PhowsynthesisResearch 48: 263-270, 1996. Q 1996 KluwerAcademicPublishers. Printedin the Netherlands. Regular paper
Excitation energy transfer in the green photosynthetic bacterium Chloroflexus aurantiacus: A specific effect of 1-hexanol on the optical properties of baseplate and energy transfer processes M a m o r u M i m u r o 1, Y o s h i n o b u N i s h i m u r a 2, I w a o Y a m a z a k i 2, M a s a y u k i K o b a y a s h i 3, Z h e n g Yu W a n g 3, T s u n e n o r i N o z a w a 3, K e i z o S h i m a d a 4 & K a t s u m i M a t s u u r a 4
1National Institute for Basic Biology, Myodaiji, Okazaki 444, Japan; 2Department of Chemical Process Engineering, Faculty of Engineering, Hokkaido University, Sapporo 060, Japan; 3Department of Biochemical Engineering, Faculty of Engineering, Tohoku University, Sendai 970, Japan; 4Department of Biology, Faculty of Science, Tokyo Metropolitan University, Hachioji 192-03, Japan Received 13 October 1995; accepted in revised form 23 January 1996
Key words: Chloroflexus aurantiacus, chlorophyll, chlorosome, energy transfer, green bacteria, photosynthesis Abstract
The effect of 1-hexanol on spectral properties and the processes of energy transfer of the green gliding photosynthetic bacterium Chloroflexus aurantiacus was investigated with reference to the baseplate region. On addition of 1hexanol to a cell suspension in a concentration of one-fourth saturation, a specific change in the baseplate region was induced: that is, a bleach of the 793-nm component, and an increase in absorption of the 813-nm component. This result was also confirmed by fluorescence spectra of whole cells and isolated chlorosomes. The processes of energy transfer were affected in the overall transfer efficiency but not kinetically, indicating that 1-hexanol suppressed the flux of energy flow from the baseplate to the B806-866 complexes in the cytoplasmic membranes. The fluorescence excitation spectrum suggests a specific site of interaction between bacteriochlorophyll (BChl) c with a maximum at 771 nm in the rod elements and BChl a with a maximum at 793 nm in the baseplate, which is a funnel for a fast transfer of energy to the B806-866 complexes in the membranes. The absorption spectrum of chlorosomes was resolved to components consistently on the basis, including circular dichroism and magnetic circular dichroism spectra; besides two major BChl c forms, bands corresponding to tetramer, dimer, and monomer were also discernible, which are supposed to be intermediary components for a higher order structure. A tentative model for the antenna system of C. aurantiacus is proposed.
Abbreviations: A 6 7 0 - a
component whose absorption maximum is located at 670 nm; (B)Chl(bacterio)chlorophyll; C D - c i r c u l a r dichroism; F 6 7 5 - a component whose emission maximum is located at 675 nm; FMO protein-Fenna-Mathews-Olson protein; LD -linear dichroism; LH-light-harvesting; M C D - m a g n e t i c circular dichroism; PS-photosystem; RC-reaction center
Introduction
An antenna system of the green photosynthetic bacterium Chloroflexus aurantiacus consists of B806-866 complexes in the cytoplasmic membranes and chlorosomes outside of the membranes (Olson 1980). The B806-866 complex directly interacts with the reaction center (RC), as in the case of light-harvesting complex (LH) 1, even though the plural spectral peaks are simi-
lar to LH 2 of purple bacteria. The chlorosome is a main light harvesting apparatus of this bacterium and is connected to the B806-866 complex through a baseplate where the bacteriochlorophyll (BChl) a protein complexes are located. Chlorosomes include highly organized molecular assemblies of BChl c (Krasnovsky et al. 1980; Holzwarth et al. 1990; Mimuro et al. 1992; Hirota et al. 1992). Proteins are not directly involved in the molecular organization of BChl c; instead, a
264 main force in the stabilization of the functional form of BChl c is an intermolecular interaction between BChl c molecules; a central Mg atom, a hydroxyl group in the C31 position, and a keto carbonyl group in the C 13 l position play an essential role for the self assemblies (Lutz and van Brakel 1988; Hildebrandt et al. 1994; Chiefari et al. 1995). The BChl c aggregates have their Qy transition moments almost parallel to the long axis of chlorosomes (van Amerongen et al. 1988; Matsuura et al. 1993; Mimuro et al. 1994), irrespective of the spectral heterogeneity of the rod elements. Treatment with alcohol induces 'monomerization' of BChl c aggregates, which is reflected by a large blue shift of the BChl c Qy transition (Brune et al. 1987a; Matsuura and Olson 1990). In the case of 1-hexanol, a saturated amount (59 mM or 0.74% at 25 °C) of alcohol induced a complete conversion of the spectral forms, and this process was shown to be fully reversible (Matsuura and Olson 1990). By slowly diluting the alcohol content from the saturated concentration, the original optical properties were recovered, but when the concentration was diluted quickly, an intermediate form or forms appeared and further red shift was no longer observed. Accompanying these spectral changes was a change in the morphology of chlorosomes (Wang et al. 1995). Based on these phenomena, we can use the treatment with 1-hexanol to monitor the aggregation behavior of the building blocks, such as monomers and oligomers Of BChl c. The spectral change is attributed to the binding of alcohol molecules to the central Mg atoms, which consequently disrupts the intermolecular interaction among BChl c molecules. Conversely, BChl a in the baseplate is known to be insensitive to 1-hexanol treatment, presumably because of a tight binding of pigments to polypeptides. Two types of the BChl a pool exist in the baseplate, as shown by the time-resolved fluorescence spectra of whole cells (Brune et al. 1987b; Mimuro et al. 1989, 1994). Several reports have been published regarding the effect of alcohol, including 1-hexanol, on isolated chlorosomes (Brune et al. 1987a; Matsuura et al. 1993; Mimuro et al. 1994). However, the effects of 1-hexanol on whole cells are scarcely known, although the potential effect is similar to the effect on chlorosomes. We therefore investigated the effect of the 1-hexanol treatment on the spectral properties and processes of energy transfer in whole cells of C. aurantiacus. We found a selective effect of 1-hexanol on the baseplate component, which is located within the main energy flow to the RC. We propose a specific interaction between
BChl c and BChl a in the funneling site to B806-866 in the membranes. Materials and methods Chloroflexus aurantiacus (OK-70) was grown photoheterotrophically at 55 °C in the medium described by Hanada et al. (1995). Cells at the late log-growth phase were harvested and chlorosomes were isolated by the method developed by Gerola and Olson (1986), using 2 M of chaotropic anion (NaSCN) and sucrose density gradient centrifugation. The absorption spectrum was measured with a Hitachi 330 spectrophotometer, the fluorescence spectrum was measured with a Hitachi 850 spectrofluorometer, and the circular dichroism (CD) spectrum was measured with a Jasco J-720 spectropolarimeter. The magnetic circular dichroism (MCD) was measured with a Jasco J-720 polarimeter equipped with an electromagnet with a field strength of 1.35 T (Nozawa et al. 1995). The time-resolved fluorescence spectra were obtained by a time-correlated single-photon counting method (Mimuro et al. 1989) using an S-1 type photomultiplier. The light source was a Ti-Sapphire laser (720 nm) with a pulse width of 200 fs (typical), the repetition rate was 4 MHz, time resolution was greater than 3 ps and spectral resolution was less than 2 nm. All of the spectral data was transferred to a microcomputer and numerically analyzed, i.e. base-line correction, difference spectrum, correction for spectral sensitivity of the fluorometer, deconvolution of spectra and decay analysis with the assumption of an exponential decay function (Mimuro et al. 1989, 1994). Results Overall energy transfer in whole cells
The overall processes of energy transfer in C. aurantiacus at - 1 9 6 °C are shown by time-resolved fluorescence spectra (Figure 1A). When BChl c was excited at 720 nm, a sequential energy flow in the order of BChl c (752 nm), baseplate BChl a (810 nm), and the B866 in the membranes (888 nm) was clearly resolved. The emission of B866 was scarcely detected at 0 ps, but within 50 ps, the intensity reached the maximum. These spectral changes were identical to previous reports (Mimuro et al. 1989, 1994); however, because of a better time-resolution, we observed a preferential emission from BChl c at the initial phase
265 500C
(A)
.
.
.
.
i
.
.
.
.
i
.
.
.
.
i
.
.
.
.
(A) Whole cells 4OO(
300( 158.4
8
105.6
200C
100C .
.
.
.
65( Wavelength (nrn)
.
'
'
700 .
.
.
.
i
'
.
.
.
.
750 .
.
.
.
i
.
.
r
800 .
.
.
.
i
850 .
.
(B) Chlorosomes
.
.
~3
200C
(B) 150C
s=
,-
I ~ A
(x20)
1+ 2 50(~
650
t~
700
........................ 750
800
850
900
950
Wavelength (nm)
. (C) A ~
750
~)']
\~
800 ":/db
850 ....
