Phowsynthesis Research 48: 385-393, 1996. ~) 1996 Kluwer Academic Publishers. Printed in the Netherlands.
Regular paper
Manipulation of the baeteriochlorophyll c homolog distribution in the green sulfur bacterium Chlorobium tepidum D o r t e B. S t e e n s g a a r d , R a y m o n d P. C o x & M e t t e Miller*
Institute of Biochemistry, Odense University, Campusvej 55, DK-5230 Odense M, Denmark; *Author for correspondence and~or reprints Received 25 January 1996; accepted in revised form 22 March 1996
Key words: bacteriochlorophyll homologs,/~-cyclodextrin, chlorosome, green bacteria
Abstract
We have shown that the green sulfur bacterium Chlorobium tepidum can be grown in batch culture supplemented with potentially toxic fatty alcohols without a major effect on the growth rate if the concentration of the alcohols is kept low either by programmed addition or by adding the alcohol as an inclusion complex with/~-cyclodextrin. HPLC and GC analysis of pigment extracts from the supplemented ceils showed that the fatty alcohols were incorporated into bacteriochlorophyll c as the esterifying alcohol. It was possible to change up to 43% of the naturally occurring farnesyl ester of bacteriochlorophyll c with the added alcohol. This change in the homolog composition had no effect on the spectral properties of the cells when farnesol was partially replaced by stearol, phytol or geranylgeraniol. However, with dodecanol we obtained a blue-shift of 6 nm of the Qy band of the bacteriochlorophyll c and a concomitant change in the fluorescence emission was observed. The possible significance of these findings is discussed in the light of current ideas about bacteriochlorophyll organization in the chlorosomes.
Abbreviations: /3-CD- ~-cyclodextrin; BChl- bacteriochlorophyll; BChl cH- bacteriochlorophyllide c; [E,M] BChl CF-8-ethyi, 12-methyl, farnesyl BChl c; [E,E] BChl CF-8-ethyl, 12-ethyl, farnesyl BChl c; [P,E] BChl CF-8-propyl, 12-ethyl, farnesyl BChl c; [I,E] BChl CF-8-isobutyl, 12-ethyl, farnesyl BChl c; Car-carotenoids Introduction
The light-harvesting pigments of the green bacteria are found in chlorosomes, which are cigar-formed structures found appressed to the cytoplasmic membrane which contains the reaction centers. These unique antenna structures enable extremely efficient light harvesting and energy transfer, allowing green sulfur bacteria to grow in very dim light (Amesz 1991). Depending on the species and strain involved, the main light harvesting pigment of green bacteria is either bacteriochlorophyll c, d, or e. A unique feature of the chlorosome is the low amount of protein associated with the antenna system compared to other known light-harvesting systems. It is believed that the bacteriochlorophylls are organized in self-aggregated rod structures, dominated by
pigment-pigment interactions (reviewed by Blankenship et al. 1995). Another characteristic feature is the presence of homologous forms of the bacteriochlorophyll which differ in molecular structure but have the same absorption properties in the monomeric form (Scheer 1991). The homologs ofchlorosome chlorophylls can vary in several ways. Homologs of BChl c found in the filamentous green bacterium Chloroflexus aurantiacus have the same substituents to the tetrapyrrole ring (8-ethyl, 12-methyl) but differ in the esterifying alcohols, which can be linear (octadecanol, hexadecanol or 9-octadecenol) or isoprenoid (phytol or geranylgeraniol) (Risch et al. 1979; Larsen et al. 1994). In green sulfur bacteria containing BChl c the substituents to the tetrapyrrole ring can be ethyl, n-propyl or iso-butyl at the 8-position and methyl or ethyl at the 12-position
386 (Smith et al. 1983), and the main esterifying alcohol is farnesol. However, in old batch cultures of Chlorobium limicola, up to 10% of the esterifying alcohols of the BChl c molecules were geranylgeraniol, tetrahydrogeranylgeraniol, phytol, cis-9-hexadecen-l-ol and 4-undecyl-2-furanmethanoi (Caple et al. 1978). In Cb. vibrioforme containing both BChl c and BChl d, 10% of the esterifying alcohols were decenol and phytol and in the BChl e containing Cb. phaeovibrioides 20% of the alcohols were cetoi, 4-undecyl-2-furanmethanol and oleyl alcohol (Otte et al. 1993). The possible biological significance of the BChl homolog distribution in the Chlorobiaceae and Chloroflexacae remains an open question. It has been shown that the culture conditions, especially the light regime during growth, influence the homolog distribution in green sulfur bacteria (Huster and Smith 1990). In 4 strains of green bacteria, a decreased light intensity was found to lead to an increase of the more alkylated BChl homologs and a concomitant red-shift of the absorption maximum of the Qy band (Borrego and GarciaGil 1995). The illumination intensity also affected the homolog distribution of BChl c in Cf. aurantiacus grown in turbidostat culture under light-limiting conditions, although no clear pattern of changes was observed (Larsen et al. 1994). Furthermore, it has been shown that Cf. aurantiacus can take up and incorporate exogenous long-chain alcohols into BChl c. Both naturally-occurring and novel BChl c homologs were formed without affecting the spectral properties of the chlorosome (Larsen et al. 1995). The green sulfur bacteria seem more fastidious in their choice of esterifying alcohol than Chloroflexus. The work described here was carried out to investigate whether, in spite of this tendency, it is possible to manipulate the composition of BChl c homologs in the green sulfur bacterium, Chlorobium tepidum. We demonstrate that Cb. tepidum can be grown in medium supplemented with long chain fatty alcohols that are potentially toxic and that/~-cyclodextrin is useful for increasing the bioavailability of the fatty alcohols. Our results show that Cb. tepidum does indeed incorporate exogenous long-chained alcohols of different kinds and that, in contrast to the case with Chloroflexus, this does lead to observable changes in the organization of BChl c in the chlorosome.
