HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 39, 36253-36261, September 27, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/39/36253    most recent M205062200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Morosinotto, T. Articles by Croce, R. Search for Related Content PubMed PubMed Citation Articles by Morosinotto, T. Articles by Croce, R. Social Bookmarking                      What's this?

Mutation Analysis of Lhca1 Antenna Complex

LOW ENERGY ABSORPTION FORMS ORIGINATE FROM PIGMENT-PIGMENT INTERACTIONS^*

Tomas Morosinotto, Simona Castelletti^§, Jacques Breton^§, Roberto Bassi^¶, and Roberta Croce

From the  Dipartimento Scientifico e Tecnologico, Università di Verona. Strada Le Grazie, 15-37234 Verona, Italy and the ^§ Service de Bioénergétique, Bâtiment 532 CEA-Saclay, 91191 Gif-sur-Yvette, France

Received for publication, May 23, 2002, and in revised form, June 13, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The light harvesting complex Lhca1, one of the four gene products comprising the photosystem I antenna system, has been analyzed by site-directed mutagenesis with the aim of determining the chromophore(s) responsible for its long wavelength chlorophyll spectral form, a specific characteristic of the LHCI antenna complex. A family of mutant proteins, each carrying a mutation at a single chlorophyll-binding residue, was obtained and characterized by biochemical and spectroscopic methods. A map of the chromophores bound to each of the 10 chlorophyll-binding sites was drawn, and the energy levels of the Q_y transition were determined in most cases. When compared with Lhcb proteins previously analyzed, Lhca1 is characterized by stronger interactions between individual chromophores as detected by both biochemical and spectroscopic methods; most mutations, although targeted to a single residue, lead to the loss of more than one chromophore and of conservative CD signals typical of chlorophyll-chlorophyll interactions. The lower energy absorption form (686 nm at 100K, 688 nm at room temperature), which is responsible for the red-shifted emission components at 690 and 701 nm, typical of Lhca1, is associated with a chlorophyll a/chlorophyll a excitonic interaction originating from a pigment cluster localized in the protein domain situated between helix C and the helix A/helix B cross. This cluster includes chlorophylls bound to sites A5-B5-B6 and a xanthophyll bound to site L2.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The light-harvesting system of higher plant photosynthesis is composed of pigment-binding proteins belonging to the Lhc multigenic family (1). Lhc polypeptides are able to bind Chl¹ a, Chl b, and xanthophyll molecules in a suitable structural and dynamic mutual arrangement, ensuring high efficiency of energy transfer processes involved in light-harvesting and photoprotection (2, 3). Among Lhc proteins, only the structure of LHCII, the main antenna complex of photosystem II, has been determined at near atomic resolution, revealing the presence of three transmembrane helices, 12 Chls, two carotenoids, and eight Chl-binding residues (4). Based on sequence comparison, it has been proposed that all Lhc proteins have a similar fold (5), characterized by a cross between helices A and B, stabilized by ionic pairs buried in the lipid bilayer, and stabilized by an helix C domain perpendicular to the membrane plane. Helix D is parallel to the lumenal surface because of its amphiphilic nature. The first (A) and third (B) transmembrane helices and the putative Chl-binding ligands are highly conserved in all family members (6). Despite high homology, each complex has specific biochemical and spectroscopic properties that, although not yet fully understood, are thought to be the basis for the physiological role of each Lhc proteins. LHCI is characterized by low energy absorption forms that are responsible for the emission at wavelengths as long as 700 nm or more (7-9). Whereas it has been proposed that the amplitude of these "red" spectral forms is maximal in the Lhca4 complex (10), this seems to be a property of all Lhca proteins, thus exhibiting absorption and fluorescence emission spectra red-shifted as compared with the Lhcb proteins (12).² Lhca1 has two spectral forms absorbing at 688 and 700 nm and fluorescing with components at 690 and 701 nm, respectively, whose amplitude accounts for nearly one Chl molecule per holoprotein. Each polypeptide binds eight Chls a, two Chls b, and three xanthophylls molecules.² This is the most stable Lhca member so far expressed and reconstituted in vitro (10),² thus making it best suitable for an analysis of the origin of red spectral forms based on mutation studies. In this work, Lhca1 has been analyzed by site-selected mutagenesis at the putative Chl-binding sites and in vitro reconstitution. Comparison with the available data on CP29 and LHCII (13, 14) allows detection of multiple pigment-pigment interactions and the identification of a pigment-protein domain responsible for the peculiar low energy spectral forms of this pigment protein.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

DNA Constructions-- Mutation analysis on Lhca1 from Arabidopsis thaliana² was performed as previously reported (14).

Isolation of Overexpressed Lhca1 Apoproteins from Bacteria-- Lhca1 WT and mutants apoproteins were isolated from the SG13009 strain of Escherichia coli transformed with constructs following a protocol described previously (15, 16).

Reconstitution and Purification of Lhca1-Pigment Complexes-- These procedures were performed as described in Ref. 17 with the following modifications. The reconstitution mixture contained 420 µg of apoprotein, 240 µg of chlorophylls, and 60 µg of carotenoids in total 1.1 ml. The Chl a/b ratio of the pigment mixture was 4.0. The pigments used were purified from spinach thylakoids.

