(Received for publication, December 5, 1996, and in revised form, January 29, 1997)
From the Botanisches Institut der Universität
München, 86038 München, Federal Republic of Germany and the
§ Department of Biology, Indiana University,
Bloomington, Indiana 47405
Genes coding for putative chlorophyll a synthase (chlG) from Synechocystis sp. PCC 6803 and bacteriochlorophyll a synthase (bchG) from Rhodobacter capsulatus were amplified by the polymerase chain reaction and cloned into T7 RNA polymerase-based expression plasmids. In vitro enzymatic assays indicated that heterologous expression of the chlG and bchG gene products in Escherichia coli conferred chlorophyll a and bacteriochlorophyll a synthase activity, respectively. Chlorophyll a synthase utilized chlorophyllide a, but not bacteriochlorophyllide a, as a substrate, whereas bacteriochlorophyll a synthase utilized bacteriochlorophyllide a, but not chlorophyllide a. Both enzymes were also observed to exhibit a marked preference for phytyl diphosphate over geranylgeranyl diphosphate.
Cyanobacteria, algae, and plants synthesize the well characterized
Mg-tetrapyrrole chlorophyll a, whereas many anoxygenic photosynthetic bacteria synthesize a related, but more reduced derivative known as bacteriochlorophyll a. Analysis of
Mg-tetrapyrrole intermediates that accumulate in various mutant strains
of photosynthetic bacteria and algae have given rise to our current
understanding of these pathways (1). As indicated in Fig.
1, both pathways utilize common intermediates from
5-aminolevulinate via Mg-protoporphyrin IX to chlorophyllide
a. At this point, these pathways diverge, with the final
step of chlorophyll a synthesis accomplished by the enzyme
chlorophyll synthase, which catalyzes esterification of chlorophyllide
a with phytyl diphosphate in green plants and with
geranylgeranyl diphosphate in greening etiolated seedlings (2, 3).
Esterification is a precondition for stabilization of apoproteins and
for accumulation of chlorophyll-protein complexes (4, 5). The
bacteriochlorophyll a biosynthetic pathway is more complex,
involving a multistep modification of the chlorophyllide a
ring, giving rise to a more reduced structure known as
bacteriochlorophyllide a (1, 6). The final step of the
bacteriochlorophyll a pathway involves esterification of
bacteriochlorophyllide a catalyzed by the enzyme
bacteriochlorophyll synthase (7). Bacteriochlorophyll a in
most photosynthetic bacteria is esterified with phytol, but in some
species, it contains geranylgeraniol instead (2). Since the
bacteriochlorophyll esterification reaction has not been previously carried out in vitro, it has not been determined whether the
diphosphate derivatives of the respective alcohols are the true
substrates in bacteria.
Genetic studies on pigment biosynthesis in Rhodobacter capsulatus initially revealed the presence of a cluster of genes involved in bacteriochlorophyll a biosynthesis (reviewed in Refs. 1 and 8). The roles of specific open reading frames in bacteriochlorophyll synthesis were determined by the construction of a defined set of interposon insertion mutations in sequenced open reading frames (6, 7, 9). One of the disruptions constructed by Bollivar et al. (6, 7) resulted in the accumulation of bacteriochlorophyllide a, indicating that orf304 (bchG) may encode for the enzyme bacteriochlorophyll a synthase, which is responsible for esterification of bacteriochlorophyllide with phytol. However, direct proof that bchG codes for a polypeptide that exhibits bacteriochlorophyll synthase activity has not been obtained. In this study, we heterologously expressed the R. capsulatus bchG gene product in Escherichia coli and demonstrate that assayable bacteriochlorophyll a synthase activity is present in cell-free extracts. We also cloned and overexpressed a bchG homolog (chlG) that was recently identified in the Synechocystis genome sequencing project (10, 11). Expression of the chlG gene product resulted in assayable chlorophyll synthase activity in E. coli extracts. These results conclusively demonstrate that the bchG and chlG genes do indeed code for bacteriochlorophyll and chlorophyll synthases, respectively.