Methods
Growth conditions Chlorobium tepidum ATCC 49652 was grown at 45 °C in the medium described by Wahlund et al. (1991). The inoculum was obtained from a light-limited continuous culture grown at 45 °C with a dilution rate 0.9 d - l . The culture was illuminated with incandescent light bulbs giving a photon flux density of 20 #mol m -2 s- l measured at the surface of the bioreactor. Growth in the presence of alcohols was carried out in 70 ml flat culture bottles. Light was provided by a Philips TLD type 29 fluorescent tube (18 W) giving a photon flux density of 20 #mol m -2 s - l at the surface of the culture bottle. In all experiments cultures were inoculated with cells from the continuous culture to give an initial concentration of 3.0 nmol BChl c ml- I. Three different protocols were used for addition of fatty alcohols. For initial experiments, 4.5, 45 or 450/zM of either phytol, geranylgeraniol, stearol or decanol or 37/zM of dodecanol were added to the growth medium before inoculation. For programmed addition, phytol, geranylgeraniol and octadecanol were added from stock emulsions containing 1.5 mM alcohol in the growth medium. Alcohol emulsions were added every 7-9 h for 3 days as follows: 1, 2, 3, 4, 5, 5, 5, 5 and 5 ml. The bottles, containing a glass bead to promote mixing, were shaken vigorously after every alcohol addition. For growth with dodecanol as an inclusion complex with/3-cyclodextrin, the alcohol and fl-CD were added the growth medium before inoculation in the molar ratio 1 alcohol: 4/3-CD, corresponding to 0.94 #M alcohol and 3.7 #M/~-CD. Growth was followed by absorbance measurements at the wavelength of the Qy band of BChl c on a Shimadzu UV-160A spectrophotometer. Four days after inoculation the cells were harvested by centrifugation for 10 min at 24000 x g and stored at - 5 0 ° C . Pigment extraction and analysis by HPLC Pigments were extracted from cell pellets in 50 volumes of acetone:methanol (7:2 v/v) by sonication for 60 s. After centrifugation at 20000 x g for 5 min, extractions were repeated until the solvent was colorless. The combined extracts were dried under a stream of N2 and traces of water were removed under reduced pressure. Dried extracts were stored under N2 at - 2 0 °C until use.
387 HPLC separation was achieved on a 300 x 3.9 mm (i.d.) octadecylsilyl bonded amorphous silica column (Nova-Pak C18, 60 lk pore size, 4 #m particles) with a precolumn (Nova-Pak C18) both from Waters Millipore Corporation, Milford, MA, USA. The HPLC protocol was a modification of that used by Borrego and Garcia-Gil (1994). Solvent A was composed of 25% H20, 33% acetonitrile and 42% methanol by volume. Solvent B was composed 30% ethyl acetate, 31% acetonitrile and 39% methanol by volume. The column temperature was 30 °C. For separation the following gradient was used at a flow rate of 1.0 ml/min: 50% solvent B at the time of injection, increasing linearly for 30 min to 100% B, then 2 min with 100% B before the system was allowed to equilibrate at 50% B for 5 min prior to a new injection. Samples were prepared by dissolving the dried pigments in solvent B. After centrifugation for 5 min at 20000 x g the supernatant was filtered through a 4.5 #m pore filter. Before injection 10% by volume of 1.0 M ammonium acetate in water was added to the pigment extract as a buffer. A Kontron 432 single wavelength detector measured absorbance at 450 nm every 0.5 s and spectra of the eluent were recorded every 3 s by a Specord S10 diode array spectrophotometer (Carl Zeiss, Jena, Germany). Spectrochromatograms were collected and analyzed by specially written software (Frigaard et al. 1996).
Gas chromatography For identification of the esterifying alcohols of the BChl c, green fractions from the HPLC eluent were collected for further analysis by gas chromatography. The eluents were dried under a stream of N2 and transesterified to give methyl bacteriochlorophyllides and free fatty alcohols essentially as described by Fages et al. (1990); ca. 0.5 pmol BChl c was dissolved in 20 ml 1% KOH in methanol, and the extract was left overnight in the dark at room temperature. Water (60 ml) was added and the esterifying alcohols of BChl c were extracted with 3 x 50 ml ice-cold diethyl ether. The combined ether phases were washed with 3 x 50 ml water to remove traces of KOH and evaporated under reduced pressure. The dry extracts containing fatty alcohols were stored under N2 at - 2 0 °C. Sampies were dissolved in hexane before injection. The GC system used was a Chrompack CP9001 with a Chrompack WCOT fused silica 25 m x 0.32 mm (i.d.) column with a coating of CP-Sil 8CB with a
film thickness of 0.25 #m. The carrier gas was helium at an inlet pressure of 100 kPa, and the temperatures of the FID-detector and the injector were both 280 °C. Separation was achieved using a linear temperature gradient starting at 150 °C at the time of injection and rising to 270 °C over 30 min. The peaks due to farnesol, geranylgeraniol, phytol, octadecanol and dodecanol were identified by comparison with standards.
