Arginine Biosynthesis in Thermotoga maritima: Characterization of the Arginine-Sensitive N-Acetyl-l-Glutamate Kinase

Abstract

To help clarify the control of arginine synthesis in Thermotoga maritima, the putative gene (argB) for N-acetyl-l-glutamate kinase (NAGK) from this microorganism was cloned and overexpressed, and the resulting protein was purified and shown to be a highly thermostable and specific NAGK that is potently and selectively inhibited by arginine. Therefore, NAGK is in T. maritima the feedback control point of arginine synthesis, a process that in this organism involves acetyl group recycling and appears not to involve classical acetylglutamate synthase. The inhibition of NAGK by arginine was found to be pH independent and to depend sigmoidally on the concentration of arginine, with a Hill coefficient (N) of ∼4, and the 50% inhibitory arginine concentration (I_0.5) was shown to increase with temperature, approaching above 65°C the I_0.50 observed at 37°C with the mesophilic NAGK of Pseudomonas aeruginosa (the best-studied arginine-inhibitable NAGK). At 75°C, the inhibition by arginine of T. maritima NAGK was due to a large increase in the K_m for acetylglutamate triggered by the inhibitor, but at 37°C arginine also substantially decreased the V_max of the enzyme. The NAGKs of T. maritima and P. aeruginosa behaved in gel filtration as hexamers, justifying the sigmoidicity and high Hill coefficient of arginine inhibition, and arginine or the substrates failed to disaggregate these enzymes. In contrast, Escherichia coli NAGK is not inhibited by arginine and is dimeric, and thus the hexameric architecture may be an important determinant of arginine sensitivity. Potential thermostability determinants of T. maritima NAGK are also discussed.

Thermotoga maritima, one of the most highly thermophilic eubacteria (optimum growth temperature, 80°C) (20) and possibly one of the deepest branching and more slowly evolving of the eubacterial lineages (2), has been the subject of an already completed genome sequencing project (33) and, given its possible evolutionary position and biotechnological potential (19), is also the target of one of the few structural genomics projects being developed now (25).

To exploit to its maximum the completeness of the genomic information and the massive structural data expected, it would be highly desirable to have a detailed knowledge of the physiology and metabolism of T. maritima rather than the patchy existing knowledge (3). Concerning the object of the present work, arginine biosynthesis, crude T. maritima extracts were shown to exhibit (36) enzyme activity for N-acetyl-l-glutamate synthase (abbreviated acetylglutamate synthase), N-acetyl-l-glutamate kinase (NAGK), N-acetyl-l-ornithine:glutamate N-acetyltransferase (or, in short, transacetylase), ornithine transcarbamylase, argininosuccinate synthetase, argininosuccinate lyase, and carbamoyl phosphate synthetase but not to exhibit acetylornithinase activity, and thus it was proposed (36) that T. maritima, like Pseudomonas aeruginosa and many other organisms but unlike Escherichia coli (7), makes arginine by using a cyclic pathway, a pathway in which the acetyl group of acetylornithine is transacetylated to glutamate (Fig. 1). Indeed, putative genes have been identified in the T. maritima genome for all the enzymes of the cyclic pathway (Fig. 1) except acetylglutamate synthase (11, 27, 33), although BLAST searches for the enzyme yielded weakly significant hits in the putative gene for NAGK, argB (11), that must reflect homology between acetylglutamate synthase and NAGK (6), since we demonstrate here that the purified protein product of the cloned and overexpressed T. maritima argB gene is a genuine, highly active, specific, and thermostable NAGK.

FIG. 1.

Genes and enzymes of arginine synthesis in T. maritima. The flow of carbon is shown with thick lines. The four N-acetylated metabolites are represented along the outer circumference, to emphasize the recycling of the acetyl group, whereas the steps after ornithine production are shown inside the circle. The de novo synthesis of acetylglutamate is shown with dashed arrows, given the purely anaplerotic role expected for this reaction. Intermediates and immediate precursors are given in bold capital letters. Enzymes and the genes that code for them are in boxes. The names of the genes given in bold type are those generally used for the equivalent genes in other bacterial species. They are followed in parentheses by the corresponding gene denomination in the T. maritima gene list (33; http://www.TIGR.org). Note that argJ is shown to encode two enzyme activities and that two genes (carA and carB) are needed for making carbamoyl phosphate synthetase. The thick empty arrow that links l-arginine with NAGK and the nearby negative sign in a circle represent the feedback inhibition by arginine of NAGK. To highlight the paramount controlling role of NAGK, this enzyme (and its gene) is printed with larger type than the other enzymes. CoA, coenzyme A.

