Biochemical and Kinetic Characterization of the Eukaryotic Phosphotransacetylase Class IIa Enzyme from Phytophthora ramorum

Abstract

Phosphotransacetylase (Pta), a key enzyme in bacterial metabolism, catalyzes the reversible transfer of an acetyl group from acetyl phosphate to coenzyme A (CoA) to produce acetyl-CoA and P_i. Two classes of Pta have been identified based on the absence (Pta^I) or presence (Pta^II) of an N-terminal regulatory domain. Pta^I has been fairly well studied in bacteria and one genus of archaea; however, only the Escherichia coli and Salmonella enterica Pta^II enzymes have been biochemically characterized, and they are allosterically regulated. Here, we describe the first biochemical and kinetic characterization of a eukaryotic Pta from the oomycete Phytophthora ramorum. The two Ptas from P. ramorum, designated PrPta^II1 and PrPta^II2, both belong to class II. PrPta^II1 displayed positive cooperativity for both acetyl phosphate and CoA and is allosterically regulated. We compared the effects of different metabolites on PrPta^II1 and the S. enterica Pta^II and found that, although the N-terminal regulatory domains share only 19% identity, both enzymes are inhibited by ATP, NADP, NADH, phosphoenolpyruvate (PEP), and pyruvate in the acetyl-CoA/P_i-forming direction but are differentially regulated by AMP. Phylogenetic analysis of bacterial, archaeal, and eukaryotic sequences identified four subtypes of Pta^II based on the presence or absence of the P-loop and DRTGG subdomains within the N-terminal regulatory domain. Although the E. coli, S. enterica, and P. ramorum enzymes all belong to the IIa subclass, our kinetic analysis has indicated that enzymes within a subclass can still display differences in their allosteric regulation.

INTRODUCTION

Acetate production has been studied for many years in bacteria but has received less attention in eukaryotic microbes, even though acetate is produced as an important end product of energy metabolism in yeasts (1–3) and protists (4–6). Four different pathways for production of acetate from acetyl-coenzyme A (CoA) have been identified in eukaryotic microbes (7). ADP-forming acetyl-CoA synthetase (Acs) (EC 6.2.1.13) (equation 1) has been implicated in acetate production in amitochondriate protists and some species of archaea.

$$\text{Acetyl-CoA}+\text{ADP}+{\text{P}}_{\text{i}}\leftrightarrow \text{acetate}+\text{ATP}+\text{CoA}$$ (1)

Acetate:succinate CoA-transferase (EC 2.8.3.8) (equation 2) is present in kinetoplastids, Trichomonas species, the trematode Fasciola hepatica, the yeast Saccharomyces cerevisiae, and the rumen fungus Neocallimastix sp. strain L2 (8–12).

$$\text{Acetyl-CoA}+\text{succinate}\leftrightarrow \text{acetate}+\text{succinyl-CoA}$$ (2)

Acetyl-CoA hydrolase (EC 3.1.2.1) (equation 3) is involved in peroxisomal acetate formation in kinetoplastids (7).

$$\text{Acetyl-CoA}+{\text{H}}_2\text{O}\leftrightarrow \text{acetate}+\text{CoA}$$ (3)

Phosphotransacetylase (Pta) (EC 2.3.1.8) (equation 4) and acetate kinase (Ack) (EC 2.7.2.1) (equation 5) form a pathway for the interconversion of acetate and acetyl-CoA that was previously thought to be limited to bacteria and one genus of archaea but has now been shown to be present in eukaryotes, such as green algae and Phytophthora (13, 14).

$$\text{Acetyl-CoA}+{\text{P}}_{\text{i}}\leftrightarrow \text{acetyl}\hspace{0.17em}\text{phosphate}+\text{CoA}$$ (4)

$$\text{Acetyl}\hspace{0.17em}\text{phosphate}+\text{ADP}\leftrightarrow \text{acetate}+\text{ATP}$$ (5)

The Pta-Ack pathway is best understood in its roles in both acetate production and assimilation in Escherichia coli and other bacteria. The pathway is responsible for the production of acetate during mixed-acid fermentation under hypoxic conditions and in a metabolic overflow mechanism in which acetyl-CoA is diverted from the tricarboxylic acid (TCA) cycle when there is an imbalance between the rapid uptake of glucose and its conversion into products (15). Under high acetate concentrations, this low-affinity pathway can also be used for assimilation of acetate by its conversion to acetyl-CoA (16).

In eukaryotes, the Pta-Ack pathway has been investigated only in the green alga Chlamydomonas, in which two parallel Pta-Ack pathways have been identified (13). Proteomic studies have suggested that the Pat1-Ack2 pathway is localized to mitochondria (note that phosphotransacetylase is designated Pat in Chlamydomonas), and the Pat2-Ack1 pathway is localized to chloroplasts (13, 17). Acetate has been found to be one of the major fermentative products excreted by Chlamydomonas during growth under dark, anoxic conditions, and ACK1, ACK2, PAT1, and PAT2 transcript levels are increased, in agreement with a role for the Pat-Ack pathway in acetate production (18). The ack1 and pat2 mutants were the most vulnerable to anoxia, and strains could not be recovered after a 24-h exposure to anoxia (19). Far less acetate was produced in each of the mutants after imposition of anoxia, yet small amounts of acetate (<20%) were still produced in the ack1-ack2 double mutant, suggesting Pat-Ack is not the only pathway for acetate production in Chlamydomonas (19).