Spectroscopy A Perkin-Elmer 330 spectrophotometer was used for taking the absorption spectra of whole cells. Absorbance values at 1.0 nm intervals with a scan speed at 30 nm min-l were collected using a microcomputer. Second derivative spectra were calculated using a 19-point window (Savitsky and Golay 1964). For fluorescence measurements a Fluorolog F2C spectrofluorometer (Spex Industries Inc., Edison, NJ, USA) was used with an integration time of 5 s nm - l . The excitation band width was 9.0 nm and the emission band width was 4.5 nm.
Materials Octadecanol from Sigma Chemical Co, St. Louis, MO, USA, a cisltrans mixture of phytol was from Aldrich-Chemic, Steinheim, BRD, decanol and dodecanol were from Nu-Chek-Prep, Elysian, MN, USA and trans, trans-geranylgeraniol was from American Radiolabeled Chemicals Inc. St. Louis, MO, USA. flCyclodextrin was obtained from Merck. Organic solvents were HPLC grade (Rathburn, Walkerburn, UK).
Results
Effect of fatty alcohols on growth Fatty alcohols are potentially toxic to cells as a result of their effects on membrane properties. They can therefore only be added to growth media in limited amounts. In initial experiments the alcohols were added to the growth medium before inoculation and toxic effects were investigated. As shown in Table 1, Cb. tepidum tolerated addition of up to 450 #M of the fatty alcohols phytol, geranylgeraniol and stearol without major effects on the growth rate. However, the shorter chain fatty alcohols had more pronounced effects on the growth rate; 40/aM dodecanol only gave a small decrease in growth rate (to 67% of the con-
388 Table 1. Effect of supplementation of the culture medium with fatty alcohols on growth rate and bacteriochlorophyll c homolog composition in Cldorobium tepidum Alcohol
Experimental protocol Addition to medium before inoculation
Programmed addition Total amount Growth ratea BChl c homologs (#mol I- l) (% of control) with added alcohol (% of ~tal)
Concentration (#M)
Growth ratea BChl c homologs (% of control) with added alcohol (% of ~tal)
Phytol
4.4 44 440
101 107 104
0 8 11
150 7300
90 92
18 20
Geranylgeraniol
4.5 45 450
101 103 108
0 19 20
150 7500
87 84
35 35
Stcarol
4.8 48 480
94 101 87
0 0 0
160 8000
98 101
7 8
37
67
0
37 190 940
92 85 87
15 43 40
Dodccanol Dodexaaol + /3-cyclodexlrin
Decanol
1.3 13
130
103 18
0b
0
0
aGrowth was linear (light-limited) in all cases. bCells lysed.
trod whereas addition of 13 #M decanol decreased the growth rate to 18% of the control value. This method of supplementation only allowed limited incorporation of the added fatty alcohols into BChl homologs: 10% for phytol and 20% for geranylgeraniol. For the fatty alcohols that did not show large inhibitory effects on growth rate, supplementation with sufficient alcohol to obtain high incorporation into BChl c could be combined with uninhibited growth if the alcohol was added stepwise during growth as described in the methods section. In this way a total amount of up to 7.5 mmol fatty alcohol per litre of culture could be added without affecting the growth rate to a major extent; for cells supplemented with phytol, geranylgeraniol and stearol, growth rates were 90, 85 and 100% of the control, respectively. The amount of BChl found to be esterified with exogenous alcohol was highest with the isoprenoid alcohol geranylgeraniol. It is of interest that the same degree of homolog
replacement was obtained with final amounts of 0.15 and 7.5 mmol 1- l suggesting that there may be an upper limit to the degree of substitution that can be obtained. To overcome the greater inhibitory effects of the addition of dodecanol we investigated the possibility of addition of the alcohols as an inclusion complex with cyclodextrin. Water solubility and bioavailability of long-chain alcohols are increased by inclusion complex formation with ~-CD, a cyclic molecule with seven glucose residues (Szente et al. 1993). As illustrated in Table 1, addition of 40/~M dodecanol as an inclusion complex with B-CD only caused a 10% inhibition of the growth rate compared to the 30% inhibition observed when this fatty alcohol was added by itself prior to inoculation. Table 1 also shows that it was possible to add 0.9 mM dodecanol as a/~-CD complex and still observe more than 85% of the growth rate of the control cells. In this case 40% of the isolated BChl c homologs were esterified with dodecanol. This
389 '
'
'
°
I
I
800 - Control
.
.
.
.
I
.
.
.
.
I
0
700
~, o @ o
600 500 400 0
=_.. 300 O) t=
>=
'
'
'
'
I
,
I
A,
!
.
800 - Supplemented
,_
.
.
I
.
,
~ . ~
~
,~
,
!
0
700
®
o
600 500 400
o
o°
300 ,
5
10
I
x .
.
.
.
I
'
15 20 Elution time (min)
~,
25
"
, ~
i30
Figure 1. Spectrochromatograms of pigment extracts from Chlorobium tepidum. Upper panel shows control cells and lower panel shows cells supplemented with a total of 150 #mol geranylgeraniol per litre of culture (programmed addition). Contours are drawn at logarithmic intervals corresponding to absorbance values of 2.0, 1.0, 0.5, 0.25 and 0.125.
