Regulatory role of the C-terminus of the epsilon subunit from the chloroplast ATP synthase

The ATP synthases from chloroplasts and Escherichia coli are regulated by several factors, one of which is the epsilon subunit. This small subunit is also required for ATP synthesis. Thylakoid membranes reconstituted with CF1 lacking the epsilon subunit (CF1-epsilon) exhibit no ATP synthesis and very high ATP hydrolysis. Either native or recombinant epsilon restores ATP synthesis and inhibits ATP hydrolysis. Previously, we showed that truncated epsilon, lacking the last 45 C-terminal amino acids, restored ATP synthesis to membranes reconstituted with CF1-epsilon but was not an efficient inhibitor of ATP hydrolysis. In this paper, we show that this truncated epsilon is unable to inhibit ATP hydrolysis when Mg(2+) is the divalent cation present, both for the enzyme in solution and on the thylakoid membrane. In addition, the rate of reduction of the disulfide bond of the gamma subunit by dithiothreitol is not decreased by truncated epsilon, although full-length epsilon greatly impedes reduction. Thylakoid membranes can synthesize ATP at the expense of proton gradients generated by pH transitions in the dark. Our reconstituted membranes are able to produce a limited amount of ATP under these "acid-bath" conditions, with approximately equal amounts produced by the membranes containing wild-type epsilon and those containing truncated epsilon. However, the membranes containing truncated epsilon exhibit much higher background ATP hydrolysis under the same acid-bath conditions, leading to the conclusion that, without the C-terminus of epsilon, the CF1CFo is unable to check unwanted ATP hydrolysis.     

Functional consequences of N- or C-terminal deletions of the delta subunit of chloroplast ATP synthase

Mutagenesis was used to generate seven truncation mutants of the spinach (Spinacia oleracea) chloroplast ATP synthase delta subunit lacking 5, 11, 17, or 35 amino acid residues from the N-terminus or 3, 9, or 15 residues from the C-terminus. Interactions between these mutants and all other subunits of the chloroplast ATPase were investigated by a yeast two-hybrid system. The results indicate that the N-terminal deletions mainly affected interactions between the delta subunit and the other part of CF(1), but did not significantly affect interactions with the CF(0) sector. In contrast, C-terminal truncations of the delta subunit mainly affected its interaction with the CF(0) sector and caused little impairment in interactions with the other part of CF(1). The conformation of the delta subunit C-terminal domain seems to be more sensitive to the truncations, as shown by minimal expression driven by C-terminal deleted (nine residues) mutants. Further studies showed C-terminal truncations of the delta subunit greatly impaired its ability to restore cyclic photophosphorylation in NaBr vesicles, whereas N-terminal truncations had little effect on the ability of delta to plug the CF(0) channel. None of the mutants impaired ATP hydrolysis by CF(1).     

Significance of the epsilon subunit in the thiol modulation of chloroplast ATP synthase

To understand the regulatory function of the gamma and epsilon subunits of chloroplast ATP synthase in the membrane integrated complex, we constructed a chimeric FoF1 complex of thermophilic bacteria. When a part of the chloroplast F1 gamma subunit was introduced into the bacterial FoF1 complex, the inverted membrane vesicles with this chimeric FoF1 did not exhibit the redox sensitive ATP hydrolysis activity, which is a common property of the chloroplast ATP synthase. However, when the whole part or the C-terminal alpha-helices region of the epsilon subunit was substituted with the corresponding region from CF1-epsilon together with the mutation of gamma, the redox regulation property emerged. In contrast, ATP synthesis activity did not become redox sensitive even if both the regulatory region of CF1-gamma and the entire epsilon subunit from CF1 were introduced. These results provide important features for the regulation of FoF1 by these subunits: (1) the interaction between gamma and epsilon is important for the redox regulation of FoF1 complex by the gamma subunit, and (2) a certain structural matching between these regulatory subunits and the catalytic core of the enzyme must be required to confer the complete redox regulation mechanism to the bacterial FoF1. In addition, a structural requirement for the redox regulation of ATP hydrolysis activity might be different from that for the ATP synthesis activity.     

Molecular dissection of the epsilon subunit of the chloroplast ATP synthase of spinach.

