(Received for publication, January 9, 1997, and in revised form, February 18, 1997)
From the Children's Hospital Oakland Research Institute, Oakland, California 94609
The structural basis for the dual specificity of
the malonyl-CoA/acetyl-CoA:acyl carrier protein
S-acyltransferase associated with the multifunctional
animal fatty acid synthase has been investigated by mutagenesis.
Arginine 606, which is positionally conserved in the transacylase
domains of all multifunctional fatty acid and polyketide synthases, was
replaced by alanine or lysine in the context of the isolated
transacylase domain, and the mutant proteins were expressed in
Escherichia coli. Malonyl transacylase activity of the
Arg-606 Ala and Arg-606
Lys mutant enzymes was reduced by
100- and 10-fold, respectively. In contrast, acetyl transacylase
activity was increased 6.6-fold in the Arg-606
Ala mutant and
1.7-fold in the Arg-606
Lys mutant. Kinetic studies revealed that
selectivity of the enzyme for acetyl-CoA was increased >16,000-fold by
the Ala mutation and 16-fold by the Lys mutation. Activity toward
medium chain length acyl thioesters was also increased >3 orders of
magnitude by mutation of Arg-606, so that the Ala-606 enzyme is an
effective medium chain length fatty acyl transacylase. These results
indicate that Arg-606 plays an important role in the binding of malonyl
moieties to the transacylase domain but is not required for binding of
acetyl moieties; these results are also consistent with a mechanism
whereby interaction between the positively charged guanidinium group of
Arg-606 and the free carboxylate anion of the malonyl moiety serves to
position this substrate in the active site of the enzyme.
The animal FAS1 consists of two
identical polypeptides, each carrying six enzymes and an acyl carrier
protein, that are juxtaposed to form two centers for the synthesis of
palmitic acid from acetyl- and malonyl-CoA (1-3). The iterative
condensation of an acetyl moiety with successive malonyl moieties and
reduction of the -keto intermediates normally results in the
formation of palmitic acid as the major product. Initiation of the
series of condensation reactions necessitates the translocation of an
acetyl moiety and subsequently seven malonyl moieties, from CoA
thioester to the 4
-phosphopantetheine of the acyl carrier protein
domain of the FAS. A single transacylase enzyme is responsible for
translocation of both substrates, and intermediate formation of
acyl-O-serine intermediates occurs at the same site (4). The
employment of a common enzyme for the loading of both substrates has
important consequences in terms of the kinetics of the FAS reaction
sequence, since substrate binding to the transacylase domain is random
and each substrate is a competitive inhibitor for the other. Thus, for
synthesis to proceed efficiently, inappropriately bound substrates must be rapidly removed by transfer back to the CoA acceptor (5). This
substrate sorting or "self-editing" process occurs extremely rapidly, with turnover numbers >100 s
1, and does not
limit the rate of acyl chain assembly, provided an appropriate
concentration of free CoA is available (5, 6).
The chain initiation mechanism for FASs that consist of separate
monofunctional proteins such as those of plants and most bacteria
differs from that of their multifunctional counterparts in that acetyl
and malonyl moieties are translocated from CoA by separate
transacylases. Thus there is a direct transacetylation from CoA to the
active-site cysteine of a unique ketoacyl synthase termed ketoacyl
synthase III, which functions solely in the initial condensation
reaction (7, 8). In these systems, malonyl moieties are transferred to
the 4-phosphopantetheine of the acyl carrier protein by a monospecific
malonyl transacylase. Recently, a crystal structure of the
malonyl-CoA:acyl carrier protein S-malonyltransferase from
Escherichia coli has been reported (9). However, the
structural information did not provide an obvious rationale for the
substrate specificity of the enzyme. We have adopted an alternative
strategy to identify residues that play a role in promoting catalysis
and in determining the substrate specificity of the malonyl/acetyl transacylase associated with the multifunctional FAS, for which no
three-dimensional structure is available. Our rationale is first to
identify from multiple sequence alignments conserved residues that
might be candidates for specific roles in catalysis and then to examine
the effect of mutating these residues. Using this approach, we have
previously identified His-6832 of the rat
FAS as playing an essential role in activation of the catalytic residue
Ser-581 (10). Only one other basic residue, Arg-606 in the rat FAS, is
universally conserved in the transacylases associated with
multifunctional and monofunctional forms of FAS and polyketide synthase
(10). Since all of these transacylases can utilize either malonyl-CoA
or methylmalonyl-CoA as a substrate, we hypothesized that this residue
might facilitate substrate binding by interaction with the free
carboxyl group of the malonyl or methylmalonyl moiety and that its
replacement might compromise binding of these substrates. A corollary
to this hypothesis predicts that, in the context of the
malonyl-CoA/acetyl-CoA:acyl carrier protein transacylase, such a
mutation would likely have no negative effect on acetyl binding. We
report here the results of experiments designed to test this
hypothesis.