clearly demonstrates the great effectivity of/3-CD in increasing the bioavailability of dodecanol. BChl c homolog distribution
Farnes01
.=. Geranylgerani01
In the HPLC method used in the present work the various BChl c homologs eluted in groups having the same esterifying alcohols but differing in the degree of alkylation of the tetrapyrrole macrocycle. The upper panel in Figure 1 shows a spectrochromatogram of a pigment extract from cells of Cb. tepidum grown in unsupplemented medium. The two major homologs of BChl c with retention times of 8.5 and 9.3 min were identified by mass spectroscopy as [E,E] BChl CF and [P,E] BChl CF, respectively (results not shown). By comparison with previous reports (Nozawa et al. 1991; Uehara et al. 1994) the two minor peaks with retention times of 7.8 and 10.0 min, comprising less than 10% of this homolog set were tentatively identified as the farnesol esters of [E,M] BChl c and [I,E] BChl c, respectively. After 16 min, BChl a eluted together with a second set of BChl c homologs with an unknown esterifying alcohol. This second set of homologs only amounted to 11% of the total amount of BChl c. The three
o o
¢3
L
|1_ '
5
10
15
'
i
20
,
,
'
•
i
,
25
Retention time (rain)
Figure 2. GC analysis of fatty acohols esterified to bacteriochlorophyll c in cells supplemented with geranylgeranioi as described in the legend to Figure 1. Alcohols were obtained by transesterification of the bacteriocholorophyll c containing fractions from HPLC.
last components to elute were carotenoids. From their absorption zpectra they were identified as different cis and trans isomers of chlorobactene in agreement with the report of Wahlund et al. (1991). The lower panel of Figure 1 shows an example of a spectrochromatogram obtained with cells supplemented geranylgeraniol. Compared to the control shown in
390 the upper panel, a new set of BChl c homologs are seen eluting between 12.5 and 14.5 min. The esterifying alcohols were analysed by gas chromatography as described in the methods section. Figure 2 shows a chromatogram of the esterifying alcohols found in the pooled HPLC eluent collected between 5 min and 23 min. Three peaks were identified as farnesol, geranyigeraniol and phytoi corresponding to the main homologs of BChl c esterified with farnesol, the new BChl c homologs esterified with the fatty alcohol added to the growth medium, and BChl a esterified with phytol. This confirms that the new pigments are due to homologs of BChl c esterified with the fatty alcohol added to the medium. Figure 3 shows a comparison of HPLC chromatograms of extracts obtained from Cb. tepidum cells grown in the presence of various types of long chain fatty alcohols. In all cases new peaks eluting at different times appear in the chromatograms and from the corresponding spectrochromatograms (not shown) these were identified as homologs of BChl c. GC analysis of the new homologs, after transesterfication as described above, confirmed that the new pigments are homologs of BChl c esterified with the fatty alcohol given in the medium (results not shown). Calculated as a percentage of the two major farnesol homologs, [E,E] BChl CF and [P,E] BChl cF, addition of dodecanol in this experiment caused 43% of the esterifying alcohols to change from farnesol to dodecanoi. The corresponding values for other alcohols were: geranylgeraniol, 35%; phytol, 20%; and ste/trol, 8%. From the chromatograms shown in Figure 3 it can be seen that, with the exception of the cultures supplemented with stearol, the ratios between the two major farnesyl BChl c homologs remained unchanged. It is noteworthy that in all our supplementation studies we have never observed BChl a to be esterified with alcohols other than phytol.
Effect of growth with fatty alcohols on carotenoids Figure 1 shows that the carotenoid content changes in cells grown in medium supplemented with fatty alcohols compared to control cells. Determination of the relative carotenoid to BChl c ratio from the peak areas in the chromatogram gave a value of 0.5 in control cells. The corresponding values were 1.7 for cells grown with geranylgeraniol and 0.9 for cells grown with dodecanol, phytol and stearol. These differences in the carotenoid content could also be seen in absorption.spectra of whole cells grown with geranylgeraniol
and dodecanol as an increase in the shoulder at 510 nm due to carotenoid (results not shown). This might suggest that exogenously added alcohols can act as precursors for the biosynthesis of carotenoid. This is supported by the appearance in cells grown with geranylgeraniol of novel carotenoids which are not found in detectable amounts in control cells. Two new peaks eluting between 28 and 29 min were seen in cells supplemented with geranylgeraniol, as shown in the lower panel of Figure 1. From the absorption spectra we have identified these pigments, which have absorbance maxima at 298 and 285 nm, as isomers of phytoene. Furthermore, trace amounts of phytofluene, with absorbance maxima at 348 and 368 nm, eluted after 27.5 min.
Effect of changes in BChl c homolog distribution on spectral characteristics Cells grown in medium supplemented with geranylgeraniol, phytol and stearol had the same spectral properties as untreated cells, as judged from the position of the Qy band at 754 nm (results not shown). However, a blue-shift of the absorption maxima of the Qy band of the BChl c was observed in cells grown in the presence of dodecanol, as shown in Figure 4. This blue shift was also observed in the second derivative absorption spectra and in the fluorescence emission spectrum. These results indicate that substitution of farnesol by dodecanol leads to a change in the organization of BChl c in the chlorosomes in such a way as to alter the spectral properties.