An important question that is clarified here is the mode of feedback control of arginine biosynthesis in T. maritima. In the organisms having a linear route of arginine biosynthesis (that is, in which acetylornithine is deacylated hydrolytically), such as E. coli, the target of feedback inhibition by arginine is the initial step, catalyzed by acetylglutamate synthase (7, 26), whereas in the organisms that recycle the acetyl group (1, 7, 8, 10, 18), such as P. aeruginosa, the key target of feedback inhibition is NAGK, since acetylglutamate synthase plays a purely anaplerotic role in these organisms (7). Thus, the report that the NAGK activity found in crude T. maritima extracts was not inhibited by arginine (36) was puzzling, since T. maritima appeared to use the cyclic route to make arginine, given the respective presence and absence of transacetylase and acetylornithinase activity in T. maritima extracts (36). We have examined here the arginine sensitivity of the purified T. maritima NAGK, finding that, indeed, this NAGK is potently and highly specifically inhibited by arginine, as expected if it belongs to a cyclic arginine biosynthetic route. We also show that the inhibition is sigmoidal and has a high Hill coefficient, that the arginine concentrations required for inhibition increase with temperature but are independent of pH, and that at the high living temperatures of T. maritima, arginine inhibits NAGK by increasing the value of the K_m for acetylglutamate.

An additional key point requiring clarification, given the existence of arginine-sensitive and insensitive NAGKs, concerns the nature of the physical determinants that render a NAGK sensitive to feedback inhibition by arginine. Although the structure at atomic resolution of E. coli NAGK was previously reported and substrate binding and catalysis by this enzyme were clarified (15, 34), E. coli NAGK is not inhibited by arginine, and thus its structure has yielded no clues concerning the physical bases of arginine inhibition. Our present results shed some new light on this question, since they strongly suggest that a hexameric quaternary structure may be a key trait of arginine-inhibitable NAGKs.

MATERIALS AND METHODS

Construction of NAGK expression vector (pNAGK-TM16).

The putative T. maritima argB gene, corresponding to gene TM1784 of the T. maritima gene list (33; http://www.TIGR.org), was PCR-cloned from the genomic DNA of T. maritima (provided by F. E. Jenney, University of Georgia) by using a high-fidelity proofreading thermostable DNA polymerase (Deep Vent; New England Biolabs), the forward primer derived from nucleotides 1760682 to 1760714 of TM1784 (5′GGAGGTACAGCATATGAGGATCGACACGGTCA3′), and the reverse primer derived from the complementary antiparallel sequence for nucleotides 1761527 to 1761539 of TM1784 (5′GTGTTCATCAGAAGCTTCTTTACCCCTCCAGTTCT3′). These primers encompass the beginning and the end of the putative open reading frame (underlined) and short flanking genomic T. maritima sequences, and they incorporate mutations (shown in italic) to introduce NdeI (direct primer) and HindIII (reverse primer) sites after the initiator and stop codons. The amplified fragment, digested with NdeI and HindIII and ligated with T4 ligase into the same sites of plasmid pET-22b (Novagen), was used to transform E. coli DH5α cells (Clontech), allowing the isolation of plasmid pNAGK-TM16, which carries in its insert the NAGK gene from T. maritima, as corroborated by automated DNA sequencing (DNA sequencing core facility, IBV-CSIC, Valencia, Spain).

Protein expression and purification.