825 . . . . . . . . . . 752
Figure 1. Time-resolvedfluorescence spectra of whole cells of C. aurantiacus at -196 °C. (A) control, and (B) with 0.4% 1-hexanol. The excitation wavelength was 720 nm. For the low-temperature spectroscopy, 15% of poly(ethylene glycol) 4000 was added to obtain homogenousice.
of excitation and a clear rise of the BChl a emission in the membranes. The emission from the baseplate was located at 810 nm, and its intensity was low throughout the transfer process. This fact indicates that the energy flow between the baseplate and B806 in the membranes is not the rate-limiting step. By assuming exponential decay, we estimated kinetic parameters on the decay curves at seven particular wavelengths from 750 to 910 nm. A fast-decaying component or components (in the range of 100 fs, Savikhin et al. 1994) was not clearly resolved due to a limit of the time-resolution; however, we found clear rise times in the acceptors. Those were 16 ps for the baseplate BChl a component (at 810 nm), 28 ps for B866 (at 888 nm) and 50 ps at 910 nm. The difference in the rise times between 888 and 910 nm clearly indicates that a long wavelength antenna is expected in the B866 pool, as similarly found with purple bacteria (Kramer et al. 1984). The 50-ps rise component corresponds to the overall energy flow from BChl c to the RC, and it was almost identical to the result obtained with purple bacteria (van Grondelle et al. 1994). The rise time of B866, i.e. 28 ps, was also
400. . . . . . . .
iO() . . . . . . . . . .600 . . . . . . . . . . . . . 700 Wavelength (nm)
Figure 2. The effect of 1-bexanol on the steady-state fluorescence
spectra of whole ceils (A) and of isolated chlorosomes (B) of C. aurantiacus at -196 °C is shown. The excitation wavelengthwas 460 nm for both cases, and because of a low sensitivitybeyond 850 nm, the corrected spectra were shownup to this wavelength. In (A), the full line shows the spectrum in the absence of 1-hexanol, the broken line, with 0.4% l-hexanol. Spectra were normalizedat 752 nm. In (B), numbers from 1 to 4 correspond to the concentrations of 1-hexanol of 0, 0.2, 0.4 and 0.6%, respectively. A magnified spectrum (20 fold) was shown for 0.6% of 1-hexanol. Spectra are shown after normalizationat 752 nm. (C) Excitation spectra were measured on isolated chlorosomes at -196 °C in the absence of 1-hexanol. Spectra are shown after normalizationat 460 nm. obtained by the time-resolved fluorescence anisotropy spectra measured on cells aligned in a polyacrylamide gel (data not shown).
The effect o f 1-hexanol on the sequence o f energy transfer and on kinetics The fluorescence properties of whole cells were strongly affected by 1-hexanol as shown in the steady-state spectrum at - 196 °C (Figure 2A). Under a low concen-
266 tration of 1-hexanol (0.4%) in which the main emission was still observed at 752 nm as in the control, the emission maximum of the baseplate was shifted from 815 to 824 nm (Figure 2A, curve 2). The difference spectrum showed a decrease in the intensity at 813 nm and an increase at 826 nm (data not shown), indicating the presence of plural components in the baseplate region. In addition, two fluorescence bands induced by the 1-hexanol treatment were discernible at 771 and 788 nm. Furthermore, a few emission bands were detected around 700 nm; these bands may correspond to monomer (675 nm), dimer (694 nm) and tetramer (717 nm) (see 'Discussion'). Time-resolved fluorescence spectra at - 1 9 6 ° C also reflected changes in spectral properties induced by the addition of 0.4% 1-hexanol (Figure 1B). The order of appearance of the emission was the same as in the control (Figure 1A), but the relative intensity of the B866 emission was lower, and the intensity of the intermediate component with the maximum at 825 nm was, in turn, high. The rise time for maximal intensity at 888-nm emission was almost the same as the rise time in the control. These changes indicate that the energy flow to the B866 was suppressed by treatment with 1-hexanol, but the kinetics were not affected. Thus, it was concluded that the flux of energy flow to the baseplate was suppressed by treatment with 1-hexanol. The emission around 825 nm was remarkable, as in the case of the steady-state measurements (Figure 2A). It is known that the 825-nm component has a long lifetime even under physiological conditions (Brune et al. 1987b; Mimuro et al. 1994) and thus is not involved in the main flow of energy (Mimuro et al. 1994). Therefore, it can be reasonably assumed that once the energy is transferred to the 825-nm emission component, it does not come back to the main flow because of a large energy gap for the uphill transfer of energy under this temperature condition. The energy flow to the B866 in the membranes was mediated through the 813-nm emission component.
The effect of 1-hexanol on fluorescence properties of chlorosomes The 1-hexanol also induced a shift of the emission from the baseplate on isolated chlorosomes (Figure 2B). Accompanying an increase in the 1-hexanol concentration, the magnitude of the shift of emission increased; at 0.6% 1-hexanol, the main emission was observed at 826 nm, and the intensity of fluorescence increased in the wavelength region from 760 to 800 nm (Figure 2B,
curve 4). This pattern was essentially the same as that observed in whole cells (Figure 2A). Emissions from intermediary components were also observed around 800 nm on chlorosomes; however, because of a selective excitation of the aggregated forms at 460 nm, the intensities from intermediary components were low. Conversely, on excitation at 415 nm, those intermediary components were preferentially excited and gave rise to a higher intensity (data not shown). We measured the fluorescence excitation spectra on chlorosomes in the absence of 1-hexanol to investigate the origin of individual bands (Figure 2C). When fluorescence was monitored at 825 nm, we found three peaks: BChl c (461 nm), carotenoids (516 nm), and BChl a (609 nm). BChl c and carotenoid peaks indicate the energy transfer from those pigments to BChl a. On the other hand, when monitored at 810 nm, we detected almost the same excitation spectrum, but with a lower contribution from BChl a compared with BChl c. The 813-nm component was assigned to BChl a bound to polypeptide and functioned as an intermediary component in the energy transfer (Figure 1A); however, it was sensitive to treatment with 1-hexanol. Thus, the process of energy transfer to the 813-nm component was affected by treatment with 1-hexanol. Our observation suggests a specific interaction between BChl c in the rod elements and BChl a in the baseplate to stabilize the 813-nm form of BChl a (see 'Discussion'). Conversely, when fluorescence was monitored at 752 nm, a contribution of BChl c and carotenoids was detected, but not of BChl a. Transfer efficiency from carotenoids to BChl c in chlorosomes was estimated to be about 60 to 70%, and the transition moment of their Sz state was almost parallel to the Qy transition of BChl c, that is, the long axis of the rod elements (K. Tuji et al., unpublished).
Building block of BChl c aggregates revealed by treatment with 1-hexanol The absorption spectra of isolated chlorosomes were affected by treatment with 1-hexanol (Figure 3A). After the addition of 1-hexanol, the main absorption band at 742 nm shifted to the blue, and the band coming from 'monomer BChl c' was finally observed at 670 nm. The difference absorption spectra (Figure 3B) indicated the components responsible for the spectral changes. After the addition of 0.2% 1-hexanol, in which the main peak was still observed at 742 nm and the difference in absorbance was only 3% of the total (data not shown), we are readily aware of the decrease
267 . . . .
J
. . . .
r
. . . .
i
. . . .
i
. . . . 1 5
•
•
•
LD ~ ~
~ 600
650
700
750
800
850
(C)
CD 0
i
. . . .
i
. . . .
i
. . . .
i
1
:',
MC
-15 600
.
650
700
.
.
.
.
750
, ,
800
Wavenumber ( 1000cm- 1) 16 15 14 13 . . . .
1
-1 850
0
12
. . . .
(B)
5o
2___
. . . .
600
L
. . . .
i
. . . .
i
. . . .
1
p
. . . .