Discussion
Our results show that it is possible to change up to 43% of the esterifying alcohols of BChl c in Cb. tepidum when the cells are grown in medium supplemented with fatty alcohols. This major change of the BChl c homolog composition had no effect on the spectral properties of the chlorosome when farnesol was substituted with stearol, phytol or geranylgeraniol. However, with dodecanol a blue-shift of 6 nm of the Qy band of BChl c and a concomitant change in the fluorescence emission were observed. It is now generally accepted that the rods observed in freeze-fracture electron micrographs of chlorosomes from green bacteria are supramolecular aggregates of bacteriochlorophylls (reviewed by Blankenship et al. 1995). However, several studies have shown that with-
391
Control
L
I
I
i
,
'
'
Dodecanol
l
I
I' '
Geranylgeraniol
o ,It el
~
~,~
,
Phytol
,
Regular paper
Manipulation of the baeteriochlorophyll c homolog distribution in the green sulfur bacterium Chlorobium tepidum D o r t e B. S t e e n s g a a r d , R a y m o n d P. C o x & M e t t e Miller*
Institute of Biochemistry, Odense University, Campusvej 55, DK-5230 Odense M, Denmark; *Author for correspondence and~or reprints Received 25 January 1996; accepted in revised form 22 March 1996
Key words: bacteriochlorophyll homologs,/~-cyclodextrin, chlorosome, green bacteria
Abstract
We have shown that the green sulfur bacterium Chlorobium tepidum can be grown in batch culture supplemented with potentially toxic fatty alcohols without a major effect on the growth rate if the concentration of the alcohols is kept low either by programmed addition or by adding the alcohol as an inclusion complex with/~-cyclodextrin. HPLC and GC analysis of pigment extracts from the supplemented ceils showed that the fatty alcohols were incorporated into bacteriochlorophyll c as the esterifying alcohol. It was possible to change up to 43% of the naturally occurring farnesyl ester of bacteriochlorophyll c with the added alcohol. This change in the homolog composition had no effect on the spectral properties of the cells when farnesol was partially replaced by stearol, phytol or geranylgeraniol. However, with dodecanol we obtained a blue-shift of 6 nm of the Qy band of the bacteriochlorophyll c and a concomitant change in the fluorescence emission was observed. The possible significance of these findings is discussed in the light of current ideas about bacteriochlorophyll organization in the chlorosomes.
Abbreviations: /3-CD- ~-cyclodextrin; BChl- bacteriochlorophyll; BChl cH- bacteriochlorophyllide c; [E,M] BChl CF-8-ethyi, 12-methyl, farnesyl BChl c; [E,E] BChl CF-8-ethyl, 12-ethyl, farnesyl BChl c; [P,E] BChl CF-8-propyl, 12-ethyl, farnesyl BChl c; [I,E] BChl CF-8-isobutyl, 12-ethyl, farnesyl BChl c; Car-carotenoids Introduction
The light-harvesting pigments of the green bacteria are found in chlorosomes, which are cigar-formed structures found appressed to the cytoplasmic membrane which contains the reaction centers. These unique antenna structures enable extremely efficient light harvesting and energy transfer, allowing green sulfur bacteria to grow in very dim light (Amesz 1991). Depending on the species and strain involved, the main light harvesting pigment of green bacteria is either bacteriochlorophyll c, d, or e. A unique feature of the chlorosome is the low amount of protein associated with the antenna system compared to other known light-harvesting systems. It is believed that the bacteriochlorophylls are organized in self-aggregated rod structures, dominated by
pigment-pigment interactions (reviewed by Blankenship et al. 1995). Another characteristic feature is the presence of homologous forms of the bacteriochlorophyll which differ in molecular structure but have the same absorption properties in the monomeric form (Scheer 1991). The homologs ofchlorosome chlorophylls can vary in several ways. Homologs of BChl c found in the filamentous green bacterium Chloroflexus aurantiacus have the same substituents to the tetrapyrrole ring (8-ethyl, 12-methyl) but differ in the esterifying alcohols, which can be linear (octadecanol, hexadecanol or 9-octadecenol) or isoprenoid (phytol or geranylgeraniol) (Risch et al. 1979; Larsen et al. 1994). In green sulfur bacteria containing BChl c the substituents to the tetrapyrrole ring can be ethyl, n-propyl or iso-butyl at the 8-position and methyl or ethyl at the 12-position
386 (Smith et al. 1983), and the main esterifying alcohol is farnesol. However, in old batch cultures of Chlorobium limicola, up to 10% of the esterifying alcohols of the BChl c molecules were geranylgeraniol, tetrahydrogeranylgeraniol, phytol, cis-9-hexadecen-l-ol and 4-undecyl-2-furanmethanoi (Caple et al. 1978). In Cb. vibrioforme containing both BChl c and BChl d, 10% of the esterifying alcohols were decenol and phytol and in the BChl e containing Cb. phaeovibrioides 20% of the alcohols were cetoi, 4-undecyl-2-furanmethanol and oleyl alcohol (Otte et al. 1993). The possible biological significance of the BChl homolog distribution in the Chlorobiaceae and Chloroflexacae remains an open question. It has been shown that the culture conditions, especially the light regime during growth, influence the homolog distribution in green sulfur bacteria (Huster and Smith 1990). In 4 strains of green bacteria, a decreased light intensity was found to lead to an increase of the more alkylated BChl homologs and a concomitant red-shift of the absorption maximum of the Qy band (Borrego and GarciaGil 1995). The illumination intensity also affected the homolog distribution of BChl c in Cf. aurantiacus grown in turbidostat culture under light-limiting conditions, although no clear pattern of changes was observed (Larsen et al. 1994). Furthermore, it has been shown that Cf. aurantiacus can take up and incorporate exogenous long-chain alcohols into BChl c. Both naturally-occurring and novel BChl c homologs were formed without affecting the spectral properties of the chlorosome (Larsen et al. 1995). The green sulfur bacteria seem more fastidious in their choice of esterifying alcohol than Chloroflexus. The work described here was carried out to investigate whether, in spite of this tendency, it is possible to manipulate the composition of BChl c homologs in the green sulfur bacterium, Chlorobium tepidum. We demonstrate that Cb. tepidum can be grown in medium supplemented with long chain fatty alcohols that are potentially toxic and that/~-cyclodextrin is useful for increasing the bioavailability of the fatty alcohols. Our results show that Cb. tepidum does indeed incorporate exogenous long-chained alcohols of different kinds and that, in contrast to the case with Chloroflexus, this does lead to observable changes in the organization of BChl c in the chlorosome.