Protein and Pigment Concentration-- HPLC analysis was as in Ref. 18. The chlorophyll to carotenoid ratio and the Chl a/b ratio were independently measured by fitting the spectrum of acetone extracts with the spectra of individual purified pigments. For determination of pigment stoichiometry of different mutants, the Chl/apoprotein ratio of each mutant was determined using WT protein as reference. Chl concentration was determined as in Ref. 19, and the apoprotein content of every mutant was determined as previously reported (20).

Spectroscopy-- The absorption spectra at room temperature and 100 K were recorded using a SLM-Aminco DK2000 spectrophotometer, in 10 mM Hepes, pH 7.5, 20% glycerol (60% at low temperature (LT)), and 0.06% n-dodecyl--D-maltopyranoside. The wavelength sampling step was 0.4 nm. Fluorescence emission spectra were measured using a Jasco FP-777 spectrofluorimeter and corrected for the instrumental response. The samples were excited at 440, 475, and 500 nm. The spectral bandwidth was 5 nm (excitation) and 3 nm (emission). Chlorophyll concentration was about 0.02 µg/ml. LD spectra were obtained as described in Refs. 21 and 22 using samples oriented by the polyacrylamide gel squeezing technique.

The CD spectra were measured at 10 °C on a Jasco 600 spectropolarimeter. The optical density of the samples was 1 at the maximum in the Q_y transition for all complexes, and the samples were in the same solution described for absorption measurements. All of the spectra presented were normalized to the polypeptide concentration based on the Chl binding stoichiometry (23). The denaturation temperature measurements were performed by following the decay of the CD signal at 459 nm while increasing the temperature from 20 to 80 °C with a time slope of 1 °C/min and a resolution of 0.2 °C. The thermal stability of the protein was determined by finding the t_1/2 of the signal decay.

Photobleaching-- The kinetics of Lhca1 WT and mutant photobleaching was measured as described (24) with a light intensity of about 6000 µE m² s¹, and a sample temperature of 10 °C. The data were normalized at 100% at time 0 (initial and maximal absorbance = 0.6).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Sequence comparison shows that all amino acid residues that have been proven to bind Chl in Lhcb1 (14) are conserved in the Lhca1 sequence. The only difference detected was the substitution of the glutamine ligand of Chl in site B6 by a glutamate (1). Glutamate has been reported as Chl ligand in site B6 of CP29, whereas substitution of Gln with Glu in Lhcb1 did not prevent Chl binding (13).

In this work, each putative Chl-binding ligand was mutated to an apolar residue, as reported in Table I, to prevent Chl coordination in each individual site. In the case of sites A1 and A4, in which Chl is coordinated by a glutamate residue charge compensated by an arginine residue, both amino acids were mutated on the basis of previous experience with CP29, showing that noncompensated charges in transmembrane helices decreased complex stability and prevented protein folding (13). In the case of site B5, whose ligand is coordinated by an intrahelix ionic pair, both the single (E110V) and the double (E110V/R113L) mutations were performed. We only report on the results obtained by using the single mutant, essentially identical to those from the mutant carrying the double mutation.

The apoproteins were expressed in E. coli and then reconstituted in vitro with pigments. All of the mutant proteins yielded a folded complex upon reconstitution, as shown by the appearance of a green band upon sucrose gradient ultracentrifugation, but mutant A1, an indication that the Glu¹⁴⁸ (helix A)-Arg⁴⁹ (helix B) ionic bridge is essential for stabilization of Lhca1, as previously observed in CP29 (13). The green bands were harvested with a syringe and submitted to anion exchange chromatography to remove unspecifically bound pigments. The eluate was concentrated, and the aggregation state of Lhca1 WT and mutant proteins was investigated by ultracentrifugation into a density gradient that was shown to be effective in separating monomers from dimers and free pigments (10). As a reference we used WT Lhca1/Lhca4 heterodimer produced by co-refolding the two apoproteins with pigments in the presence of lipids (10).² In agreement with previous work (10), recombinant WT Lhca1 was monomeric and did not form homodimers or higher order aggregates. Fig. 1 shows the result of glycerol gradient ultracentrifugation of WT and three representative Lhca1 mutants, showing that they all have a sedimentation rate corresponding to monomeric aggregation state as was the case for the all mutant proteins not shown in the figure. The yield of reconstitution was differentially affected by each mutation, ranging from 90% of WT yield in the case of mutant B3 to 20% of WT yield in the case mutant A3 (A2, 55%; A4, 25%; A5, 75%; B5, 50%; and B6, 75%). Irrespective of the yield, the complex obtained showed the same mobility in the glycerol gradient, suggesting that no important changes were produced by the mutation in the overall structure of the complex. When native LHCI from Arabidopsis was ultracentrifuged in the same conditions, it exhibited the same sedimentation rate as the recombinant Lhca1/Lhca4 heterodimer (not shown) (25).²

View larger version (34K): [in this window] [in a new window]   Fig. 1.   Aggregation state of Lhca1 mutants. The mobility of WT Lhca1 and selected mutants (B3, B5, and A4) upon ultracentrifugation into a density gradient is compared with that of (monomeric) Lhca1 WT and the (dimeric) Lhca1-Lhca4 complex.

The stability of the reconstituted holoproteins, as studied by the kinetic of CD signal decay with increasing temperature, is reported in Table II. Mutant at the site B3 ligand did not show a significant decrease in stability with respect to WT. The remaining mutations all produced a decrease in stability in the range of 20-30% as compared with WT.