Analysis of bacterial and archaeal Pta sequences revealed two classes (20). Pta^I enzymes consist of a single catalytic domain, whereas the Pta^II enzymes have an additional N-terminal regulatory domain (20, 21). The Pta from the archaeon Methanosarcina thermophila is the best-studied class I enzyme (22–25), and several structures have been solved (26, 27). A ternary-complex mechanism based on kinetic and structural studies has been proposed for the enzyme (24). Two Pta^II enzymes have been characterized, one from Salmonella enterica (SePta^II) and one from E. coli (EcPta^II) (20, 21); however, a structure for a Pta^II has not been reported. The N-terminal regulatory domains of EcPta^II and SePta^II contain two recognizable subdomains designated the P-loop and the DRTGG subdomains (21). Truncations of the N-terminal domain of EcPta^II revealed that the P-loop subdomain is required for regulation of the enzyme by NADH, ATP, phosphoenolpyruvate (PEP), and pyruvate and that the DRTGG subdomain is vital for the sigmoidal response that is observed in allosteric enzymes (20, 21).

Here, we report the first biochemical and kinetic investigation of a eukaryotic Pta, the Pta^II enzyme from Phytophthora ramorum, a pathogenic oomycete that causes sudden oak death (28). P. ramorum has a single open reading frame (ORF) that encodes Ack and two ORFs that encode class II Ptas (here designated PrPta^II1 and PrPta^II2). Our characterization of PrPta^II1 demonstrates that the enzyme strongly prefers the acetyl-CoA/P_i-forming direction, unlike the S. enterica and E. coli enzymes. PrPta^II1 displays substrate cooperativity for acetyl phosphate and CoA and is allosterically regulated through inhibition by ATP, AMP, NADP, NADH, PEP, and pyruvate. Our phylogenetic analysis of the Pta family, the first reported that includes eukaryotic sequences, indicates there are four different subclasses of Pta^II based on differences in the N-terminal regulatory domain. A comparison of the bacterial and eukaryotic enzymes and the phylogenetic diversity suggests that the N-terminal domain and its regulatory role have evolved throughout the domains Bacteria and Eukarya.

MATERIALS AND METHODS

Materials.

Chemicals were purchased from Sigma-Aldrich, VWR, Fisher Scientific, and Gold Biotechnology. Oligonucleotide primers were purchased from Integrated DNA Technologies. A codon-optimized gene encoding P. ramorum PrPta^II1 (JGI 78441; http://genome.jgi-psf.org/Phyra1_1/Phyra1_1.home.html) was synthesized by GenScript and supplied in the E. coli expression vector pET21b, which provides for addition of a C-terminal His tag for use in nickel affinity column purification. Plasmid pPTA69 (kindly provided by Jorge Escalante-Semerena, University of Georgia) encodes S. enterica Pta^II with a His₆ tag fused to the N terminus of the protein (20).

Phylogenetic analysis of Pta.

BLASTP and TBLASTN (29, 30) were used to search the sequence databases at the National Center for Biotechnology Information (NCBI) (http://www.ncbi.nlm.nih.gov), the Broad Institute (http://www.broadinstitute.org/), and the U.S. Department of Energy Joint Genome Institute (http://genome.jgi-psf.org/) for putative Pta amino acid sequences, using M. thermophila Pta^I as the query sequence. Sequences were aligned using ClustalW with the Gonnet protein weight matrix, a gap-opening penalty of 10.0, and a gap extension penalty of 0.2 (31). Phylogenetic analysis of the aligned sequences was performed with the MEGA 5.1 program (32) using the neighbor-joining algorithm with partial deletion estimates. Five hundred bootstrap replicates were executed, and bootstrap values of 75% or greater are shown. The EMBOSS Needle Pairwise Sequence Alignment program (http://www.ebi.ac.uk) was used to determine sequence identity and similarity (33). The presence or absence of the P-loop and DRTGG subdomains in each Pta^II sequence was analyzed using the Pfam Protein Families Database (http://pfam.xfam.org/search) with an E value cutoff of E−10 (34).

Heterologous production and purification of Pta.

Recombinant plasmids were introduced into E. coli Rosetta 2(DE3) pLysS, and cells were grown in Luria-Bertani broth with 50 μg/ml ampicillin and 34 μg/ml chloramphenicol at 37°C with shaking at 200 rpm until an optical density at 600 nm (OD₆₀₀) of 0.6 to 0.8 was reached. Production of recombinant Pta was induced by addition of isopropyl-β-d-thiogalactopyranoside to a final concentration of 1 mM.