The gene encoding the epsilon subunit (atpE) of the chloroplast ATP synthase of Spinacia oleracea has been overexpressed in Escherichia coli. The recombinant protein can be solubilized in 8 M urea and directly diluted into buffer containing ethanol and glycerol to obtain epsilon that is as biologically active as epsilon purified from chloroplast-coupling factor 1 (CF1). Recombinant epsilon folded in this manner inhibits the ATPase activity of soluble and membrane-bound CF1 deficient in epsilon and restores proton impermeability to thylakoid membranes reconstituted with CF1 deficient in epsilon. Site-directed mutagenesis was used to generate truncations and single amino acid substitutions in the primary structure of epsilon. In the five mutants tested, alterations that weaken ATPase inhibition by recombinant epsilon affect its ability to restore proton impermeability to a similar extent, with one exception. Substitution of histidine-37 with arginine appears to uncouple ATPase inhibition and the restoration of proton impermeability. As in the case of E. coli, it appears that N-terminal truncations of the epsilon subunit have more profound effects than C-terminal deletions on the function of epsilon. Recombinant epsilon with six amino acids deleted from the C terminus, which is the only region of significant mismatch between the epsilon of spinach and the epsilon of Pisum sativum, inhibits ATPase activity with a reduced potency similar to that of purified pea epsilon. Four of the six amino acids are serine or threonine. These hydroxylated amino acids may be important in epsilon-CF1 interactions.     

Energy-dependent changes in the conformation of the epsilon subunit of the chloroplast ATP synthase

Rabbit antiserum raised against the isolated native epsilon subunit of the chloroplast coupling factor 1 activated the ATPase activity of coupling factor 1 in solution by removing the epsilon subunit. Incubation of thylakoid membranes with the antiserum in the dark had no effect on photophosphorylation or on the dithiothreitol-induced Mg2+-ATPase activity. Incubation with the antiserum during illumination, however, strongly inhibited both activities and caused the membranes to become leaky to protons. The results indicate that the formation of a proton gradient across the thylakoid membrane induces a change in conformation of the epsilon subunit of the ATP synthase such that it becomes susceptible to attack and removal by the antibodies. This change may be a part of the mechanism that results in energy-dependent activation of the ATP synthase.     

The C-terminal domain of the epsilon subunit of the chloroplast ATP synthase is not required for ATP synthesis

The epsilon subunit of the ATP synthases from chloroplasts and Escherichia coli regulates the activity of the enzyme and is required for ATP synthesis. The epsilon subunit is not required for the binding of the catalytic portion of the chloroplast ATP synthase (CF1) to the membrane-embedded part (CFo). Thylakoid membranes reconstituted with CF1 lacking its epsilon subunit (CF1-epsilon) have high ATPase activity and no ATP synthesis activity, at least in part because the membranes are very leaky to protons. Either native or recombinant epsilon subunit inhibits ATPase activity and restores low proton permeability and ATP synthesis. In this paper we show that recombinant epsilon subunit from which 45 amino acids were deleted from the C-terminus is as active as full-length epsilon subunit in restoring ATP synthesis to membranes containing CF1-epsilon. However, the truncated form of the epsilon subunit was significantly less effective as an inhibitor of the ATPase activity of CF1-epsilon, both in solution and bound to thylakoid membranes. Thus, the C-terminus of the epsilon subunit is more involved in regulation of activity, by inhibiting ATP hydrolysis, than in ATP synthesis.     

The carboxyl terminus of the epsilon subunit of the chloroplast ATP synthase is exposed during illumination

The epsilon subunit of the chloroplast ATP synthase is an inhibitor of activity of the enzyme. Recombinant forms of the epsilon subunit from spinach chloroplasts lacking the last 10, 32, or 45 amino acids were immobilized onto activated Sepharose. A polyclonal antiserum raised against the epsilon subunit was passed over these immobilized protein columns, and the purified antibodies which were not bound recognized the portions of the epsilon subunit missing from the recombinant form present on the column. The full polyclonal antiserum can strip the epsilon subunit from the ATP synthase in illuminated thylakoid membranes [Richter, M. L., and McCarty, R. E. (1987) J. Biol. Chem. 262, 15037-15040]. Exposure of illuminated thylakoid membranes to antibodies recognizing the last 32 amino acids of the epsilon subunit collapses the proton gradient and hinders ATP synthesis with similar efficiency as the full polyclonal preparation. These results indicate that antibodies against the last 32 amino acids of the epsilon subunit are capable of stripping the subunit from the ATP synthase in illuminated membranes. Neither of these effects was seen when the membranes were exposed to the antibodies in the dark. This is direct evidence that the chloroplast ATP synthase undergoes a conformational shift during its activation by the electrochemical proton gradient which specifically alters the conformation of the carboxyl-terminal domain of the epsilon subunit from protected to solvent-exposed. The relation between this shift and activation of the enzyme by the electrochemical proton gradient is discussed.     

