What is the purpose of the stacks of thylakoid membrane within the chloroplast?

(a) Organization of the thylakoid membranes in higher plant chloroplasts

The thylakoid membranes of higher plant chloroplasts have a unique organization into stacks of flattened disk-shaped sacs or lamellae in which the membranes of adjacent lamellae are closely appressed (granal lamellae) and lamellae in which the membranes are not appressed and thus their outer surfaces are in direct contact with the stroma (stromal lamellae) (Fig. 2.16). The two membrane systems and the lumenal spaces they enclose are continuous.

The thylakoid membrane lipids comprise about 80% glycolipids, with neutral hydrophobic head groups, consisting of monogalactosyldiacyglycerol and digalactosyldiacyglycerol in a 2:1 ratio. The remainder of the lipid is mainly phospholipid (10%) and sulfolipid (5%), both of which carry a negative charge at neutral pH. Linolenic acid (C18:3) is the predominant fatty acid (∼70%) and its high degree of unsaturation ensures that the membrane is very fluid. There is a partially asymmetric distribution of the various lipids between appressed and non-appressed membranes (Gounaris et al., 1983) but the functional significance of this is unknown.

The complex thylakoid membrane organization is reflected in a marked lateral heterogeneity in the distribution of the membrane–protein complexes (Andersson & Anderson, 1980; Plate 5a). The ATP synthase and PSI with its associated LHCI are located only in the non-appressed membrane regions, whereas the majority of the PSII and LHCII are located in the appressed membranes. This pool of PSII is referred to as PSIIα. A minor pool of PSII centers with a smaller population of LHCII per PSII (PSIIβ) is located in the stromal lamellae (Melis & Anderson, 1983). The cytochrome b6f complex is distributed throughout the thylakoid membrane system (Anderson, 1982). The stoichiometry of complexes in the membrane is discussed in section 2.2.8.

Membrane appression is thought to result from Van der Waal's interactions between LHCII complexes in opposing membranes and between LHCII and PSII in the same membrane (reviewed in Allen, 1992a,b). In shade-grown plants, the amount of LHCII per PSII increases, leading to an increase in the extent of the appressed granal membrane system (see also section 2.2.8). Although the membrane lipid itself is fluid, the high concentration of integral membrane-protein complexes, particularly in the appressed membranes (where they occupy ∼50% of the membrane area; Plate 5b), severely restricts diffusion of these complexes within the plane of the membrane. The diffusion coefficient of phosphorylated LHCII (see section 2.2.5c) in the appressed membranes has been estimated to be in the range (2–4) × 10−12cm2 s−1 (Drepper et al., 1993) which is two orders of magnitude lower than that reported for membrane proteins of similar size in artificial lipid bilayers.

1.21.2 Energy Absorption, Trapping, Conversion, and Storage

The thylakoid membrane systems of cyanobacteria and of plant and algal chloroplasts contain the specialized photosystems, photosystem I (PSI), and photosystem II (PSII), in which the reaction centers photochemically transform light energy into useable chemical energy in the form of ATP and NADPH through photosynthetic electron transport (PET). The major light-harvesting complexes (LHCs) associated with PSI and PSII of plants and green algae are supramolecular, integral thylakoid membrane, pigment–protein complexes that absorb sunlight and then transfer the absorbed light energy to the reaction centers.10,11 Within PSII reaction centers, photochemistry traps this energy through the formation of a charge-separated state such that P680 is photo-oxidized (P680  +  photon  →  P680+  +  e−). Subsequently, P680+, the strongest oxidizing agent in nature, oxidizes water and generates O2 as a byproduct, which is required by all other aerobic organisms. Consequently, the advent of photosynthesis and photoautotrophy had a global effect by polluting the atmosphere with O2 over a period of billions of years and converting the abiotic environment on this Earth from reducing to oxidizing conditions.