650 700 750 800 Wavelength (nm)
850
600
650 700 750 800 Wavelength (nm)
850
Figure 3. The l-hexanol effect on the absorption spectra of chlorosomes isolated from C. aurantiacus at 25 °C is shown. (A) Absorption spectra, (B) difference absorption spectra, (C) LD linear dichroism, CD, and MCD spectra of chlorosomes without 1-hexanol and (D) deconvolution of absorption spectrum. In (A), spectra 1 to 3 correspond to a concentration of 1-hexanol of 0, 0.6 and 0.8%, respectively. In (B), the spectra 1 to 3 correspond to the difference in spectra between 0 and 0.2%, between 0 and 0.4%, and between 0.6 and 0.8% 1-hexnol, respectively. In (C), the LD spectrum is given as AAJA. The CD and MCD spectra were drawn on the left-hand side scale, and Abs and LD were drawn on the right-hand side scale. For deconvolution of the absorption spectrum (D), CD and MCD spectra were also studied to determine the band locations and band widths.
in the absorbance at 793 nm and the increase at 813 nm with an isosbetstic point at 803 nm, and of a bleach of a band around 750 nm, which was red-shifted compared with the major component at 742 nm (Figure 3B, curve 1). These observances clearly indicate the presence of very sensitive components in chlorosomes, i.e., the baseplate and the specific component in the rod elements. Changes in the baseplate are consistent with the observation by fluorescence spectra (Figure 2B). This change was also confirmed by CD and MCD spectra (data not shown, see Figure 3C). When the concentration of 1-hexanol increased to 0.4%, the main absorption band at 742 nm preferentially disappeared (data not shown), and the difference in spectrum indicated the appearance of a few bands in the short wavelength region (Figure 3B, curve 2). Those bands were resolved into three bands at 709 nm (tetramer), 686 nm (dimer), and 670 nm (monomer) (Olson and Cox 1991, also see Figure 3D). When the 1-hexanol concentration increased up to 0.6%, the absorption maximum was observed no longer at 742 nm but at 726 nm (Figure 3A, curve 2). In the process of a shift of the 726-nm peak to monomer, the 686-nm and 709-nm components also bleached (Figure 3B, curve 3); however,
we could not detect the appearance of an intermediary component or components. This spectral behavior of components suggests that 1-hexanol induced the disruption of higher aggregates in chlorosomes and that smaller aggregates, such as tetramer or dimer, were formed as intermediary components in the conversion process from the 741-nm component. These dimers and tetramers were found as transient species in the recovery process from monomer to aggregate after the treatment with hexanol (R.P. Cox et al., unpublished). We deconvoluted the absorption spectrum on the basis of two kinds of experimental results: the consistency of the spectra, including CD and MCD (Figure 3C); and the other spectral changes induced by treatment with 1-hexanol (Figures 2B and 3B). We found several components, among which were two main BChl c components at 726 and 741 nm, as indicated previously (Matsuura et al. 1993), and a few minor components at 771,751,709, 686 and 670 nm, which we could not assign to specific components in our previous study (Matsuura et al. 1993). The 771-nm band corresponded to the BChl c component, which is assigned to a component specifically interacting with BChl a to form a funnel to the B806-866 (see 'Discus-
268 sion'). The absorption m a x i m u m o f BChl C771 is very close to that o f free BChl a; however, we could not detect free BChl a in our preparation by the fluorescence spectrum excited at various wavelengths from 360 to 630 nm. The three components, tetramer at 709 nm, dimer at 686 nm, and monomer at 670 nm (Olson and Cox 1991), are probably located in the rod elements. Thus, we were able to clarify the relationship between the absorption component at 25 °C and the fluorescence components at - 196 °C shown in this study: BChl c m o n o m e r (A670 and F675), BChl c dimer (A686 and F694), BChl c tetramer (A709 and F717), BChl c (A726 and corresponding fluorescence was not detected, a major component), BChl c (A741 and F752, a major component), BChl c (A751 and corresponding fluorescence might be F771), BChl c (A771 and F788), BChl a (A793 and F813, a functional baseplate component) and BChl a (A813 and F826, additional baseplate component), BChl a806 (A806 and emission m a x i m u m was unresolved), BChl a866 (A866 and F888) and BChl a888 (absorption m a x i m u m at 888 nm was suggested by fluorescence polarization spectrum and F910) (see also 'Discussion').
Discussion
1-Hexanol effect on the baseplate It is known that 1-hexanol as well as other alcohols induce the dissociation o f BChl c aggregates in the rod elements (Brune et al. 1987a; Matsuura and Olson 1990), giving rise to the spectra corresponding to m o n o m e r at the concentration o f near saturation. According to Wang et al. (1995), molecules o f alcohol that exceed a saturated amount can partition in water and lipid components, including BChl. In the present study, we observed a rather sensitive response of the baseplate to 1-hexanol. After the addition of 0.2% 1-hexanol, a shift of the absorption spectrum of a baseplate component was observed, although the main absorption peak remained the same as in the control. This shift was also confirmed by the fluorescence spectra (Figure 2), and under these conditions the processes of energy transfer were also affected, i.e. the suppression of energy flow to B806-866. Therefore, the effect of 1-hexanol on the baseplate component is more sensitive, reflecting a smaller amount o f such components than those of BChl c. In comparison, 1% of methanol corresponds to approximately 250 mM, but in many cases we have not experienced any damage to sam-
galactolipid
renpmsm
RC
B888
B806-866
Figure 4. A model for the BChl c aggregates in chlorosomes and the antenna system of C aurantiacus is shown. Q in RC refers to a quinone molecule as the second electron acceptor. Squares in the rod elements indicate BChl c aggregates, squares in the B806-866 complex, BChl a and a shaded pair, and a special long wavelength component (B888). BChl c refers to a specific BChl c component interacting with BChl a. In the baseplate, two kinds of BChl a components are drawn; one is square in shape, which stands for the functional BChl a component (A793 and F813), and the other is rounded in shape, indicating nonfunctionalBChl a (A813 and F826) (see text). ples, for example, denaturation or inactivation, at this concentration of methanol. A model f o r an antenna system ofchlorosomes in Chloroflexus In Figure 4, our current model for the antenna system in C. aurantiacus is shown. We propose the specific interaction between a special component of BChl c in the rod elements and BChl a in the baseplate based on the 1 - h e x a n o l effect. In the baseplate, two types of BChl a pool exist, even at physiological temperature (Mimuro et al. 1989, 1994); one is a functional form, and the other is not involved in the energy flow to the membranes (Mimuro et al. 1994). The absorption spectrum (Figure 3D) indicates that the former is a major component. 1-Hexanol induced the transformation from the former to the latter, as indicated by fluorescence (Figures 2 A and 2B) and absorption (Figures 3 A and 3B) spectra. Assuming the same extinction coefficient of BChl a793 and BChl a813, we estimated fractions of the baseplate components: BChl a793 converting to BChl a813 accounted for about 15%, the original BChl a813 accounted for 25% and BChl a793, which was not converted, remained at 60%. This composition might correspond to a relatively lower (approximately 70%) overall transfer efficiency from chlorosomes to RC (Wittmershaus et al. 1988). 1-Hexanol at low concentration may not cause direct damage to the pigment-protein complex; thus, it is reasonable to conclude that the process of energy transfer to the former was impaired by 1-hexanol. A probable reason for that impairment is interruption of the interaction between
269 BChl c and BChl a or spatial separation, which indicates that the functional form of BChl a might be stabilized by interaction with BChl c. A bathochromic shift due to interruption of interaction is, however, not common in the pigment system. Under physiological conditions, two kinds of BChl a793 pool might be equilibrated; however, it is not clear whether or not the energy flow occurs from BChl a793 to BChl a813. Time-resolved fluorescence spectra upon excitation of BChl c did not necessarily give an indication on this process, and the issue may need to be resolved further. Conversely, Savihkin et al. (1994) measured the transient absorption changes on chlorosomes in the fs time range and found a rapid bleach of BChl a upon excitation of BChl c. Based on this observation, they also proposed a new component in which BChl a and BChl c are coupled to each other. This interpretation is in line with our own findings. Two independent analyses yielded essentially the same conclusion regarding the specific interaction between BChl c and BChl a. This consensus indicates a structural heterogeneity in rod elements in addition to the spectral heterogeneity mentioned in our previous study (Matsuura et al. 1993). A component, consisting of a coupled BChl a and BChl c, functions as the energy sink in chlorosomes and as a funnel between the chlorosomes and the B806-866 in the membranes. At the moment, we do not understand the structure of this new component; a possible architecture is a tight coupling between the baseplate BChl a and a special form of BChl c oligomer located in the vicinity of baseplate. This BChl c oligomer might have an absorption maximum that is at a longer wavelength than that of the major component (741 nm); thus, the most likely candidates are BChl c75~ and BChl c771. The nature of the BChl c751 is not clear at this experimental stage. In our previous study (Matsuura et al. 1993; Mimuro et al. 1994), we did not discover any evidence for the presence of this component; however, as indicated by the difference in the absorption spectrum (Figure 3B, curve 1), some part of this component was sensitive to treatment with 1-hexanol. This component was also suggested by the MCD spectrum (data not shown). Thus, it is reasonable to assume that this component is present, but we need further confirmation to identify its chemical nature. The other long wavelength component of BChl c has been known in Chloroflexus chlorosomes (771 nm, Matsuura et al. 1993; Mimuro et al. 1994), and it is assigned to this specific component. How-
ever, the 771-nm component was not affected by the 1-hexanol treatment, as indicated by a difference in the absorption spectrum in the initial stage of transformation (Figure 3B). The special form of BChl c might be stabilized by a membrane component or components in the envelope or formed by binding with a specific peptide in the envelope or baseplate. In the case of Chlorobium chlorosomes, a long wavelength emission of BChl c was observed at 795 nm at room temperature and was insensitive to treatment with 1-hexanol (Hirota et al. 1992). The difference between Chloroflexus and Chlorobium in the molecular species in the baseplate region, including FMO proteins, might be the cause for such a difference. We also noticed that 1-hexanol-induced changes in the spectra of Chlorobium chlorosomes are not necessarily identical to those of Chloroflexus chlorosomes (data not shown). An additional difference exists between this model and the previous one (Feick and Fuller 1984). B888 was also suggested to be present in the B806-866 complex by the difference in rise time of energy transfer. The B806-866 complex, which is comparable to LH 1, interacts with the RC complex, even though the plural peaks are similar to LH 2 in the purple bacteria. In the instance of LH 1, a long wavelength antenna is spectrally known as B896 in Rhodobacter sphaeroides (Kramer et al. 1984) and as B888 in an aerobic photosynthetic bacterium Roseobacter denitrificance (Shimada et al. 1990). The B 888 in the B 806-866 complex in C. aurantiacus corresponds to the characteristics of LH 1. The B888 was also detected by fluorescence polarization spectra (data not shown). As was pointed out by Chiefari et al. (1995), a termination of BChl c aggregates in rod elements should be considered; according to their interpretation, a few molecular species are located at the terminal parts of the rod elements, each of which shows specifi c Raman lines. By deconvolution of absorption spectra, we also suggest the presence of tetramer, dimer, and even monomer in the isolated chlorosomes (Figure 3D); however, at the moment, we do not have data enough for identification of their locations in the rod elements. This point is closely related to the aggregation number of BChl c of major components, which is not yet elucidated, although we have recently proposed a new model for molecular arrangement of BChl c in the rod structure (Mimuro et al., 1995).