Expression of the cloned putative argB gene was performed as described previously for human ornithine transcarbamylase (31), using overnight induction at 25°C with 0.01 mM isopropyl-β-d-thiogalactoside and E. coli BL21 (DE3) cells (Novagen) cotransformed with pNAGK-TM16 and with plasmid pGroESL (a pACYC184-derived expression plasmid encoding the E. coli chaperonins GroES and GroEL [16], provided by A. E. Gatenby, DuPont de Nemours, Wilmington, Delaware). The cells, harvested by centrifugation from 1.5-liter cultures, were resuspended at 4°C in 15 ml of 0.1 M Na phosphate-0.2 mM dithioerythritol, pH 7.0, per gram of cells and were disrupted by sonication on ice. The lysate was centrifuged for 30 min at 4°C and 35,000 × g, and the supernatant was incubated for 9 min at 80°C, chilled, and centrifuged again for 15 min at 4°C and 18,000 × g. All subsequent steps were done at 4°C. The supernatant was dialyzed overnight against a column buffer consisting of 20 mM Na phosphate-1 mM dithioerythritol, pH 8.0. It was then loaded onto a Q-Sepharose Fast Flow column (1 by 18 cm; Amersham Biosciences) preequilibrated with the buffer, the column was washed with 100 ml of buffer, and then a 400-ml linear gradient of 0 to 0.5 M NaCl in the same buffer was applied. The overexpressed, partially pure protein (monitored by sodium dodecyl sulfate-polyacrylamide gel electrophoresis [SDS-PAGE] as a band of 30 kDa) was eluted at approximately 0.15 M NaCl. Fractions containing the protein were pooled, dialyzed overnight against 20 mM Na phosphate-1 mM dithioerythritol-20 mM MgCl₂, pH 7.0, and loaded onto an Affigel Blue column (2 by 20 cm; Bio-Rad) preequilibrated with the same buffer. After the column was washed with 180 ml of buffer, a 400-ml linear gradient of 0.5 to 3 M NaCl in the same buffer devoid of MgCl₂ was applied to the column. The >95% pure protein was eluted at approximately 1.5 M NaCl, concentrated to >2.5 mg/ml by ultrafiltration (Amicon stirred cell and YM10 membrane), and supplemented with 10% (vol/vol) glycerol before storage at −20°C. The preparation of P. aeruginosa and E. coli NAGKs has been reported previously (12, 14).

Enzyme activity assays.

NAGK activity was measured with the hydroxylamine-containing colorimetric assay of Haas and Leisinger (17), which detects at 540 nm the formation of acetylglutamyl hydroxamate (extinction coefficient, 456 M⁻¹ cm⁻¹ [17]), although when indicated, ADP was measured (35) in samples taken from this assay mixture. To prevent enzyme inactivation associated with extreme dilution of the enzyme, 0.02 mg of bovine serum albumin (shown in preliminary tests to prevent dilution inactivation)/ml was added in all assays involving extreme dilutions. Unless otherwise indicated, the enzyme was incubated for 10 min at 37°C in the assay medium at pH 7.5, and the concentrations of ATP, acetylglutamate, and MgCl₂ were 10, 40, and 20 mM, respectively. When a pH of 6.5 was used, the assay mixture was brought to this pH with HCl. To determine the effect on the activity (assayed at either 37°C or 75°C) of various ATP and acetylglutamate concentrations, the concentration of acetylglutamate was varied between 2 and 80 mM at five fixed concentrations of ATP (range, 2.5 to 25 mM), keeping a constant excess of 10 mM MgCl₂ over the concentration of the nucleotide. One enzyme unit is the amount of enzyme that generates 1 μmol of product in 1 min. The program GraphPad Prism (GraphPad Software, San Diego, Calif.) was used for fitting to the arginine inhibition data the sigmoidal curves generated with the nonlogarithmic form of the Hill equation for inhibition, v_Arg = v_Arg=0 × {1 − [Arg^N/(I_0.5^N + Arg^N)]}, where N is the Hill coefficient, I_0.5 is the concentration of arginine that yields 50% inhibition, v_Arg is the velocity at a given concentration of arginine, and v_Arg=0 is the velocity in the absence of arginine. The same program was used for hyperbolic fitting of the substrate saturation data and for estimating V_max and K_m values from secondary hyperbolic plots of apparent V_max values for one substrate versus the concentration of the other substrate. In this way, the K_m value given for each substrate has been estimated at infinite concentration of the other substrate, and the V_max is the velocity extrapolated at infinite concentration of both substrates.

