INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

The phosphoenolpyruvate (PEP):carbohydrate phosphotransferase system (PTS) catalyzes the uptake of carbohydrate concomitant with its phosphorylation. The phosphoryl group is transferred from PEP via the general energy-coupling proteins enzyme I and HPr to the sugar-specific phosphoryl transfer protein/domain IIA; phosphorylated IIA (IIA~P) transfers the phosphoryl group to the sugar-specific IIB protein domain which phosphorylates the sugar that is translocated via the sugar-specific IIC protein domain (23). Apart from its function in the uptake and phosphorylation of sugars, the PTS regulates transport and subsequent metabolism of non-PTS carbohydrates. In gram-negative enteric bacteria, this regulation is mediated by the phosphorylation state of the Glc-specific IIA (IIA^Glc), which is determined by the relative rates of phosphorylation by HPr~P and dephosphorylation by IICB^Glc. For instance, IIA^Glc~P is involved in the stimulation of adenylate cyclase, whereby the expression of many catabolic enzymes is regulated through changes in cyclic AMP (cAMP) levels (1).

Unphosphorylated IIA^Glc, on the other hand, binds directly to several transporters and enzymes of carbohydrate metabolism, and thereby inhibits their activities, via a phenomenon called inducer exclusion (23). The interaction of IIA^Glc with one of its targets, glycerol kinase (GlpK), has been elucidated by analyzing the crystal structure of Escherichia coli glycerol kinase in complex with E. coli IIA^Glc (7). This study revealed that IIA^Glc binds to glycerol kinase at a region that is distant from the catalytic site of glycerol kinase, which suggests that long-range conformational changes mediate the inhibition of glycerol kinase by IIA^Glc.

In gram-positive bacteria, not only can HPr be phosphorylated by PEP/enzyme I on a histidine residue (His-15), but also a metabolite-activated ATP-dependent protein kinase can phosphorylate a serine residue at position 46 (2). The serine-phosphorylated form of HPr [HPr(Ser-P)] seems to control carbohydrate metabolism both at the protein level and at the gene level, i.e., transport activities (inducer exclusion of both PTS and non-PTS sugars and/or inducer expulsion) and transcription (28).

There is no evidence for the involvement of IIA^Glc or IIA-like proteins in PTS-mediated regulation in gram-positive bacteria. However, several non-PTS sugar transporters have a carboxyl-terminal domain that is homologous to IIA^Glc of E. coli (18, 20). The best-characterized system of this family of transporters with a two-domain structure is the lactose transport protein (LacS) of Streptococcus thermophilus. This protein is, among others, also homologous to the melibiose transport proteins of Salmonella typhimurium and E. coli, which lack a IIA-like domain but are regulated by IIA^Glc (18, 41). The carboxyl-terminal IIA domain of the LacS protein of S. thermophilus (IIA^LacS) is located in the cytoplasm and has 34% residue identity with IIA^Glc of E. coli (20). IIA^LacS is phosphorylated, most likely at His-552, by HPr~P, which inhibits the transport activity of LacS (17, 19). This histidine residue corresponds with His-90 of IIA^Glc in E. coli, which has been shown to be the phosphoryl-accepting site (4, 24) (Fig. 1). The phosphorylation of LacS by HPr~P has been assessed only qualitatively, and it is not known whether the IIA^LacS domain has phosphoryl transfer activity equivalent to that of IIA^Glc.

In this study, we investigated (i) the kinetics of (de)phosphorylation of the IIA^LacS domain of S. thermophilus and (ii) the ability of the protein to carry out phosphoryl transfer to IICB^Glc of E. coli (PTS-mediated glucose transport) and to inhibit glycerol kinase. Information about the (de)phosphorylation kinetics is relevant because the only known function of the IIA domain in LacS, and homologous transporters, involves the regulation of lactose-H⁺ symport activity, for which a very rapid phosphoryl transfer may not be critical. For this study, we expressed IIA^LacS in E. coli and S. typhimurium and constructed two mutants in which either the proposed phosphoryl-accepting histidine was replaced by Arg [IIA^LacS(H552R)] or two residues near the putative phosphorylation site were replaced by the equivalent residues conserved in all PTS members of the IIA^Glc family [IIA^LacS(I548E/G556D)] (Fig. 1). The crystal structures of IIA^Glc from E. coli and B. subtilis show that the Glu residue is exposed to the surface of the molecule and may be critical for the interaction of IIA with its partner molecules; the Asp residue is close to the active site in the tertiary structure (12, 40). For the in vitro phosphorylation assays, we purified each of the mutant proteins and assessed phosphorylation by HPr~P from E. coli, B. subtilis, and S. thermophilus.