270
Acknowledgements T h i s w o r k w a s s u p p o r t e d in part b y g r a n t s - i n - a i d s f r o m the M i n i s t r y o f E d u c a t i o n , S c i e n c e , S p o r t s a n d Culture, J a p a n ( G r a n t No. 0 7 8 3 9 0 1 9 to M M , 06226211 and 0 6 5 5 5 2 5 7 to T N , 0 3 N P 0 3 0 1 to I Y a n d 0 7 8 3 9 0 1 4 to K M ) . M M also is g r a t e f u l for the financial aid g i v e n by the C i b a - G e i g y S c i e n c e F o u n d a t i o n (Japan) for the Promotion of Science.
References Brune DC, Nozawa T and Blankenship RE (1987a) Antenna organization in green photosynthetic bacteria. 1. Oligomeric bacteriochlorophyll c as a model for the 740 nm absorbing bacteriochlorophyli c in Chloroflexus aurantiacus chlorosomes. Biochemistry 26:8644-8652 Brune DC, King GH, Infosio A, Steiner T, Thewalt MLW and Blankenship RE (1987b) Antenna organization in green photosynthetic bacteria. 2. Excitation transfer in detached and membrane-bound chlorosomes from Chloroflexus aurantiacus. Biochemistry 26:8652-8658 Chiefari J, Griebenow K, Griebenow N, Balaban TS, Holzwarth AR and Schaffner K (1995) Models for the pigment organization in the chlorosomes of photosynthetic bacteria: Diastereoselective control of in vitro bacteriochlorophyll cs aggregation. J Phys Chem 99:1357-1365 Feick RG and Fuller RC (1984) Topology of the photosynthetic apparatus of Chloroflexus aurantiacus. Biochemistry 23: 36933700 Gerola PD and Olson JM (1986) A new bacteriochlorophyll aprotein complex associated with chlorosomes of green sulfur bacteria. Biochim Biophys Acta 848:69-76 Hanada S, Hiraishi A, Shimada K and Matsuura K (1995) Isolation of Chloroflexus aurantiacus and related thermophilic phototrophic bacteria from Japanese hot springs using an improved isolation procedure. J Gen Appl Microbio141:119-130 Hildebrandt P, Tamiaki H, Holzwarth AR and Schaffner K (1994) Resonance Raman spectroscopic study of metallochlorin aggregates. Implication for the supramolecular structure in chlorosomal BChl c antenna of green bacteria. J Phys Chem 98:2192-2197 Hirota M, Moriyama T, Shimada K, Miller M, Olson JM and Matsuura K (1992) High degree of organization of bacteriochlorophyll c in chlorosomes-like aggregates spontaneously assembled in aqueous solution. Biochim Biophys Acta 1099:271-274 Holzwarth AR, Griebenow K and Schaffner K (1990) A photosynthetic antenna system which contains a protein-free chromophore aggregate. Z Naturforsch 45c: 203-296 Kramer HJM, Pennoyer L, van Grondelle R, Westerhuis WHJ, Niederman RA and Amesz J (1984) Low temperature optical properties and pigment organization of the B875 light-harvesting bacteriochlorophyll-protens complex of purple photosynthetic bacteria. Biochim Biophys Acta 767:335-344 Krasnovsky AA and Bystrova MI (1980) Self-assembly of chlorophyll aggregated structures. BioSystems 12:181-194 Lutz M and van Brakel G (1988) Ground-state molecular interaction of bacteriochlorophyll c in chlorosomes of green bacteria and in model system: a resonance Raman study. In: Olson JM, Ormerod JG, Amesz J, Stackerbrandt E and Truper HG (eds) Green Pho-
tosynthetic Bacteria, pp 23-34. Plenum Press, New York and London Matsuura K and Olson JM (1990) Reversible conversion of aggregated bacteriochlorophyll c to the monomeric form by 1-hexanol in chlorosomes from Chlorobium and Chloroflexus. Biochim Biophys Acta 1019:233-238 Matsuura K, Hirota M, Shimada K and Mimuro M (1993) Spectral forms and orientation of bacteriochlorophyUs c and a in chlorosomes of the green photosynthetic bacterium Chloroflexus aurantiacus. Photochem Photobiol 57:92-97 Mimuro M, Nozawa T, Tamai N, Shimada K, Yamazaki I, Lin S, Knox RS, Wittmershaus BP, Brune DC and Blankenship RE (1989) Excitation energy flow in chlorosome antennas of green photosynthetic bacteria. J Phys Chem 93:7503-7509 Mimuro M, Matsuura K, Nishimura Y, Shimada K and Yamazaki I (1992) Excitation energy transfer processes in green photosynthetic bacteria: Analysis in picosecond time domain. In: Murata N (ed) Research in Photosynthesis, Vol I, pp 17-24. Kluwer Academic Publishers, Dordrecht, The Netherlands Mimuro M, Nozawa T, Tamai N, Nishimura Y and Yamazaki I (1994) Presence and significance of minor antenna components in the energy transfer sequence of a green photosynthetic bacterium Chloroflexus aurantiacus. FEBS Lett 340:167-172 Mimuro M, Matsuura K, Shimada K, Nishimura Y, Yamazaki I, Kobayashi M, Wang Z-Y, and Nozawa T (1995) Molecular networks and funneling process of energy transfer in green photosynthetic bacteria. In: Mathis P (ed) Photosynthesis: From Light to Biosphere, Vol I, pp 41--46. Kluwer Academic Publishers, Dordreeht, The Netherlands Nozawa T, Kobayashi M, Wang Z-Y, Itoh S, lwaki M, Mimuro M and Satoh K (1995) Magnetic circular dichroism investigation on chromophores in reaction centers of Photosystem I and II of green plant photosynthesis. Spectrochim Acta 51A: 125-134 Olson JM (1980) Chlorophyll organization in green photosynthetic bacteria. Biochim Biophys Acta 594:33-51 Olson JM and Cox RP (1991) Monomers, dimers and tetramers of 4-n-propyl-5-ethyl farnesyl bacteriochlorophyll c in dichloromethane and carbon tetrachloride. Photosynth Res 30: 35-43 Savikhin S, Zhu Y, Lin S, Blankenship RE and Struve WS (1994) Femtosecond spectroscopy of chlorosome antennas from the green photosynthetic bacterium Chloroflexus aurantiacus. J Phys Chem 98:10322-10334 Shimada K, Yamazaki I, Tamai N and Mimuro M (1990) Excitation energy flow in a photosynthetic bacterium lacking B850: Fast energy transfer from B806 to B870 in Erythrobacter sp. strain OCh 114. Biochim Biophys Acta 1016:266-271 van Amerongen H, Vasmel H and van Grondelle R (1988) Linear dichroism of chlorosomes from Chloroflexus aurantiacus in compressed gel and electric fields. Biophys J 54:65-76 van Grondelle R, Sundstrom V, Dekker J and Gillbro T (1994) Energy transfer and trapping in photosynthesis. Biochim Biophys Acta 1065:1-65 Wang Z-Y, Marx G, Umetsu M, Kobayashi M, Mimuro M and Nozawa T (1995) Morphology and spectroscopy of chlorosomes from Chlorobium tepidurn by alcohol treatments. Biochim Biophys Acta 1232:187-196 Wittmershaus BP, Brune DC and Blankenship BE (1988) Energy transfer in Chloroflexus aurantiacus: Effect of temperature and anaerobic conditions. In: Scheer H and Schneider S (eds) Photosynthetic Light Harvesting systems: Structure and Function, pp 543-554. Walter de Gruyter, Berlin
Excitation energy transfer in the green photosynthetic bacterium Chloroflexus aurantiacus: A specific effect of 1-hexanol on the optical properties of baseplate and energy transfer processes M a m o r u M i m u r o 1, Y o s h i n o b u N i s h i m u r a 2, I w a o Y a m a z a k i 2, M a s a y u k i K o b a y a s h i 3, Z h e n g Yu W a n g 3, T s u n e n o r i N o z a w a 3, K e i z o S h i m a d a 4 & K a t s u m i M a t s u u r a 4
1National Institute for Basic Biology, Myodaiji, Okazaki 444, Japan; 2Department of Chemical Process Engineering, Faculty of Engineering, Hokkaido University, Sapporo 060, Japan; 3Department of Biochemical Engineering, Faculty of Engineering, Tohoku University, Sendai 970, Japan; 4Department of Biology, Faculty of Science, Tokyo Metropolitan University, Hachioji 192-03, Japan Received 13 October 1995; accepted in revised form 23 January 1996
Key words: Chloroflexus aurantiacus, chlorophyll, chlorosome, energy transfer, green bacteria, photosynthesis Abstract
The effect of 1-hexanol on spectral properties and the processes of energy transfer of the green gliding photosynthetic bacterium Chloroflexus aurantiacus was investigated with reference to the baseplate region. On addition of 1hexanol to a cell suspension in a concentration of one-fourth saturation, a specific change in the baseplate region was induced: that is, a bleach of the 793-nm component, and an increase in absorption of the 813-nm component. This result was also confirmed by fluorescence spectra of whole cells and isolated chlorosomes. The processes of energy transfer were affected in the overall transfer efficiency but not kinetically, indicating that 1-hexanol suppressed the flux of energy flow from the baseplate to the B806-866 complexes in the cytoplasmic membranes. The fluorescence excitation spectrum suggests a specific site of interaction between bacteriochlorophyll (BChl) c with a maximum at 771 nm in the rod elements and BChl a with a maximum at 793 nm in the baseplate, which is a funnel for a fast transfer of energy to the B806-866 complexes in the membranes. The absorption spectrum of chlorosomes was resolved to components consistently on the basis, including circular dichroism and magnetic circular dichroism spectra; besides two major BChl c forms, bands corresponding to tetramer, dimer, and monomer were also discernible, which are supposed to be intermediary components for a higher order structure. A tentative model for the antenna system of C. aurantiacus is proposed.