Substitutions of the Conserved Gly47 Affect the CF1 Inhibitor and Proton Gate Functions of the Chloroplast ATP Synthase ε Subunit

The conserved residue Gly47 of the chloroplast ATP synthase ε subunit was substituted with Leu, Arg, Ala and Glu by site-directed mutagenesis. This process generated the mutants εG47L, εG47R, εG47A and εG47E, respectively. All the ε variants showed lower inhibitory effects on the soluble CF1(-ε) Ca^2 -ATPase compared with wild-type ε. In reduced conditions, εG47E and εG47R had a lower inhibitory effect on the oxidized CF1(-ε) Ca^2 -ATPase compared with wild-type ε. In contrast, εG47L and εG47Aincreased the Ca^2 -ATPase activity of soluble oxidized CF1(-ε). The replacement of Gly47 significantly impaired the interaction between the subunit ε and γ in an in vitro binding assay. Further study showed that all ε variants were more effective in blocking proton leakage from the thylakoid membranes. This enhanced ATP synthesis of the chloroplast and restored ATP synthesis activity of the reconstituted membranes to a level that was more efficient than that achieved by wild-type ε. These results indicate that the conserved Gly47 residue of the ε subunit is very important for maintaining the structure and function of the ε subunitand may affect the interaction between the ε subunit, β subunit of CF1 and subunit Ⅲ of CF0, therebyregulating the ATP hydrolysis and synthesis, as well as the proton translocation role of the subunit Ⅲ of CF0.     

Important subunit interactions in the chloroplast ATP synthase

General structural features of the chloroplast ATP synthase are summarized highlighting differences between the chloroplast enzyme and other ATP synthases. Much of the review is focused on the important interactions between the epsilon and gamma subunits of the chloroplast coupling factor 1 (CF(1)) which are involved in regulating the ATP hydrolytic activity of the enzyme and also in transferring energy from the membrane segment, chloroplast coupling factor 0 (CF(0)), to the catalytic sites on CF(1). A simple model is presented which summarizes properties of three known states of activation of the membrane-bound form of CF(1). The three states can be explained in terms of three different bound conformational states of the epsilon subunit. One of the three states, the fully active state, is only found in the membrane-bound form of CF(1). The lack of this state in the isolated form of CF(1), together with the confirmed presence of permanent asymmetry among the alpha, beta and gamma subunits of isolated CF(1), indicate that ATP hydrolysis by isolated CF(1) may involve only two of the three potential catalytic sites on the enzyme. Thus isolated CF(1) may be different from other F(1) enzymes in that it only operates on 'two cylinders' whereby the gamma subunit does not rotate through a full 360 degrees during the catalytic cycle. On the membrane in the presence of a light-induced proton gradient the enzyme assumes a conformation which may involve all three catalytic sites and a full 360 degrees rotation of gamma during catalysis.     

Resonance energy transfer between tryptophan 57 in the epsilon subunit and pyrene maleimide labeled gamma subunit of the chloroplast ATP synthase

The intrinsic fluorescence of the catalytic portion of the chloroplast ATP synthase (CF1) is quenched when cysteine 322, the penultimate amino acid of the gamma subunit, is specifically labeled with pyrene maleimide (PM). The epsilon subunit of CF1 contains the only two residues of tryptophan, which dominate the intrinsic fluorescence of unlabeled CF1. CF1 deficient in the epsilon subunit (CF1-epsilon) was reconstituted with mutant epsilon subunits in which phenylalanine replaced tryptophan at position 15 (epsilonW15F) and position 57 (epsilonW15/57F). CF1(epsilonW15F) containing a single tryptophan, epsilonW57, was labeled with PM at gammaC322. Resonance energy transfer (RET) from epsilonW57 to PM on gammaC322 occurred with an efficiency of energy transfer of 20%. RET was also observed from epsilonW57 to PM attached to the disulfide thiols of the gamma subunit (gammaC199,205) with an efficiency of approximately 45%. The R(o) (the distance at which the efficiency of energy transfer is 50%) for the epsilonW57 and PM donor/acceptor pair is 30 A, indicating that both gammaC322 and gammaC199,205 must be within 40 A of epsilonW57. These RET measurements show that both gammaC322 and gammaC199,205 are located near the base of the alpha/beta hexamer. This places the C-terminus of CF1 gamma much closer to epsilon than hypothesized based on homology to crystal structures of mitochondrial F1. These new RET measurements also allow the alignment of the predicted epsilon subunit structure. The orientation is similar to that predicted from cross-linking and mutational studies for the epsilon subunit of Escherichia coli F1.     