Pigment Composition-- The pigment composition of the mutant holoproteins was analyzed by HPLC and fitting of the spectra of the acetone extracts with the spectra of the individual pigments. The results are summarized in Table II. Interestingly, most of the mutants showed an increase of the Chl a/b ratio compared with WT. This was surprising; an opposite effect was expected because Lhca1 binds eight Chls a and two Chls b, and therefore the majority of the mutations should affect Chl a-binding sites and thus lead to a decrease in Chl a/b ratio. Nevertheless, a clear loss of Chl a was only observed in case of mutations at sites A3 and B3, whereas the mutations at sites A4, A5, B5, and B6 showed a loss of Chl b. It is interesting to note that most of the complexes were also affected in xanthophyll binding as shown by an increase of the Chl/carotenoid ratio in five mutants of seven. The violaxanthin/Lu ratio also changed in most cases, with a preferential loss of violaxanthin in mutants A5, B5, and B6 and of lutein in mutant A3, indicating at least partial selectivity of the carotenoid-binding sites.

The above results imply that both Chls and carotenoids were lost as a consequence of the mutations at most of the Chl-binding residues. To support this finding, pigment/protein stoichiometry was measured for each mutant protein. The results are summarized in Table II. It clearly appears that only in two mutant proteins, namely B3 and B6, a single chlorophyll was lost, whereas mutations at sites A2, A3, A4, A5, and B5 brought the loss of two or three Chls. It was also confirmed that all of the mutants except for the one in site B3 lost part of their xanthophyll complement together with Chls ligands (Table II).

Absorption Spectra-- Binding to individual sites has been shown to tune transition energy of each chromophore to different levels that determine pathway of energy equilibration within Lhc proteins. Further complexity is added to the system by pigment/pigment interactions producing additional energy levels. To determine the absorption properties of Chls in individual binding sites, the absorption spectra of WT and mutant Lhca proteins were recorded at 100 K (Fig. 2) and at room temperature (not shown). At least five Chl a (662, 669, 676, 682, and 688 nm) and two Chls b (644 and 651 nm) absorption forms have been reported as components of Lhca1 spectrum.² The main absorption peak was found at 680.5 and 678.5 nm at room temperature and 100 K, respectively, whereas a shoulder from Chl b absorption was detected at around 645 nm. Mutant proteins showed a blue shift of the major peak, thus implying that most of the mutations affected the environment of Chl a chromophores. In particular, WT minus mutant difference spectra (e.g. Figs. 6, 7, and 8A) indicated that spectral forms at 680 nm were associated to both sites A2 and A3, whereas a 663-nm form was clearly lacking in the mutant at site B3. A special feature of LHCI proteins is the presence of low amplitude, red-shifted, absorption forms. These are responsible for long wavelength fluorescence emission and can be detected at the long wavelength edge of the spectrum. Mutations at sites A5, B5, and B6 clearly showed a reduction in this red tail. These mutants also showed a decreased absorption at 644 nm, indicating a loss of Chl b.

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Absorption spectra at 100 K of all reconstituted complexes. All the spectra are normalized to the maximum. In all of the panels the spectrum of the WT is reported (solid lines). A, dashed line, A3; dotted line, B3; dashed and dotted line, A5. B, dashed line, A2; dotted line, A4. C, dashed line, B5; dotted line, B6.

Linear Dichroism-- Absorption spectral components of Lhca1 are not readily associated to individual binding sites as previously obtained with Lhcb4 (13). This is due to the loss of more than one Chl upon mutation of a single Chl-binding ligand. An additional source of complexity is that pigments bound to mutant proteins may undergo frequency shift with respect to their absorption in the WT protein because of the loss of interacting neighbor pigments. In this case, two spectral effects combine: (i) the loss of forms originating from pigment-pigment interactions and (ii) the appearance of shifted absorption(s) from the residual noninteracting partner of a disrupted pair. In an attempt to gather more detailed information on the spectroscopic effects of mutations, we have analyzed LD spectra of WT versus mutant Lhca1 proteins. LD signals depend not only on absorption wavelength but also on the orientation of the dipole transition moment, thus allowing discrimination of individual chromophores with similar transition energy but different orientations. The LD spectra of WT and mutants at 100 K are reported in Fig. 3 as normalized to the Chl a Q_y peak. The LD spectrum of WT Lhca1 holoprotein is characterized by a negative signal at 642 nm and two positive components at 652 and 679.8 nm. Only mutants B5 and B6 are affected in the amplitude of the 642 nm signal. The positive 680-nm signal from the Chl a Q_y transition was essentially unaffected by most mutations, showing at the most a minor red shift, except in the case of the A2 mutant whose spectrum was shifted to 677 nm. This indicates that a Chl a ligand in site A2 is responsible for most of the 680-nm LD signal. Mutants A4 and A5 lost positive LD around 677 nm, and their peaks are slightly red-shifted (by 0.8 and 0.4 nm, respectively), suggesting that the missing chromophores contribute to the blue edge of the main Chl a LD peak. Changes were observed in the 660-670-nm region for all of the mutant proteins. In particular, mutant A5 showed a strong negative contribution at 670 nm, whereas a negative contribution around 665 nm can be detected in mutants A3 and B3.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Linear dichroism spectra at 100 K of Lhca1 complexes. All the spectra are normalized to the Chl a peak. In all of the panels the spectrum of the WT is reported (solid line). A, dashed line, A3; dotted line, B3. B, dashed line, A2; dotted line, A4; dashed and dotted line, A5. C, dashed line, B5; dotted line, B6.