Abbreviations: A 6 7 0 - a
component whose absorption maximum is located at 670 nm; (B)Chl(bacterio)chlorophyll; C D - c i r c u l a r dichroism; F 6 7 5 - a component whose emission maximum is located at 675 nm; FMO protein-Fenna-Mathews-Olson protein; LD -linear dichroism; LH-light-harvesting; M C D - m a g n e t i c circular dichroism; PS-photosystem; RC-reaction center
Introduction
An antenna system of the green photosynthetic bacterium Chloroflexus aurantiacus consists of B806-866 complexes in the cytoplasmic membranes and chlorosomes outside of the membranes (Olson 1980). The B806-866 complex directly interacts with the reaction center (RC), as in the case of light-harvesting complex (LH) 1, even though the plural spectral peaks are simi-
lar to LH 2 of purple bacteria. The chlorosome is a main light harvesting apparatus of this bacterium and is connected to the B806-866 complex through a baseplate where the bacteriochlorophyll (BChl) a protein complexes are located. Chlorosomes include highly organized molecular assemblies of BChl c (Krasnovsky et al. 1980; Holzwarth et al. 1990; Mimuro et al. 1992; Hirota et al. 1992). Proteins are not directly involved in the molecular organization of BChl c; instead, a
264 main force in the stabilization of the functional form of BChl c is an intermolecular interaction between BChl c molecules; a central Mg atom, a hydroxyl group in the C31 position, and a keto carbonyl group in the C 13 l position play an essential role for the self assemblies (Lutz and van Brakel 1988; Hildebrandt et al. 1994; Chiefari et al. 1995). The BChl c aggregates have their Qy transition moments almost parallel to the long axis of chlorosomes (van Amerongen et al. 1988; Matsuura et al. 1993; Mimuro et al. 1994), irrespective of the spectral heterogeneity of the rod elements. Treatment with alcohol induces 'monomerization' of BChl c aggregates, which is reflected by a large blue shift of the BChl c Qy transition (Brune et al. 1987a; Matsuura and Olson 1990). In the case of 1-hexanol, a saturated amount (59 mM or 0.74% at 25 °C) of alcohol induced a complete conversion of the spectral forms, and this process was shown to be fully reversible (Matsuura and Olson 1990). By slowly diluting the alcohol content from the saturated concentration, the original optical properties were recovered, but when the concentration was diluted quickly, an intermediate form or forms appeared and further red shift was no longer observed. Accompanying these spectral changes was a change in the morphology of chlorosomes (Wang et al. 1995). Based on these phenomena, we can use the treatment with 1-hexanol to monitor the aggregation behavior of the building blocks, such as monomers and oligomers Of BChl c. The spectral change is attributed to the binding of alcohol molecules to the central Mg atoms, which consequently disrupts the intermolecular interaction among BChl c molecules. Conversely, BChl a in the baseplate is known to be insensitive to 1-hexanol treatment, presumably because of a tight binding of pigments to polypeptides. Two types of the BChl a pool exist in the baseplate, as shown by the time-resolved fluorescence spectra of whole cells (Brune et al. 1987b; Mimuro et al. 1989, 1994). Several reports have been published regarding the effect of alcohol, including 1-hexanol, on isolated chlorosomes (Brune et al. 1987a; Matsuura et al. 1993; Mimuro et al. 1994). However, the effects of 1-hexanol on whole cells are scarcely known, although the potential effect is similar to the effect on chlorosomes. We therefore investigated the effect of the 1-hexanol treatment on the spectral properties and processes of energy transfer in whole cells of C. aurantiacus. We found a selective effect of 1-hexanol on the baseplate component, which is located within the main energy flow to the RC. We propose a specific interaction between
BChl c and BChl a in the funneling site to B806-866 in the membranes. Materials and methods Chloroflexus aurantiacus (OK-70) was grown photoheterotrophically at 55 °C in the medium described by Hanada et al. (1995). Cells at the late log-growth phase were harvested and chlorosomes were isolated by the method developed by Gerola and Olson (1986), using 2 M of chaotropic anion (NaSCN) and sucrose density gradient centrifugation. The absorption spectrum was measured with a Hitachi 330 spectrophotometer, the fluorescence spectrum was measured with a Hitachi 850 spectrofluorometer, and the circular dichroism (CD) spectrum was measured with a Jasco J-720 spectropolarimeter. The magnetic circular dichroism (MCD) was measured with a Jasco J-720 polarimeter equipped with an electromagnet with a field strength of 1.35 T (Nozawa et al. 1995). The time-resolved fluorescence spectra were obtained by a time-correlated single-photon counting method (Mimuro et al. 1989) using an S-1 type photomultiplier. The light source was a Ti-Sapphire laser (720 nm) with a pulse width of 200 fs (typical), the repetition rate was 4 MHz, time resolution was greater than 3 ps and spectral resolution was less than 2 nm. All of the spectral data was transferred to a microcomputer and numerically analyzed, i.e. base-line correction, difference spectrum, correction for spectral sensitivity of the fluorometer, deconvolution of spectra and decay analysis with the assumption of an exponential decay function (Mimuro et al. 1989, 1994). Results Overall energy transfer in whole cells
The overall processes of energy transfer in C. aurantiacus at - 1 9 6 °C are shown by time-resolved fluorescence spectra (Figure 1A). When BChl c was excited at 720 nm, a sequential energy flow in the order of BChl c (752 nm), baseplate BChl a (810 nm), and the B866 in the membranes (888 nm) was clearly resolved. The emission of B866 was scarcely detected at 0 ps, but within 50 ps, the intensity reached the maximum. These spectral changes were identical to previous reports (Mimuro et al. 1989, 1994); however, because of a better time-resolution, we observed a preferential emission from BChl c at the initial phase
265 500C
(A)
.
.
.
.
i
.
.
.
.
i
.
.
.
.
i
.
.
.
.
(A) Whole cells 4OO(
300( 158.4
8
105.6
200C
100C .
.
.
.
65( Wavelength (nrn)
.
'
'
700 .
.
.
.
i
'
.
.
.
.
750 .
.
.
.
i
.
.
r
800 .
.
.
.
i
850 .
.
(B) Chlorosomes
.
.
~3
200C
(B) 150C
s=
,-
I ~ A
(x20)
1+ 2 50(~
650
t~
700
........................ 750
800
850
900
950
Wavelength (nm)
. (C) A ~
750
~)']
\~
800 ":/db
850 ....