Overexpression of chlG and
bchG was accomplished using the T7 RNA polymerase expression
system described by Studier et al. (12). A chlG
expression plasmid was constructed by polymerase chain reaction
(PCR)1 amplification of the sequenced
chlG gene from Synechocystis sp. PCC 6803 (11).
Chromosomal DNA for use as a PCR template was isolated from lysed
Synechocystis cells as described by Williams (13). The
forward primer (5-CCTCTGACACACAAAATACCG) was designed to contain an NdeI restriction site (underlined) at
the chlG translation start site. The reverse primer
(5
-CGTTCAAATCCCCGCATGGCC) was designed to contain a
BamHI restriction site (underlined) downstream from the end
of the chlG open reading frame. PCR conditions were as
follows: 10 mM Tricine buffer (14), 0.3 mM
dATP, 0.3 mM dGTP, 0.3 mM dTTP, 0.3 mM dCTP, 20 µM forward primer, 20 µM reverse primer, and 1 µg of genomic chromosomal DNA.
The PCR buffer cycle (PTC-100 programmable thermal controller, MJ
Research, Inc.) was initiated with a "hot start," which involved
preincubation of the PCR buffer at 98 °C for 4 min, followed by a
reduction of reaction temperature to 80 °C, at which point 5 units
of Taq DNA polymerase were added to the reaction buffer. PCR
amplification subsequently involved 30 cycles of 98 °C for 45 s, 65 °C for 2 min, and 72 °C for 1 min. The DNA segments were
then blunt-ended by a 20-min incubation period at 65 °C. The
PCR-amplified chlG gene was purified by agarose gel
electrophoresis and cloned into the PCR cloning vector pCRII
(Invitrogen) following the manufacturer's instructions. The
chlG gene was then subsequently subcloned as an
NdeI-BamHI restriction fragment into the T7
polymerase expression vector pET28a+ (Novagen Inc.),
forming the construct pChlG.
Construction of a bchG expression vector was accomplished
essentially as described for the chlG expression vector. As
a template for amplification, the plasmid pDB26 (6) was used at a
concentration of 0.01 µg. The forward primer
(5-GACAGTGCCCAAGACCTTTCG) contained an
NdeI restriction site, and the reverse primer
(5
-ACAGCCAGCCCAGCGTCATACCG) contained an
EcoRI restriction site. The amplified bchG gene
was initially cloned into pCRII and subsequently subcloned into the expression vector pT7-7 (15), forming the expression plasmid pBchG.
E.
coli strain BL21(DE3)/pLysS (12) was transformed with plasmids
pChlG and pBchG, with the resulting constructs BL21(DE3)/pLysS/pChlG and BL21(DE3)/pLysS/pBchG, which were stored in 20% glycerol at 80 °C. For overexpression of chlG and bchG,
10 ml of overnight cultures of the appropriate overexpression strain
were subcultured into 1 liter of prewarmed Terrific Broth (15)
supplemented with chloramphenicol at 35 µg/ml and ampicillin at 200 µg/ml and grown at 37 °C with vigorous shaking to a density of 0.5 A650. T7 RNA polymerase expression was then
induced by the addition of
isopropyl-
-D-thiogalactopyranoside to a final
concentration of 0.4 mM. Subsequent chlG or
bchG expression was allowed to proceed for 3 h prior to
harvesting cells by centrifugation at 10,000 × g for
10 min at 4 °C. The cells were washed by resuspension of the cell
pellet in 10 mM Tris-HCl, pH 8.0, followed by
centrifugation as described above. Cell pellets were stored at
80 °C.
Chlorophyllides a and b were prepared from leaves of Ailanthus altissima using the chlorophyllase reaction (16). Chlorophyllide a was purified by preparative column chromatography on Silica Gel C18 (reverse phase; Waters) with 60% acetone and 40% buffer (50 mM Hepes/KOH, pH 7.8, and 1 mM Na2S2O4) according to Helfrich et al. (17). Bacteriochlorophyllide a was prepared according to Fiedor et al. (18) with some modifications. As a source of chlorophyllase, leaves of A. altissima were used.