Acetyl-S-pantetheine was prepared by acetylation of pantetheine with acetic anhydride and removal of the acetic acid formed by extraction with diethyl ether. The sources of other materials used were reported previously (10).
Construction of Plasmids and Expression of Transacylase ProteinConstruction of the transacylase expression plasmid
pET23a/388-MAT has been described previously (10).
Oligonucleotide-directed in vitro mutagenesis based on the
method described by Kunkel et al. (11) was adopted to
generate Arg-606 Ala and Arg-606
Lys mutants. Correct
introduction of mutations and authenticity of the DNA was confirmed by
DNA sequencing. Plasmids carrying the mutations were used to transform
E. coli BL21(DE3) cells, and expression of the recombinant
proteins was induced by 1 mM isopropyl-
-D-thiogalactoside. The recombinant
transacylases were recovered from inclusion bodies and refolded
in vitro (10, 12).
The Arg-606 Ala and Arg-606
Lys mutant
transacylases were incubated with malonyl-CoA (100 µM) at
0 °C for 1 h, and the amount of free CoA formed was determined
as described earlier (10).
Transacylase activity was routinely determined at 0 °C using either [1-14C]acetyl-CoA or [2-14C]malonyl-CoA as the substrate and pantetheine as the model acceptor (10). In some experiments acetyl-S-pantetheine was used as the acetyl donor and CoASH as the acceptor; the acetyl-CoA product was isolated and quantitated by high performance liquid chromatography on a C4 reversed phase column (Vydac) as described previously (10). One unit of activity is the amount of enzyme catalyzing the utilization of 1 µmol of substrate/min.
Assay of Medium Acyl Chain Length Transacylase ActivityTransfer of medium chain acyl moieties from a pantetheine donor to a CoASH acceptor was determined by separating the substrates and products using high performance liquid chromatography. Wild type and mutant transacylases were incubated with 400 µM acyl-S-pantetheine thioesters and 1 mM CoASH in 100 mM potassium phosphate buffer, pH 6.8 (50-µl volume), at room temperature for 1-15 min, depending on the activity of the preparation. Reactions were stopped by the addition of 1.05 ml of buffer A (25 mM potassium phosphate, pH 5.4, 5% acetonitrile), and the mixture was immediately injected onto a reversed phase high performance liquid chromatography column (Intertsil Phenyl, 5 µm, 4.6 × 150 mm, MetaChem, Torrance, CA) maintained at 45 °C. The column was washed with buffer A at 1.5 ml/min for 5 min followed by a three-step linear gradient: to 25% acetonitrile over 3 min, to 55% acetonitrile over 10 min, and to 80% acetonitrile over 2 min. The amount of acyl-CoA eluted was determined from the observed absorbance at 258 nm using decanoyl-CoA as the standard. Saturating concentrations of donor and acceptor substrates were employed, and the reaction rates were directly proportional to time and enzyme concentration.
Assay of Protein ConcentrationProtein concentration was calculated from the absorbance at 280 nm; 1 mg/ml transacylase gives an absorbance of 1.07.
Arg-606 was mutated to Ala and Lys in the
context of the independently expressed transacylase domain representing
residues 429-815 of the rat FAS. The mutant proteins were expressed in E. coli, solubilized in urea, renatured, and purified.
Purity of the preparations was estimated to be >95% by
SDS-polyacrylamide gel electrophoresis (Fig.
1A). The molecular masses of the proteins were approximately 42 kDa, as predicted, and both mutant proteins were
recognized by rabbit anti-rat liver FAS antibodies (Fig. 1B).