825 . . . . . . . . . . 752
Figure 1. Time-resolvedfluorescence spectra of whole cells of C. aurantiacus at -196 °C. (A) control, and (B) with 0.4% 1-hexanol. The excitation wavelength was 720 nm. For the low-temperature spectroscopy, 15% of poly(ethylene glycol) 4000 was added to obtain homogenousice.
of excitation and a clear rise of the BChl a emission in the membranes. The emission from the baseplate was located at 810 nm, and its intensity was low throughout the transfer process. This fact indicates that the energy flow between the baseplate and B806 in the membranes is not the rate-limiting step. By assuming exponential decay, we estimated kinetic parameters on the decay curves at seven particular wavelengths from 750 to 910 nm. A fast-decaying component or components (in the range of 100 fs, Savikhin et al. 1994) was not clearly resolved due to a limit of the time-resolution; however, we found clear rise times in the acceptors. Those were 16 ps for the baseplate BChl a component (at 810 nm), 28 ps for B866 (at 888 nm) and 50 ps at 910 nm. The difference in the rise times between 888 and 910 nm clearly indicates that a long wavelength antenna is expected in the B866 pool, as similarly found with purple bacteria (Kramer et al. 1984). The 50-ps rise component corresponds to the overall energy flow from BChl c to the RC, and it was almost identical to the result obtained with purple bacteria (van Grondelle et al. 1994). The rise time of B866, i.e. 28 ps, was also
400. . . . . . . .
iO() . . . . . . . . . .600 . . . . . . . . . . . . . 700 Wavelength (nm)
Figure 2. The effect of 1-bexanol on the steady-state fluorescence
spectra of whole ceils (A) and of isolated chlorosomes (B) of C. aurantiacus at -196 °C is shown. The excitation wavelengthwas 460 nm for both cases, and because of a low sensitivitybeyond 850 nm, the corrected spectra were shownup to this wavelength. In (A), the full line shows the spectrum in the absence of 1-hexanol, the broken line, with 0.4% l-hexanol. Spectra were normalizedat 752 nm. In (B), numbers from 1 to 4 correspond to the concentrations of 1-hexanol of 0, 0.2, 0.4 and 0.6%, respectively. A magnified spectrum (20 fold) was shown for 0.6% of 1-hexanol. Spectra are shown after normalizationat 752 nm. (C) Excitation spectra were measured on isolated chlorosomes at -196 °C in the absence of 1-hexanol. Spectra are shown after normalizationat 460 nm. obtained by the time-resolved fluorescence anisotropy spectra measured on cells aligned in a polyacrylamide gel (data not shown).
The effect o f 1-hexanol on the sequence o f energy transfer and on kinetics The fluorescence properties of whole cells were strongly affected by 1-hexanol as shown in the steady-state spectrum at - 196 °C (Figure 2A). Under a low concen-
266 tration of 1-hexanol (0.4%) in which the main emission was still observed at 752 nm as in the control, the emission maximum of the baseplate was shifted from 815 to 824 nm (Figure 2A, curve 2). The difference spectrum showed a decrease in the intensity at 813 nm and an increase at 826 nm (data not shown), indicating the presence of plural components in the baseplate region. In addition, two fluorescence bands induced by the 1-hexanol treatment were discernible at 771 and 788 nm. Furthermore, a few emission bands were detected around 700 nm; these bands may correspond to monomer (675 nm), dimer (694 nm) and tetramer (717 nm) (see 'Discussion'). Time-resolved fluorescence spectra at - 1 9 6 ° C also reflected changes in spectral properties induced by the addition of 0.4% 1-hexanol (Figure 1B). The order of appearance of the emission was the same as in the control (Figure 1A), but the relative intensity of the B866 emission was lower, and the intensity of the intermediate component with the maximum at 825 nm was, in turn, high. The rise time for maximal intensity at 888-nm emission was almost the same as the rise time in the control. These changes indicate that the energy flow to the B866 was suppressed by treatment with 1-hexanol, but the kinetics were not affected. Thus, it was concluded that the flux of energy flow to the baseplate was suppressed by treatment with 1-hexanol. The emission around 825 nm was remarkable, as in the case of the steady-state measurements (Figure 2A). It is known that the 825-nm component has a long lifetime even under physiological conditions (Brune et al. 1987b; Mimuro et al. 1994) and thus is not involved in the main flow of energy (Mimuro et al. 1994). Therefore, it can be reasonably assumed that once the energy is transferred to the 825-nm emission component, it does not come back to the main flow because of a large energy gap for the uphill transfer of energy under this temperature condition. The energy flow to the B866 in the membranes was mediated through the 813-nm emission component.
The effect of 1-hexanol on fluorescence properties of chlorosomes The 1-hexanol also induced a shift of the emission from the baseplate on isolated chlorosomes (Figure 2B). Accompanying an increase in the 1-hexanol concentration, the magnitude of the shift of emission increased; at 0.6% 1-hexanol, the main emission was observed at 826 nm, and the intensity of fluorescence increased in the wavelength region from 760 to 800 nm (Figure 2B,
curve 4). This pattern was essentially the same as that observed in whole cells (Figure 2A). Emissions from intermediary components were also observed around 800 nm on chlorosomes; however, because of a selective excitation of the aggregated forms at 460 nm, the intensities from intermediary components were low. Conversely, on excitation at 415 nm, those intermediary components were preferentially excited and gave rise to a higher intensity (data not shown). We measured the fluorescence excitation spectra on chlorosomes in the absence of 1-hexanol to investigate the origin of individual bands (Figure 2C). When fluorescence was monitored at 825 nm, we found three peaks: BChl c (461 nm), carotenoids (516 nm), and BChl a (609 nm). BChl c and carotenoid peaks indicate the energy transfer from those pigments to BChl a. On the other hand, when monitored at 810 nm, we detected almost the same excitation spectrum, but with a lower contribution from BChl a compared with BChl c. The 813-nm component was assigned to BChl a bound to polypeptide and functioned as an intermediary component in the energy transfer (Figure 1A); however, it was sensitive to treatment with 1-hexanol. Thus, the process of energy transfer to the 813-nm component was affected by treatment with 1-hexanol. Our observation suggests a specific interaction between BChl c in the rod elements and BChl a in the baseplate to stabilize the 813-nm form of BChl a (see 'Discussion'). Conversely, when fluorescence was monitored at 752 nm, a contribution of BChl c and carotenoids was detected, but not of BChl a. Transfer efficiency from carotenoids to BChl c in chlorosomes was estimated to be about 60 to 70%, and the transition moment of their Sz state was almost parallel to the Qy transition of BChl c, that is, the long axis of the rod elements (K. Tuji et al., unpublished).
Building block of BChl c aggregates revealed by treatment with 1-hexanol The absorption spectra of isolated chlorosomes were affected by treatment with 1-hexanol (Figure 3A). After the addition of 1-hexanol, the main absorption band at 742 nm shifted to the blue, and the band coming from 'monomer BChl c' was finally observed at 670 nm. The difference absorption spectra (Figure 3B) indicated the components responsible for the spectral changes. After the addition of 0.2% 1-hexanol, in which the main peak was still observed at 742 nm and the difference in absorbance was only 3% of the total (data not shown), we are readily aware of the decrease
267 . . . .
J
. . . .
r
. . . .
i
. . . .
i
. . . . 1 5
•
•
•
LD ~ ~
~ 600
650
700
750
800
850
(C)
CD 0
i
. . . .
i
. . . .
i
. . . .
i
1
:',
MC
-15 600
.
650
700
.
.
.
.
750
, ,
800
Wavenumber ( 1000cm- 1) 16 15 14 13 . . . .
1
-1 850
0
12
. . . .
(B)
5o
2___
. . . .
600
L
. . . .
i
. . . .
i
. . . .
1
p
. . . .