Methods
Growth conditions Chlorobium tepidum ATCC 49652 was grown at 45 °C in the medium described by Wahlund et al. (1991). The inoculum was obtained from a light-limited continuous culture grown at 45 °C with a dilution rate 0.9 d - l . The culture was illuminated with incandescent light bulbs giving a photon flux density of 20 #mol m -2 s- l measured at the surface of the bioreactor. Growth in the presence of alcohols was carried out in 70 ml flat culture bottles. Light was provided by a Philips TLD type 29 fluorescent tube (18 W) giving a photon flux density of 20 #mol m -2 s - l at the surface of the culture bottle. In all experiments cultures were inoculated with cells from the continuous culture to give an initial concentration of 3.0 nmol BChl c ml- I. Three different protocols were used for addition of fatty alcohols. For initial experiments, 4.5, 45 or 450/zM of either phytol, geranylgeraniol, stearol or decanol or 37/zM of dodecanol were added to the growth medium before inoculation. For programmed addition, phytol, geranylgeraniol and octadecanol were added from stock emulsions containing 1.5 mM alcohol in the growth medium. Alcohol emulsions were added every 7-9 h for 3 days as follows: 1, 2, 3, 4, 5, 5, 5, 5 and 5 ml. The bottles, containing a glass bead to promote mixing, were shaken vigorously after every alcohol addition. For growth with dodecanol as an inclusion complex with/3-cyclodextrin, the alcohol and fl-CD were added the growth medium before inoculation in the molar ratio 1 alcohol: 4/3-CD, corresponding to 0.94 #M alcohol and 3.7 #M/~-CD. Growth was followed by absorbance measurements at the wavelength of the Qy band of BChl c on a Shimadzu UV-160A spectrophotometer. Four days after inoculation the cells were harvested by centrifugation for 10 min at 24000 x g and stored at - 5 0 ° C . Pigment extraction and analysis by HPLC Pigments were extracted from cell pellets in 50 volumes of acetone:methanol (7:2 v/v) by sonication for 60 s. After centrifugation at 20000 x g for 5 min, extractions were repeated until the solvent was colorless. The combined extracts were dried under a stream of N2 and traces of water were removed under reduced pressure. Dried extracts were stored under N2 at - 2 0 °C until use.
387 HPLC separation was achieved on a 300 x 3.9 mm (i.d.) octadecylsilyl bonded amorphous silica column (Nova-Pak C18, 60 lk pore size, 4 #m particles) with a precolumn (Nova-Pak C18) both from Waters Millipore Corporation, Milford, MA, USA. The HPLC protocol was a modification of that used by Borrego and Garcia-Gil (1994). Solvent A was composed of 25% H20, 33% acetonitrile and 42% methanol by volume. Solvent B was composed 30% ethyl acetate, 31% acetonitrile and 39% methanol by volume. The column temperature was 30 °C. For separation the following gradient was used at a flow rate of 1.0 ml/min: 50% solvent B at the time of injection, increasing linearly for 30 min to 100% B, then 2 min with 100% B before the system was allowed to equilibrate at 50% B for 5 min prior to a new injection. Samples were prepared by dissolving the dried pigments in solvent B. After centrifugation for 5 min at 20000 x g the supernatant was filtered through a 4.5 #m pore filter. Before injection 10% by volume of 1.0 M ammonium acetate in water was added to the pigment extract as a buffer. A Kontron 432 single wavelength detector measured absorbance at 450 nm every 0.5 s and spectra of the eluent were recorded every 3 s by a Specord S10 diode array spectrophotometer (Carl Zeiss, Jena, Germany). Spectrochromatograms were collected and analyzed by specially written software (Frigaard et al. 1996).
Gas chromatography For identification of the esterifying alcohols of the BChl c, green fractions from the HPLC eluent were collected for further analysis by gas chromatography. The eluents were dried under a stream of N2 and transesterified to give methyl bacteriochlorophyllides and free fatty alcohols essentially as described by Fages et al. (1990); ca. 0.5 pmol BChl c was dissolved in 20 ml 1% KOH in methanol, and the extract was left overnight in the dark at room temperature. Water (60 ml) was added and the esterifying alcohols of BChl c were extracted with 3 x 50 ml ice-cold diethyl ether. The combined ether phases were washed with 3 x 50 ml water to remove traces of KOH and evaporated under reduced pressure. The dry extracts containing fatty alcohols were stored under N2 at - 2 0 °C. Sampies were dissolved in hexane before injection. The GC system used was a Chrompack CP9001 with a Chrompack WCOT fused silica 25 m x 0.32 mm (i.d.) column with a coating of CP-Sil 8CB with a
film thickness of 0.25 #m. The carrier gas was helium at an inlet pressure of 100 kPa, and the temperatures of the FID-detector and the injector were both 280 °C. Separation was achieved using a linear temperature gradient starting at 150 °C at the time of injection and rising to 270 °C over 30 min. The peaks due to farnesol, geranylgeraniol, phytol, octadecanol and dodecanol were identified by comparison with standards.