Effect of Mutation of Arg-606 on Transacylase Activity
Replacement of Arg-606 with Ala or Lys decreased malonyl transacylase activity by 99 and 91%, respectively, but increased acetyl transacylase activity by 6.6- and 1.7-fold, respectively (Table I). These results demonstrated clearly that Arg-606 is required for optimal transmalonylation but not for transacetylation.
|
Kinetic analysis of the mutant
transacylases revealed that the Km for malonyl-CoA
was increased 8.5-fold by the Arg-606 Ala mutation but was
relatively unaffected by the Arg-606
Lys mutation. The value for
kcat was lowered about 100-fold by the Arg-606
Ala mutation and about 9-fold by the Arg-606
Lys mutation
(Table I). With acetyl-CoA as the substrate,
kcat increased significantly as a result of both
mutations, and the Km decreased slightly. As a
result of the Arg-606
Ala mutation, the calculated specificity
constant (kcat/Km) of the transacylase was decreased 3 orders of magnitude with malonyl-CoA as
the substrate but was increased almost 2 orders of magnitude with
acetyl-CoA as the substrate. On the other hand, replacement of Arg-606
with Lys had a much less drastic effect on the properties of the
enzyme: the specificity constant for malonyl-CoA was reduced <1 order
of magnitude, whereas that for acetyl-CoA was increased about 2-fold.
Thus, based on the calculated values for relative specificity
constants, the selectivity of the enzyme for acetyl-CoA was increased
>16,000-fold by the Ala mutation and 16-fold by the Lys mutation.
To assess the possibility that the observed lower activity toward
malonyl-CoA exhibited by the Arg-606 Ala and Arg-606
Lys mutant
enzymes might be due to increased hydrolase activity, we measured the
ability of these mutant enzymes to hydrolyze malonyl-CoA. However, the
activity of all enzyme preparations was extremely low: wild type, 5.1 milliunits/mg; Arg-606
Ala, 1.8 milliunits/mg; Arg-606
Lys, 2.4 milliunits/mg.
In the absence of a thiol
acceptor, almost 80% of the transacylase can be acetylated in the case
of both the wild type (10) and Arg-606 mutant enzymes (see Table III).
When the transacylase was incubated with equimolar amounts of
acetyl-CoA and malonyl-CoA, again almost 80% of both wild type and
mutant enzymes was acylated. Malonyl moieties accounted for 39% of the
occupied sites in the wild type transacylase, but only 4 and 11% in
the Arg-606 Ala and Arg-606
Lys mutant enzymes, respectively
(see Table III).
|
The surprising
observation that the catalytic efficiency of the transacylase toward
acetyl-CoA was actually increased by the replacement of Arg-606 with
either Ala or Lys prompted us to examine the effect of these mutations
on the ability of the enzyme to handle longer chain length acyl
moieties. In these experiments, we used acyl-S-pantetheine
thioesters of different chain lengths as substrates and CoASH as the
acceptor. With acetyl moieties as the translocated species, the
activity with CoA as the donor and pantetheine as the acceptor was
about 1 order of magnitude higher than when pantetheine was the donor
and CoA the acceptor (Table II). Under saturating
substrate conditions, the wild type enzyme exhibited low medium chain
transacylase activity, and the activity decreased as the chain length
of the acyl moiety was increased from 8 to 10 carbon atoms (Table II).
The Arg-606 Ala mutant exhibited transacylase activity >3 orders
of magnitude higher than that of the wild type enzyme. Replacement of
Arg-606 with Lys had a less dramatic effect, increasing octanoyl and
decanoyl transacylase activities approximately 2 orders of
magnitude.