650 700 750 800 Wavelength (nm)
850
600
650 700 750 800 Wavelength (nm)
850
Figure 3. The l-hexanol effect on the absorption spectra of chlorosomes isolated from C. aurantiacus at 25 °C is shown. (A) Absorption spectra, (B) difference absorption spectra, (C) LD linear dichroism, CD, and MCD spectra of chlorosomes without 1-hexanol and (D) deconvolution of absorption spectrum. In (A), spectra 1 to 3 correspond to a concentration of 1-hexanol of 0, 0.6 and 0.8%, respectively. In (B), the spectra 1 to 3 correspond to the difference in spectra between 0 and 0.2%, between 0 and 0.4%, and between 0.6 and 0.8% 1-hexnol, respectively. In (C), the LD spectrum is given as AAJA. The CD and MCD spectra were drawn on the left-hand side scale, and Abs and LD were drawn on the right-hand side scale. For deconvolution of the absorption spectrum (D), CD and MCD spectra were also studied to determine the band locations and band widths.
in the absorbance at 793 nm and the increase at 813 nm with an isosbetstic point at 803 nm, and of a bleach of a band around 750 nm, which was red-shifted compared with the major component at 742 nm (Figure 3B, curve 1). These observances clearly indicate the presence of very sensitive components in chlorosomes, i.e., the baseplate and the specific component in the rod elements. Changes in the baseplate are consistent with the observation by fluorescence spectra (Figure 2B). This change was also confirmed by CD and MCD spectra (data not shown, see Figure 3C). When the concentration of 1-hexanol increased to 0.4%, the main absorption band at 742 nm preferentially disappeared (data not shown), and the difference in spectrum indicated the appearance of a few bands in the short wavelength region (Figure 3B, curve 2). Those bands were resolved into three bands at 709 nm (tetramer), 686 nm (dimer), and 670 nm (monomer) (Olson and Cox 1991, also see Figure 3D). When the 1-hexanol concentration increased up to 0.6%, the absorption maximum was observed no longer at 742 nm but at 726 nm (Figure 3A, curve 2). In the process of a shift of the 726-nm peak to monomer, the 686-nm and 709-nm components also bleached (Figure 3B, curve 3); however,
we could not detect the appearance of an intermediary component or components. This spectral behavior of components suggests that 1-hexanol induced the disruption of higher aggregates in chlorosomes and that smaller aggregates, such as tetramer or dimer, were formed as intermediary components in the conversion process from the 741-nm component. These dimers and tetramers were found as transient species in the recovery process from monomer to aggregate after the treatment with hexanol (R.P. Cox et al., unpublished). We deconvoluted the absorption spectrum on the basis of two kinds of experimental results: the consistency of the spectra, including CD and MCD (Figure 3C); and the other spectral changes induced by treatment with 1-hexanol (Figures 2B and 3B). We found several components, among which were two main BChl c components at 726 and 741 nm, as indicated previously (Matsuura et al. 1993), and a few minor components at 771,751,709, 686 and 670 nm, which we could not assign to specific components in our previous study (Matsuura et al. 1993). The 771-nm band corresponded to the BChl c component, which is assigned to a component specifically interacting with BChl a to form a funnel to the B806-866 (see 'Discus-
268 sion'). The absorption m a x i m u m o f BChl C771 is very close to that o f free BChl a; however, we could not detect free BChl a in our preparation by the fluorescence spectrum excited at various wavelengths from 360 to 630 nm. The three components, tetramer at 709 nm, dimer at 686 nm, and monomer at 670 nm (Olson and Cox 1991), are probably located in the rod elements. Thus, we were able to clarify the relationship between the absorption component at 25 °C and the fluorescence components at - 196 °C shown in this study: BChl c m o n o m e r (A670 and F675), BChl c dimer (A686 and F694), BChl c tetramer (A709 and F717), BChl c (A726 and corresponding fluorescence was not detected, a major component), BChl c (A741 and F752, a major component), BChl c (A751 and corresponding fluorescence might be F771), BChl c (A771 and F788), BChl a (A793 and F813, a functional baseplate component) and BChl a (A813 and F826, additional baseplate component), BChl a806 (A806 and emission m a x i m u m was unresolved), BChl a866 (A866 and F888) and BChl a888 (absorption m a x i m u m at 888 nm was suggested by fluorescence polarization spectrum and F910) (see also 'Discussion').
Discussion
1-Hexanol effect on the baseplate It is known that 1-hexanol as well as other alcohols induce the dissociation o f BChl c aggregates in the rod elements (Brune et al. 1987a; Matsuura and Olson 1990), giving rise to the spectra corresponding to m o n o m e r at the concentration o f near saturation. According to Wang et al. (1995), molecules o f alcohol that exceed a saturated amount can partition in water and lipid components, including BChl. In the present study, we observed a rather sensitive response of the baseplate to 1-hexanol. After the addition of 0.2% 1-hexanol, a shift of the absorption spectrum of a baseplate component was observed, although the main absorption peak remained the same as in the control. This shift was also confirmed by the fluorescence spectra (Figure 2), and under these conditions the processes of energy transfer were also affected, i.e. the suppression of energy flow to B806-866. Therefore, the effect of 1-hexanol on the baseplate component is more sensitive, reflecting a smaller amount o f such components than those of BChl c. In comparison, 1% of methanol corresponds to approximately 250 mM, but in many cases we have not experienced any damage to sam-
galactolipid
renpmsm
RC
B888
B806-866
Figure 4. A model for the BChl c aggregates in chlorosomes and the antenna system of C aurantiacus is shown. Q in RC refers to a quinone molecule as the second electron acceptor. Squares in the rod elements indicate BChl c aggregates, squares in the B806-866 complex, BChl a and a shaded pair, and a special long wavelength component (B888). BChl c refers to a specific BChl c component interacting with BChl a. In the baseplate, two kinds of BChl a components are drawn; one is square in shape, which stands for the functional BChl a component (A793 and F813), and the other is rounded in shape, indicating nonfunctionalBChl a (A813 and F826) (see text). ples, for example, denaturation or inactivation, at this concentration of methanol. A model f o r an antenna system ofchlorosomes in Chloroflexus In Figure 4, our current model for the antenna system in C. aurantiacus is shown. We propose the specific interaction between a special component of BChl c in the rod elements and BChl a in the baseplate based on the 1 - h e x a n o l effect. In the baseplate, two types of BChl a pool exist, even at physiological temperature (Mimuro et al. 1989, 1994); one is a functional form, and the other is not involved in the energy flow to the membranes (Mimuro et al. 1994). The absorption spectrum (Figure 3D) indicates that the former is a major component. 1-Hexanol induced the transformation from the former to the latter, as indicated by fluorescence (Figures 2 A and 2B) and absorption (Figures 3 A and 3B) spectra. Assuming the same extinction coefficient of BChl a793 and BChl a813, we estimated fractions of the baseplate components: BChl a793 converting to BChl a813 accounted for about 15%, the original BChl a813 accounted for 25% and BChl a793, which was not converted, remained at 60%. This composition might correspond to a relatively lower (approximately 70%) overall transfer efficiency from chlorosomes to RC (Wittmershaus et al. 1988). 1-Hexanol at low concentration may not cause direct damage to the pigment-protein complex; thus, it is reasonable to conclude that the process of energy transfer to the former was impaired by 1-hexanol. A probable reason for that impairment is interruption of the interaction between
269 BChl c and BChl a or spatial separation, which indicates that the functional form of BChl a might be stabilized by interaction with BChl c. A bathochromic shift due to interruption of interaction is, however, not common in the pigment system. Under physiological conditions, two kinds of BChl a793 pool might be equilibrated; however, it is not clear whether or not the energy flow occurs from BChl a793 to BChl a813. Time-resolved fluorescence spectra upon excitation of BChl c did not necessarily give an indication on this process, and the issue may need to be resolved further. Conversely, Savihkin et al. (1994) measured the transient absorption changes on chlorosomes in the fs time range and found a rapid bleach of BChl a upon excitation of BChl c. Based on this observation, they also proposed a new component in which BChl a and BChl c are coupled to each other. This interpretation is in line with our own findings. Two independent analyses yielded essentially the same conclusion regarding the specific interaction between BChl c and BChl a. This consensus indicates a structural heterogeneity in rod elements in addition to the spectral heterogeneity mentioned in our previous study (Matsuura et al. 1993). A component, consisting of a coupled BChl a and BChl c, functions as the energy sink in chlorosomes and as a funnel between the chlorosomes and the B806-866 in the membranes. At the moment, we do not understand the structure of this new component; a possible architecture is a tight coupling between the baseplate BChl a and a special form of BChl c oligomer located in the vicinity of baseplate. This BChl c oligomer might have an absorption maximum that is at a longer wavelength than that of the major component (741 nm); thus, the most likely candidates are BChl c75~ and BChl c771. The nature of the BChl c751 is not clear at this experimental stage. In our previous study (Matsuura et al. 1993; Mimuro et al. 1994), we did not discover any evidence for the presence of this component; however, as indicated by the difference in the absorption spectrum (Figure 3B, curve 1), some part of this component was sensitive to treatment with 1-hexanol. This component was also suggested by the MCD spectrum (data not shown). Thus, it is reasonable to assume that this component is present, but we need further confirmation to identify its chemical nature. The other long wavelength component of BChl c has been known in Chloroflexus chlorosomes (771 nm, Matsuura et al. 1993; Mimuro et al. 1994), and it is assigned to this specific component. How-
ever, the 771-nm component was not affected by the 1-hexanol treatment, as indicated by a difference in the absorption spectrum in the initial stage of transformation (Figure 3B). The special form of BChl c might be stabilized by a membrane component or components in the envelope or formed by binding with a specific peptide in the envelope or baseplate. In the case of Chlorobium chlorosomes, a long wavelength emission of BChl c was observed at 795 nm at room temperature and was insensitive to treatment with 1-hexanol (Hirota et al. 1992). The difference between Chloroflexus and Chlorobium in the molecular species in the baseplate region, including FMO proteins, might be the cause for such a difference. We also noticed that 1-hexanol-induced changes in the spectra of Chlorobium chlorosomes are not necessarily identical to those of Chloroflexus chlorosomes (data not shown). An additional difference exists between this model and the previous one (Feick and Fuller 1984). B888 was also suggested to be present in the B806-866 complex by the difference in rise time of energy transfer. The B806-866 complex, which is comparable to LH 1, interacts with the RC complex, even though the plural peaks are similar to LH 2 in the purple bacteria. In the instance of LH 1, a long wavelength antenna is spectrally known as B896 in Rhodobacter sphaeroides (Kramer et al. 1984) and as B888 in an aerobic photosynthetic bacterium Roseobacter denitrificance (Shimada et al. 1990). The B 888 in the B 806-866 complex in C. aurantiacus corresponds to the characteristics of LH 1. The B888 was also detected by fluorescence polarization spectra (data not shown). As was pointed out by Chiefari et al. (1995), a termination of BChl c aggregates in rod elements should be considered; according to their interpretation, a few molecular species are located at the terminal parts of the rod elements, each of which shows specifi c Raman lines. By deconvolution of absorption spectra, we also suggest the presence of tetramer, dimer, and even monomer in the isolated chlorosomes (Figure 3D); however, at the moment, we do not have data enough for identification of their locations in the rod elements. This point is closely related to the aggregation number of BChl c of major components, which is not yet elucidated, although we have recently proposed a new model for molecular arrangement of BChl c in the rod structure (Mimuro et al., 1995).