View larger version (48K): [in this window] [in a new window]   FIG. 1.   Alignment of the active site regions of PTS IIA and non-PTS IIA protein domains. *, conserved residue; :, similar residue; s, salt bridge with residue in glycerol kinase; h, hydrophobic interaction with glycerol kinase; z, coordination of Zn(II) (7); @, charged group near phosphorylation site. Only a portion of each PTS IIA protein domain in the SwissProt database is shown. Two-letter suffixes denote S. thermophilus (St), Pseudomonas aeruginosa (Pa), Lactobacillus bulgaricus (Lb), Leuconastoc lactis, (Ll), S. typhimurium (Sty), E. coli (Ec), B. subtilis (Bs), Klebsiella pneumoniae (Kp), Brevibacterium lactofermentum (Bl), and Corynebacterium glutamicum (Cg).

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

Bacterial strains and growth conditions. The strains used were E. coli DH5 [deoR endA1 gyrA961 hsdR17 (r_K m_K⁺) recA rel-1 supE44 thi-1 (lacZYA-arg-169) Ø80lacZ] (8), BL21 [hsdS gal( cIts857) ind-1 sam-7 nin-5 lacUV5-T7 gene1] (32), LM1 [crr-1 thi-1 his-1 argG6 metB galT rpsL ptsM manI nagE] (11), and MC1061 [(lacIPOZYA) F araD139 (ara-leu)7697 galU galK r_K⁺ m_K strA] (8) and S. typhimurium PP2178 [crr-307::Tn10 nagE142 trpB223] (36). To isolate plasmid DNA, the cells were grown in Luria broth under vigorous aeration at 37°C (29). For the transport assays, the cells were grown in minimal salts medium A (31) with glycerol (0.2% [wt/vol]) as carbon and energy source as described by van der Vlag et al. (37). For large-scale protein purification, the cells were grown in Luria broth in a 10-liter fermentor (Bio Bench ADI 1065; Applicon, Inc.) with oxygen supply (50% saturation) and pH control (pH 7.0). Growth on succinate was performed on agar plates containing minimal salts medium A supplemented with succinate (0.5%) and the essential nutrients as indicated by the autotrophic markers. When necessary, carbenicillin (50 µg/ml), chloramphenicol (10 µg/ml), tetracycline (12.5 µg/ml), or isopropyl--D-thiogalactopyranoside (1 mM) was added to the medium.

DNA manipulations. DNA modifications were performed as described by Sambrook et al. (29) unless indicated otherwise. Subcloning of plasmids into E. coli strains was performed via E. coli DH5. Plasmids used for the expression of proteins in E. coli and S. typhimurium are listed in Table 1.

Nomenclature. IIA^LacS refers to the IIA domain of LacS in general, whereas constructs representing protein fragments of specific lengths are indicated by numbers between brackets (e.g., IIA^LacS[181] denotes a domain of 181 residues). The His-tagged IIA^LacS[173] protein, which was used in most of the experiments, is referred to as IIA^LacS-6H; when appropriate, mutations are indicated between parentheses [e.g., IIA^LacS-6H(H552R) and IIA^LacS-6H(I548E/G556D)]. Ile-548, His-552, and Gly-556 denote residue positions in LacS; the same numbering is used to indicate these positions in IIA^LacS.

Protein purification. All purification procedures were carried out at 4°C unless indicated otherwise. Enzyme I and HPr of B. subtilis and E. coli were purified as described previously (26, 27, 38)). For the isolation and purification of HPr from S. thermophilus, the cells were lysed after lysozyme treatment (6). For the removal of lysozyme, the supernatant of the cell lysate fraction was diluted 1× in Milli Q water and incubated with S-Sepharose (10 ml/g of lysozyme) for 1 h at 4°C. Fresh S-Sepharose was added twice, after removal of old resin by decanting the supernatant, after which the cell lysate fraction was incubated for another hour at 4°C. The proteins were then precipitated by addition of ammonium sulfate to 80% (wt/vol), and incubation overnight on ice water. After centrifugation (45 min at 70,000 × g), the pellet was dissolved in 20 mM Tris-HCl (pH 8.5) and loaded on a DEAE-Sepharose fast flow column (1.6 by 40 cm; Pharmacia Biotech Inc.) that had been equilibrated with 20 mM Tris-HCl (pH 8.5). The column was washed with 10 column volumes of 20 mM Tris-HCl (pH 7.0). Proteins were eluted with 10 column volumes of 20 mM Tris-HCl (pH 7.0)-40 mM NaCl and precipitated by 80% ammonium sulfate as described above. The pellet was dissolved in 20 mM sodium acetate (pH 4.0) and desalted on a PD-10 column (Pharmacia Biotech). The resulting fraction was loaded onto an S-Sepharose fast flow column (HR 5/5; Pharmacia Biotech) that had been equilibrated with 20 mM sodium acetate (pH 4.0). The proteins were eluted with a 250-ml gradient of 0 to 250 mM NaCl in 20 mM sodium acetate (pH 4.0). The fractions containing HPr were pooled and concentrated by 80% ammonium sulfate as described above. The pellet was dissolved in 50 mM potassium phosphate (KP_i; pH 7.0) to a concentration of 2 mg/ml.

