Enzyme I from Escherichia coli A NOVEL ENZYME OF THE PHOSPHOENOLPYRUVATE-DEPENDENT PHOSPHOTRANSFERASE SYSTEM EXHIBITING STRICT SPECIFICITY FOR ITS PHOSPHORYL ACCEPTOR, NPr*

The phosphoenolpyruvate (PEP)-dependent phosphotransferase system (PTS) phosphorylates sugars and regulates cellular metabolic processes using a phosphoryl transfer chain including the general energy coupling proteins, Enzyme I (EI) and HPr as well as the sugarspecific Enzyme II complexes. Analysis of the Escherichia coli genome has revealed the presence of 5 paralogues of EI and 5 paralogues of HPr, most of unknown function. The ptsP gene encodes an EI paralogue designated Enzyme I (EI), and two genes located in the rpoN operon encode PTS protein paralogues, NPr and IIA, both implicated in the regulation of s activity. The ptsP gene was polymerase chain reaction amplified from the E. coli chromosome and cloned into an overexpression vector allowing the overproduction and purification of EI. EI was shown to phosphorylate NPr in vitro using either a [P]PEP-dependent protein phosphorylation assay or a quantitative sugar phosphorylation assay. EI phosphorylated NPr but not HPr, whereas Enzyme I exhibited a strong preference for HPr. These two pairs of proteins (EI/NPr and EI/HPr) thus exhibit little cross-reactivity. Phosphoryl transfer from PEP to NPr catalyzed by EI has a pH optimum of 8.0, is dependent on Mg, is stimulated by high ionic strength, and exhibits two Km values for NPr (2 and 10 mM) possibly because of negative cooperativity. The results suggest that E. coli possesses at least two distinct PTS phosphoryl transfer chains, EI3 NPr3 IIA and EI 3 HPr 3 IIA. Sequence comparisons allow prediction of residues likely to be important for specificity. This is the first report demonstrating specificity at the level of the energy coupling proteins of the PTS.

The bacterial phosphoenolpyruvate(PEP) 1 :sugar phosphotransferase system (PTS) is a complex protein system that mediates uptake and concomitant phosphorylation of carbohydrates (1)(2)(3). Characterized PTS proteins include the cytoplasmic Enzyme I and HPr, which lack sugar specificity, and membranous Enzyme II complexes, each specific for one or a few sugars. The latter complexes usually consist of three proteins or protein domains that are designated IIA, IIB, and IIC (4). The phosphoryl relay proceeds sequentially from PEP to Enzyme I, HPr, IIA, IIB, and finally to the incoming sugar that is transported across the membrane and concomitantly phosphorylated by IIC. The PTS is present in a wide variety of Grampositive and Gram-negative bacteria, but PTS protein homologues have not been found in archaea or eukaryotes. In addition to its primary functions in sugar transport, sugar phosphorylation, and chemoreception, the PTS is involved in regulatory processes such as catabolite repression and inducer exclusion (5,6).
Novel PTS proteins, NPr and IIA Ntr (paralogues of HPr and IIA Fru , respectively), are encoded by the npr and ptsN genes, respectively, localized to the rpoN operon of Escherichia coli, which also encodes the nitrogen-related factor, 54 (7)(8)(9)(10). ptsN deletion mutants lacking IIA Ntr exhibit a growth defect in the presence of an organic nitrogen source and a sugar or tricarboxylic acid cycle intermediate. Protein phosphorylation involving NPr and IIA Ntr was suggested to function in linking carbon and nitrogen metabolism, and IIA Ntr has also been implicated in the regulation of the essential GTPase, Era, which appears to function in cell cycle progression and the initiation of cell division (10,11). IIA Ntr homologues have been identified in numerous Gram-negative bacteria (10), and a link between the ptsN gene and nitrogen regulation has been suggested for Rhizobium etli (12), Pseudomonas aeruginosa (13), and Klebsiella pneumoniae (14). The crystal structure of Enzyme IIA Ntr has recently been determined (15).