Spectroscopy A Perkin-Elmer 330 spectrophotometer was used for taking the absorption spectra of whole cells. Absorbance values at 1.0 nm intervals with a scan speed at 30 nm min-l were collected using a microcomputer. Second derivative spectra were calculated using a 19-point window (Savitsky and Golay 1964). For fluorescence measurements a Fluorolog F2C spectrofluorometer (Spex Industries Inc., Edison, NJ, USA) was used with an integration time of 5 s nm - l . The excitation band width was 9.0 nm and the emission band width was 4.5 nm.
Materials Octadecanol from Sigma Chemical Co, St. Louis, MO, USA, a cisltrans mixture of phytol was from Aldrich-Chemic, Steinheim, BRD, decanol and dodecanol were from Nu-Chek-Prep, Elysian, MN, USA and trans, trans-geranylgeraniol was from American Radiolabeled Chemicals Inc. St. Louis, MO, USA. flCyclodextrin was obtained from Merck. Organic solvents were HPLC grade (Rathburn, Walkerburn, UK).
Results
Effect of fatty alcohols on growth Fatty alcohols are potentially toxic to cells as a result of their effects on membrane properties. They can therefore only be added to growth media in limited amounts. In initial experiments the alcohols were added to the growth medium before inoculation and toxic effects were investigated. As shown in Table 1, Cb. tepidum tolerated addition of up to 450 #M of the fatty alcohols phytol, geranylgeraniol and stearol without major effects on the growth rate. However, the shorter chain fatty alcohols had more pronounced effects on the growth rate; 40/aM dodecanol only gave a small decrease in growth rate (to 67% of the con-
388 Table 1. Effect of supplementation of the culture medium with fatty alcohols on growth rate and bacteriochlorophyll c homolog composition in Cldorobium tepidum Alcohol
Experimental protocol Addition to medium before inoculation
Programmed addition Total amount Growth ratea BChl c homologs (#mol I- l) (% of control) with added alcohol (% of ~tal)
Concentration (#M)
Growth ratea BChl c homologs (% of control) with added alcohol (% of ~tal)
Phytol
4.4 44 440
101 107 104
0 8 11
150 7300
90 92
18 20
Geranylgeraniol
4.5 45 450
101 103 108
0 19 20
150 7500
87 84
35 35
Stcarol
4.8 48 480
94 101 87
0 0 0
160 8000
98 101
7 8
37
67
0
37 190 940
92 85 87
15 43 40
Dodccanol Dodexaaol + /3-cyclodexlrin
Decanol
1.3 13
130
103 18
0b
0
0
aGrowth was linear (light-limited) in all cases. bCells lysed.
trod whereas addition of 13 #M decanol decreased the growth rate to 18% of the control value. This method of supplementation only allowed limited incorporation of the added fatty alcohols into BChl homologs: 10% for phytol and 20% for geranylgeraniol. For the fatty alcohols that did not show large inhibitory effects on growth rate, supplementation with sufficient alcohol to obtain high incorporation into BChl c could be combined with uninhibited growth if the alcohol was added stepwise during growth as described in the methods section. In this way a total amount of up to 7.5 mmol fatty alcohol per litre of culture could be added without affecting the growth rate to a major extent; for cells supplemented with phytol, geranylgeraniol and stearol, growth rates were 90, 85 and 100% of the control, respectively. The amount of BChl found to be esterified with exogenous alcohol was highest with the isoprenoid alcohol geranylgeraniol. It is of interest that the same degree of homolog
replacement was obtained with final amounts of 0.15 and 7.5 mmol 1- l suggesting that there may be an upper limit to the degree of substitution that can be obtained. To overcome the greater inhibitory effects of the addition of dodecanol we investigated the possibility of addition of the alcohols as an inclusion complex with cyclodextrin. Water solubility and bioavailability of long-chain alcohols are increased by inclusion complex formation with ~-CD, a cyclic molecule with seven glucose residues (Szente et al. 1993). As illustrated in Table 1, addition of 40/~M dodecanol as an inclusion complex with B-CD only caused a 10% inhibition of the growth rate compared to the 30% inhibition observed when this fatty alcohol was added by itself prior to inoculation. Table 1 also shows that it was possible to add 0.9 mM dodecanol as a/~-CD complex and still observe more than 85% of the growth rate of the control cells. In this case 40% of the isolated BChl c homologs were esterified with dodecanol. This
389 '
'
'
°
I
I
800 - Control
.
.
.
.
I
.
.
.
.
I
0
700
~, o @ o
600 500 400 0
=_.. 300 O) t=
>=
'
'
'
'
I
,
I
A,
!
.
800 - Supplemented
,_
.
.
I
.
,
~ . ~
~
,~
,
!
0
700
®
o
600 500 400
o
o°
300 ,
5
10
I
x .
.
.
.
I
'
15 20 Elution time (min)
~,
25
"
, ~
i30
Figure 1. Spectrochromatograms of pigment extracts from Chlorobium tepidum. Upper panel shows control cells and lower panel shows cells supplemented with a total of 150 #mol geranylgeraniol per litre of culture (programmed addition). Contours are drawn at logarithmic intervals corresponding to absorbance values of 2.0, 1.0, 0.5, 0.25 and 0.125.