Phosphorylation and dephosphorylation assays. For the phosphorylation of IIA^LacS-6H, 5.8 µM purified IIA^LacS-6H was incubated in 50 mM Tris acetate (pH 7.5) containing 1 mM dithiothreitol (DTT), 2 mM MgCl₂, 0.8 mM purified enzyme I, 10 mM PEP, and HPr at concentrations ranging from 1 to 90 µM. The phosphorylation reactions were carried out at 10°C in a total volume of 10 µl. The reactions were stopped by addition of 10 µl of 2× sodium dodecyl sulfate (SDS) sample buffer (29), and the samples were stored on ice. For the dephosphorylation of IIA^LacS-6H~P, IIA^LacS-6H was first phosphorylated by PEP, enzyme I, and HPr, after which these components were removed by binding IIA^LacS-6H~P to the Ni-NTA resin. Briefly, 70 µl of reaction mixture was mixed with 40 µl of Ni-NTA that had been equilibrated with 50 mM KP_i (pH 7.0). Following a wash with 2 ml of 50 mM KP_i, IIA^LacS-6H~P was eluted with 120 µl of 50 mM KP_i (pH 7.0)-500 mM imidazole. For the dephosphorylation reactions, IIA^LacS-6H~P (10 mM or as indicated otherwise) was incubated in 50 mM Tris acetate (pH 7.5)-1 mM DTT-2 mM MgCl₂ and HPr in the range of 1 to 100 µM in a total volume of 10 ml at 10°C. The reaction was monitored as described above. The amount of (de)phosphorylated protein was determined by SDS-polyacrylamide gel electrophoresis (PAGE) analysis (15% polyacrylamide [10]) and Coomassie brilliant blue staining (29), and the amounts of IIA^LacS-6H and IIA^LacS-6H~P were determined by densitometry using a Dextra DF2400T scanner (Dextra Technology, Inc.).

Immunological methods. Immunodetection of wild-type and mutant IIA^LacS was performed with antibodies raised against a peptide corresponding to the carboxyl-terminal 17 residues of LacS (17) or antibodies raised against purified IIA^LacS-6H (this work). Immunodetection of HPr from S. thermophilus was performed with antibodies raised against HPr of S. salivarius. The proteins were separated by SDS-PAGE (15% polyacrylamide gel) and transferred to polyvinylidene difluoride membranes by semidry electrophoretic blotting. A Western-light chemiluminescence detection kit (Tropix Inc.) was used to visualize the proteins.

Miscellaneous. Uptake of labeled carbohydrates in intact cells was carried out as described by Postma (21). Protein quantification was performed by the D_c protein assay (Bio-Rad), using bovine serum albumin as the standard. N-terminal sequencing of proteins was performed by Eurosequence, Inc., Groningen, The Netherlands.

Materials. D-[U-¹⁴C]glucose (293 mCi/mmol) and [U-¹⁴C]glycerol (150 mCi/mmol) were obtained from the Radiochemical Centre, Amersham, United Kingdom. QIA quick spin columns and Ni-NTA resin were purchased from Qiagen, the Bio-Spin columns were from Bio-Rad Laboratories, PD-10 and MonoQ columns (HR 5/5) were from Pharmacia Biotech, 2-deoxy-D-glucose grade II and polyethyleneimine were from Sigma, and the enzymes needed for DNA manipulations were obtained from Boehringer Mannheim. All other materials were reagent grade and obtained from commercial sources.