Energy-dependent conformational changes in the epsilon subunit of the chloroplast ATP synthase (CF0CF1)

Incubation of spinach chloroplast thylakoids with pyridoxal 5'-phosphate modified the epsilon subunit of ATP synthase (CF0CF1). Illumination of thylakoids stimulated the modification of one specific amino acid residue of the epsilon subunit by a factor of 3. Endoproteinase Glu-C treatment of the isolated epsilon subunit and fractionation of the peptides by high performance liquid chromatography revealed a major fluorescent peptide with the sequence GKRQKIE. Further treatment of this peptide with endoproteinase Arg-C gave a strongly fluorescent tripeptide (GXR). From the primary structure of the epsilon subunit, the specifically modified residue was deduced to be Lys-109. This suggests the energy-dependent conformational changes in the epsilon subunit which change the surroundings of Lys-109 and alter the reactivity of this residue.     

Characterisation of subunit III and its oligomer from spinach chloroplast ATP synthase

Proton ATP synthases carry out energy conversion in mitochondria, chloroplasts, and bacteria. A key element of the membrane integral motor CFO in chloroplasts is the oligomer of subunit III: it converts the energy of a transmembrane electrochemical proton gradient into rotational movement. To enlighten prominent features of the structure-function relationship of subunit III from spinach chloroplasts, new isolation methods were established to obtain highly pure monomeric and oligomeric subunit III in milligram quantities. By Fourier-transform infrared (FTIR) and CD spectroscopy, the predominantly alpha-helical secondary structure of subunit III was demonstrated. For monomeric subunit III, a conformational change was observed when diluting the SDS-solubilized protein. Under the same conditions the conformation of the oligomer III did not change. A mass of 8003 Da for the monomeric subunit III was determined by MALDI mass spectrometry (MALDI-MS), showing that no posttranslational modifications occurred. By ionisation during MALDI-MS, the noncovalent homooligomer III14 disaggregated into its III monomers.     

Recombinant production and purification of the subunit c of chloroplast ATP synthase

The unique immunoglobulin idiotype expressed on the surface of B lymphoma cells can be used as an effective antigen in tumor-specific vaccines when fused to immunostimulatory proteins and cytokines. A DNA vaccine encoding for an idiotype antibody single chain Fv (scFv) fragment fused to the Tetanus Toxin Fragment C (TTFrC) has been shown to induce protective anti-tumor responses. Protein-based strategies may be more desirable since they provide greater control over dosage, duration of exposure, and in vivo distribution of the vaccine. However, production of fusion protein vaccines containing complex disulfide bonded idiotype antibodies and antibody-derived fragments is challenging. We use an Escherichia coli-based cell-free protein synthesis platform as well as high-level expression of E. coli inclusion bodies followed by refolding for the rapid generation of an antibody fragment – TTFrC fusion protein vaccine. Vaccine proteins produced using both methods were shown to elicit anti-tumor humoral responses as well as protect from tumor challenge in an established B cell lymphoma mouse model. The development of technologies for the rapid production of effective patient-specific tumor idiotype-based fusion protein vaccines provides opportunities for clinical application.     

C-Terminal mutations in the chloroplast ATP synthase gamma subunit impair ATP synthesis and stimulate ATP hydrolysis

Two highly conserved amino acid residues, an arginine and a glutamine, located near the C-terminal end of the gamma subunit, form a "catch" by hydrogen bonding with residues in an anionic loop on one of the three catalytic beta subunits of the bovine mitochondrial F1-ATPase [Abrahams, J. P., Leslie, A. G., Lutter, R., and Walker, J. E. (1994) Nature 370, 621-628]. The catch is considered to play a critical role in the binding change mechanism whereby binding of ATP to one catalytic site releases the catch and induces a partial rotation of the gamma subunit. This role is supported by the observation that mutation of the equivalent arginine and glutamine residues in the Escherichia coli F1 gamma subunit drastically reduced all ATP-dependent catalytic activities of the enzyme [Greene, M. D., and Frasch, W. D. (2003) J. Biol. Chem. 278, 5194-5198]. In this study, we show that simultaneous substitution of the equivalent residues in the chloroplast F1 gamma subunit, arginine 304 and glutamine 305, with alanine decreased the level of proton-coupled ATP synthesis by more than 80%. Both the Mg2+-dependent and Ca2+-dependent ATP hydrolysis activities increased by more than 3-fold as a result of these mutations; however, the sulfite-stimulated activity decreased by more than 60%. The Mg2+-dependent, but not the Ca2+-dependent, ATPase activity of the double mutant was insensitive to inhibition by the phytotoxic inhibitor tentoxin, indicating selective loss of catalytic cooperativity in the presence of Mg2+ ions. The results indicate that the catch residues are required for efficient proton coupling and for activation of multisite catalysis when MgATP is the substrate. The catch is not, however, required for CaATP-driven multisite catalysis or, therefore, for rotation of the gamma subunit.     