CD Spectra-- The above results strongly suggest closer packing of pigments in Lhca1 with respect to Lhcb proteins. This should yield excitonic interactions that can be studied by circular dichroism. To test the presence of pigment/pigment interactions, CD spectra of all complexes were measured and are reported in Fig. 4. On the basis of the CD pattern in the 600-750-nm range, the mutants can be classified in two categories: (i) mutants A2, A3, and A4 showed the loss of a 683-nm () contribution and an increase of signal at 672 nm (+) (Fig. 4A) and (ii) mutants A5, B5, and B6 showed an increased amplitude of a 685 nm () signal and the disappearance of a contribution at 668 nm () (Fig. 4B). In the Soret region, changes in the CD spectrum were also detected; mutants A5 and B5 lost the major positive signal at 412 nm. They also lost a negative contribution at 501 nm, probably associated to a carotenoid molecule. This signal is possibly correlated to a contribution (+) at 430 nm, which decreases in the same mutants, suggesting the presence of a xanthophyll/Chl a interaction. An opposite effect was detected in the CD spectrum of the A2 mutant, where an increase of both signals was observed along with a shift to the blue of the 497 nm () contribution. The signal in the 470-520-nm region is possibly the result of different effects; in the A3 mutant, the new positive contribution at 490 nm is probably related to the absence of a negative signal at this wavelength. Based on the biochemical data, which indicated loss of lutein in the A3 mutant, the negative signal at 490 nm can be attributed to this xanthophyll species. Also in this case, a decrease of the 430 nm (+) contribution can be observed, thus suggesting a lutein/Chl a interaction.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Circular dichroism spectra at 10 °C of Lhca1 complexes. The spectra are normalized to the polypeptide concentration. In both panels the WT spectrum is presented (solid line). A, dashed line, A2; dotted line, A3; dashed and dotted line, A4. B, dashed line, A5; dotted line, B5; dashed and dotted line, B6.

Fluorescence Emission Spectra-- The fluorescence emission spectrum of WT Lhca1 shows a major peak at 685 nm, whereas two components at 690 and 701 nm are detectable toward longer wavelengths (10).² Mutants A2, A3, A4, and B3 show emission peak wavelengths identical to that of WT. The red-shifted component, however, has a higher amplitude in mutants A3 and A4, thus suggesting the loss of bulk Chl a without affecting the pigment-binding domain responsible for the low energy forms. The second group of mutants, namely A5, B5, and B6, showed a blue shift of the major peak by 2-3 nm and a decreased amplitude of low energy emission. The emission spectra of selected mutants, i.e. A4 and A5, at 100 K are presented in Fig. 5, together with the spectrum of WT Lhca1.

View larger version (12K): [in this window] [in a new window]   Fig. 5.   Fluorescence emission spectra at 100 K. Lhca1 WT (dotted line), A4 (solid line), and A5 (dashed line) are shown. The spectra were recorded at 0.01 optical density. The spectra are normalized to the maximum.

Photobleaching-- The effect of mutations on photoprotection by quenching of triplet chlorophyll states was assessed by studying the kinetics of photobleaching (24). The kinetic of Q_y band absorption decay, upon strong illumination in the presence of O₂, showed two kind of different behavior; mutants A2, A3, and B3 were similar to WT, whereas mutants A5, B5, and B6 were more sensitive to photobleaching (Fig. 6). Mutant A4 showed a somehow intermediate behavior; although the initial rate (0-5 min) of decay was as fast as those of the most sensitive mutants, longer times of exposure yielded a slow bleaching similar to that of WT.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Photobleaching of Lhca1 mutants. The decay curves show the total Qy absorption relative to a 100% initial value. Points refer to experimental data.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Energy transfer within Lhc proteins is defined by the distance between chromophores, by their transition energy, and by the mutual orientation of transition vectors (26, 27). Determination of these major parameters has been difficult to obtain in higher plant antenna proteins by spectroscopic methods because of the high number of bound chromophores having their spectra overlapping the Q_y band. Recently, introduction of mutation analysis of Chl-binding residues allowed the interchange of individual amino acid residues, coordinating Chls through their central Mg²⁺ atom, and determination of individual chromophore spectra by difference spectroscopy. The case of CP29 (Lhcb4) was the simplest. This protein binds eight Chls and two xanthophylls, which have, if any, very weak excitonic interactions between each other (23). Each mutation led to the removal of a single Chl, and an individual spectral form was detected in the absorption difference spectrum (13, 28). A somewhat more complex situation was found in the major antenna protein LHCII (Lhcb1) that binds 12 Chl/polypeptide and shows excitonic interactions between a subset of its chromophores. Some Lhcb1 mutations caused removal of single Chls from the protein and of individual spectral forms. Other mutations led to the loss of multiple Chls and of excitonic interactions as detected by both absorption band shifts and by the disappearance of conservative CD signals (14). In the present study, mutation analysis was applied to Lhca1, a protein binding ten Chls and three xanthophylls.² This pigment protein was stable upon reconstitution in vitro from the apoprotein overexpressed in E. coli and purified pigments. Moreover, the recombinant protein showed characteristics corresponding to those of the native protein as judged by its capacity of forming heterodimers with Lhca4, which exhibited spectral features typical of native LHCI (10).² All of the Lhca1 mutants could refold in the presence of pigment with the exception of mutant A1, and the mutant proteins showed the same (monomeric) aggregation state as WT Lhca1 (10). Much of the data obtained with Lhca1 indicate an higher level of pigment/pigment interactions: (i) only two mutations (B3 and B6) caused the loss of a single Chl, whereas two Chls were lost in most cases. (ii) All of the mutations, except for the one in site B3, caused loss of xanthophylls, whereas in CP29 and LHCII this effect was found only for few specific mutations. (iii) WT minus mutant difference CD spectra showed the loss of conservative signals with the typical shape of an excitonic interaction in all mutants (e.g. Fig. 10B).