Following overnight incubation at ambient temperature, cells were harvested by centrifugation. The cells were resuspended in buffer A (25 mM Tris, 150 mM NaCl, 25 mM imidazole, 10% glycerol, pH 7.5), disrupted by three passages through a chilled French pressure cell at 138 mPa, and centrifuged at 100,000 × g for 90 min at 4°C. The supernatant was applied to a HisTrap HP Ni²⁺ affinity column (GE Healthcare, Inc.) equilibrated with buffer A. The protein was eluted using a linear gradient from 25 mM to 500 mM imidazole in 25 mM Tris-HCl, 150 mM NaCl, and 10% glycerol (pH 7.5). Fractions with Pta activity were pooled and dialyzed against buffer containing 25 mM Tris-HCl and 10% glycerol (pH 7.5). The recombinant enzyme was determined to be electrophoretically pure by SDS-PAGE. The protein concentration for purified PrPta^II1 was calculated from the absorbance at 280 nm using an extinction coefficient of 36,330 M⁻¹ cm⁻¹.

Activity assays.

PrPta^II1 activity in the acetyl-CoA-forming direction was measured by monitoring the increase in absorbance at 233 nm due to thioester bond formation (ε₂₃₃ = 5.55 mM⁻¹ cm⁻¹) (20, 21). The reaction mixture contained 50 mM Tris (pH 7.5), 20 mM KCl, 20 mM NH₄Cl, and 1 mM dithiothreitol (DTT), and the concentrations of acetyl phosphate and CoA were varied. Activity in the acetyl phosphate-forming direction was measured with two different assays. The thioester assay monitors the decrease in absorbance at 233 nm due to the release of CoA (ε₂₃₃ = 4.44 mM⁻¹ cm⁻¹) (22), and Ellman's thiol assay monitors the increase in absorbance at 412 nm due to the formation of the thiophenolate anion with DTNB [5′,5-dithiobis(2-nitrobenzoic acid)] (ε₂₃₃ = 14,150 M⁻¹ cm⁻¹) (20, 21). The reaction mixtures for both assays contained 50 mM Tris (pH 7.5), 20 mM KCl, 20 mM NH₄Cl, 1 mM DTT, and 1 mM DTNB, and the concentrations of acetyl-CoA and P_i were varied. The reaction mixtures for all assays were preincubated for 3 min at 37°C. Reactions were initiated by the addition of enzyme and were performed in triplicate.

SePta^II activity in the acetyl-CoA-forming direction was measured by monitoring the increase in absorbance at 233 nm. The reaction mixture contained 50 mM Tris (pH 7.5), 40 mM NH₄Cl, 1 mM CoA, and 3 mM acetyl phosphate. Reaction mixtures containing acetyl phosphate and enzyme were preincubated for 1 min at 37°C, and reactions were initiated by the addition of CoA. All reactions were performed in triplicate.

Assays were performed in 96-well plates, and the absorbance was monitored using the Synergy HT Multi-mode Microplate Reader (BioTek Instruments, Inc.). Data are expressed as means ± standard deviations (SD).

Kinetic analysis.

To determine apparent kinetic parameters for PrPta^II1 in the acetyl-CoA-forming direction, the concentration of one substrate was varied while the second substrate was held constant at saturating concentration (determined to be 4 mM for CoA and 5 mM for acetyl phosphate). KaleidaGraph (Synergy Software) was used to fit the data to the Michaelis-Menten equation (equation 6), where V₀ is the initial velocity, [S] is the substrate concentration, V is the maximum velocity, and K_m is the Michaelis constant.

$$V_0=V\times \left[S\right]/(K_m+\left[S\right])$$ (6)

When acetyl phosphate was varied, the data displayed positive cooperativity when fitted to the Hill equation (equation 7) (35, 36), where K_0.5 is the substrate concentration at half-maximal velocity and h is the Hill constant.

$$V_0=V\times {\left[S\right]}^h/({K_{0.5}}^h+{\left[S\right]}^h)$$ (7)

Determining the IC₅₀.

Metabolic intermediates, coenzymes, and nucleotide triphosphates were tested as allosteric effectors of PrPta^II1 and SePta^II. Substrate concentrations were held at saturating levels, and effector molecule concentrations were varied from 3 μM to 3 mM. Half-maximal inhibitory concentrations (IC₅₀s) were determined for all PrPta^II1 and SePta^II allosteric inhibitors by measuring the decrease in activity as a function of increasing inhibitor concentration. IC₅₀s were determined using GraphPad Prism 5 (GraphPad Software, Inc.) by fitting the data with a log [inhibitor]-versus-response curve.

Site-directed mutagenesis.

The QuikChange Lightning site-directed mutagenesis kit (Stratagene, Inc.) was used for mutagenesis according to the manufacturer's instructions. The primers used in the alteration of the Gly-300 codon were as follows: 5′PrPTAG300D, CACCTGAAAAAATATAAAGACGACGCGATGATTATCACCAGTGGT; 3′PrPTAG300D, ACCACTGGTGATAATCATCGCGTCGTCTTTATATTTTTTCAGGTG; 5′PrPTAG300A, CACCTGAAAAAATATAAAGACGCGGCGATGATTATCACCAGTGGT; 3′PrPTAG300A, ACCACTGGTGATAATCATCGCCGCGTCTTTATATTTTTTCAGGTG. Alterations were confirmed by sequencing at the Clemson University Genomic Institute (CUGI).

