INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

The Escherichia coli BglF protein (EII^bgl), an enzyme II (EII) of the phosphoenolpyruvate (PEP)-dependent carbohydrate phosphotransferase system (PTS), catalyzes concomitant transport and phosphorylation of -glucosides (11). In addition, BglF regulates bgl operon expression by controlling the activity of the transcriptional regulator BglG. In the absence of -glucosides, BglF phosphorylates BglG, thus inactivating it; in the presence of -glucosides, BglF dephosphorylates BglG, which can then function as a transcriptional antiterminator and enable bgl operon expression (1, 2, 3, 24). Thus, BglF is the -glucoside phosphotransferase, the BglG kinase, and the phosphorylated BglG (BglG-P) phosphatase. A dimeric form of BglF can catalyze all these activities (7).

Like other EIIs of the PTS, BglF is composed of three domains. IIA^bgl possesses the first phosphorylation site, His-547 (site 1), which is phosphorylated by HPr; IIB^bgl possesses the second phosphorylation site, Cys-24 (site 2), which accepts the phosphoryl group from IIA^bgl and transfers it to -glucosides; and IIC^bgl, the membrane-spanning domain, presumably forms the sugar translocation channel and at least part of the sugar-binding site (9, 25). The order of these domains in BglF is IIBCA^bgl (reviewed in references 17 and 20). BglF uses site 2, Cys-24 on the IIB^bgl domain, to phosphorylate the two substrates, -glucosides and BglG (9), and to dephosphorylate BglG-P (10). A rearranged BglF protein which contains the three domains in the order IICBA^bgl (scrambled-BglF) catalyzes BglG phosphorylation but fails to carry out the sugar-induced reactions, i.e., sugar phosphotransfer and BglG-P dephosphorylation (8). These findings suggest that the structural requirements for the sugar-induced functions differ from those for the noninduced function. The key to BglF stimulation, i.e., the sugar-induced change which switches it from a BglG kinase mode to a BglG-P phosphatase and sugar phosphotransferase mode, is not understood.

To define the domain(s) required for catalysis of the different functions of BglF, we subcloned and expressed the three individual domains, IIA^bgl, IIB^bgl, and IIC^bgl, as well as truncated BglF proteins which lack one domain (IIBC^bgl and IICA^bgl). We show that phosphorylated IIB^bgl alone can phosphorylate BglG in vitro and negatively regulate its activity as a transcriptional antiterminator in vivo. In contrast, IIBC^bgl is required for -glucoside phosphorylation in vitro and for -glucoside utilization in vivo. The same holds true for BglG-P activation by dephosphorylation. These results suggest that the immediate environment of the BglF active site changes upon sugar stimulation. This change induces specific interactions between the active site-containing domain, IIB^bgl, and the membrane-bound domain, IIC^bgl.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

Strains. The following E. coli K-12 strains were used. K38 (HfrC trpR thi ⁺) was obtained from C. Richardson. LM1 contains mutations in the nagE and crr genes, which code for EII^nag and IIA^glc, respectively (13). ZSC112G contains a deletion of the ptsG gene, which codes for IICB^glc (6). PPA501 contains a mutation in the bglF gene and carries a bgl'-lacZ fusion on its chromosome (9). SG13009 was obtained from Qiagen.

Plasmids. The plasmids that encode the BglF derivatives used in this study are listed in Table 1. Plasmid pACYC184 was obtained from New England Biolabs. Plasmid pQE-30, which contains a translation start site followed by a sequence coding for six histidines and a multicloning site, and plasmid pREP4, which carries the lacI gene encoding the lac repressor, were obtained from Qiagen. Plasmid pT712, which contains the phage T7 late promoter, and plasmid pGP1-2, which carries the T7 RNA polymerase gene under the control of the cI857 repressor, were obtained from Bethesda Research Laboratories. Plasmid pT7OAC-F carries the entire bglF gene cloned downstream of the T7 promoter in pT712 (1). Plasmid pT7CQ-F1, a derivative of pT7OAC-F, encodes BglF with His-547 mutated to Arg (H547R) (9). Plasmid pMN5 carries the entire bglF gene cloned in pBR322 (14). Plasmids pT7CQ-F3, pT7CQ-F4, pT7CQ-F5, pT7CQ-F6, and pT7CQ-F8 are derivatives of pT712 which code for IIB^bgl, IIC^bgl, IIA^bgl, IICA^bgl, and IIBC^bgl, respectively, from the T7 promoter (7). Additional plasmids were constructed as described below.

