less barley mutant. Eur. J.Biochem. 165: 531-535.White, M.J. and B.R. Green (1988). Intermittent-light chioroplasts are not developmentallyequivalent to chiorina f2 chloroplasts in barley. Photosyn. Res. 15: 195-203.Wiedemann, I., C. Wilhelm and A. Wild (1983). Isolation of chlorophyll-protein complexes andquantification of electron transport components in Synura petersenii and Tribonerna aequale.Photosyn. Res. 4: 3 17-329.Wilhelm, C., C. Büchel and B. Rousseau (1988). The molecular organization of chlorophyll-protein complexes in the Xanthophycean alga Pleurochioris meiringensis. Biochim. Biophys.Acta 934: 220-226.Winhauer, T., S. Jager, K. Valentine and K. Zetsche (1991). Structural similarities betweenpsbA genes from red and brown algae. Curr. Genet. 20: 177-180.Witt, D. and E. Stackebrandt (1988). Disproving the hypothesis of a common ancestry for theOchromonas danica chrysoplast and Heliobacterium chiorum. Arch. Microbiol. 150: 244-248.Wolfe, G.R., F.X. Cunningham, D.G. Durnford, B.R.G. Green and E. Gantt (1994). Evidencefor a common origin of chloroplasts with light-harvesting complexes of differentpigmentation. Nature 367: 566-568.Wolfe, G.R., F.X. Cunningham and E. Gantt (1992). In the red alga Porphiridium cruentumphotosystem I is associated with a putative LHCI complex. Research in photosynthesis.Proc. IX mt. Cong. Photosynthesis, Nagoya, Japan. Dordrecht, Kiuwer. 315-318.Wolfe, G.R., F.X. Cunningham, B. Grabowski and E. Gantt (1994b). Isolation andcharacterization of photosystem I and II from the red alga Porphyridium cruentum. Biochim.Biophys. Acta 1188: 357-366.Yu, L., J. Zhao, U. Mühlenhoff, D.A. Bryant and J.H. Golbeck (1993). PsaE is required for invivo cyclic electron flow around photosystem I in the cyanobacterium Synechococcus sp.PCC 7002. Plant. Physiol. 103: 171-180.Yurina, N.P., G.V. Karakashev, N.V. Karapetyan and M.S. Odintsova (1991). Composition andbiosynthesis of thylakoid membrane polypeptides in the red alga Cyanidium caldarium:Comparison with the thylakoid polypeptide composition of higher plants and cyanobacteria.Photosyn. Res. 30: 15-23.Zhou, Y.H. and M.A. Ragan (1993). cDNA cloning and characterization of the nuclear geneencoding chloroplast glyceraldehyde-3-phosphate dehydrogenase from the marine red algaGracilaria verrucosa. Curr. Genet. 23: 483-489.Zuber, H. (1986). Primary structure and function of the light-harvesting polypeptides fromcyanobacteria, red algae, and purple photosynthetic bacteria. Photosynthesis III..Photosynthetic membranes and light harvesting systems. New York, Springer-Verlag. 238-251.196"@en ; edm:hasType "Thesis/Dissertation"@en ; vivo:dateIssued "1995-11"@en ; edm:isShownAt "10.14288/1.0088173"@en ; dcterms:language "eng"@en ; ns0:degreeDiscipline "Botany"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "University of British Columbia"@en ; dcterms:rights "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en ; ns0:scholarLevel "Graduate"@en ; dcterms:title "An analysis of the fucoxanthin-chlorophyll proteins and the genes encoding them in the unicellular marine raphidophyte, Heterosigma carterae: characterization and evolution"@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/7251"@en .