The involvement of most Lhca1 pigments into tight interactions is evidenced by the unexpected effect of mutations on the protein stability as assessed by thermal denaturation. In CP29 and LHCII, having a low level of pigment/pigment interactions, each mutation affected protein stability to a different extent, because of specific roles of each protein domain in maintaining protein folding (13, 24). The opposite was found in Lhca1; all of the mutations, except for A1, which prevented the formation of a stable pigment-protein complex, and B3, which has the same stability as WT, led to a similar decrease of protein stability. However, the effect was rather small and of similar amplitude for mutations in different structural domains, suggesting a more crucial role for interacting pigments in maintaining folding stability.

Lhca1 exhibits two major spectroscopic differences with respect to CP29 and LHCII: (i) a red-shifted "bulk" absorption of the Q_y transition as shown by the 680.5-nm peak and (ii) the presence of "minor" spectral forms further shifted to lower energies. It is interesting to discuss how pigment/pigment interactions contribute to each of these two effects.

In the following, we summarize the effects of individual mutations with the aim of identifying the chromophore bound to each individual site, its spectroscopic characteristics, and the interactions with neighbor pigments. In doing so we first analyze chromophores that only contribute to the bulk absorption band (sites A3, B3, A2, and A4) followed by those which also contribute to the low energy absorption forms (sites A5, B5, and B6).

Mutants Specifically Affecting Bulk Absorption

Mutants A3 and B3-- The mutation on Chl B3 ligand has the smallest influence on the properties of the pigment-protein complex. Biochemical analysis clearly indicated the loss of one Chl a (0.9 Chl a and 0.1 Chl b). Nevertheless, two peaks with similar intensity, at 663 and 679 nm, were present in the WT minus B3 difference absorption spectrum (LT) (Fig. 7).

View larger version (14K): [in this window] [in a new window]   Fig. 7.   Spectroscopic analysis of A3 and B3 mutants. Absorption difference spectra at 100 K of WT minus A3 (solid line), WT-B3 (dashed line), and B3-A3 (dotted line). The spectra were normalized to the Chl concentration before subtraction.

The mutation on Chl A3 ligand influences the binding of at least two Chls ligands. Again the LT (RT) difference spectrum shows two peaks at 663 (663.6) and 679 (681.6) nm, with a relative intensity of 1:3 (Fig. 7).

The above data strongly suggest the presence of an interaction between two Chls a bound to these two sites, in agreement with their close location in LHCII (4, 29). According to this figure, the WT minus A3 difference spectrum shows the absorption contribution of the two interacting Chls. Because the B3 mutant does not lose Chl A3 but loses the interaction, the B3 minus A3 difference spectrum reveals the contribution of the Chl A3 monomer. As shown in Fig. 7, this corresponds to a single band peaking at 678.5 nm. This absorption band is blue-shifted by 0.5-1 nm with respect to the low energy band of the interaction. The absorption of B3 monomer is thus expected to peak around 663 nm, similar to what is observed in LHCII (14). The small red shift of Chl A3 in the dimer is as expected for an interaction between two Chls rather distant in energy (around 350 cm¹), because of the small value of the overlap integral between these two chromophores (23). The redistribution of the oscillating strength between the two monomers is rather strong, depending only on the angle between the dipole moments of the two pigments (30). A further result, highlighting a stronger pigment-pigment interaction in Lhca1 with respect to CP29 and LHCII, is the loss of one lutein molecule as an effect of mutation in site A3, whereas xanthophylls were not affected by the homologous mutation in LHCII and CP29. This is likely to be the lutein in site L1. In fact, site L1 has been found to be specific for lutein in all Lhc proteins so far analyzed (31-33) and is the only xanthophyll site located in close contact to Chl A3 (4). The loss of lutein in L1 can explain the high similarity of CD and LD spectra of mutants A3 and A2, suggesting that the loss of this xanthophyll may affect the organization of the A2-B2 domain, which is responsible for both the strong positive LD signal at 681 nm and the 684-672 nm excitonic interaction visible in the CD spectrum (see below).