Bacteriochlorophyll a from Rhodobacter sphaeroides (30 nmol), dissolved in 50 µl of acetone, was incubated with 500 µl of chlorophyllase extract, 2500 µl of 50 mM phosphate buffer, pH 7.0, containing 50 mM KCl and 0.24% Triton X-100, 75 µl of freshly prepared 500 mM sodium ascorbate in phosphate buffer, and 3 µl of pyridine for 60 min at 37 °C. The reaction was stopped by the addition of acetone to a final concentration of 80%. The mixture was centrifuged for 20 min at 20,000 × g. The clear supernatant was extracted once with 0.5 volume of n-hexane to remove unreacted bacteriochlorophyll. The bacteriochlorophyllide was then extracted with ethyl acetate and used without further purification.
Analytical HPLC was performed on a Gynkotek Model 480 equipped with a 20-µl injection loop, a Rosil C18 column (5 µm, 250 × 4.6 mm), an ultraviolet-visible light source, and a fluorescence detector. The sample was injected and then eluted for 2 min with an acetone/water solvent mixture (70:30), followed by three different linear gradients (within 2 min to acetone/water (82:18), within 11 min to acetone/water (88:12), and within 4 min to 100% acetone). The column was then flushed with 100% acetone for 5 min and returned to the original acetone/water mixture (70:30) within 5 min. Fluorescence detection was set at 405 nm for excitation and at 675 nm for emission.
Chlorophyll and Bacteriochlorophyll Synthase AssaysFrozen
(80 °C) cell paste of the various E. coli strains
(1-1.5 ml) was thawed, and the material was then incubated with 20 µl of DNase (10 mg/ml in 5 mM MgCl2) for 60 min at room temperature, followed by two cycles of freezing and
thawing. Aliquots of this bacterial lysate containing ~8 mg of
protein were diluted with 200 µl of reaction buffer (120 mM potassium acetate, 10 mM magnesium acetate, 50 mM Hepes/KOH, pH 7.6, 14 mM
mercaptoethanol, and 10% glycerol), 30 µl of 5 mM ATP,
and 10 µl of 4 mM geranylgeranyl diphosphate or phytyl
diphosphate. The reaction was then started by the addition of 10 µl
of 0.1 mM chlorophyllide a or
bacteriochlorophyllide a (in acetone). After vigorous
mixing, the samples were incubated for 90 min at 26-28 °C. The
reaction was stopped by the addition of 750 µl of acetone, followed
by centrifugation for 10 min at 14,000 × g. The clear
supernatants were then used to quantitate esterified chlorophyll or
bacteriochlorophyll by HPLC or by phase separation with
n-hexane as described previously (17).
A series of experiments were performed to assay the
esterification activity of protein extracts derived from E. coli strain BL21(DE3)/pLysS that was transformed either with
plasmid pChlG, which expresses the presumed chlorophyll synthase enzyme
from Synechocystis, or with pBchG, which expresses the
putative bacteriochlorophyll synthase from R. capsulatus.
Esterification of chlorophyllide a or bacteriochlorophyllide
a does not change the UV-visible spectral properties, but
does decrease the polarity of these pigments. This difference in
polarity can be used for separation of product from substrate by HPLC
or by phase separation (17). To avoid allomerization or other
modification of the product, the central magnesium was removed by
acidification (forming pheophytin a from chlorophyll
a) prior to HPLC analysis for esterification (19). HPLC-separated reaction products can easily identify the non-esterified pheophorbide a, pheophytin a esterified with
geranylgeraniol, or pheophytin a esterified with phytol.