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

Expression of the IIA^LacS. To study the functional properties of the IIA^LacS, this portion of the protein was expressed separately from the carrier domain. Although IIA^LacS is homologous to E. coli IIA^Glc and various other IIA proteins, the similarity at the amino-terminal end is not significant and the start of the linker region, connecting the carrier and IIA domain of LacS, is not well defined (20, 33). Therefore, we selected three translation initiation sites near the linker region such that IIA^LacS proteins 181, 173, and 154 amino acids long were obtained (Fig. 2). The individual proteins were tested for the ability to be phosphorylated by HPr~P, which was determined by monitoring the migration of the proteins on an SDS-polyacrylamide gel (4). The IIA^LacS proteins present in the cell extracts were detected by immunoblotting using an antibody directed against the carboxyl terminus of LacS. On an SDS-polyacrylamide gel, the IIA^LacS proteins migrated at a somewhat higher apparent molecular mass than predicted from the deduced amino acid sequence, which has also been observed for other IIA proteins (27). For IIA^LacS[173] and IIA^LacS[181], but not for IIA^LacS[154], a shift in the migration of the protein upon SDS-PAGE was observed after phosphorylation, indicating that IIA^LacS[173] and IIA^LacS[181] are phosphorylated by HPr~P. Further experiments were performed with IIA^LacS[173], i.e., the smallest fragment that could be phosphorylated and having an amino terminus of the same length as that of IIA^Glc.

View larger version (26K): [in this window] [in a new window]   FIG. 2.   Construction of gene fragments specifying IIALacS. The 4.1-kb EcoRI chromosomal DNA fragment of pSKE8, containing the lacS gene of S. thermophilus, is shown. Only relevant restriction sites are indicated. Shown are the lacS promoter (Ps), the lacS gene (arrow), wild-type protein and portions of the IIA domain (bars). The putative linker and flanking regions are indicated by their amino acid sequences. Amino acids that frequently occur in a Q linker are marked by asterisks, and the residues following the Met of IIALacS[181], IIALacS[173], and IIALacS[154] are indicated by the corresponding numbers. Phosphorylation of the protein was performed in the presence of enzyme I, HPr, and PEP (for details, see Materials and Methods). The ability of LacS to become phosphorylated by HPr~P was taken from reference 19.

Purification of IIA^LacS-6H. For the purification of IIA^LacS-6H, E. coli DH5/PSKIIAhis cells were grown to late exponential growth phase (Fig. 3, lane 2). After the cells were broken in a French pressure cell and cell debris/membranes and DNA were removed (Fig. 3, lane 3), the His-tagged protein could be isolated with a purity of 90% by nickel chelate affinity chromatography (lane 4). Further purification was achieved by anion-exchange chromatography using a MonoQ column (lane 5). The protein was eluted at approximately 150 mM NaCl. From 1 liter of cells grown to an A₆₆₀ of 4.5 in a computer-controlled fermentor and under vigorous aeration, 1.6 mg of IIA^LacS-6H was obtained. The amino-terminal sequence of the purified protein was Met-Glu-Leu-Glu-His-Arg, which is identical to the anticipated amino acid sequence (20); Glu-462 in LacS was replaced by Met to obtain the initiation codon of the IIA domain. Purification of the His-tagged IIA^LacS(H552R) and IIA^LacS(I548E/G556D) mutant proteins was performed exactly as for IIA^LacS.

View larger version (52K): [in this window] [in a new window]   FIG. 3.   Purification of IIALacS-6H. Shown is a Coomassie brilliant blue-stained SDS-15% polyacrylamide gel. Lane 1, protein marker; lane 2, cell extract of E. coli DH5/pSKIIAhis (50 µg of protein); lane 3, cytosolic fraction of E. coli DH5/pSKIIAhis (50 µg of protein); lane 4, IIALacS-6H after nickel chelate affinity chromatography (~10 µg of protein); lane 5, IIALacS-6H after nickel chelate and anion-exchange chromatography (~7 µg of protein).

Purification of HPr. HPr from S. thermophilus was analyzed in column fractions by native PAGE (15% polyacrylamide gel) and immunodetection using antibodies raised against HPr of S. salivarius. HPr was isolated from cells of S. thermophilus ST11 (lacS)/pGKhis that were grown to late exponential phase. After the cells were broken, most of the lysozyme present in the cell lysate was removed by adsorption to S-Sepharose. Subsequently, HPr was purified to near homogeneity in two steps, involving anion-exchange and cation-exchange chromatography. On an SDS-polyacrylamide gel, HPr migrated at ~13 kDa. This apparent molecular mass of HPr from S. thermophilus is in the same range as found for HPr purified from other streptococci, 6.7 to 17 kDa, while the molecular mass from the nucleotide sequence is 8.9 kDa (35).