The electrons generated by PSII photochemistry are used to reduce plastoquinone (PQ), a small, lipophilic electron carrier, to plastoquinol (PQH2) via electron transfer from QA to QB that are bound to the D1 and D2 reaction center polypeptides, respectively. The oxidation of PQH2 is coupled to the reduction of the lumenal copper protein, plastocyanin, via another major integral thylakoid membrane protein complex, the Cyt b6/f complex, which is a PQH2–plastocyanin oxidoreductase. The oxidation of PQH2 is considered to be the rate-limiting step of linear electron transport and occurs on a timescale of milliseconds.12 Reduced plastocyanin is oxidized by P700+ generated by the photochemical conversion of P700 to P700+ + e−. The electrons generated photochemically by PSI are used for the reduction of the stromal localized, FeS protein, ferredoxin, which, in turn, is oxidized by the enzyme ferredoxin-NADP oxidoreductase to form NADPH. The release of free energy as electrons flow from PSII reaction centers to P700+ generates a PMF (Table 2) across the thylakoid membrane used by the fourth major protein complex, the thylakoid ATP synthase, to generate ATP through the process of photophosphorylation. Crystallization of PSI13 and PSII reaction centers14 coupled with the crystal structure data for the Cyt b6/f complex15 and partial crystal structure data for the ATP synthase complex10 has provided major advances into the molecular and biophysical mechanisms that underlie energy transfer, energy trapping, and energy conversion in the photosynthetic apparatus of photosynthetic bacteria, cyanobacteria, and terrestrial plants.

The generation of NADPH and ATP is coupled to energy storage initiated through the fixation and reduction of CO2 by the Calvin cycle (Table 2). However, the initial CO2 fixation reaction catalyzed by the chloroplast enzyme, Rubisco (CO2  +  RuBP  →  2PGA), is exergonic and thus does not require any input of energy. NADPH and ATP are required for the reduction of CO2 to create an essential stromal pool of triose-P. This critical metabolic intermediate has three important fates. First, it is converted via the Calvin cycle to regenerate the metabolite RuBP. This consumes additional ATP and is an absolute prerequisite for the continuous fixation of CO2 in the light. Thus, CO2 assimilation is light dependent because of the requirement for the constant regeneration of RuBP. Second, triose-P is exported from the chloroplast to the cytosol via the triose-P translocator for sucrose biosynthesis via the cytosolic hexose-P pool and the action of the rate-controlling cytosolic enzyme, sucrose phosphate synthase (SPS). Third, triose-P can be converted to starch via the stromal hexose-P pool and the action of the rate-controlling chloroplast enzyme, ADP-glucose pyrophosphorylase (Table 2).

Reduced ferredoxin represents an important branch point in the flow of electrons and thus redox potential energy in photoautotrophic metabolism. Not only does its redox potential govern the biosynthesis of NADPH, but it is also critical for the reduction of NO−3 to ammonium required for amino acid biosynthesis via glutamine synthetase/glutamate synthase (GS/GOGAT). In addition, reduced ferredoxin is critical for the regulation of several light-activated enzymes of the Calvin cycle including chloroplastic GAPDH, FBPase, SBPase and R5P kinase. In addition to C, N, and S reduction, photosynthetically generated electrons and metabolic carbon intermediates feed into the biosynthetic pathways for cellulose, lipids and fatty acids, nucleic acids, as well as the complex pathways for the myriad secondary metabolites present in photoautotrophs. Thus, it is clear that photosynthesis is the major process that generates and regulates cellular energy flow in photoautotrophs.

II Introduction

Chloroplast thylakoid membranes form the internal membrane system in chloroplasts that function as a quantum‐, electron‐, and proton‐transfer machine, essential for sustaining life on earth. The energy‐generating capabilities of these membranes stem from the action of four different supramolecular protein complexes, each assembled from 14 to 26 different protein subunits originating from both plastid and nuclear genomes. The shear abundance of thylakoid membranes and photosynthetic complexes that operate in these membranes requires the thylakoid to be one of the major protein export sites of the photosynthetic cell. At least four different thylakoid export pathways originate in the stroma, which is also the site of transcription and translation for plastid‐encoded proteins. In that context, the stroma is much like the bacterial cytosol or mitochondrial matrix. Yet the function of thylakoid export pathways is best understood for nuclear‐encoded thylakoid proteins, which are expressed in the cytoplasm as full‐length precursors and imported into the stroma where they gain access to conserved thylakoid export systems.