|
Arg-606 is positionally conserved in all transacylases that function in the substrate-loading reaction of fatty acid and polyketide synthesis, regardless of whether the enzymes are specific for malonyl-CoA or methylmalonyl-CoA or possess dual specificity for acetyl- and malonyl-CoA. Conceivably, a conserved arginine residue might function to enhance catalytic activity by interacting with the substrate carbonyl, increasing its electrophilicity and promoting the nucleophilic attack by the active-site serine residue. Alternatively, by neutralizing the carboxylate anion of the malonyl moiety, the strongly basic guanidinium group could serve to anchor the malonyl moiety in the active-site pocket. If Arg-606 functioned in the first role, then it should participate in both the malonyl transacylase and acetyl transacylase reactions. Since mutation of Arg-606 negatively impacts only the malonyl transacylase activity, this scenario can be entirely discounted. The selective effect of Arg-606 mutation on malonyl transacylase activity is clearly most compatible with the second possibility, namely that this residue specifically forms a salt bridge with the free carboxylate group of the malonyl moiety. Since the arginine is also positionally conserved in polyketide synthases regardless of whether they preferentially utilize malonyl-CoA or methylmalonyl-CoA, it is likely that this residue plays the same role in facilitating the binding of both malonyl and methylmalonyl moieties. Recruitment of a methylmalonyl or malonyl extender unit by the polyketide synthases is clearly regulated at the level of the transacylase as illustrated recently by transacylase domain-swapping experiments (13). Nevertheless, malonyl- and methylmalonyl-specific transacylases share significant sequence similarity, and the structural basis for their different specificities has yet to be elucidated.
The rate of acylation of the active-site serine residue by acetyl and
malonyl moieties is too fast to be measured by conventional methods.
Earlier quenched flow studies on the whole fatty acid synthase have
yielded first order rate constants of ~150 s1 for
acetylation of the active-site serine; however, even at equilibrium only about 50% of the serine sites were occupied (14). Somewhat higher
stoichiometries for acylation of the isolated transferase domain
were obtained in this study, approximately 0.8 mol of substrate/mol of
enzyme. When both acetyl and malonyl moieties are offered
simultaneously to the wild type transacylase, malonyl moieties compete
effectively for binding to the active-site serine. However, in the
Arg-606
Ala enzyme, and to a lesser extent in the Arg-606
Lys
enzyme, the ability of malonyl moieties to compete with acetyl moieties is severely compromised (Table III). This finding is
consistent with the proposed role for Arg-606 in facilitating the
initial binding of malonyl moieties to the enzyme.
Removal of the positive charge at residue 606 by replacement with a neutral alanine also had the effect of increasing the acetyl transacylase activity of the enzyme. Indeed the activity toward other uncharged and more hydrophobic acyl moieties was increased even more dramatically so that the acyl chain specificity of the Ala-606 mutant is extended to medium chain lengths. Replacement of Arg-606 with the less basic, slightly smaller Lys residue produced an enzyme with substrate specificity intermediate between wild type and Ala-606 enzymes. Thus, the Lys-606 enzyme exhibited increased activity toward acetyl and medium chain acyl substrates without completely compromising malonyl transacylase activity.
The experiments described provide a rationalization for the ability of the FAS-related transacylases to catalyze the malonyl transfer reaction. Clearly, in the dual specificity malonyl-CoA/acetyl-CoA:acyl carrier protein S-acyltransferase associated with the multifunctional animal FASs, an additional residue, or residues, must serve to stabilize the acetyl moiety at the substrate binding site. We are presently attempting to identify these amino acids using the same strategy of combining sequence comparisons with mutagenesis experiments.
The finding that a single amino acid replacement, for example Arg-606
Lys, can effectively extend the acyl chain length specificity of
the transacylase invites speculation as to whether such mutations may
have occurred in nature. It has not escaped our attention that the
ruminant FASs are able to release medium chain length acyl moieties as
products by direct transfer to a CoA acceptor (15-18). This
transacylation reaction is catalyzed by the same transacylase domain of
the FAS responsible for substrate loading (19). This unique property of
the ruminant FASs has been attributed to the unusually permissive
substrate specificity of the transacylase domain. On the basis of the
results of our mutagenesis experiments, it is tempting to speculate
that the broadening of the acyl chain length specificity of the
ruminant FASs may have been achieved in evolution by a relatively
simple mutational event. This possibility can be evaluated by
comparison of the amino acid sequences of the transacylase domains of
FASs from ruminant and non-ruminant sources.
We are grateful to Dr. Andrzej Witkowski for providing the medium chain length acyl-S-pantetheine substrates and for helpful discussions. We thank Dr. Babak Oskouian for performing the photographic work.