270
Acknowledgements T h i s w o r k w a s s u p p o r t e d in part b y g r a n t s - i n - a i d s f r o m the M i n i s t r y o f E d u c a t i o n , S c i e n c e , S p o r t s a n d Culture, J a p a n ( G r a n t No. 0 7 8 3 9 0 1 9 to M M , 06226211 and 0 6 5 5 5 2 5 7 to T N , 0 3 N P 0 3 0 1 to I Y a n d 0 7 8 3 9 0 1 4 to K M ) . M M also is g r a t e f u l for the financial aid g i v e n by the C i b a - G e i g y S c i e n c e F o u n d a t i o n (Japan) for the Promotion of Science.
References Brune DC, Nozawa T and Blankenship RE (1987a) Antenna organization in green photosynthetic bacteria. 1. Oligomeric bacteriochlorophyll c as a model for the 740 nm absorbing bacteriochlorophyli c in Chloroflexus aurantiacus chlorosomes. Biochemistry 26:8644-8652 Brune DC, King GH, Infosio A, Steiner T, Thewalt MLW and Blankenship RE (1987b) Antenna organization in green photosynthetic bacteria. 2. Excitation transfer in detached and membrane-bound chlorosomes from Chloroflexus aurantiacus. Biochemistry 26:8652-8658 Chiefari J, Griebenow K, Griebenow N, Balaban TS, Holzwarth AR and Schaffner K (1995) Models for the pigment organization in the chlorosomes of photosynthetic bacteria: Diastereoselective control of in vitro bacteriochlorophyll cs aggregation. J Phys Chem 99:1357-1365 Feick RG and Fuller RC (1984) Topology of the photosynthetic apparatus of Chloroflexus aurantiacus. Biochemistry 23: 36933700 Gerola PD and Olson JM (1986) A new bacteriochlorophyll aprotein complex associated with chlorosomes of green sulfur bacteria. Biochim Biophys Acta 848:69-76 Hanada S, Hiraishi A, Shimada K and Matsuura K (1995) Isolation of Chloroflexus aurantiacus and related thermophilic phototrophic bacteria from Japanese hot springs using an improved isolation procedure. J Gen Appl Microbio141:119-130 Hildebrandt P, Tamiaki H, Holzwarth AR and Schaffner K (1994) Resonance Raman spectroscopic study of metallochlorin aggregates. Implication for the supramolecular structure in chlorosomal BChl c antenna of green bacteria. J Phys Chem 98:2192-2197 Hirota M, Moriyama T, Shimada K, Miller M, Olson JM and Matsuura K (1992) High degree of organization of bacteriochlorophyll c in chlorosomes-like aggregates spontaneously assembled in aqueous solution. Biochim Biophys Acta 1099:271-274 Holzwarth AR, Griebenow K and Schaffner K (1990) A photosynthetic antenna system which contains a protein-free chromophore aggregate. Z Naturforsch 45c: 203-296 Kramer HJM, Pennoyer L, van Grondelle R, Westerhuis WHJ, Niederman RA and Amesz J (1984) Low temperature optical properties and pigment organization of the B875 light-harvesting bacteriochlorophyll-protens complex of purple photosynthetic bacteria. Biochim Biophys Acta 767:335-344 Krasnovsky AA and Bystrova MI (1980) Self-assembly of chlorophyll aggregated structures. BioSystems 12:181-194 Lutz M and van Brakel G (1988) Ground-state molecular interaction of bacteriochlorophyll c in chlorosomes of green bacteria and in model system: a resonance Raman study. In: Olson JM, Ormerod JG, Amesz J, Stackerbrandt E and Truper HG (eds) Green Pho-
tosynthetic Bacteria, pp 23-34. Plenum Press, New York and London Matsuura K and Olson JM (1990) Reversible conversion of aggregated bacteriochlorophyll c to the monomeric form by 1-hexanol in chlorosomes from Chlorobium and Chloroflexus. Biochim Biophys Acta 1019:233-238 Matsuura K, Hirota M, Shimada K and Mimuro M (1993) Spectral forms and orientation of bacteriochlorophyUs c and a in chlorosomes of the green photosynthetic bacterium Chloroflexus aurantiacus. Photochem Photobiol 57:92-97 Mimuro M, Nozawa T, Tamai N, Shimada K, Yamazaki I, Lin S, Knox RS, Wittmershaus BP, Brune DC and Blankenship RE (1989) Excitation energy flow in chlorosome antennas of green photosynthetic bacteria. J Phys Chem 93:7503-7509 Mimuro M, Matsuura K, Nishimura Y, Shimada K and Yamazaki I (1992) Excitation energy transfer processes in green photosynthetic bacteria: Analysis in picosecond time domain. In: Murata N (ed) Research in Photosynthesis, Vol I, pp 17-24. Kluwer Academic Publishers, Dordrecht, The Netherlands Mimuro M, Nozawa T, Tamai N, Nishimura Y and Yamazaki I (1994) Presence and significance of minor antenna components in the energy transfer sequence of a green photosynthetic bacterium Chloroflexus aurantiacus. FEBS Lett 340:167-172 Mimuro M, Matsuura K, Shimada K, Nishimura Y, Yamazaki I, Kobayashi M, Wang Z-Y, and Nozawa T (1995) Molecular networks and funneling process of energy transfer in green photosynthetic bacteria. In: Mathis P (ed) Photosynthesis: From Light to Biosphere, Vol I, pp 41--46. Kluwer Academic Publishers, Dordreeht, The Netherlands Nozawa T, Kobayashi M, Wang Z-Y, Itoh S, lwaki M, Mimuro M and Satoh K (1995) Magnetic circular dichroism investigation on chromophores in reaction centers of Photosystem I and II of green plant photosynthesis. Spectrochim Acta 51A: 125-134 Olson JM (1980) Chlorophyll organization in green photosynthetic bacteria. Biochim Biophys Acta 594:33-51 Olson JM and Cox RP (1991) Monomers, dimers and tetramers of 4-n-propyl-5-ethyl farnesyl bacteriochlorophyll c in dichloromethane and carbon tetrachloride. Photosynth Res 30: 35-43 Savikhin S, Zhu Y, Lin S, Blankenship RE and Struve WS (1994) Femtosecond spectroscopy of chlorosome antennas from the green photosynthetic bacterium Chloroflexus aurantiacus. J Phys Chem 98:10322-10334 Shimada K, Yamazaki I, Tamai N and Mimuro M (1990) Excitation energy flow in a photosynthetic bacterium lacking B850: Fast energy transfer from B806 to B870 in Erythrobacter sp. strain OCh 114. Biochim Biophys Acta 1016:266-271 van Amerongen H, Vasmel H and van Grondelle R (1988) Linear dichroism of chlorosomes from Chloroflexus aurantiacus in compressed gel and electric fields. Biophys J 54:65-76 van Grondelle R, Sundstrom V, Dekker J and Gillbro T (1994) Energy transfer and trapping in photosynthesis. Biochim Biophys Acta 1065:1-65 Wang Z-Y, Marx G, Umetsu M, Kobayashi M, Mimuro M and Nozawa T (1995) Morphology and spectroscopy of chlorosomes from Chlorobium tepidurn by alcohol treatments. Biochim Biophys Acta 1232:187-196 Wittmershaus BP, Brune DC and Blankenship BE (1988) Energy transfer in Chloroflexus aurantiacus: Effect of temperature and anaerobic conditions. In: Scheer H and Schneider S (eds) Photosynthetic Light Harvesting systems: Structure and Function, pp 543-554. Walter de Gruyter, Berlin