Analysis of the E. coli genome revealed a gene, ptsP, encoding Enzyme I Ntr (EI Ntr ), consisting of 2 domains, an N-terminal domain of 127 amino acids homologous to the N-terminal "sensory" domain of the NifA protein of Azotobacter vinelandii (16) and a C-terminal domain of 578 amino acids homologous to all currently sequenced enzymes I. EI Ntr was suggested to serve a sensory function linking carbon and nitrogen metabolism (17). A mutation in the orthologous EI Ntr -encoding ptsP gene of A. vinelandii resulted in impaired metabolism of poly-␤-hydroxybutyrate as well as diminished respiratory protection of nitrogenase under carbon-limiting conditions (18).
In the present study we report the cloning and overexpression of the ptsP gene from E. coli, the purification of EI Ntr , and the characterization of its biochemical activities. The results of these studies establish for the first time the existence of parallel but independent PTS phosphoryl transfer chains in which distinct Enzyme I paralogues exhibit specificity for their cognate HPr paralogues. Residues are identified that may account for this specificity.
Plasmid pJRENtr used for overexpression of EI Ntr is described below. Purified ppGpp was kindly provided by Mike Cashel (National Institutes of Health, Bethesda, MD). [U-14 C]D-Mannitol (specific activity 32 mCi/mmol) and [␥-32 P]ATP (specific activity Ͼ4000 Ci/mmol) were obtained from ICN (Costa Mesa, CA). Other chemicals were of analytical grade.
Cloning of ptsP and Purification of EI Ntr -A DNA fragment containing the ptsP gene encoding EI Ntr was amplified by polymerase chain reaction using E. coli genomic DNA. The top strand primer (Ntr1) contained a NdeI site within the initiation codon (underlined) of ptsP, 5Ј-ACACGAATTCCATATGCTCACTCGCCTGCGCGAAATAG-3Ј. The bottom strand (Ntr2) contained a SalI site (underlined) 5Ј-CTAAGTC-GACATGATCCGCGCTATAACCCTCCGCGAA-3Ј. Amplification was performed with a Hybaid thermal reactor (Hybaid Ltd., Teddigton, Middlessex, United Kingdom) in a reaction mixture containing Taq DNA polymerase, 100 mM Tris/HCl, pH 8.8, 15 mM MgCl 2 , and 250 mM KCl (Stratagene, La Jolla, CA) in a total volume of 100 l. The amplification mixture was overlaid with 50 l of mineral oil and subjected to 30 cycles of amplification as follows: 1-min denaturation at 94°C, 1-min annealing at 55°C, and 2-min extension at 72°C. The polymerase chain reaction-amplified DNA was ligated to pCR2.1 TM (Invitrogen), and the NdeI-SalI fragment encompassing the complete ptsP gene was then excised from this plasmid and cloned between the NdeI and SalI sites of the overexpression vector pROEX TM -1 (Life Technologies, Inc.) to create the plasmid pJRENtr. Cloning of ptsP was confirmed by nucleotide sequencing using polymerase chain reaction dye terminator sequencing on an ABI 373 sequencer (Applied Biosystems) with a series of oligonucleotides specific for ptsP. Synthetic oligonucleotides were obtained commercially from either GenSet (La Jolla, CA) or Research Genetics, Inc. (Huntsville, AL).