clearly demonstrates the great effectivity of/3-CD in increasing the bioavailability of dodecanol. BChl c homolog distribution
Farnes01
.=. Geranylgerani01
In the HPLC method used in the present work the various BChl c homologs eluted in groups having the same esterifying alcohols but differing in the degree of alkylation of the tetrapyrrole macrocycle. The upper panel in Figure 1 shows a spectrochromatogram of a pigment extract from cells of Cb. tepidum grown in unsupplemented medium. The two major homologs of BChl c with retention times of 8.5 and 9.3 min were identified by mass spectroscopy as [E,E] BChl CF and [P,E] BChl CF, respectively (results not shown). By comparison with previous reports (Nozawa et al. 1991; Uehara et al. 1994) the two minor peaks with retention times of 7.8 and 10.0 min, comprising less than 10% of this homolog set were tentatively identified as the farnesol esters of [E,M] BChl c and [I,E] BChl c, respectively. After 16 min, BChl a eluted together with a second set of BChl c homologs with an unknown esterifying alcohol. This second set of homologs only amounted to 11% of the total amount of BChl c. The three
o o
¢3
L
|1_ '
5
10
15
'
i
20
,
,
'
•
i
,
25
Retention time (rain)
Figure 2. GC analysis of fatty acohols esterified to bacteriochlorophyll c in cells supplemented with geranylgeranioi as described in the legend to Figure 1. Alcohols were obtained by transesterification of the bacteriocholorophyll c containing fractions from HPLC.
last components to elute were carotenoids. From their absorption zpectra they were identified as different cis and trans isomers of chlorobactene in agreement with the report of Wahlund et al. (1991). The lower panel of Figure 1 shows an example of a spectrochromatogram obtained with cells supplemented geranylgeraniol. Compared to the control shown in
390 the upper panel, a new set of BChl c homologs are seen eluting between 12.5 and 14.5 min. The esterifying alcohols were analysed by gas chromatography as described in the methods section. Figure 2 shows a chromatogram of the esterifying alcohols found in the pooled HPLC eluent collected between 5 min and 23 min. Three peaks were identified as farnesol, geranyigeraniol and phytoi corresponding to the main homologs of BChl c esterified with farnesol, the new BChl c homologs esterified with the fatty alcohol added to the growth medium, and BChl a esterified with phytol. This confirms that the new pigments are due to homologs of BChl c esterified with the fatty alcohol added to the medium. Figure 3 shows a comparison of HPLC chromatograms of extracts obtained from Cb. tepidum cells grown in the presence of various types of long chain fatty alcohols. In all cases new peaks eluting at different times appear in the chromatograms and from the corresponding spectrochromatograms (not shown) these were identified as homologs of BChl c. GC analysis of the new homologs, after transesterfication as described above, confirmed that the new pigments are homologs of BChl c esterified with the fatty alcohol given in the medium (results not shown). Calculated as a percentage of the two major farnesol homologs, [E,E] BChl CF and [P,E] BChl cF, addition of dodecanol in this experiment caused 43% of the esterifying alcohols to change from farnesol to dodecanoi. The corresponding values for other alcohols were: geranylgeraniol, 35%; phytol, 20%; and ste/trol, 8%. From the chromatograms shown in Figure 3 it can be seen that, with the exception of the cultures supplemented with stearol, the ratios between the two major farnesyl BChl c homologs remained unchanged. It is noteworthy that in all our supplementation studies we have never observed BChl a to be esterified with alcohols other than phytol.
Effect of growth with fatty alcohols on carotenoids Figure 1 shows that the carotenoid content changes in cells grown in medium supplemented with fatty alcohols compared to control cells. Determination of the relative carotenoid to BChl c ratio from the peak areas in the chromatogram gave a value of 0.5 in control cells. The corresponding values were 1.7 for cells grown with geranylgeraniol and 0.9 for cells grown with dodecanol, phytol and stearol. These differences in the carotenoid content could also be seen in absorption.spectra of whole cells grown with geranylgeraniol
and dodecanol as an increase in the shoulder at 510 nm due to carotenoid (results not shown). This might suggest that exogenously added alcohols can act as precursors for the biosynthesis of carotenoid. This is supported by the appearance in cells grown with geranylgeraniol of novel carotenoids which are not found in detectable amounts in control cells. Two new peaks eluting between 28 and 29 min were seen in cells supplemented with geranylgeraniol, as shown in the lower panel of Figure 1. From the absorption spectra we have identified these pigments, which have absorbance maxima at 298 and 285 nm, as isomers of phytoene. Furthermore, trace amounts of phytofluene, with absorbance maxima at 348 and 368 nm, eluted after 27.5 min.
Effect of changes in BChl c homolog distribution on spectral characteristics Cells grown in medium supplemented with geranylgeraniol, phytol and stearol had the same spectral properties as untreated cells, as judged from the position of the Qy band at 754 nm (results not shown). However, a blue-shift of the absorption maxima of the Qy band of the BChl c was observed in cells grown in the presence of dodecanol, as shown in Figure 4. This blue shift was also observed in the second derivative absorption spectra and in the fluorescence emission spectrum. These results indicate that substitution of farnesol by dodecanol leads to a change in the organization of BChl c in the chlorosomes in such a way as to alter the spectral properties.
Discussion
Our results show that it is possible to change up to 43% of the esterifying alcohols of BChl c in Cb. tepidum when the cells are grown in medium supplemented with fatty alcohols. This major change of the BChl c homolog composition had no effect on the spectral properties of the chlorosome when farnesol was substituted with stearol, phytol or geranylgeraniol. However, with dodecanol a blue-shift of 6 nm of the Qy band of BChl c and a concomitant change in the fluorescence emission were observed. It is now generally accepted that the rods observed in freeze-fracture electron micrographs of chlorosomes from green bacteria are supramolecular aggregates of bacteriochlorophylls (reviewed by Blankenship et al. 1995). However, several studies have shown that with-
391
Control
L
I
I
i
,
'
'
Dodecanol
l
I
I' '
Geranylgeraniol
o ,It el
~
~,~
,
Phytol
,