Gel filtration chromatography.

The native molecular mass of recombinant PrPta^II1 was examined by gel filtration chromatography using an Äkta fast protein liquid chromatography (FPLC) system with a Superose 12 column (GE Healthcare). The gel filtration column was calibrated with cytochrome c (12.4 kDa), carbonic anhydrase (29 kDa), albumin (66 kDa), amylase (200 kDa), apoferritin (443 kDa), and thyroglobulin (669 kDa) (Sigma-Aldrich Co.). The column was equilibrated with buffer containing 50 mM Tris and 150 mM KCl (pH 7.5) and developed at a rate of 0.5 ml/min.

RESULTS

Discovery of the different subclasses of Pta^II.

The Pta enzyme family was previously divided into two classes (20). Pta^I enzymes are approximately 350 amino acids long and consist of only a catalytic domain. Pta^II enzymes are approximately twice that size, with a C-terminal catalytic domain and an N-terminal regulatory domain (20, 21). Although Pta was commonly considered to be a bacterial enzyme, sequences have now been identified in the Eukarya (14). To determine how widespread the Pta enzyme family is in this domain, searches of the sequence databases were performed using the M. thermophila Pta^I sequence as the query. Pta sequences were identified in a number of eukaryotes, such as green algae, lycophytes, moss, and oomycetes, but were absent in fungi, diatoms, higher plants, and metazoans. Every completed eukaryotic genome that has a Pta sequence also has an ORF encoding a product with identity to Ack, consistent with these enzymes acting as a pathway, as in bacteria. In a phylogeny of Pta sequences constructed based on the catalytic domain, the Pta^II sequences form a separate clade (Fig. 1). All of the eukaryotic sequences belong to the Pta^II class, except for those from Perkinsus marinus, Emiliana huxleri, and Thecomonas trahens. The P. marinus Pta^I is a single-domain enzyme, whereas the E. huxleyi and T. trahens Pta sequences (shown in red in Fig. 1) are multidomain enzymes but are considered to be a Pta^I rather than Pta^II because the Pta catalytic domain is fused to Ack rather than an N-terminal regulatory domain, as for the Pta^II enzymes. The E. huxleyi enzyme has just the Pta and Ack domains, whereas the T. trahens Pta has four domains: a domain with unknown function, an Ack domain, a Pta domain, and a poly(R)-hydroxyalkanoic acid synthase domain.

FIG 1.

Phylogeny of the Pta^I and Pta^II family. The phylogenetic tree was constructed based on the sequences of the Pta catalytic domains. Pta^I sequences are shown in black, Pta^IIa sequences are shown in blue, Pta^IIb sequences are shown in purple, Pta^IIc sequences are shown in green, and Pta^IId sequences are shown in pink. The scale bar indicates the expected number of amino acid replacements per site. Only bootstrap values of 75% or higher are shown.

The N-terminal regulatory domains of EcPta^II and SePta^II contain two recognizable subdomains designated the P-loop and the DRTGG subdomains (21). The P-loop nucleoside triphosphatase (NTPase) subdomain contains a conserved nucleotide triphosphate-binding motif similar to that found in enzymes involved in translation, transcription, intracellular trafficking, membrane transport, and DNA replication and repair (37). The DRTGG subdomain has an unknown function and is named after some of its most conserved residues. This domain is related to the cystathionine-beta-synthase (CBS) domain that exists in both membrane-bound and cytosolic proteins and is known to function in eukaryotes, prokaryotes, and archaea (38).

Although the Pta^II N-terminal domains are similar in size, analysis of their sequences has revealed four subclasses based on the presence or absence of the P-loop and DRTGG subdomains (Fig. 2). The Pta^IIa N-terminal domain includes both subdomains, whereas the Pta^IIb, Pta^IIc, and Pta^IId subclasses lack one or both subdomains. The majority of Pta^II sequences belong to the Pta^IIa subclass, with Pta^IIb, Pta^IIc, and Pta^IId sequences distributed infrequently throughout the phylogeny. Surprisingly, the two Chlamydomonas reinhardtii Pta^II enzymes belong to two different subclasses, Pta^IId and Pta^IIc.

FIG 2.

Subdomain structures of the Pta^I and Pta^II classes. Pta^I enzymes have only a catalytic domain; Pta^II enzymes have a catalytic domain and an N-terminal domain. Four subclasses of Pta^II enzymes have been identified based on the presence or absence of the P-loop and DRTGG subdomains, as shown. The domains are not drawn to scale.

Purification and molecular properties of PrPta^II1.