E. coli DH10B was transformed with the expression vector pJRENtr encoding EI Ntr . DH10B cells bearing the overexpression plasmid were grown overnight at 37°C in LB containing ampicillin (100 g/ml) and diluted 50 -100-fold into fresh LB medium containing the same concentration of ampicillin. When the optical density of the culture growing at 37°C reached an A 600 of 0.5-1.0, isopropyl-1-thio-␤-D-galactopyranoside was added to a final concentration of 0.1 mM, and incubation was continued for an additional 3 h. Cells were collected by centrifugation, resuspended in TP buffer (50 mM Tris/HCl, pH 7.4, containing 0.2 mM phenylmethylsulfonyl fluoride), and ruptured as described previously (20). The crude extract was centrifuged for 20 min at 16,000 rpm (SS34 rotor, RC5-centrifuge; Sorvall-DuPont, Wilmington, DE), and the overexpressed EI Ntr , which was found primarily in the pellet, was solubilized with 6 M guanidinium hydrochloride. Following centrifugation for 20 min at 16,000 rpm (SS34 rotor, RC5-centrifuge), the supernatant containing the overexpressed protein was applied to a nickel-nitrilotriacetic acid resin (Novagen, Inc., Madison, WI) column equilibrated with 20 mM Tris/HCl (pH 7.9) containing 0.5 M NaCl and 6 M guanidinium hydrochloride. The column was washed with 10 volumes of 20 mM Tris/HCl, pH 7.9, containing 0.5 M NaCl, 20 mM imidazole, and 6 M guanidinium hydrochloride, and EI Ntr was eluted with 50 mM Tris/HCl buffer, pH 7.9, containing 1 M NaCl, 0.3 M imidazole, 10% glycerol, and 6 M guanidinium hydrochloride. The fraction containing EI Ntr was then dialyzed against 50 mM Tris/HCl buffer, pH 7.5, containing 0.1 mM phenylmethylsulfonyl fluoride, 0.1 mM EDTA, 1 mM dithiothreitol, and 10% glycerol.
Concentrations of soluble PTS proteins, EI Ntr , EI, NPr, and HPr, were determined according to the method described by Bradford (25). Protein concentrations in butanol-urea-extracted membranes were measured using the DC Protein Assay as described by the manufacturer (Bio-Rad). Bovine plasma ␥ globulin was used as a standard. Membrane samples and standards were boiled for 5 min in the presence of 1% SDS or N-octyl-3-D-glucopyranoside prior to the addition of the reagents.
Preparation of Butanol-Urea-extracted Membranes-Membranes containing high levels of Enzyme IICBA Mtl , used for assaying EI and EI Ntr , were prepared from strain SB2950, which lacks EI, HPr, and IIA Glc (26). Batch cultures (5 liters) of strain SB2950 were grown overnight in nutrient broth (Difco). Cultures were centrifuged at 16,000 ϫ g (GSA rotor, RC-5 centrifuge) for 5-10 min at 4°C, washed in mineral medium (50 mM potassium phosphate buffer, pH 7.5, containing 15 mM (NH 4 ) 2 SO 4 and 1.7 mM MgSO 4 ), and resuspended in 25 ml of 50 mM potassium phosphate buffer, pH 7.5, containing 2 mM DTT, 1 mM EDTA, and 10 g/ml DNase I. The cells were ruptured at 10,000 p.s.i. in a French pressure cell, and intact cells and cell debris were removed by centrifugation at 14,500 ϫ g (SS34 rotor, RC-5 centrifuge) for 10 min at 4°C. Membranes were recovered from the supernatant by centrifugation at 100,000 ϫ g (Ti rotor, L7-65 Ultracentrifuge, Beckman Instruments) and resuspended in a few ml of 50 mM potassium phosphate buffer, pH 7.5, containing 2 mM DTT and 1 mM EDTA. These membranes were extracted with urea and 1-butanol (27). Urea (480 mg/ml) was added and dissolved by stirring on ice. Then 1-butanol (40 l/ml) was added, and this solution was gently stirred for 2 h on ice. Mem- branes were recovered by centrifugation at 100,000 ϫ g (Ti rotor) for 4 h at 4°C. Extracted membranes from 5 liters of stationary phase cells were resuspended in 10 ml of 25 mM Tris/HCl, pH 7.5, containing 1 mM DTT and 1 mM EDTA to a final protein concentration of 11-32 mg/ml and transferred to boiled dialysis tubes (MWCO: 12-14.000, Spectrum Medical Industries Inc., Houston, TX). Membranes were dialyzed 3 times for 8 h against the same buffer and were then aliquoted in small plastic vials for storage at Ϫ20°C. Extraction of peripheral proteins from the membranes eliminated background sugar phosphorylation activity.