Analytical gel filtration chromatography.

Chromatography on a Superdex 200HR (10/30) column mounted on an ΔKTA fast protein liquid chromatography system (Amersham Biosciences) was carried out at 24°C at a flow rate of 0.25 ml/min, monitoring the optical density at 280 nm of the effluent. The same solution was used for running the column and for suspending the protein samples (0.2- to 0.3-ml samples containing 0.02 to 0.23 mg of protein), consisting of either 50 mM Tris-HCl-0.15 M NaCl, pH 7.5, or 0.1 M K phosphate, pH 7.0. When indicated, 10 mM arginine or a mixture of 40 mM acetylglutamate-10 mM ATP-20 mM MgCl₂ was included in the solutions, and in these cases the enzyme was incubated for 15 min at 24°C with these components before gel filtration. Molecular weight marker proteins, either commercial (Amersham Biosciences or Sigma) or produced in our laboratory (Pyrococcus furiosus carbamate kinase [35] and intact or truncated E. coli aspartokinase III [29]), were used for column calibration.

Other methods.

Cross-linking with dimethyl suberimidate (Pierce) and electrophoretic analysis of the cross-linked protein by SDS-PAGE in phosphate buffer were carried out as reported previously (9), using an NAGK concentration of 0.5 mg of protein ml⁻¹. In all other instances, SDS-PAGE was performed according to procedures described by Laemmli (23) with gels of 15% polyacrylamide concentration. Densitometry analysis of the digitized images of the Coomassie blue-stained gels was carried out with the program Sigmagel (Jandel Scientific). Protein concentrations were determined by the method described by Bradford (5) using a commercial reagent (Bio-Rad) and bovine serum albumin as a standard. Preparation of N-acetyl-d-glutamate from the racemic mixture by selective aminoacylase deacylation of the l form has been reported previously (4).

RESULTS

Expression and purification.

The T. maritima gene TM1784 (nucleotides 1760682 to 1761527 of the T. maritima chromosome), PCR-cloned into the pET-22b-derived plasmid pNAGK-TM16, yielded insoluble protein when overexpressed in E. coli BL21 (DE3) cells under standard induction conditions (12). However, when the expression was carried out using cells that also overexpressed the chaperonins GroES and GroEL from the cotransforming plasmid pGroESL, a large fraction of the product was soluble if the concentration of isopropyl-β-d-thiogalactoside was low and the induction temperature was 25°C (Fig. 2). The relatively high abundance of the overexpressed protein and its thermostability (see below) simplified its purification in only three steps (see Materials and Methods): (i) heat treatment, causing threefold purification with 95% yield; (ii) ion-exchange chromatography, resulting in 16-fold purified protein and 70% yield relative to the initial extract; and (iii) dye-affinity chromatography, giving >95% pure protein (measured by SDS-PAGE) (Fig. 2) in a yield of approximately 10 mg liter⁻¹ of initial culture. The mass of the isolated protein, determined by matrix-assisted laser desorption ionization-time of flight mass spectrometry, was 30,341 Da, in excellent agreement with the sequence-deduced polypeptide mass (30,344), and its N-terminal sequence determined experimentally in 35 cycles of Edman sequencing agreed exactly with the gene-deduced sequence. These findings indicated that the purified protein was the genuine product of the cloned gene, that it had no significant posttranslational modifications, and that it preserved uncleaved the N-terminal methionine.

FIG. 2.

NAGK purification. Shown are SDS-PAGE analyses after the various steps of the purification and of the purified and concentrated protein. Molecular weight marker proteins were from Sigma (Dalton Mark VII-L).

The purified protein is a highly thermostable NAGK.

As expected for NAGK, the purified protein yielded color in the classical hydroxylamine-coupled NAGK enzyme activity assay (17), corresponding to an amount of acetylglutamyl phosphate synthesized that was equivalent, within experimental error, to the amount of ADP produced (45 μmol of protein min⁻¹ mg⁻¹ at 37°C). Both color production and ADP release were strictly dependent on the addition of N-acetyl-l-glutamate to the assay, and this substrate could not be replaced by either (Table 1) its d isomer (the small amount of activity observed with this isomer was due to traces of the l isomer [4]), its one-less-carbon analog N-acetyl-l-aspartate, or the nonacetylated amino acid, l-glutamate, proving the exquisite specificity of the enzyme for N-acetyl-l-glutamate, as expected for genuine NAGK (18). Concerning the nucleotide substrate (Table 1), the enzyme used GTP much less effectively than ATP, and the pyrimidine nucleotides UTP and CTP were very poor substrates.