The genome of P. ramorum (http://genome.jgi-psf.org/Phyra1_1/Phyra1_1.home.html) has two Pta^IIa ORFs, designated Pta^II1 (protein identifier [ID] 78441) and Pta^II2 (protein ID 78440). An E. coli codon-optimized gene encoding PrPta^II1 (GenScript Inc.) was cloned into pET21b (C-terminal His tag), and the recombinant enzyme was produced in E. coli and purified by nickel affinity chromatography to electrophoretic homogeneity. In determining the optimal reaction conditions for PrPta^II1, we tested the requirement for KCl and NH₄Cl, as other Pta^II enzymes have been shown to have increased activity in the presence of one or both salts (20, 21). Maximum activity was observed in the presence of 20 mM NH₄Cl and 20 mM KCl, and both salts are required. Increasing the concentration of either or both salts does not further increase enzyme activity. Optimal activity also requires the presence of 1 mM (final concentration) DTT. PrPta^II1 has the highest activity at 37°C; therefore, all experiments were conducted at 37°C. Due to aggregation issues, the molecular mass of the enzyme could not be determined by gel filtration. Brinsmade and Escalante-Semerena (20) encountered similar discrepancies with the S. enterica Pta.

Kinetic analysis of PrPta^II1.

In determining kinetic parameters for PrPta^II1 in the acetyl-CoA/P_i-forming direction, plots of substrate concentration versus velocity were sigmoidal. Apparent kinetic parameters were determined by fitting the experimental data to the Hill equation (equation 7), in which a Hill constant (h) greater than 1.0 represents positive cooperativity and a Hill constant less than 1.0 represents negative cooperativity (39, 40). PrPta^II1 exhibited positive cooperativity (h = 1.75 ± 0.18) with acetyl phosphate and slight positive cooperativity (h = 1.04 ± 0.05) with CoA (Table 1). The maximum activity observed in the acetyl phosphate/CoA-forming direction is approximately 3-fold lower than that in the acetyl-CoA/P_i-forming direction (8.9 ± 0.5 μmol min⁻¹ mg⁻¹ versus 23.8 ± 0.7 μmol min⁻¹ mg⁻¹, respectively), but kinetics parameters could not be determined in this direction due to an inability to reach saturation with inorganic phosphate.

TABLE 1.

Kinetic parameters for wild-type PrPta^II and the G³⁰⁰D variant in the acetyl-CoA-forming direction

Enzyme	K0.5 (mM)		kcat/K0.5 (s−1 mM−1)		Hill constant (h)
	CoA	AcP	CoA	AcP	CoA	AcP
Wild type	1.05 ± 0.31	1.85 ± 0.14	35.1 ± 6.0	18.04 ± 1.26	1.04 ± 0.05	1.75 ± 0.18
G300D	0.24 ± 0.05	1.75 ± 0.10	0.32 ± 0.06	0.043 ± 0.001	1.50 ± 0.05	6.50 ± 1.76

Allosteric regulation of the Pta^IIa enzyme family.

Although the catalytic domains of all Pta enzymes share strong identity (e.g., the catalytic domains of PrPta^II1 and SePta^II share 52.0% identity), the N-terminal domains within a subclass can differ substantially. The PrPta^II1 N-terminal domain shares only 19% identity with the N-terminal domains of the E. coli and S. enterica Pta^IIa enzymes, raising the question of whether this eukaryotic Pta is subject to allosteric regulation similar to that of the bacterial Pta^II enzymes. We tested nucleotides, coenzymes, and metabolic intermediates from both the glycolytic pathway and the citric acid cycle as effector molecules of PrPta^II1. NAD⁺, NADH, NADP, and NADPH were potent inhibitors, with NADP having the strongest effect (Fig. 3A). ATP, ADP, AMP, PEP, and pyruvate were also found to inhibit activity (Fig. 3B), but to a lesser extent than the nicotinamide derivatives. Fructose-1,6-bisphosphate (1 mM), oxaloacetate (1 mM), α-ketoglutarate (1.5 mM), and citrate (1.5 mM) had no effect on PrPta^II1 activity.

FIG 3.

Regulation of PrPta^II1 by allosteric effectors. PrPta^II1 activity in the acetyl-CoA-forming direction was monitored in the absence and presence of allosteric effector molecules. All data were normalized to the control, which represents the enzymatic activity observed in the absence of an effector molecule. (A) The results are displayed as percent activity in the presence of 100 μM and 750 μM NAD⁺, NADH, NADP, and NADPH. (B) The results are displayed as percent activity in the presence of 100 μM and 750 μM ATP, ADP, and AMP and 100 μM and 1,000 μM PEP and pyruvate. The data are expressed as means ± SD.

Although SePta^II and EcPta^II were shown to be allosterically regulated (20, 21), only ATP, NADH, PEP, and pyruvate were examined for EcPta^II, and only pyruvate and NADH were tested in the acetyl phosphate-forming direction for SePta^II. IC₅₀s were not reported for EcPta^II, and the IC₅₀s for SePta^II were reported only in the acetyl phosphate/CoA-forming direction. To allow direct comparison between the eukaryotic and bacterial enzymes, recombinant SePta^II was produced and purified; the effects of ATP, AMP, NADP, NADH, and PEP on SePta^II enzymatic activity were determined in the acetyl-CoA/P_i-forming direction; and IC₅₀s were determined (Table 2). Our results show that both SePta^II and PrPta^II1 are regulated by the same allosteric effectors, although the potencies differed. Because NADP was the most potent inhibitor of PrPta^II1, we evaluated its effect on substrate affinity. The K_m values for CoA were relatively unchanged by the presence of increasing concentrations of NADP, but the V_max decreased (Fig. 4A), suggesting mixed inhibition by NADP toward CoA. The K_0.5 values for acetyl phosphate increase and the cooperativity decreases as the NADP concentration is increased (Fig. 4B), indicating that NADP has a direct effect on the affinity of PrPta^II1 for acetyl phosphate.