Sugar Phosphorylation Assay-The [ 32 P]PEP-dependent protein phosphorylation assay described above does not provide a rate but instead represents an equilibrium situation. To allow estimation of relative rates of phosphoryl transfer via various PTS proteins, a quantitative assay was required. We found that phosphorylation of [ 14 C]mannitol (or another PTS sugar) could be used for this purpose although the rate of phosphoryl transfer involving EI Ntr and NPr was very low relative to that involving EI and HPr. [ 14 C]Mannitol was selected as the phosphoryl acceptor because this sugar is PTS-specific,

FIG. 3. Dependence of EI Ntr activity on the concentration of PEP.
A, activity of EI Ntr was determined using the sugar phosphorylation assay as described under "Materials and Methods" with potassium phosphate as the buffer. Three experiments were carried out with different amounts of NPr added to the reaction mixture: ࡗ, 0.5 g; f, 1 g; and OE, 5 g. B, in these double-reciprocal plots (1/v versus 1/S), two straight lines can be drawn through the data for each concentration of NPr used, one corresponding to a low K m , low V max value and one corresponding to a high K m , high V max value. C, the intercepts on the y axis from B were plotted versus 1/NPr to give the absolute K m and V max values of 2 M and 0.3 mol [ 14 C]mannitol-phosphate/mg EI Ntr for the high affinity, low velocity curve (ࡗ) and values of 10 M and 1 mol [ 14 C]mannitol-phosphate/ mg EI Ntr for the low affinity, high velocity curve (f).

FIG. 4. Time courses of EI Ntr activity revealing that preincubation of EI Ntr with
NPr eliminates the lag phase. Activity of EI Ntr was determined using the sugar phosphorylation assay as described under "Materials and Methods." HEPES (250 mM) was used as the buffer. Reaction mixtures were preincubated for 1 h at 37°C in the absence of one or more selected component(s), and then at zero time the missing component(s) were added. The components that were missing during the preincubation are indicated to the right of the curves. In the controls, all components were present, but there was no preincubation period. and no phosphatase cleaving the product ([ 14 C]mannitol-1-phosphate) is known. The standard assay for PTS sugar phosphorylation employs membranes isolated from disrupted E. coli cells that overproduce several Enzyme II complexes because of a pts operon deletion (28). Because the inverted membrane preparations are "leaky," protein components and sugar phosphate products are not compartmentalized. Thus, this assay provides a useful measure of relative rates of overall phosphoryl transfer employing EI Ntr and NPr as well as EI and HPr, even though the former reaction is not physiologically significant.
The sugar phosphorylation assay used was modified from the method described by Kundig

pH and Divalent Cation Dependence of the EI Ntr -catalyzed
Reaction-To study the properties of EI Ntr , we developed a sugar phosphorylation assay coupling phosphoryl transfer from PEP via purified EI Ntr , NPr, and Enzyme IICBA Mtl to [ 14 C]mannitol (see "Materials and Methods"). Under the condition described by Kundig and Roseman (27) (20 mM potassium phosphate buffer, pH 7.5), EI Ntr showed only marginal activity, but activity was greatly enhanced using 250 mM potassium phosphate, pH 8.0. Potassium phosphate buffer could be replaced by HEPES but not by MOPS, TES, TRIZMA, or Tris. The HEPES-based buffer system was used when precipitation of cations by phosphate was problematic.
The pH range for EI Ntr was examined using the phosphatebuffered assay system. In contrast to Enzyme I, which has a pH optimum of 7.0 -7.5 (29,30), EI Ntr activity was optimal at pH 8.0 (Fig. 1). The dependence of EI Ntr activity on divalent cations was tested with the HEPES-buffered system (31,32). EI Ntr required Mg 2ϩ with saturation being observed at a concentration of about 2 mM (Fig. 2). Mn 2ϩ allowed activity of EI Ntr only at low concentrations (Fig. 2). In the presence of low concentrations of Co 2ϩ or Ni 2ϩ , activity could be measured, but at concentrations above 0.2 mM, strong inhibition was observed.