Mutant A2-- The mutation at the Chl A2 ligand induces the loss of two Chl molecules: 1.5 Chl a and 0.5 Chl b. The second lost Chl is here suggested to be Chl B2, because in the homologous protein LHCII B2 is the nearest neighbor to A2 (4) and has been shown to be bound through Chl A2 (14). WT minus A2 difference spectrum at LT (RT), shows a major Chl a peak at 679.5 (682) nm (Fig. 8A), whereas the Chl b contribution at LT (RT) is found at 649 (651) nm. A negative component in this difference spectrum is also observed at 670 nm and is attributed either to a shift induced by the mutation in the absorption of a neighbor Chl or to the absorption of a monomeric Chl upon disruption of an excitonic interaction (with Chl A4, see below). The presence of an excitonic interaction between Chls A2 and B2 is supported by the WT - mutant difference CD spectrum showing a conservative signal (Fig. 4), with contributions at 672 (+) and at 685 nm (). The same signal disappears in the A4 mutant. The organization of the A2/B2/A4 domain in Lhca1 is thus homologous to that in LHCII, as suggested by the very similar effect of the mutations in sites A2 and A4. The only difference is that although site B2 was specific for Chl b in LHCII (14), it has a mixed Chl a/Chl b occupancy in Lhca1. The presence of a Chl ligand in site B2 is also supported by the ligation of Chl A2 by Asn (like in LHCII and CP26) rather than by His (as in CP29). This is consistent with the large amplitude of the LD signal of Chl A2 (Fig. 8B), characteristic of the specific orientation assumed by Chl A2 when binding Chl B2 (33). On the basis of difference spectra and of Chl a binding to site A2 in CP29 and LHCII (14), we assign the 679.5 nm (682 nm at room temperature) absorption to Chl A2, whereas in site B2 Chl b absorbs at 650 nm and Chl a absorbs at 682 nm.

View larger version (15K): [in this window] [in a new window]   Fig. 8.   Spectroscopic analysis of A2 mutant. Absorption (A) and LD (B) spectra at 100K of Lhca1-WT (solid lines) and A2 mutant (dashed lines) along with the difference spectrum (dotted lines). The absorption spectra are normalized to the polypeptide concentration before subtraction, whereas the linear dichroism spectra are normalized as reported in Ref. 27.

In both CP29 and LHCII, Chl A2 was the red-most absorbing pigment (680 and 681 nm, respectively) and thus the principal fluorescence emitter. This is not the case in Lhca1; although the absorption properties of Chl A2 seem to be conserved across the family, the fluorescence spectrum of A2 mutant peaks at 685 nm as in the WT, and the low energy emission tail is still present, thus indicating that the red-most absorption in Lhca1 is not associated to this Chl.

Mutant A4-- Mutant A4 looses two Chl molecules: 1.2 Chl a and 0.8 Chl b. The WT minus mutant difference spectrum shows two major peaks at 679 and 646 nm representing the absorption of Chl a and Chl b, respectively, but minor contributions at 684 (+), 674 (+), and around 660-670 nm were also detected (Fig. 9). These features clearly indicate that the mutations do not influence only the target site but also the spectroscopic characteristics of neighbor pigments. The difference LD spectrum shows a positive contribution at 678 nm, whereas no changes were observed in the Chl b absorption region (Fig. 9). The CD spectrum of A4 mutant is identical to the spectrum of A2 mutant, in agreement with the results on LHCII (see the A2 mutant discussion), thus indicating the loss of the same excitonic interaction. On the basis of the homology with LHCII, we suggest that site A4 is occupied by a Chl a molecule peaking at 678 nm, whereas the second lost Chl is probably accommodated in B2 site, which shows mixed occupancy.

View larger version (11K): [in this window] [in a new window]   Fig. 9.   Spectroscopic analysis of A4 mutant. The absorption (dashed lines) and LD (solid lines) difference spectra at 100 K of A4 mutant are shown. The spectra are normalized to the polypeptide concentration before subtraction, whereas the linear dichroism spectra are normalized as reported in Ref. 27.

Mutants Affecting the Low Energy Absorption Forms

Mutant A5-- This mutant looses 1.2 Chl a and 0.8 Chl b with respect to WT Lhca1. It is interesting to consider that only one Chl a was lost upon mutation on A5 ligand in both CP29 and LHCII, suggesting that one additional Chl, held in place by interactions with other Chls rather than with a specific amino acid residue, is present in Lhca1. The difference absorption spectrum (Fig. 10A) shows several contributions: a major 679-nm signal and a shoulder at 686 nm in Chl a region, whereas a Chl b contribution is observed at 644 nm. In agreement with previous finding in CP29 and LHCII, we suggest that site A5 binds Chl a and that a Chl b site is also affected by the mutation. Based on the LHCII structure, the only site located in close proximity to Chl A5 is site B5. However, it is quite unlikely that mutant A5 looses also Chl in site B5, which is bound in a different protein domain and has a defined ligand on the protein. Moreover, pigment analysis and LD spectra clearly show that A5 and B5 mutants do not loose the same Chls; the 642 nm () signal disappears in mutant B5, whereas it is conserved in the LD spectrum of mutant A5. On this basis we propose the presence of a new site in Lhca1. This site should have mixed Chl a/Chl b occupancy but higher affinity for Chl b (0.8 versus 0.2). The bound Chl should be located in the proximity of Chl A5 and should be oriented close to the magic angle as suggested by the absence of a detectable LD signal. We cannot exclude that this site corresponds to site B1 in LHCII. Nevertheless, the distance (13.4 Å) appears to be too large for a proximity effect similar to that described for the loss of Chl B2 in the mutant A2. The determination of the energy transition of Chl a in site A5 is not straightforward; the WT minus A5 difference absorption spectrum shows a major contribution at 679 nm, whereas the LD spectrum shows changes in shape at around 675 nm (Fig. 3). The WT minus A5 difference CD spectrum shows the presence of excitonic interactions adding incertitude to the attribution (see below). At the moment, we cannot explain this discrepancy and are thus forced to leave this question open.