TABLE 1.

Substrate specificity of T. maritima acetylglutamate kinase

Experiment	Substrate or analog	Activity (%)
Experiment Ia	N-Acetyl-l-glutamate	100
	N-Acetyl-d-glutamate	3.5
	N-Acetyl-l-aspartate	0
	l-Glutamate	0
Experiment IIb	ATP	100
	GTP	16
	UTP	6
	CTP	4

^aActivity was assayed in the hydroxylamine-based assay at 37°C in the presence of 10 mM ATP and 3 mM N-acetyl-l-glutamate or the indicated analog.

^bActivity was assayed as in experiment I by using 40 mM N-acetyl-l-glutamate and 3 mM concentrations of the indicated nucleotides.

Given the hyperthermophilic character of T. maritima, the NAGK of this organism should be highly thermostable. This stability was confirmed by the experiments shown in Fig. 3A. Whereas E. coli and P. aeruginosa NAGKs were completely inactivated at 70°C in 10 and 30 min, respectively, the T. maritima enzyme retained approximately 100, 80, and 60% of its activity when heated for 1 h at 70, 80, and 85°C, respectively. Only at 90°C was the enzyme inactivated at a rate comparable to that of P. aeruginosa NAGK at 70°C. The enzyme exhibited an optimum assay temperature of 80°C (Fig. 3B), the optimal growth temperature of T. maritima (20), and although at 37°C its specific activity (45 U mg⁻¹) was somewhat lower than those of the NAGKs of P. aeruginosa (130 U mg⁻¹) and E. coli (64 U mg⁻¹ [14]) at the optimal growth temperatures of these organisms, the activity of the T. maritima enzyme was by far the highest (677 U mg⁻¹ at 80°C). An Arrhenius plot of activity versus the reciprocal of absolute temperature (Fig. 3B, inset) failed to reveal any break in the 25 to 80°C temperature range that might reflect a drastic phase transition.

FIG. 3.

Effect of temperature on the stability (A) and the activity (B) of T. maritima acetylglutamate kinase. (A) The purified T. maritima (Tma) enzyme, at a concentration of approximately 1 mg ml⁻¹ in 20 mM sodium phosphate-1 mM dithioerythritol, pH 7, was incubated at the temperatures specified, and samples were taken after the indicated periods for assays of enzyme activity. For comparison, similar incubations and assays were carried out at 70°C with the pure P. aeruginosa (Paer) and E. coli (Eco) acetylglutamate kinases. (B) Influence of the assay temperature on the enzyme activity of the T. maritima enzyme (given as a percentage of the activity at 80°C, 677 U of protein mg⁻¹). The inset gives the Arrhenius plot for the activity of the T. maritima enzyme in the temperature range 25 to 80°C. From the plot, an activation energy of 54 kJ mol⁻¹ K⁻¹ is calculated.

Inhibition of T. maritima NAGK by arginine.

Figure 4 shows that T. maritima NAGK is potently inhibited by arginine. Similar to what is observed with P. aeruginosa NAGK (24) (Fig. 4B), the inhibition of T. maritima NAGK was virtually complete and was sigmoidally dependent on the concentration of the inhibitor, but the concentration of arginine needed for inhibition at 37°C was considerably lower for T. maritima NAGK than for P. aeruginosa NAGK. However, as the temperature was increased, the concentrations of arginine needed to inhibit T. maritima NAGK also increased (Fig. 4A), and thus, at the high living temperatures of T. maritima, the arginine sensitivity of this NAGK resembled closely that of the NAGK of P. aeruginosa at 37°C.

FIG. 4.