Kinetic Analysis of PEP-dependent EI Ntr -catalyzed NPr Phosphorylation-Previous reports on the Mg 2ϩ dependence of EI have shown that the divalent cation is required for the PEP-dependent autophosphorylation of EI but not for phosphoryl transfer between EI and HPr (31,32). The dependence of EI Ntr activity on the concentration of PEP was examined using the [ 14 C]mannitol phosphorylation assay in phosphate buffer with three different concentrations of NPr (Fig. 3).
Regardless of the concentration of NPr used, EI Ntr proved to be saturated at PEP concentrations higher than 10 mM. The data shown in Fig. 3A were converted to a double reciprocal plot (1/v versus 1/S, Fig. 3B). As can be seen, a single straight line could not be drawn through the data obtained with any one NPr concentration, but they could be fitted to two distinct lines with differing slopes. This feature is characteristic of an enzyme population that either consists of two distinct enzyme species with differing kinetic properties or of a single enzyme species exhibiting the property of negative cooperativity.
The y intercepts (1/V max apparent) shown in Fig. 3B were plotted versus 1/NPr for the two parts of the curves (Fig. 3C). Straight lines were observed in each case, and these lines extrapolated to give V max and K m values, respectively, of 1 mol of [ 14 C]mannitol phosphate/mg of EI Ntr /min and 10 M for the high velocity, low affinity curve, and 0.3 mol of [ 14 C]mannitol phosphate/mg of EI Ntr /min and 2 M for the low velocity, high affinity curve.
Time-dependent, NPr-dependent activation of EI Ntr -Analysis of time courses for EI Ntr activity showed that the enzyme exhibited a lag phase of 30 -40 min under our standard assay conditions (Fig. 4). When the otherwise complete reaction mixtures were preincubated for 1 h without either [U-14 C]D-mannitol or the extracted membranes, no lag for sugar phosphorylation was observed (Fig. 4). However, the absence of extracted membranes during preincubation yielded lower activity than when mannitol was omitted, suggesting that the membranes might facilitate EI Ntr -NPr association. In contrast, when either EI Ntr or NPr was absent during the preincubation, a lag was still observed. Preincubation of EI Ntr with NPr was therefore required for optimal activity, suggesting that a slow association of these two proteins accounts for the lag phase observed in the control sample.
Specificity of EI Ntr for NPr and of EI for HPr-Selective phosphoryl transfer by EI and EI Ntr could be demonstrated using the 32 P-protein phosphorylation assay. NPr was found to be a specific protein substrate of EI Ntr as the latter could not phosphorylate HPr (Fig. 5A, lanes 3 and 4). In contrast, Enzyme I was found to be specific for HPr and could barely phosphorylate NPr (Fig. 5A, lanes 1 and 2). These observations were confirmed using the [ 14 C]mannitol phosphorylation assay.