View larger version (16K): [in this window] [in a new window]   Fig. 10.   Spectroscopic analysis of A5, B5 and B6 mutants. The absorption difference spectra at 100 K (A) and circular dichroism difference spectra (B) of WT-A5 (solid lines), WT-B5 (dashed lines), and WT-B6 (dotted lines) are shown. The spectra before subtraction are normalized to the polypeptide concentration.

Mutants B5-- The ligand of site B5 is constituted by a ionic pair that has the double function of coordinating Chl B5 and providing end-capping of helix C. In Lhca1, the B5 mutant looses two Chls (0.9 Chl a and 1.1 Chl b) with respect to WT. It is interesting to observe that in LHCII, the double mutation induced the loss of additional Chls (namely A6 and A7), whereas in Lhca1, the E110V/R113L double mutation yields a complex with the same biochemical and spectroscopic characteristics as the single (E110V) mutant. This suggests that sites A6 and A7 are not present in the Lhca1 structure. The absorption difference spectrum is almost identical to the one of A5 mutant in which positive contributions at 644, 678.5, and 686 nm were detected (Fig. 10A). In the LD spectrum, the negative signal associated to a 643-nm Chl b disappears. Among all mutants, only B5 and B6 loose the Chl b LD signal (Fig. 3), thus implying that the second Chl lost upon mutation in site B5 is Chl B6. On the basis of the pigment analysis, which indicates that site B6 is a mixed site with 65% Chl b occupancy (see below), we can conclude that site B5 also shows mixed occupancy (1:1, Chl a:Chl b).

Mutant B6-- The pigment analysis of B6 mutant indicates the loss of one Chl: 0.35 Chl a and 0.65 Chl b, thus attributing mixed occupancy to the B6 site. This is consistent with previous results with CP29. In fact, glutamate as ligand for Chl in B6 confers mixed occupancy to this site (13). On the contrary, the spectroscopic effects are strikingly different with respect to CP29; the absorption difference spectrum shows at least three positive contributions peaking at 643, 673, and 686 nm and two negatives at 665 and 680 nm (Fig. 10A), suggesting that the B6 ligand is involved in tight interactions with neighbor pigments. The nearest neighbors are sites A5 and B5 and mutations at these three sites lead to the loss of the 686-nm absorption form and of the red-shifted emission contributions at 690 and 701 nm. The difference CD spectrum supports this view: an excitonic signal (685 nm (+)/668 nm ()) is lost in the B6 mutant, similar to what is observed for A5 and B5 complexes (Fig. 10B). We conclude that interactions between Chls bound to the protein domain in between helix C and the helix A-helix B crossing are responsible for the low energy absorption forms in Lhca1. However, a detailed interpretation is not straightforward. A possible explanation considers the presence of an excitonic interaction between two of the Chls in sites A5, B5, and B6. Excitonic interactions have been so far detected between Chl A2/B2 (14) and Chl A3/B3 (this work) having edge to edge distances below 5 Å. The distances between the components of the A5/B5/B6 cluster are of 4.8, 7.4, and 10.5Å for the A5/B5, B5/B6, and A5/B6 pairs, respectively, based on LHCII structure (4). Our best hypothesis is that the Chl a/Chl a excitonic interaction originates from interactions between Chl A5 and the Chl a component in site B5. Yet mutation at site B6 disrupts the interaction. The influence of Chl B6 on the putative Chl A5/Chl B5 interaction may well be mediated by the xanthophyll in site L2, which is lost upon mutation in site B6. Chl A5 is in close contact with the xanthophyll in site L2, and its absence could change the orientation/distance of Chl A5 with respect to Chl B5. This scheme is similar to that described above for the effect of the A3 mutation on the orientation of Chl A2 through the loss of lutein in site L1. The violaxanthin in site L2 is somewhat special, having its S₀-S_2,0 transition at 501 nm, thus red-shifted by 28 nm with respect to its absorption in acetone. Because violaxanthin bound to sites L1, L2, and V1 of LHCII is shifted by 19, 19, and 12 nm, respectively (31, 34), it appears that in Lhca1, violaxanthin undergoes a special interaction that yields into the red-most absorption observed for a violaxanthin molecule in a protein environment. This violaxanthin spectral form is lost in all three mutants.

Whereas the details are not completely clear, we propose that the red-most absorption of the Lhca1 complex originates from an excitonic interaction leading to a high energy band around 668 nm and a low energy contribution at 686 nm, which involves the domain composed by Chls A5, B5, and B6 and carotenoid L2. The absorption wavelength of the band indicates that two Chls a and possibly a carotenoid molecule are involved in the interaction. We suggest that red forms originate in Lhca4 from a similar chromophore organization. The requisite for Chl b to generate the red emission in Lhca4 (35) may thus consist of the occupancy of site B6 by Chl b.