TABLE 2.

IC₅₀s for Pta^II enzymes in the acetyl-CoA-forming direction

Effector molecule	IC50 (μM)
	PrPta	PrPtaG300D	SePta
ATP	736 ± 6	133 ± 8	364 ± 19
AMP	287 ± 4	53 ± 2	NDa
NADP	135 ± 4	81 ± 3	97 ± 8
NADH	275 ± 21	52 ± 2	24 ± 6
PEP	1,018 ± 128	74 ± 4	67 ± 3

^aND, IC₅₀ could not be determined due to the biphasic response.

FIG 4.

Effects of NADP on CoA and acetyl phosphate utilization by PrPta^II1. (A) Enzymatic activity was measured in the acetyl-CoA-forming direction in the presence of 5 mM acetyl phosphate with various concentrations of NADP using the thioester-forming assay. ●, 0 μM NADP; ■, 67.1 μM NADP; ◆, 134.2 μM NADP. (B) Enzymatic activity in the acetyl-CoA-forming direction was measured in the presence of 4 mM CoA with various concentrations of NADP using the thioester-forming assay. ●, 0 μM NADP; ■; 67.1 μM NADP; ◆, 134.2 μM NADP.

For PrPta^II1, activity steadily decreases as the AMP concentration is increased, similar to the behavior observed with other PrPta^II1 effectors. AMP has a much different effect on SePta^II, resulting in activation of the enzyme when present at low concentrations and inhibition at higher concentrations. SePta^II has 28% higher activity in the presence versus absence of 30 μM AMP (Fig. 5); however, as the concentration of AMP was increased to 100 μM, the activity returned to the level observed in the absence of effector. Activity is inhibited further as the AMP concentration is increased, with approximately 50% inhibition in the presence of 1 mM and complete loss of activity at 3 mM AMP. Pyruvate also has an unusual biphasic effect on SePta^II activity (Fig. 5). In the presence of 50 μM pyruvate, activity was 133% that observed in the absence of pyruvate. However, as the pyruvate concentration was increased further to 100 μM, activity began to decrease, and inhibition was observed in the presence of 500 μM pyruvate.

FIG 5.

AMP and pyruvate activate and inhibit SePta^II in the acetyl-CoA-forming direction. SePta^II activity in the acetyl-CoA-forming direction was determined in the presence of increasing concentrations of AMP (■) and pyruvate (●). All data were normalized to the control, which represents the enzymatic activity observed in the absence of an effector molecule. The data are expressed as means ± SD.

Analysis of PrPta1^G300D and PrPta1^G300A.

Brinsmade and Escalante-Semerena (20), using an S. enterica Acs knockout mutant, screened for Pta variants that supported the growth of the mutant on 10 mM acetate, conditions under which Acs would normally be responsible for acetate activation to acetyl-CoA. They isolated 10 single-amino-acid variants, for which the mutations all mapped to the N-terminal regulatory domain of Pta^II. Three of the variants (R²⁵²H, G²⁷³D, and M²⁹⁴I) displayed altered responses to at least one of the allosteric effectors. Differences in substrate affinities between the wild-type and variant enzymes were minor. A G²⁷³D variant had approximately 2-fold-increased V_max in the acetyl-CoA/P_i-forming direction and 3-fold-increased V_max in the acetyl phosphate/CoA-forming direction. This variant also showed approximately 3-fold stimulation by pyruvate versus 1.2-fold stimulation for the wild-type enzyme but remained subject to strong inhibition by NADH (20). This Gly residue is conserved in both PrPta^II1 and PrPta^II2, unlike the other two residues identified, which altered the wild-type SePta^II response to allosteric regulators.

To investigate the role of the conserved Gly residue in allosteric regulation of PrPta^II1, we targeted the corresponding residue, Gly³⁰⁰, for alteration. The G³⁰⁰A variant was soluble but was not active in either direction of the reaction. This result suggests either that this alteration resulted in complete loss of activity or that the structural integrity of the protein was compromised but not so severely that it was no longer soluble. Kinetic analysis of the PrPta^II1 G³⁰⁰D variant revealed a nearly 450-fold reduction in the catalytic rate in the acetyl-CoA-forming direction (Table 1), and activity was completely abolished in the acetyl phosphate-forming direction, indicating the importance of this residue in catalysis. A slight decrease in the K_0.5 for acetyl phosphate and a 4-fold decrease in the K_0.5 for CoA were observed for the G³⁰⁰D variant versus the wild-type enzyme. The k_cat/K_0.5 value for each substrate decreased significantly, with a 420-fold decrease for acetyl phosphate and a 110-fold decrease observed for CoA (Table 1). This alteration resulted in increased substrate cooperativity for both CoA and acetyl phosphate. The Hill constant for CoA increased by half, and that for acetyl phosphate increased 3.7-fold versus the wild type (Table 1).