Search for Specific Inhibitors of EI Ntr -In a previous report, a regulatory function was suggested for the N-terminal domain of EI Ntr (17). We employed the sugar phosphorylation assay to screen various compounds for effects on the NPr-dependent phosphoryl transfer activity of EI Ntr . The following salts increased the activity of EI Ntr when added to the potassium phosphate (50 mM, pH 8.0) -buffered reaction mixture: (NH 4 ) 2 SO 4 , NH 4 Cl, Na 2 SO 4 , NaCl, K 2 SO 4 , KCl, and LiCl (tested in a concentration range of 10 -400 mM). This demonstrated that EI Ntr activity requires high ionic strength, but no evidence for a specific effect by any one ionic species was obtained. Other compounds (concentrations between 0.1 and 10 mM) tested in the presence of 20 mM Mg 2ϩ (to avoid Mg 2ϩ limitation because of chelation) were L-glutamine, L-glutamate, ␣-ketoglutarate, ATP, ADP, AMP, GTP, GDP, GMP, UDP, ppGpp, and adenosine 3Ј-phosphate 5Ј-phophosulfate. Only GDP and ppGpp at concentrations of 5 mM or higher had effects on the activity of EI Ntr (50 -60% inhibition at 10 mM). The inhibitory effect of GDP was confirmed using the [ 32 P]PEP-dependent protein phosphorylation assay. EI Ntr -dependent phosphorylation of NPr was significantly reduced although the amount of phosphorylated Enzyme I Ntr remained similar (Fig. 5B, lanes 1 and  2). Although phosphorylation of EI was not affected, the inhibition of NPr phosphorylation is not likely to be of physiological FIG. 7. Two parallel phosphorelay chains of the E. coli phosphotransferase system. The top scheme shows the classical PTS that is responsible for sugar group translocation. The bottom scheme shows the parallel sequence of phosphoryl transfer events indicated by the novel PTS constituents (EI Ntr and NPr) studied in this report that presumably have regulatory consequences. All of the arrows shown represent physiologically reversible phosphoryl transfer reactions except for IIC-catalyzed sugar phosphorylation. Some degree of nonspecificity has been demonstrated between the HPr and IIA protein constituents of these chains. IIA Ntr and IIA Mtl , though homologous, exhibit significant conformational differences that may account for their dissimilar functions (47). significance. Cyclic AMP and cyclic GMP were without effect in the concentration range 0.01-1 mM.
A number of redox cofactors were examined with respect to potential regulatory effects. The following compounds were tested in the concentration range of 1-10 mM: NAD, NADH, NADP, NADPH, and FAD. Of these compounds, only FAD was inhibitory (50% with 1 mM FAD). EI was similarly inhibited. DISCUSSION EI Ntr resembles the classical EI except for the N-terminal NifA-like putative sensory domain of EI Ntr . Both enzymes catalyze PEP-dependent phosphoryl transfer to an HPr-like protein in a pH-, salt-, and Mg 2ϩ -or Mn 2ϩ -dependent process (30,33,34). The effects of ionic strength may be because of stabilization of the dimeric forms of these enzymes, and Mg 2ϩ may similarly promote association (35)(36)(37). However, EI Ntr exhibits biphasic kinetics with respect to PEP concentration, whereas EI exhibits monophasic kinetics; EI Ntr exhibits essentially absolute specificity for NPr, whereas EI could phosphorylate NPr at a rate that was only a small fraction of that at which it phosphorylates HPr. The turnover number (38) for EI Ntr using the sugar phosphorylation assay proved to be 0.8 pmol/pmol EI Ntr /min, whereas that for EI is 1.34 nmol/pmol EI/min (39). This 1000-fold difference explains, in part, why EI is absolutely required for sugar phosphorylation under in vivo conditions. ptsI deletion mutants can be mutated so as to express an EI paralogue that can phosphorylate HPr (40), but preliminary evidence suggests that this enzyme is not EI Ntr . 2 The nearly absolute specificity of EI Ntr for NPr and of EI for HPr provides the first evidence that two different EIs in a single organism exhibit specificity at the level of their phosphoryl acceptor PTS proteins. This finding clearly leads to both functional and mechanistic predictions. EI Ntr does not appear to function in sugar phosphorylation and may function exclusively in regulation, possibly controlling the activities of NPr and IIA Ntr . These latter proteins are encoded within the rpoN operon of E. coli and have been implicated in the regulation of 54 -dependent transcriptional initiation of genes concerned with organic nitrogen utilization (10). Moreover, in A. vinelandii, the EI Ntr -encoding ptsP gene has been shown to play a role in poly-␤-hydroxybutyrate metabolism and respiratory protection of nitrogenase under carbon-limiting conditions (18). In K. pneumoniae, P. aeruginosa, and R. etli, IIA Ntr has been shown to function in capacities similar to those demonstrated for E. coli (12)(13)(14)41). EI Ntr , NPr, and IIA Ntr have all been identified in the fully sequenced genome of P. aeruginosa. 3 As for E. coli, these Pseudomonas proteins presumably comprise a phosphoryl transfer chain that functions in parallel with EI, HPr, and various IIA sugar proteins with entirely different physiological consequences (see Fig. 7).