Based on the evolutionary origin of organelles, it is not surprising that protein export from the stroma to the thylakoid resembles export to the endoplasmic reticulum (ER), to the mitochondrial inner membrane, and to the bacterial cytoplasmic membrane [1]. For example, proteins that must cross the thylakoid to reach their functional location on the luminal side of the membrane are transported by homologous twin‐arginine‐targeting (TAT) or secretion (Sec) transport systems, similar to those that translocate bacterial proteins across the cytoplasmic membrane. Integral thylakoid proteins present a uniquely different set of localization issues that stem, in part, from their propensity to form aggregates in solution. For many proteins that must integrate into the ER membrane, the mitochondrial inner membrane, or the bacterial cytoplasmic membrane, aggregation is avoided by mechanisms that promote membrane insertion as the polypeptide is being translated. Similar cotranslational insertion mechanisms appear to support integration of chloroplast‐synthesized thylakoid proteins, the majority of which are integral membrane proteins. However, the nuclear‐encoded light‐harvesting chlorophyll‐binding proteins (LHCPs) are often the most abundant integral membrane proteins in chloroplast thylakoids and must enter conserved protein export pathways following their import into the chloroplast.

The question of how LHCPs find their way posttranslationally from the chloroplast envelope to the thylakoid and subsequently insert into the membrane has been the subject of research for more than 20 years. Hence, this question is a central theme of this chapter owing to the finding that a chloroplast signal recognition particle (cpSRP) and SRP receptor homologue (cpFtsY) function in LHCP routing to the thylakoid where the protein Albino3 (Alb3) is required for stable integration of LHCPs. The SRP transport pathway in chloroplasts also serves to export chloroplast‐encoded proteins to the thylakoid by an overlapping cotranslational mechanism. However, details of the localization mechanism are sparse for this set of proteins owing to difficulties associated with reconstituting their localization into isolated thylakoids. Nevertheless, based on a combination of genetic, biochemical, and structural studies, as well as studies of protein export in mitochondria, bacteria, and through the ER, a more detailed model of cpSRP‐mediated protein export has begun to emerge. The model reflects both conserved and unique functions of an archaic targeting/integration mechanism that evolved to meet protein‐sorting requirements associated with the endosymbiotic event that gave rise to chloroplasts from a cyanobacterial progenitor.

II RECONSTITUTION OF O2 EVOLUTION IN CHOLATE-EXTRACTED THYLAKOIDS

Broken thylakoid membranes prepared from class 2 spinach chloroplasts were used as the starting material for our reconstitution experiments. Details of the reconstitution procedure appeared elsewhere [3,8]. Briefly, specific proteins were extracted from UP-10 by incubating it in buffer solution A (0.2 M sucrose, 3 mM MgCl2, 1 M NaCl and 20 mM tricine pH 8.4) containig, additionally, 50 mM sodium cholate with stirring for 15 min at 4°C followed by centrifugation. This process was repeated once or twice, if necessary. The reconstitution was done 1 hr after incubation of the extracted UP-10 (the pellet after cholate extraction) with total cholate extracts or other test protein samples, or without any additive, in buffer solution B (0.2 M sucrose, 3 mM MgCl2 and 20 mM MOPS pH 7.0) containing, additionally, 20 mM sodium cholate and 25% glycerol by volume. After incubation, the mixture was diluted 50 times by buffer solution B and centrifuged. The pellet, collected as reconstituted UP-10 (RUP-10), was resuspended in buffer solution B without MgCl2 and submitted to examination of photoinduced electron transfer and oxygen evolution.

Electron micrographs of negatively stained preparations of UP-10, the cholate-extracted UP-10 and RUP-10 formed in the presence of glycerol are shown in Fig. 1. Vesicular membrane structures which were observed in UP-10 disappeared during cholate extraction and were reformed in RUP-10. The size of the vesicles of UP-10 and RUP-10 were 0.1–1.0 μm.

Photoinduced electron transfer from diphenylcarbazide (DPC) to 2,6-dichlorophenolindophenol (DCIP) and from water to phenyl-p-benzoquinone (oxygen evolution) by RUP-10 prepared under different conditions were measured and compared with those of UP-10. All RUP-10 preparations, including the samples prepared without any additive (control RUP-10), exhibited reasonable rates of PS II activity measured as electron transfer from DPC to DCIP, suggesting that components essential for electron transfer are only slightly solubilized from UP-10 by the cholate-treatment. In contrast, recovery of oxygen evolution in RUP-10 required some particular components in the cholate-extracts. Fig. 2 shows an example of the reconstitution of oxygen evolution in RUP-10. Control RUP-10, prepared from twice extracted UP-10 without any additive, had no oxygen evolving activity, while RUP-10 prepared from twice extracted UP-10 and the total cholate-extracts (total protein/chlorophyll = 3 in wt. in the incubation medium) usually presented a 70 to 80% recovery of the oxygen evolution activity of the original UP-10. The recovery capacity for the total cholate-extract as a function of the amount of protein added is shown in Fig. 2-b. These results clearly revealed that some particular components closely related to the water splitting process were solubilized by the cholate-treatment from thylakoids and successfully reconstituted in RUP-10. It should be mentioned here that the presence of glycerol in the incubation medium during the reconstitution process was one of the essential conditions for successfully restoring oxygen evolution activity in RUP-10.