Inhibition of T. maritima acetylglutamate kinase by arginine and the influence of the temperature (A) and pH (B) of the assay on this inhibition. (A) The indicated temperatures are given in degrees Celsius, and the curves are fitted to sigmoidal inhibition (see Materials and Methods) for the values of the Hill coefficient, N, and I_0.5 for arginine given in the inset. (B) The temperature used was 37°C, and the filled circles and squares denote the use of pH 7.5 and 6.5, respectively, in the assay of the T. maritima (Tma) enzyme. For comparison, the results of similar assays at these two pHs with the P. aeruginosa (Paer) enzyme are shown.

In contrast to the large increase with temperature of the I_0.5 for arginine (Fig. 4A, inset), the sigmoidicity of the inhibition was little affected by the temperature, as reflected in the relatively small change in the value of the Hill coefficient, N (see Materials and Methods), from 4.25 at 37°C to 3.45 at 75°C (Fig. 4A, inset). The sigmoidal character and the value of N (∼4) indicate that each enzyme molecule contains at least four equivalent sites for arginine to which the effector binds cooperatively (30).

Arginine inhibition of P. aeruginosa NAGK was previously reported (18) and has been confirmed here (Fig. 4B) to be strongly influenced by pH changes in the 6.5 to 7.5 range, as expected if arginine binding involves an ionizable group with a pK value within this pH range (possibly a histidine). No group with such characteristics appears to be involved in the inhibition by arginine of T. maritima NAGK, given the lack of influence on this inhibition of similar pH changes (Fig. 4B).

The replacement of l-arginine by arginine analogs revealed that arginine is a highly specific inhibitor of T. maritima NAGK (data not shown). Whereas the addition of 1 mM l-arginine caused 95% inhibition (assayed at 37°C), the same concentration of d-arginine or of agmatine (the product of arginine decarboxylation) caused no inhibition. The inability of agmatine to inhibit indicates that the α-carboxylate group of arginine is essential. However, the negative charge on the α-carboxylate appears unessential, since a 1 mM concentration of the methyl ester of l-arginine was strongly inhibitory (80% inhibition). In contrast, neither 1 mM citrulline nor 1 mM l-canavanine caused any substantial inhibition, and similarly, l-lysine, l-ornithine, putrescine, the guanidinium ion, and urea, tested at 1 mM concentrations, failed to inhibit the enzyme.

The enzyme exhibited hyperbolic kinetics for its two substrates irrespective of the assay temperature (tested at 37 and 75°C) and of the presence or absence of arginine (data not shown). At 75°C, arginine caused an important increase in the K_m of the enzyme for acetylglutamate (K_m^NAG), the only detrimental effect on enzyme functionality (Fig. 5, top panel), and thus, this mechanism may be the basic mechanism for the inhibition in vivo of T. maritima NAGK by arginine. However, at 37°C, in addition to increasing the K_m^NAG, arginine substantially decreased the V_max, and at high concentrations it also modestly increased the K_m^ATP (Fig. 5, bottom panel).

FIG. 5.

Influence of the concentration of arginine on the kinetic constants of T. maritima acetylglutamate kinase at 75°C (top panel) and 37°C (bottom panel). The kinetic constants were determined as described in Materials and Methods.

T. maritima NAGK appears to be hexameric, and arginine does not alter the aggregation state of this enzyme.

Cross-linking of T. maritima NAGK with dimethyl suberimidate resulted in the generation of multiple bands in SDS-PAGE (Fig. 6A) of decreasing intensity with increasing mass, migrating as expected for oligomers of 2, 3, 4, 5, and 6 subunits (the last is too faint to be seen in Fig. 6A). The decrease in the intensity with the increasing number of cross-linked subunits is to be expected for partial efficiency for the cross-linking of every pair of interacting chains. Thus, these results suggest that T. maritima NAGK oligomerizes at least to hexamers.

FIG. 6.