Chemicals. [-³²P]ATP (7,000 Ci/mmol) was obtained from ICN. [³⁵S]methionine (1,200 Ci/mmol) was obtained from Du Pont. PEP, pyruvic acid, and pyruvate kinase were obtained from Sigma. [³²P]PEP was prepared and separated from [³²P]ATP as described before (1). Purified enzyme I (EI) and HPr were obtained from J. Reizer. Maltose binding protein (MBP)-BglG was purified as described previously (9). Ni-nitrilotriacetic acid (NTA) resin was obtained from Qiagen.

Molecular cloning and -galactosidase assay. All manipulations with recombinant DNA were carried out by standard procedures (21). Assays for -galactosidase activity were carried out as described by Miller (16). Cells were grown in minimal medium supplied with 0.4% succinate as a carbon source.

[³⁵S]methionine labeling of BglF and its derivatives. Cells of strain K38, containing plasmid pGP1-2 and one of the plasmids carrying the bglF gene or its derivatives under the control of the phage T7 promoter (pT7OAC-F, pT7CQ-F3, pT7CQ-F4, pT7CQ-F6, or pT7CQ-F8), were induced and labeled with [³⁵S]methionine in the presence of rifampin (Sigma) as described previously (26). To study the stability of plasmid-encoded BglF derivatives, unlabeled methionine was added to a final concentration of 0.5 mg/ml to the growth medium (chase) following 2 min of pulse-labeling with [³⁵S]methionine, and aliquots were removed at various times for autoradiographic analysis.

Preparation of membrane fractions and cell extracts. Membrane fractions enriched for wild-type BglF or one of its derivatives containing the membrane-bound IIC domain (H547R, IICA^bgl, IIBC^bgl, or IIC^bgl) were prepared as described previously (1). Cell extract enriched for the soluble IIB^bgl polypeptide was prepared as described previously for BglG-enriched extract (1). The proteins were expressed from their respective genes cloned under T7 promoter control in plasmids pT7OAC-F, pT7CQ-F1, pT7CQ-F6, pT7CQ-F8, pT7CQ-F4, and pT7CQ-F3, respectively. Expression of T7 RNA polymerase from compatible pGP1-2 was induced thermally. Strain LM1 was used as a host in most cases. When needed, strain ZSC112G (with a deletion of the ptsG gene) was used as a host for IIBC^bgl overproduction from pT7CQ-F8. Membrane fractions lacking BglF were prepared from LM1/pGP1-2/pT712 and ZSC112G/pGP1-2/pT712 and were used in control experiments.

Purification of IIA^bgl. Expression and purification of IIA^bgl were carried out essentially as recommended by Qiagen with some modifications. One liter of SG13009/pREP4/pQE-F5 culture was grown to an optical density at 600 nm of 0.5 in Luria broth containing 200 µg of ampicillin and 30 µg of kanamycin per ml at 30°C. Isopropyl--D-thiogalactopyranoside (IPTG) was added to a final concentration of 0.1 mM, and growth was continued for 5 h. Cells were harvested by centrifugation at 4,000 × g for 20 min in the cold. The pelleted cells were resuspended in buffer S (50 mM sodium phosphate [pH 7.8], 300 mM NaCl, 0.2 mM phenylmethylsulfonyl fluoride) supplemented with 5 µg of DNase and 20 µg of RNase per ml and were broken by passing the suspension twice through a French pressure cell at 12,000 lb/in². Unbroken cells were removed by centrifugation at 10,000 × g for 20 min in the cold. The supernatant was mixed gently at 4°C for 16 h with 2 ml of a 50% slurry of Ni-NTA resin preequilibrated in buffer S. The resin was then packed in a column. The column was washed once with buffer S, once with buffer S containing 40 mM imidazole and 10% (vol/vol) glycerol, and once with buffer S containing 250 mM imidazole. Fractions collected during the last wash were analyzed on tricine-sodium dodecyl sulfate (SDS)-polyacrylamide gels, and those containing histidine-tagged IIA^bgl were dialyzed against buffer S to remove the imidazole. The protein concentration was determined by the Bradford assay with a kit purchased from Bio-Rad Laboratories.