Mutant A1-- The mutant affected at site A1 does not fold in vitro, confirming the primary role of the Glu¹⁴⁸-Arg⁴⁹ ionic bridge, not only as a ligand for Chl A1 but also in the stabilization of protein folding. The same effect was previously observed in both CP29 and LHCII (13, 14). By computation of the Chl a versus Chl b occupancy in the other sites, it clearly appears that site A1 accommodates a Chl a molecule as was the case in CP29 and LHCII. A description of site occupancy in Lhca1 and of the wavelength absorption peak for individual Chls is given in Fig. 11 and Table III.

View larger version (33K): [in this window] [in a new window]   Fig. 11.   Model of Lhca1 structure. Chl and carotenoid binding site occupancy were modified from Ref. 4. Chl a and mixed sites are indicated in blue and green, respectively; lutein and violaxanthin are in orange and purple, respectively. The wavelengths of absorption peak for individual chlorophylls are indicated in Table III.

Lhca1 in the Lhca1-4 Heterodimer

The comparison between the CD spectra of Lhca1 and Lhca4 monomers with Lhca1-4 heterodimer showed that, in the dimer, the negative contribution at 669 nm, typical of Lhca1, is lost.² This suggests that the pigment(s) responsible for this signal in Lhca1 are involved in the dimerization process. Because this signal is clearly associated with Chls A5, B5, and B6, we suggest that the C helix domain is involved in dimerization.

Carotenoid-binding Sites in Lhca1

Lhca1 has three xanthophyll-binding sites that can accommodate both violaxanthin and lutein when refolded in vitro.² Most mutants are affected in the carotenoid binding, thus allowing localization into individual sites. Mutation A3 shows that lutein is preferentially bound to L1 site and has an S2_0,0 transition at 492 nm, as detected by absorption difference spectra (not shown). The remaining two xanthophylls, 1.2 violaxanthin and 0.8 lutein, are located in sites L2 and N1. From the analysis of the mutant A5 (located in close proximity of L2) and B5 (close to site N1), it can be concluded that both sites bind lutein and violaxanthin. We thus propose that the affinity of the three xanthophyll-binding sites is as follows: Lu in L1, Lu + violaxanthin in L2 in agreement with results with Lhcb1, and Lu + violaxanthin + neoxanthin traces in N1 (Fig. 11). An important issue derives from the photobleaching experiments; removal of lutein from site L1 did not lead to a substantial increase of photobleaching. This is in contrast with previous results with LHCII (24) showing that lutein in site L1 was essential for Chl triplet quenching. We suggest that L2 rather than L1 is the primary ³Chl* quencher in Lhca1. This might be due to a different conformation of Lhca versus Lhcb protein as discussed below.

Lhca versus Lhcb

A major difference between Lhca and Lhcb proteins consists of their different fluorescence yield, which is lower in Lhca. Fluorescence decay kinetic shows three lifetime components in Lhca1: 0.35, 1.7, and 3.6 ns that account, respectively, for 27, 53, and 20% of the total decay (36). The same lifetime components have been found in Lhcb proteins, although with different relative amplitudes, the 3.6-ns lifetime having the highest amplitude (37, 38). Lhcb proteins may undergo conformational changes to a low fluorescence conformation in excess light conditions upon triggering of NPQ (37) and/or incorporation of zeaxanthin into the L2 site (11, 24). Lhca proteins appear to assume a constitutive "dissipative" conformation, thus suggesting that the "quenched" conformation of Lhcb proteins is similar to that of Lhca proteins. The study of Lhca proteins may therefore be important not only for the understanding of their role as antenna for photosystem I but also as models for the mechanism of nonradiative dissipation in photoprotection.

	FOOTNOTES

^* This work was founded by CNR "agenzia 2000" and by "MIUR" Progetto FIRB n.RBAU01ECX.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be addressed. Tel.: 390458027916; Fax: 390458027929; E-mail: bassi@sci.univr.it.

Present address: Istituto di biofisica, CNR Milano, c/o Dipartimento di Biologia, Via Celoria 26, 20133 Milano, Italy.

Published, JBC Papers in Press, July 2, 2002, DOI 10.1074/jbc.M205062200

² Croce, R., Morosinotto, T., Castelletti, S., Breton, J., and Bassi, R. (2002) Biochim. Biophys. Acta, in press.

	ABBREVIATIONS

The abbreviations used are: Chl, chlorophyll; HPLC, high performance liquid chromatography; LD, linear dichroism; Lhc, light-harvesting complex; LHCI and LHCII, light-harvesting complexes of photosystems I and II; Lu, lutein; WT, wild type.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				G. Bonente, B. D. Howes, S. Caffarri, G. Smulevich, and R. Bassi Interactions between the Photosystem II Subunit PsbS and Xanthophylls Studied in Vivo and in Vitro J. Biol. Chem., March 28, 2008; 283(13): 8434 - 8445. [Abstract] [Full Text] [PDF]

				R. Croce, A. Chojnicka, T. Morosinotto, J. A. Ihalainen, F. van Mourik, J. P. Dekker, R. Bassi, and R. van Grondelle The Low-Energy Forms of Photosystem I Light-Harvesting Complexes: Spectroscopic Properties and Pigment-Pigment Interaction Characteristics Biophys. J., October 1, 2007; 93(7): 2418 - 2428. [Abstract] [Full Text] [PDF]