The 20 C-terminal amino acid residues of the chloroplast ATP synthase gamma subunit are not essential for activity.

It has been suggested that the last seven to nine amino acid residues at the C terminus of the gamma subunit of the ATP synthase act as a spindle for rotation of the gamma subunit with respect to the alpha beta subunits during catalysis (Abrahams, J. P., Leslie, A. G. W., Lutter, R., and Walker, J. E. (1994) Nature 370, 621-628). To test this hypothesis we selectively deleted C-terminal residues from the chloroplast gamma subunit, two at a time starting at the sixth residue from the end and finishing at the 20th residue from the end. The mutant gamma genes were overexpressed in Escherichia coli and assembled with a native alpha3beta3 complex. All the mutant forms of gamma assembled as effectively as the wild-type gamma. Deletion of the terminal 6 residues of gamma resulted in a significant increase (>50%) in the Ca-dependent ATPase activity when compared with the wild-type assembly. The increased activity persisted even after deletion of the C-terminal 14 residues, well beyond the seven residues proposed to form the spindle. Further deletions resulted in a decreased activity to approximately 19% of that of the wild-type enzyme after deleting all 20 C-terminal residues. The results indicate that the tip of the gammaC terminus is not essential for catalysis and raise questions about the role of the C terminus as a spindle for rotation.     

Nucleotide binding to the isolated beta subunit of the chloroplast ATP synthase.

The beta subunit isolated from the chloroplast ATP synthase F1 (CF1) has a single dissociable nucleotide binding site, consistent with the proposed function of this subunit in nucleotide binding and catalysis. The beta subunit bound the nucleotide analogs trinitrophenyl-ATP (TNP-ATP) or trinitrophenyl-ADP (TNP-ADP) with nearly equal affinities (Kd = 1-2 microM) but did not bind trinitrophenyl-AMP. Both ATP and ADP effectively competed with TNP-ATP for binding. Other nucleoside triphosphates were also able to compete with TNP-ATP for binding to beta; their order of effectiveness (ATP greater than GTP, ITP greater than CTP) mimicked the normal substrate specificity of CF1. The single nucleotide binding site on the isolated beta subunit very closely resembles the low affinity catalytic site (site 3) of CF1 (Bruist, M.F., and Hammes, G. G. (1981) Biochemistry 20, 6298-6305), suggesting that tight nucleotide binding by other sites on the enzyme involves other CF1 subunits in addition to the beta subunit. The results are inconsistent with an earlier report (Frasch, W.D., Green, J., Caguial, J., and Mejia, A. (1989) J. Biol. Chem. 264, 5064-5069), which suggested more than one nucleotide binding site per beta subunit. Binding of nucleotides to the isolated beta subunit was eliminated by a brief heat treatment (40 degrees C for 10 min) of the protein. A small change in the circular dichroism spectrum of beta accompanied the heat treatment indicating that a localized (rather than global) change in the folding of beta, involving at least part of the nucleotide binding domain, had occurred. Also accompanying the loss of nucleotide binding was a loss of the reconstitutive capacity of the beta subunit. ATP protected against the effects of the heat treatment.     

Structural mapping of cysteine-63 of the chloroplast ATP synthase beta subunit.

The single sulfhydryl residue (cysteine-63) of the beta subunit of the chloroplast ATP synthase F1 (CF1) was accessible to labeling reagents only after removal of the beta subunit from the enzyme complex. This suggests that cysteine-63 may be located at an interface between the beta and the alpha subunits of CF1, although alternative explanations such as a conformational change in beta brought about by its release from CF1 cannot be ruled out. Cysteine-63 was specifically labeled with [(diethylamino)methylcoumarinyl]-maleimide, and the distance between this site and trinitrophenyl-ADP at the nucleotide binding site on beta was mapped using fluorescence resonance energy transfer. Cysteine-63 is located in a hydrophobic pocket, 42 A away from the nucleotide binding site on beta.