In vitro phosphorylation. In vitro phosphorylation experiments were carried out essentially as described by Chen et al. (9). Briefly, membranes with a final protein concentration of 0.9 mg/ml were labeled by incubation at 30°C in a mixture containing 10 µg of EI per ml, 40 µg of HPr per ml, 10 µM [³²P]PEP, and PLB buffer (50 mM Na₂HPO₄ [pH 7.4], 0.5 mM MgCl₂, 1 mM NaF, 2 mM dithiothreitol). When indicated, cell extracts enriched for IIB^bgl and/or purified histidine-tagged IIA^bgl were added at final protein concentrations of 0.8 mg/ml and 72 µg/ml, respectively. After incubation for 10 min, reactions were either terminated by the addition of electrophoresis sample buffer or further incubated as described below. To study dephosphorylation by -glucosides, salicin was added to a final concentration of 0.2%, and incubation was continued at 30°C for 5 min. To study BglG phosphorylation, purified MBP-BglG in PLB buffer was added to a final concentration of 10 µM, and incubation was continued at 30°C for 15 min.

Electrophoresis and autoradiography. Electrophoresis of proteins was carried out on SDS-polyacrylamide gels (12) or on tricine-SDS-polyacrylamide gels (23). Samples were fractionated next to prestained low- or mid-range-molecular-weight markers (Amersham). After electrophoresis, gels were stained with Coomassie blue or directly dried and exposed to Kodak XAR-5 X-ray film at 70°C.

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

To define the domains required for the different functions of BglF, we subcloned the individual domains, IIA^bgl, IIB^bgl, and IIC^bgl, as well as the domain pairs, IICA^bgl and IIBC^bgl. We then examined the ability of the different polypeptides (i) to be phosphorylated in vitro in the presence of PEP and the general PTS proteins enzyme I (EI) and HPr; (ii) once phosphorylated, to transfer the phosphoryl group to -glucosides and to BglG in vitro; and (iii) to mediate -glucoside utilization and modulate BglG activity in vivo.

The stability of the different plasmid-encoded polypeptides was assayed with pulse-chase experiments. The proteins were labeled with [³⁵S]methionine and chased with unlabeled methionine. Samples removed at different times were analyzed by SDS-polyacrylamide gel electrophoresis. The results (Fig. 1) demonstrate that IIB^bgl, IIC^bgl, IICA^bgl, and IIBC^bgl expressed from the heat-inducible T7 promoter are very stable, like wild-type BglF, although they are produced at variable levels.

View larger version (36K): [in this window] [in a new window]   FIG. 1.   Individual BglF domains and domain pairs are stable. Expression of wild-type BglF, IIBbgl (B), IICbgl (C), IICAbgl (CA), and IIBCbgl (BC), was induced from pT7OAC-F, pT7CQ-F3, pT7CQ-F4, pT7CQ-F6, and pT7CQ-F8, respectively, in E. coli K38 cells harboring pGP1-2. The plasmid-encoded proteins were pulse-labeled with [35S]methionine for 2 min and chased by the addition of unlabeled methionine to the growth medium. Aliquots, removed at the times indicated, were analyzed on SDS-10% polyacrylamide gels (A) or on tricine-SDS-16.5% polyacrylamide gels (B). Autoradiograms of the gels are shown. An unstable protein (a truncated BglG protein) was used as a control for chase success (data not shown). Molecular masses of protein standards are given in kilodaltons.

Due to the low level of IIA^bgl produced under the control of the T7 promoter, we engineered histidine-tagged IIA^bgl expressed from an IPTG-inducible promoter (see Materials and Methods). The His-tagged IIA^bgl polypeptide was purified almost to homogeneity by affinity chromatography (Fig. 2A). The ability of His-tagged IIA^bgl to accept a phosphoryl group from HPr and to deliver it to site 2 of BglF was tested in vitro as follows. Purified His-tagged IIA^bgl was incubated with [³²P]PEP, purified EI and HPr, and membranes of strain LM1 (with deletions of the crr and nagE genes) enriched for a BglF mutant protein which lacks phosphorylation site 1 (H547R). The results presented in Fig. 2B demonstrated that His-tagged IIA^bgl enabled efficient phosphorylation of H547R (lane 5). H547R was not phosphorylated in this in vitro phosphorylation system when His-tagged IIA^bgl was omitted (Fig. 2B, lane 2). His-tagged IIB^bgl was engineered and purified in a similar manner.