Phylogenetic data have shown that NPr is a distant homolog of HPr (10), whereas EI Ntr is a distant homolog of EI (17). The mechanistic implications of our biochemical observations are that specific residues in EI versus EI Ntr and/or HPr versus NPr must control the interactions and phosphoryl transfer reactions between these proteins. Alignments of two segments of all sequenced EI Ntr s (Fig. 6A, top) with representative EIs (Fig.  6A, bottom) as well as of recognized NPrs (Fig. 6B, top) and HPrs (Fig. 6B, bottom) revealed such candidate residues. Residues that are fully conserved among all homologous proteins are presented in bold print, whereas residues that are fully conserved in one group, but of a different type in the other group, are shaded. In comparing EI Ntr with EI, we found that residues conserved in one of these two sets of proteins but not the other were scattered unevenly throughout the alignment. The greatest abundance of such residues was found to immediately surround the active site histidine (Fig. 6A, top), a region of catalytic importance. This fact is particularly significant as the active site region is well conserved among either the EIs or the EI Ntr s. The second region of the multiple alignment exhibiting a high frequency of residues conserved in only one set of these two proteins occurred far downstream of the active site histidine in a region of unknown function (Fig. 6A, bottom). Neither of these two regions is at the EI-HPr interface (42). In the region that defines the EI-HPr interaction surface, two residues were markedly different in the EIs versus the EI Ntr s. These residues are the (A/G)H dipeptide in the EIs, which corresponds to the (V/L)Y dipeptide in the EI Ntr s. Fig. 6B shows the full alignment of NPr sequences with HPr sequences. As for the EIs, the greatest abundance of residues fully conserved in the NPrs, but of a different nature in the HPrs, surrounds the active site histidine. These dissimilar residues may control relative rates of phosphoryl transfer. A second region exhibiting this characteristic is found C-terminal to the regulatory serine (43). Both of these regions comprise interaction sites with EI (42). Of possible significance to the specificity of EI Ntr for NPr is the AXXM sequence in the NPrs that is lacking in the HPrs (positions 49 -52 in the E. coli HPr). The C termini of NPrs additionally possess a fully conserved NXXFDE sequence that is not found in any of the HPrs. These observations provide a guide for site-directed mutagenic studies aimed at defining their functional differences.
The presence of a NifA protein-like putative sensory domain in EI Ntr led to the postulate that this domain might sense the availability of a ligand that could control the phosphoryl transfer activity of the enzyme. Our search for such a small compound has been unsuccessful until now. Thus, GDP exerted an inhibitory effect but only at superphysiological concentrations. FAD also exerted an inhibitory effect at superphysiological concentrations, but because this effect was shared with EI, it cannot be attributed to the presence of the N-terminal NifAlike domain. It is possible that we have not identified the correct effector molecule, but it is equally possible that the effector we seek is a macromolecule such as the PII nitrogen regulatory protein (44) rather than a small metabolite. Aravind and Ponting (45) have noted that the NifA domain is homologous to a family of so-called GAF domains found in a variety of signal-transducing proteins, but the significance of this observation to EI Ntr function is not known.
In summary, we have found that EI Ntr exhibits virtually absolute specificity for NPr, whereas EI is highly specific for HPr (Figs. 5 and 7). This finding demonstrates for the first time that a single organism can possess two complete and independent PEP-dependent phosphoryl transfer chains, which presumably serve completely different physiological functions. The fact that E. coli possesses 5 paralogues of both EI and HPr (46) now leads to the possibility of multiple parallel and independently functioning phosphoryl transfer chains, each playing a different physiological role in the cell. Determination of the physiological functions of these PTS phosphoryl transfer chains provides exciting challenges for the future.