Investigation of the oligomeric state of T. maritima acetylglutamate kinase. (A) Cross-linking with dimethyl suberimidate (lane under the +) reveals bands by SDS-PAGE of up to five cross-linked subunits, with the band corresponding to the cross-linking of six subunits being too faint for reproduction (indicated with a 6 in parentheses). The lane under the − illustrates the migration of the enzyme when dimethyl suberimidate was omitted. (B) Gel filtration chromatography of the NAGKs from T. maritima (continuous line), P. aeruginosa (dotted line), and E. coli (dashed line), using a running solution of 50 mM Tris-HCl-0.15 M NaCl, pH 7.5. The minor peak observed with T. maritima NAGK that elutes approximately at the position of the E. coli enzyme does not consist of NAGK, as shown by SDS-PAGE of the collected fractions. (C) Semilogarithmic plot of molecular mass versus elution volume from the Superdex 200HR column. The filled circles correspond to the following protein standards: cytochrome C (12.3 kDa), lactalbumin (14.2 kDa), carbonic anhydrase (29.0 kDa), ovalbumin (42.7 kDa), bovine serum albumin (66.4 kDa), the dimer of bovine serum albumin (132.9 kDa), Pyrococcus furiosus carbamate kinase (68.8 kDa), intact (97.1 kDa) and truncated (31.9 kDa) aspartokinase III of E. coli, alcohol dehydrogenase (146.8 kDa), aldolase (156.8 kDa), amylase (223.8 kDa), catalase (230.3 kDa), ferritin (440 kDa), and thyroglobulin (669 kDa). The open circles and arrows denote the position of elution of the major peak of the acetylglutamate kinases of P. aeruginosa (Paer), T. maritima (Tma), and E. coli (Eco), assuming that the P. aeruginosa and T. maritima enzymes are hexamers (190.5 and 182.1 kDa, respectively) and that the E. coli enzyme is a dimer (54.3 kDa).

An independent confirmation that T. maritima NAGK is hexameric was obtained by using size exclusion chromatography on Superdex at 24°C, which revealed the elution of the enzyme as an approximately symmetrical peak (Fig. 6B) at a position fitting the expectation for a hexamer (Fig. 6C). Similar to T. maritima NAGK, the NAGK from P. aeruginosa behaved in gel filtration experiments as expected for a hexamer, whereas E. coli NAGK, which is insensitive to arginine and which was shown by X-ray crystallography to be dimeric (15, 34), was eluted, as expected for a dimer, much later than the T. maritima and P. aeruginosa NAGKs (Fig. 6B and C). The same elution patterns and positions were observed with T. maritima and P. aeruginosa NAGKs when the buffer used in the gel filtration experiments was 50 mM Tris-HCl-0.15 M NaCl, pH 7.5, as shown in Fig. 6, or when it was 0.1 M K phosphate, pH 7.0 (data not shown), as used previously in gel filtration experiments with P. aeruginosa NAGK (17). Similarly, the elution of these enzymes was not altered substantially by the addition to the enzyme solution and to the buffers of either 10 mM arginine or a mixture of 40 mM acetylglutamate, 10 mM ATP, and 20 mM MgCl₂ (data not shown). In the latter case, the nucleotide gave a high background optical absorption, but the position of elution of the enzymes was corroborated by collecting fractions and measuring the protein in the fractions with the Bradford assay.

DISCUSSION

Our demonstrations that gene TM1784 encodes an arginine-inhibitable NAGK, of the apparent absence from the T. maritima genome of a gene for acetylglutamate synthase, and of the likely possibility that, as in Thermotoga neapolitana (28), argJ encodes in T. maritima a bifunctional transacetylase/acetylglutamate synthase add up to strongly suggest that T. maritima represents the first example of a variant cyclic route of arginine synthesis (Fig. 1) that unites the characteristics of being controlled on the short-term by feedback arginine inhibition of NAGK and of using as an exclusive maker of acetylglutamate (either by recycling or de novo synthesis) the product of gene argJ.

The sigmoidal inverse dependency with a relatively high Hill coefficient between NAGK activity and arginine concentration is well suited for allowing arginine synthesis at basal cell concentrations of arginine and for stopping this synthesis rather abruptly above a certain arginine threshold, provided that this threshold is adequately set. The increase in the I_0.5 of arginine for T. maritima NAGK appears to be a necessary adaptation to set this threshold, at the high living temperatures of T. maritima, within the same range of controlling arginine concentrations as that for P. aeruginosa at 37°C. The increase in the I_0.5 with increasing temperature can be explained if the temperature influences the equilibrium between the high- and low-affinity conformations in which the arginine site must exist according to classical allosteric theory (30), with the high temperatures favoring the occurrence of the low-affinity conformation. This effect of temperature appears not to be an exclusive property of T. maritima NAGK, since the K_i value of arginine for Chlamydomonas reinhardtii NAGK was reported to increase when the temperature was raised from 15 to 37°C (10).