View larger version (32K): [in this window] [in a new window]   FIG. 2.   Histidine-tagged IIAbgl is functional. (A) Expression of His-tagged IIAbgl was induced by IPTG from plasmid pQE-F5 in strain SG13009 harboring plasmid pREP4 (lane 1). His-tagged IIAbgl was purified on an Ni-NTA column (lane 2) (see Materials and Methods). Samples were analyzed on tricine-SDS-16.5% polyacrylamide gels, followed by Coomassie blue staining. (B). Membranes of LM1 cells that overproduced wild-type (WT) BglF or H547R proteins were incubated with [32P]PEP and purified EI and HPr for 10 min without (lanes 1 and 2) or with (lanes 4 and 5) His-tagged IIAbgl. No BglF, membranes from cells which did not overproduce BglF but which were otherwise identical to the other membrane preparations used in this experiment. Samples were analyzed on SDS-10% polyacrylamide gels, followed by autoradiography. Molecular masses of protein standards are given in kilodaltons.

IIB^bgl is sufficient for BglG phosphorylation, whereas IIBC^bgl is required for -glucoside phosphorylation in vitro. It was previously shown that His-547 in IIA^bgl accepts the phosphoryl group from HPr and transfers it to Cys-24 in IIB^bgl, which can then deliver it to -glucosides or to BglG (9). It has also been demonstrated that the dephosphorylation of BglF in the presence of -glucosides in vitro is a good indication of the ability of BglF to transfer the phosphoryl group to the sugar (1). We therefore tested the ability of different combinations of BglF domains to be phosphorylated in vitro and then to be dephosphorylated by the -glucoside salicin.

View larger version (39K): [in this window] [in a new window]   FIG. 3.   Phosphorylated IIBCbgl is required for the phosphorylation of -glucosides. Wild-type BglF and IIBCbgl (BC) were overproduced in ZSC112G, a ptsG strain. IIBbgl (B), IICbgl (C), and IICAbgl (CA) were overproduced in LM1, a crr and nagE strain. His-tagged IIAbgl (A) was purified on an Ni-NTA column. Mixtures of the indicated proteins were incubated with [32P]PEP and purified EI and HPr for 10 min. Incubation was continued with (+) or without () 0.2% salicin for 5 min. Samples were analyzed on SDS-10% polyacrylamide gels (A) or on tricine-SDS-8 to 20% gradient polyacrylamide gels (B). Autoradiograms are presented. Molecular masses of protein standards are given in kilodaltons. Control, membranes from cells which did not overproduce BglF or any of its domains but which were otherwise identical to the other membrane preparations used in this experiment. EI and BglF comigrated in the gel system used in panel A.

View larger version (52K): [in this window] [in a new window]   FIG. 4.   Phosphorylated IIBbgl is sufficient for the phosphorylation of BglG. Wild-type (WT) BglF, IIBbgl (B), IICbgl (C), IIBCbgl (BC), and IICAbgl (CA) were overproduced in strain LM1. His-tagged IIAbgl was purified on an Ni-NTA column. Mixtures of the indicated proteins were labeled as described in the legend to Fig. 3 and further incubated with MBP-BglG for 15 min. Proteins were fractionated on SDS-5 to 12% gradient polyacrylamide gels, followed by autoradiography. Molecular masses of protein standards are given in kilodaltons. No BglF, see the legend to Fig. 2. EI and BglF comigrated in this gel system.

IIB^bgl can negatively regulate BglG antitermination activity, whereas IIBC^bgl is required for -glucoside utilization and for the relief of BglG inhibition in vivo. To substantiate our in vitro results by in vivo studies, we analyzed the ability of different combinations of BglF domains to transfer -glucosides into the cell while phosphorylating them and to negatively regulate BglG activity as a transcriptional antiterminator.

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

The ability of the same active site on BglF to phosphorylate -glucosides and to reversibly phosphorylate BglG prompted us to investigate the structural basis for sugar-controlled differential phosphorylation. Dimerization of BglF could not explain the -glucoside-induced switch in BglF activity, since -glucoside does not affect the BglF dimeric state (7). In addition, both the sugar-induced and noninduced functions could be catalyzed by BglF dimers, which seem to form spontaneously (7).