4.2.1.1.2 Photosystems

On the thylakoid membranes, certain pigments and associated proteins are packed together to form units called photosystems. There are two types of photosystems, designated Photosystem I (or PSI) and Photosystem II (or PSII). Each photosystem has a critical pigment for photosynthesis (called chlorophyll a) and accessory pigments, e.g., chlorophyll b and carotenoids. The critical chlorophyll a pigment is designated P700 for PSI and P680 for PSII, indicating the light wavelength that is absorbed most efficiently by these pigments. The chlorophyll pigments absorb mainly red and blue light wavelengths, and reflect and transmit green wavelengths, so leaves are green. The accessory pigments absorb light of slightly different wavelengths than chlorophyll a, and channel this energy to chlorophyll a. In higher plants, both photosystems must cooperate in carrying out photosynthesis.

Unicellular Organisms

Excellent preservation of the thylakoid membranes from Zygnema cells in TEM studies have been obtained in our studies (Fig. 1D) by following a modification of fixation and staining protocols of McLean and Pessoney (1970). We removed the growth medium/buffer from Zygnema filaments by filtering out the wash buffer as shown in Fig. 2A. The thylakoid membranes of Zygnema at the end of the logarithmic phase of growth display lamellar-cubic transition (Fig. 1D).

Conversely, amoeba Chaos cells that were simply detached from the culture dish using a fine-tip plastic pipette and transferred to 2.5% GA for fixation shows the downside of fixing specimen with too much additional liquid. The introduction of some amoeba medium when fixing the cells caused the mitochondrial outer membrane to swell and subsequent disintegration of inner membranes (Fig. 3A).

Figure 3B demonstrates comparatively better preserved mitochondrial inner membranes. However, in this example certain sections of the mitochondrial inner membranes appear to harbor cubic morphology with continuous membranes while others appear to be discontinuous (refer to section IIIC for further discussion). This observation, however, is not unusual since cubic membrane regions may develop only locally in mitochondria and do not necessarily occupy the entire organelle (Landh, 1995). This phenomenon was also described and confirmed by Deng and Mieczkowski (1998) through TEM serial sectioning and EM tomography techniques.

On another note, starved amoeba Chaos mitochondria shown in Fig. 1A–C displayed the different facets of cubic membranes when the mitochondria were sliced at different directions. This illustrates an important notion that the membrane arrangements that do exist as a complex 3D configurations may yield multiple 2D “faces” in TEM projections, dependent on the section width and direction of (random!) cutting.

8.1.2.10 Light-driven Electron Transfer in Eukaryotes

A key role of the thylakoid membranes in chloroplasts is the generation of a proton electrochemical gradient. This is achieved in two ways. The first is through the sequential action of Photosystem II, which splits water and releases oxygen, the cytochrome b6f complex, and Photosystem I. Although neither of the photosystems directly moves protons across the membrane they are important examples of proteins which principally contribute to the generation of the proton electrochemical gradient as a result of moving electrons across the membrane. The second mode of gradient formation involves only the cytochrome b6f complex and photosystem I in the process known as cyclic electron transfer. These systems are described in more detail in Chapter 8.5 and Chapter 8.7.

A. Overview

A diagrammatic representation of the thylakoid membrane components involved in the light reactions of photosynthesis is shown in Figure 6.1. Two photosystems, each with a light-harvesting antenna complex, act in series through an intermediary complex containing cytochromes b6 and f. The net result is that the redox state of electrons passing through the chain is elevated from a level that brings about the oxidation of water to a level capable of NADP reduction. The two photosystems are physically distinct and can be recognized in freeze-fracture photographs as discrete particles associated with the thylakoid membranes. ATP is generated via the membrane-associated CF0CF1 ATP synthase complex. The enzymes that fix carbon from CO2 into inorganic compounds (the “dark reactions” or Calvin cycle) are not membrane bound, and they appear in the soluble supernatant phase when cells or chloroplasts are fractionated.