Fig. 2A shows a typical HPLC separation
profile of an assay involving chlorophyllide a and phytyl
diphosphate incubated with an E. coli extract that was
derived from the strain expressing the chlG gene product. This elution profile contains a large fraction of pheophytin esterified with phytol, which indicates that the extract indeed contains chlorophyll synthase activity. A similar experiment using
geranylgeranyl diphosphate instead of phytyl diphosphate yielded a HPLC
profile of pheophytin a esterified with geranylgeraniol
(Fig. 2B). The enzyme activity is clearly due to the
presence of the expressed chlG gene product in that no
esterification was found with the same substrates that were incubated
with an extract obtained from the nontransformed E. coli
strain (Table I, control strain).
|
Shown in Fig. 3 are the results of assays involving
cell-free extract from the E. coli strain containing the
expression plasmid pBchG. The HPLC elution profile demonstrates that
this enzyme is capable of using either phytyl diphosphate (Fig.
3A) or geranylgeraniol diphosphate (Fig. 3B) as a
substrate for esterifying bacteriochlorophyllide a. Since
this enzyme reaction has never been shown in vitro, we also
checked the identity of the products by co-chromatography with
authentic bacteriochlorophyll a esterified with either
phytol or geranylgeraniol, the results of which indicated that there was indeed coelution (data not shown). Furthermore, we also observed identical absorption maxima at 359-361, 579-582, and 775-777 nm for
both authentic bacteriochlorophyll a and the
bacteriochlorophyll products generated from the esterification
reactions (data not shown). This indicates that dehydrogenation to the
corresponding chlorophyll derivatives did not take place.
In an independent assay, reaction samples containing a mixture of non-esterified pigment (substrate) and esterified pigment (product) were phase-separated with hexane after treatment of the reaction mixture with ion-exchange resin (see Ref. 17). The esterified pigment that was partitioned into the hexane phase was then assayed by spectrophotometry. The data in Table I also clearly show that considerable esterification of chlorophyllide a was obtained with E. coli extracts harboring plasmid pChlG. Phase separation assays of reactions involving extracts from the E. coli strain containing pBchG indicated considerable bacteriochlorophyll a synthase activity (Table I).
Chlorophyll and Bacteriochlorophyll Synthases Exhibit a High Degree of Substrate SpecificitySubstrate specificity of the
heterologously expressed enzymes was also investigated. In accordance
with previous results on chlorophyll synthase from etiolated oat
seedlings (20), the chlG gene product accepts chlorophyllide
a, but not bacteriochlorophyllide a, as a
substrate. Vice versa, the bchG gene product accepts
bacteriochlorophyllide a, but not chlorophyllide
a (Table I). The remarkable degree of substrate specificity
was even more obvious when the reaction contained a 1:1 mixture of
chlorophyllide a and bacteriochlorophyllide a
(Fig. 4). In this case, extracts from cells expressing
chlG esterified only chlorophyllide a (Fig.
4B), whereas extracts from cells expressing bchG
esterified only bacteriochlorophyllide a (Fig.
4C). The marginal esterification of the "wrong"
substrate detected when only this substrate was offered (see Table I)
was not observed here, probably because the "true" substrate
displaced the wrong substrate from the active center of the enzyme.
We observed in a series of experiments that phytyl diphosphate is a better substrate than geranylgeranyl diphosphate for both chlorophyll and bacteriochlorophyll synthases (Table II). We also found that bacteriopheophorbide a (the magnesium-free derivative of bacteriochlorophyllide a) is not a suitable substrate for esterification with geranylgeranyl diphosphate or phytyl diphosphate by bacteriochlorophyll synthase (data not shown). Thus, the substrate specificity of both enzymes resembles that observed for chlorophyll a synthase from green plants (21). This is distinctly different from the activity observed with etiolated seedlings, which shows a marked preference for geranylgeranyl diphosphate (19).