At 75°C, arginine is a K-type allosteric inhibitor (30) of T. maritima NAGK, since its only detrimental effect on the activity of the enzyme is to increase the K_m^NAG. In other well-studied examples of arginine-sensitive NAGKs (10, 18), the apparent affinity for acetylglutamate is also decreased by arginine, although in the case of P. aeruginosa NAGK this effect results from a change in the kinetics for acetylglutamate from hyperbolic to sigmoidal (18). The latter observation, the multiplicity of the kinetic effects of arginine at 37°C in T. maritima NAGK (decreased V_max, somewhat increased K_m^ATP, and increased K_m^NAG; see above), and the structural and chemical characteristics of the acetylglutamate site (characterized structurally in E. coli NAGK [15, 34]) appear to exclude direct physical competition between arginine and acetylglutamate for the same site as the mechanism by which arginine decreases the affinity for acetylglutamate. Indeed, the present and previous (24) findings highlight the exquisite specificity of arginine-sensitive NAGKs for the inhibitor, strongly supporting the existence of a distinct arginine site which, in accordance with the sigmoidicity and value of the Hill coefficient, must be present in T. maritima NAGK with a multiplicity of at least four nonindependent sites per enzyme molecule and which, given the results with arginine analogs (reference 24 and present results), must bind arginine in a highly stereospecific way according to the principle of three-point attachment (13).

Since from the present data the arginine-sensitive NAGKs from T. maritima and P. aeruginosa appear hexameric, whereas the arginine-insensitive NAGK of E. coli is dimeric, the hexameric architecture may be a key determinant of arginine sensitivity among NAGKs, in addition to providing the structural basis for the sigmoidal nature and high Hill coefficient of the inhibition by arginine. In earlier gel filtration experiments (17) performed at 4°C and taking many hours, arginine triggered the partial dissociation of P. aeruginosa NAGK oligomers (which in retrospect may be reinterpreted as hexamers). However, we find at 24°C and on a much shorter time scale (<1 h) no hexamer dissociation triggered by arginine with P. aeruginosa or T. maritima NAGK, and thus the differences between active and inactive NAGK must be subtler than hexamer disaggregation.

The present study reveals that the thermostability among the three NAGKs that have been studied here increases in the order E. coli < P. aeruginosa < T. maritima, conforming with the recent proposal that a high degree of aggregation is an important determinant of enzyme thermostability (38), since the less stable of these three NAGKs is a dimer whereas the other two appear hexameric (Fig. 3A). Comparison of the two hexameric NAGKs for classical markers of thermostability (21, 22, 32, 37) reveals that the T. maritima enzyme, which withstands temperatures 15 to 20°C higher than P. aeruginosa NAGK, has, relative to the latter enzyme, fewer glutamine residues (three versus eight; glutamines can be deamidated at high temperature) and a larger proportion of aliphatic (30 versus 25%) and charged (28 versus 23%) residues. T. maritima NAGK also differs in these traits from E. coli NAGK (which has 11 glutamines and 26% aliphatic and 21% charged residues), supporting the importance of such traits for NAGK thermostability. From these observations, it appears that in T. maritima NAGK the risk of thermal deamidation is decreased, the compactness and hydrophobicity of the enzyme interior is increased by the elevation in the proportion of aliphatic apolar residues, the number of ion pairs is also increased as revealed by the elevated number of charged residues, and the new intersubunit interactions found in the hexamer but not present in the dimeric E. coli NAGK strengthen and give further resilience to the structure. For a more direct analysis of the importance of these factors for the thermostability of NAGK, and for understanding in physical terms the mechanism of NAGK inhibition by arginine, the determination of the three-dimensional structures of the P. aeruginosa and T. maritima NAGKs appears necessary. Crystallographic studies with these enzymes are currently being carried out in our laboratory and might soon provide this much-sought structural information.