Each of these complexes has been thoroughly investigated in Chlamydomonas, beginning in the 1960s with isolation of non-photosynthetic mutants and with studies of chloroplast development in the y1 mutant, which is unable to form chlorophyll in the dark. The literature is huge, and has been reviewed at intervals over the years. For the current state of research, the reader should see Volume 2. Work through the 1990s was covered in the book edited by Rochaix et al. (1998) and by Grossman (2000b). The first edition of The Chlamydomonas Sourcebook (Harris, 1989) provided a history of the early studies, including descriptions of many mutants. For the historically minded reader in search of a broader perspective, volume 76 of Photosynthesis Research (2003) is also highly recommended. The discussion to follow here therefore provides only a brief introduction to the components of photosynthesis, with a focus on features that are unique to Chlamydomonas, or particularly well studied in this alga, and the usual emphasis on mutants.

4.3 TPP

TPP, found in the thylakoid membrane, belongs to the type I signal peptidase family. The signal peptidase is called SPase I. Its orthologues are found in bacteria archaea, fungi, plants, and animals (Paetzel et al., 2002). SPase I is a serine-type protease using a Ser–Lys or Ser–His catalytic dyad. It is typically anchored to the membrane with one or two transmembrane domains. At the transport of proteins across the membranes, SPase I functions to cleave the signal peptide from the immature protein (Tuteja, 2005).

In chloroplasts, TPP uses the Ser–Lys catalytic dyad mechanism that functions in the prokaryotes and mitochondria, but not the eukaryotic endoplasmic reticulum (Chaal et al., 1998). The hydropathy profile indicated that TPP has a single transmembrane region. Its catalytic center is located at the thylakoid lumen face (Chaal et al., 1998). The substrate specificity of TPP is similar to that of other SPase I (Shackleton and Robinson, 1991). It conforms to the well-known (-3, -1) rule, which states that the residues at the -3 and -1 positions are related to the cleavage site (von Heijne, 1983). These motifs are typically located in the N-terminal region of the signal peptide of lumenal proteins, which are imported via the Tat pathway into the thylakoid lumen.

3.1 Generating fusions to fluorescent proteins

The natural autofluorescence from the photosynthetic thylakoid membranes in cyanobacteria overlaps spectrally with emissions from fluorophores that emit in the red/orange spectrum, including mCherry, precluding their use. In our experience, GFP, ZsGreen, YFP, and ECFP, as well as other variants in the green/yellow color spectrum, are expressed well and easily differentiated from cellular autofluorescence with the appropriate filters. Fluorescent tags are often appended to either the N- or C-terminus of the protein of interest (POI). In some cases, the fluorescent tag may also be inserted into an internal loop, such that each domain is allowed to fold properly and not affect the function of either GFP or the POI. If possible, structural information can be used to make an informed decision about the placement of a fluorescent tag, although even well-guided guesses must be vetted. We have used both N- and C-terminal fusions to KaiC to observe details of subcellular localization. While N-terminal fusions to YFP fully complemented a kaiC null stain, C-terminal fusions to either YFP or ECFP display a long-period phenotype, extended by ~ 5 h (Cohen et al., 2014).

A flexible linker is often introduced between the POI and the fluorescent protein to avoid steric hindrance and allow each domain to fold properly. Glycine, having the smallest side chain, allows for the greatest degree of flexibility (Campbell & Davidson, 2010). We have been successful in using short linkers (2–3 amino acids) composed of either Glycine or Alanine to generate fusions to KaiA. For KaiC, we used a longer linker (17 amino acids) composed of Glycine interspersed with Serine (Cohen et al., 2014) that additionally functions to improve solubility (Campbell & Davidson, 2010). Linkers should be optimized for every application, and in our experience, it is best to initially test multiple fusion proteins to compare N- and C-terminal fusions as well as vary linker lengths and test multiple fluorophores before settling on one fusion protein with which to proceed.

While many exciting discoveries have been made using fluorescent tags, be wary of potential localization artifacts. Examples include clustering artifacts that resulted in ClpX foci that were later found to be not biologically relevant, as well as helical cables observed for MreB that were later found to be an artifact of the high expression of the YFP tag (Landgraf, Okumus, Chien, Baker, & Paulsson, 2012; Margolin, 2012; Swulius & Jensen, 2012).