The requirements for -glucoside phosphorylation and for BglG dephosphorylation, both stimulated by the sugar, are expected to be similar and to differ from the requirements for BglG phosphorylation, which occurs in the absence of the sugar. In light of our finding that only the sugar-stimulated activities are sensitive to changing the domain order within BglF (8), we speculated that the domains required for the sugar-stimulated functions might differ from those required for the nonstimulated functions. Of course, the IIB domain, which contains the active site, is expected to be required for all functions. We therefore attempted to define the domains which are required for the different functions by assaying the catalytic activities of individual domains or combination of domains, either covalently linked or produced in trans. Our results (summarized in Table 3) showed that intact IIBC^bgl is critical for the ability to implement -glucoside phosphorylation and BglG dephosphorylation, whereas BglG phosphorylation can be catalyzed by IIB^bgl alone. IIC^bgl could not rescue IIB^bgl and enable it to catalyze the sugar-stimulated functions, unless it was linked to it, i.e., IIBC^bgl. Nonetheless, the domain requirements for the nonstimulated function are quite loose; i.e., the domain which contains the active site, IIB^bgl, is sufficient. Phosphotransfer from IIA^bgl to IIB^bgl is the same in the presence and absence of IIC^bgl, suggesting that the activity of IIB^bgl as a phosphoryl acceptor is not affected by IIC^bgl.

These results agree with our previous observation that the order of the domains in BglF (BCA) is critical for -glucoside phosphorylation and BglG dephosphorylation, whereas BglG phosphorylation can be catalyzed by a scrambled-BglF derivative with the domain order CBA (8). One possible explanation for our previously and currently reported results is that -glucosides induce an interaction between the IIB^bgl and IIC^bgl domains. Such an interaction could induce a conformational change in BglF and might be the key to the sugar-induced signal transduction pathway that leads to the expression of the bgl operon. In fact, the catalysis of alternative functions due to different interactions between protein domains might be a general theme in the stimulation and/or regulation of diverse functions which are dictated by environmental or cellular conditions. Another possible explanation for our results is suggested by the reversible nature of the BglG phosphorylation reaction (10). The sugar substrate, by dephosphorylating BglF-P, shifts the equilibrium and leads to dephosphorylation of BglG-P. The role of the IIC domain in this process might be to present the sugar substrate to the phosphorylated IIB domain. The apparent requirement for a covalent connection between the two domains might reflect a requirement that they be in proximity and properly oriented for sugar phosphorylation to occur efficiently. The two explanations are not mutually exclusive. It is possible that a sugar-induced conformation of BglF orients IIC and IIB properly for sugar phosphorylation and BglG dephosphorylation.

A number of successful gene dissection and complementation experiments have been reported for other PTS sugar permeases. For example, IIA^mtl could restore the activity of a IICBA^mtl protein which lacks the first phosphorylation site (27), IIA^glc from Bacillus subtilis complemented E. coli IICB^glc (18), and IIA^glc complemented a BglF mutant lacking phosphorylation site 1 (9, 25). Separation of the B domain of the mannitol permease or of the glucose permease from the respective C domain had a more drastic effect, yet catalytic activity was retained to some extent upon their coexpression (6, 19, 27). Moreover, a circularly permuted derivative of the E. coli glucose permease in which the order of the domains is BC rather than CB, as in the wild-type protein, has activity comparable to that of the wild-type protein (15). Nevertheless, a mixture of IIA^mtl, IIB^mtl, and IIC^mtl showed 4% the activity of the mannitol permease, which contains all three domains (19). Similarly, a mixture of IIA^glc, IIB^glc, and IIC^glc showed 2% sugar phosphorylation activity (6). A fusion protein which incorporates all proteins and protein domains of the glucose-specific PTS into a single polypeptide chain with the domain order IIC^glc-IIB^glc-IIA^glc-HPr-EI increased phosphotransfer activity over that of an equimolar mixture of the isolated subunits. Therefore, the linking of functional domains confers certain advantages to a specific PTS permease by rendering it more efficient (15). We do not have enough evidence yet to decide whether sensitivity to domain splitting and splicing is a characteristic of permeases which recognize similar sugars, of permeases which have more than one function, or of permeases which are related, due to a different, yet unknown, reason.