|
This study provides the first direct experimental evidence that the bchG and chlG genes code for bacteriochlorophyll a and chlorophyll a synthases, respectively. In many respects, heterologously expressed chlorophyll a synthase from Synechocystis exhibits an activity that is indistinguishable from that reported for green plants (21). This includes the preferred specificity for phytyl diphosphate over geranylgeranyl diphosphate as a substrate for esterification of chlorophyllide a. It remains to be shown whether the differing specificity in etiolated plants is the result of a different environment of the same enzyme or whether another chlorophyll synthase exists in etiolated plants. Specificity for phytyl diphosphate was also observed with heterologously expressed bacteriochlorophyll a synthase. This latter result is surprising given that Bollivar et al. (7) observed that a mutation in the bchP gene, which is located just downstream from bchG, resulted in the exclusive accumulation of geranylgeraniol-esterified bacteriochlorophyll. It was concluded that bacteriochlorophyll a synthase most likely utilized geranylgeranyl diphosphate as a substrate, creating geranylgeraniol esterified bacteriochlorophyll a, and that the product of the bchP gene subsequently reduced bacteriochlorophyll to bacteriochlorophyll a via a three-step reduction reaction. Given our results, which indicate that bacteriochlorophyll a synthase prefers phytyl diphosphate as a substrate, the scenario outlined by Bollivar et al. (7) must be modified. We propose that bchP codes for an enzyme that is responsible for reducing geranylgeraniol (or geranylgeranyl diphosphate) to phytol (or phytyl diphosphate) before, as well as possibly after (22), esterification. In this scenario, a disruption of the bchP gene would result in the absence of a pool of phytyl diphosphate for use as a substrate in this reaction. This would give rise to geranylgeraniol esterified bacteriochlorophyll a as a consequence of geranylgeranyl diphosphate being the only available substrate for bacteriochlorophyll synthase. This hypothesis can be tested biochemically by heterologous expression of the bchP gene product in E. coli, followed by assaying cell-free extracts for the ability to reduce geranylgeraniol. In this regard, our control reactions (Table I) confirm that the E. coli extracts contain no significant pool of geranylgeranyl diphosphate or phytyl diphosphate (23). As such, heterologous expression of bchP should be a feasible approach for studying its function.
Another interesting observation is the remarkable specificity of these two enzymes for their respective Mg-tetrapyrrole substrates. These two enzymes exhibit a significant degree of sequence divergence (63.9% non-identity), which must provide enough difference in protein structure to differentiate different Mg-tetrapyrroles as appropriate substrates. Given the fact that chlorophyllide a is an intermediate in bacteriochlorophyll a biosynthesis (Fig. 1) and must be modified to bacteriochlorophyllide a before esterification, it is understandable that the bchG gene product must be able to differentiate between chlorophyllide and bacteriochlorophyllide. Our observation that these enzymes can faithfully distinguish these substrates also supports the finding of Lopez et al. (24) that Chloroflexis aurantiacus contains two homologs of bchG, one of which is proposed to be involved in bacteriochlorophyll a synthesis and the other in bacteriochlorophyll c synthesis. The additional specificity for bacteriochlorophyllide over bacteriopheophorbide indicates that bacteriopheophytin, which is present in purple bacterial reaction centers, must be derived by loss or extraction of magnesium from esterified bacteriochlorophyll a.
Heterologous expression of bacteriochlorophyll (or chlorophyll) biosynthesis genes in E. coli, coupled with demonstration of enzyme activity in cell-free extracts, provides unequivocal proof that previously genetically characterized open reading frames do indeed code for predicted enzymes. The results of this study complement those of Gibson et al. (25), who also utilized heterologous expression to demonstrate that the bchD, bchH, and bchI genes code for subunits of Mg-chelatase, and those of Bollivar et al. (26), who demonstrated that bchM codes for Mg-protoporphyrin IX monomethyltransferase. Several genes implicated in additional steps in the Mg-tetrapyrrole pathway, such as cyclopentone ring formation and 4-vinyl reduction, await similar analysis.
We thank C. Blank for construction of pBchG and Dr. Scheer for samples of bacteriochlorophyll.