We analyzed the effects of several allosteric inhibitors on the PrPta^II1 G³⁰⁰D variant and found that the IC₅₀s for ATP, AMP, and NADH decreased approximately 5-fold versus the wild type, whereas only a 1.6-fold decrease was observed for NADP (Table 2). A nearly 14-fold decrease was observed in the IC₅₀ for PEP (Table 2). Interestingly, nearly 50% inhibition was observed with 50 μM pyruvate, but activity increased as the pyruvate concentration was increased. At 1 mM pyruvate, enzyme activity had reached the uninhibited level, and by 3 mM pyruvate concentration, the activity was 124% that of the control (Fig. 6). The EC₅₀ for pyruvate was determined to be 503 ± 24 μM.

FIG 6.

Alteration of the Gly³⁰⁰ residue in PrPta^II1 changes the effect of pyruvate. Enzyme activity was determined in the presence of increasing amounts of pyruvate in the acetyl-CoA-forming direction for PrPta1 (●) and PrPta1^G300D (■). The data are expressed as means ± SD.

DISCUSSION

Phylogenetic analysis of Pta sequences shows that Pta^I and Pta^II sequences form separate clades (Fig. 1). Notably, nearly all of the eukaryotic enzymes belong to the Pta^II class, whereas both Pta^I and Pta^II enzymes are well represented in bacteria. In fact, E. coli and S. enterica possess both Pta^I and Pta^II enzymes. Based on these distributions and what is currently known about the roles of Pta in various bacteria and Chlamydomonas, it is difficult to determine what dictates whether an organism is more likely to have a Pta^I or a Pta^II enzyme.

Our sequence analysis of the regulatory domain of Pta^II enzymes has revealed four subclasses based on the presence or absence of the P-loop and DRTGG subdomains. Our phylogeny indicates that the eukaryotic enzymes group together within the Pta^II clade and that the eukaryotic Pta^IIa sequences are clustered together. Both of the P. ramorum Pta^II sequences belong to this subclass and fall within this cluster.

Like P. ramorum, C. reinhardtii and Volvox carteri also have two Pta^II enzymes. However, Pat2 from C. reinhardtii lacks the DRTGG subdomain and thus belongs to the Pta^IIc subclass, whereas Pat1 lacks identity to either subdomain within the N terminus and is classified as a Pta^IId enzyme. The V. carteri Pta^II sequences both belong to Pta^IIc and are most closely related to the P. ramorum enzymes. The Pta^IIa subclass is the most prevalent, followed by Pta^IId. The infrequent occurrence of Pta^IIb and Pta^IIc sequences and their distribution throughout the phylogeny suggest these subclasses arose recently and sporadically through the loss of domains.

Yang et al. (19) have shown that the C. reinhardtii Pta enzymes have different localization, with Pat1 localized to mitochondria and Pat2 localized to chloroplasts. The fact that these two enzymes belong to different subclasses most likely reflects different regulatory requirements in each organelle. The presence of two Pta enzymes in P. ramorum may also indicate different cellular localization. However, the fact that they both belong to the IIa subclass suggests that they are similarly regulated.

It is interesting that although Pta has been identified in the oomycete Phytophthora and in the green alga Chlamydomonas, it has not been identified in other stramenopiles, such as the diatoms. A search of diatom sequences revealed an Ack sequence in Phaeodactylum tricornutum; however, the only gene encoding an Ack partner enzyme present in that genome is that of xylulose 5-phosphate/fructose 6-phosphate phosphoketolase (Xfp), the partner enzyme for Ack in euascomycete and basidiomycete fungi. No ORFS for Ack or Pta have been identified in any other diatom sequences available.

Eukaryotic Pta^II versus bacterial Pta^II.

Our characterization of PrPta^II1 allowed us to investigate the similarities and differences between bacterial and eukaryotic Pta^II enzymes and provides us with a better overall picture of the regulation of this subclass. PrPta^II1, SePta^II, and EcPta^II all display positive cooperativity, but for different substrates. SePta^II and PrPta^II1 display positive cooperativity for acetyl phosphate but little or no cooperativity for CoA, whereas EcPta^II displays positive cooperativity for CoA but not for acetyl phosphate.

As EcPta^II and SePta^II are both allosterically regulated (20, 21), we examined whether PrPta^II1 is also subject to allosteric regulation and if the same metabolites are effectors for all three enzymes. ATP, NADH, PEP, and pyruvate were previously shown to be inhibitors of EcPta^II (21), and we have now shown that they are also inhibitors of SePta^II and PrPta^II1, although the order of potency of the inhibitors differs. We have also found that AMP influences both SePta^II and PrPta^II1 activities, but in different ways. AMP inhibits PrPta^II1 but activates SePta^II at lower concentrations and then inhibits activity at higher concentrations. AMP has a similar effect on glycogen phosphorylase (41). At low concentrations, AMP acts as a strong activator by binding to a site close to the subunit interface and increasing the release of glucose 1-phosphate from glycogen (42, 43). At higher concentrations, AMP binds to the ATP inhibitor site (44, 45) located at the entrance to the channel to the catalytic site (45), resulting in inhibition. By analogy, this may suggest that AMP and ATP bind at separate effector sites on SePta^II.

It is surprising that both Pta enzymes studied here display negative regulation by both NAD(P) and NAD(P)H and ATP, ADP, and AMP. Typically, an enzyme would be expected to respond differently to signals of high versus low cellular energy or the redox indicator pairs rather than to all of them. One possible explanation is that the regulatory domain signals a response to one set of signals (e.g., ATP and NADH) and the C-terminal catalytic domain is regulated by the opposite signals. However, no such regulation of class I Pta enzymes that consist of just the catalytic domain have been reported, so it is difficult to speculate further on this at this time.

The role of the P-loop and DRTGG subdomains in catalysis and allosteric regulation.

Through analysis of EcPta^II truncations, Campos-Bermudez et al. (21) demonstrated that the N-terminal domain is required for maximal catalytic activity and that the P-loop subdomain is required for the regulation of the enzyme by metabolic effector molecules. Using a positive-selection method to identify Pta variants that allow S. enterica growth at low acetate concentrations in an acs mutant, Brinsmade and Escalante-Semerena (20) identified three single-amino-acid variants altered in the N-terminal domain for which regulation by allosteric effectors differs. Of particular interest is the G²⁷³D SePta^II variant, which displayed much stronger activation by pyruvate than the wild-type enzyme but similar inhibition by NADH (20). This glycine is located in the DRTGG subdomain and is conserved among all Pta^IIa and Pta^IIb enzymes, including both enzymes from P. ramorum.

We altered Gly³⁰⁰, the equivalent residue in PrPta^II1, to Ala and Asp to investigate whether the enzyme is regulated similarly to SePta^II. The G³⁰⁰A variant lacked activity in either direction, and the G³⁰⁰D variant had greatly reduced activity in the acetyl-CoA/P_i-forming direction and no activity in the acetyl phosphate/CoA-forming direction. With the exception of pyruvate, each allosteric effector had an effect on the activity of the PrPta^II1 G³⁰⁰D variant similar to that on the wild type, although the IC₅₀s were reduced. Remarkably, pyruvate enhanced the activity of the variant in the acetyl-CoA/P_i-forming direction. However, activity was still very low in the opposite direction. In the corresponding SePta^II variant, this alteration greatly stimulated activity in the acetyl phosphate/CoA-forming direction. This result demonstrates that residue G³⁰⁰ in the DRTGG subdomain of the N-terminal regulatory domain is important across both bacteria and eukaryotes in determining how the enzyme reacts to the presence and absence of pyruvate. In addition, our results suggest that residues within the regulatory domain may also play a catalytic role, since the G³⁰⁰A variant lacks activity and the G³⁰⁰D variant has reduced activity in one direction and activity in the other direction was abolished, both in the absence of any effector molecules.

A possible physiological role for Pta in Phytophthora.

Our knowledge of Pta and its physiological role has largely been limited to bacterial enzymes. The Pta-Ack pathway has been shown to be essential for growth and invasion in several pathogenic bacteria, including Vibrio cholerae (46), uropathogenic E. coli (47), S. enterica (48), and Listeria monocytogenes (49). A eukaryotic Pta has not previously been characterized, and the only report on the physiological role of a nonbacterial Pta is from the green alga Chlamydomonas (19).

An important aspect of adapting to changing environments is the ability to use a variety of carbon sources, including acetate. In addition to having the Ack-Pta pathway utilized by bacteria to activate acetate under high-acetate concentrations, P. ramorum also has the ubiquitous AMP-forming Acs that a number of eukaryotic microbes use for acetate activation. The fact that PrPta^II1 preferentially operates in the acetyl-CoA/P_i-forming direction suggests the physiological role of the Ack-Pta^II1 pathway is in utilization of acetate as a carbon source. Perhaps, these pathways function in Phytophthora as in the bacteria E. coli, in which the high-affinity Acs pathway functions in acetate activation at low acetate concentrations and the low-affinity Ack-Pta pathway functions at high acetate concentrations (50, 51). However, there are no reports in the literature investigating the ability of Phytophthora to utilize acetate as a carbon source, leaving this question unanswered.

The presence of a second Pta raises the possibility that the two enzymes are specialized, with PrPta^II2 operating in the acetyl phosphate/CoA-forming direction. Alternatively, PrPta^II2 may catalyze both directions of the reaction, and metabolic effectors may regulate these activities, as well as the activity of PrPta^II1. The two P. ramorum Pta sequences share 82.3% identity overall, with 91% identity between the catalytic domains and 75.6% identity in the N-terminal domains. Thus, specialization in terms of directionality may not be as likely, and the presence of two Pta enzymes may instead reflect localization to different compartments within the cell, as seen in Chlamydomonas.

Concluding remarks.

The Pta^II class is much more complex than previously thought, with four subclasses, and Pta^IIa enzymes are subject to allosteric regulation by a number of effectors (52, 53). Regulation of enzymes from the other three subclasses and further study of the N-terminal domain are needed. An understanding of the regulation exerted by this domain would be greatly facilitated by having a structure of a Pta^II enzyme.