Phosphorylation of IIA^LacS-6H by PTS-mediated enzymes. Phosphorylation of IIA^LacS-6H was analyzed by SDS-PAGE. Figure 4A shows the relative electrophoretic mobilities of purified IIA^LacS-6H in the absence (lane 2) and presence of PEP, enzyme I, and HPr of S. thermophilus (lane 4), B. subtilis (lane 6), and E. coli (lane 8). The phosphorylated form of IIA^LacS-6H exhibited a somewhat lower electrophoretic mobility than the nonphosphorylated protein (compare lanes 2 with lanes 4, 6, and 8). These experiments clearly indicate that the IIA domain of LacS can be phosphorylated via PEP and the general PTS energy-coupling proteins enzyme I and HPr of S. thermophilus as well as HPr of both E. coli and B. subtilis. Figure 4B shows the electrophoretic mobilities of the purified mutants IIA^LacS-6H(H552R) and IIA^LacS-6H(I548E/G556D). The mobility of IIA^LacS-6H(H552R) is similar to that of the wild type, whereas IIA^LacS-6H(I548E/G556D) exhibited a significantly lower mobility (lane 5). Upon addition of PEP, enzyme I of B. subtilis, and HPr of S. thermophilus, the migration of IIA^LacS-6H(H552R) was not affected (compare lanes 1, 2, and 3), whereas IIA^LacS-6H(I548E/G556D) migrated more slowly (compare lanes 5 and 7). Similar results were obtained with PEP, enzyme I, and HPr of E. coli and B. subtilis (data not shown). These results indicate that IIA^LacS-6H(I548E/G556D) but not IIA^LacS-6H(H552R) is capable of accepting the phosphoryl group from HPr~P.

View larger version (30K): [in this window] [in a new window]   FIG. 4.   Phosphorylation of IIALacS-6H, IIALacS-6H(H552R), and IIALacS-6H(I548E/G556D). The Coomassie brilliant blue-stained SDS-15% polyacrylamide gel represents samples containing 0.8 µM enzyme I (EI) from B. subtilis, 12.5 µM HPr, 5.8 µM IIALacS-6H, and/or 10 µM PEP, as indicated at the bottom. The phosphorylation reactions (at 37°C for 15 min) were carried out in 50 mM Tris-acetate (pH 7.5)-1 mM DTT-2 mM MgCl2. (A) IIALacS. (B) Lanes 1 to 3, IIALacS(H552R); lanes 4 to 6, IIALacS-6H(I548E/G556D). The sources of HPr (S. thermophilus, B. subtilis, and E. coli) are indicated above the lanes.

Kinetics of phosphorylation and dephosphorylation of IIA^LacS-6H and IIA^LacS-6H(I548E/G556D). The phosphorylation kinetics of IIA^LacS-6H by HPr~P from S. thermophilus is shown in Fig. 5. To determine the phosphorylation rates more precisely, the experiments were performed at 10°C, which decreased the phosphorylation rate approximately 1 order of magnitude compared to the level at 37°C. In these experiments, the concentration of HPr~P was 5- to 50-fold greater than that of IIA^LacS-6H, and the kinetics of IIA^LacS-6H phosphorylation could be approximated as a first-order process (Fig. 5A). The initial rates of phosphorylation of IIA^LacS-6H at different HPr~P concentrations and for HPr of S. thermophilus, E. coli, and B. subtilis are presented in Fig. 5B. Clearly, IIA^LacS-6H was phosphorylated approximately 1 and 2 orders of magnitude faster by HPr~P from S. thermophilus than by HPr~P from E. coli and B. subtilis, respectively. The derived rate constants obtained from these data are summarized in Table 2 (first column).

View larger version (18K): [in this window] [in a new window]   FIG. 5.   Phosphoryl transfer from HPr~P of S. thermophilus, B. subtilis, and E. coli to IIALacS-6H and IIALacS-6H(I548E/G556D). (A) Time course of the phosphoryl transfer reaction between 25 µM HPr~P from S. thermophilus and IIALacS-6H. (B) Phosphorylation rate of IIALacS-6H by HPr~P from S. thermophilus (), B. subtilis, (), and E. coli (). (C) Phosphorylation rate of IIALacS-6H(I548E/G556D) by HPr~P from S. thermophilus (), B. subtilis (), and E. coli (). The phosphorylation reaction was started by adding 5.8 µM purified IIALacS-6H or 4.1 µM IIALacS-6H(I548E/G556D) to 50 mM Tris acetate (pH 7.5) containing 1 mM DTT, 2 mM MgCl2, 0.8 µM purified enzyme I, 10 mM PEP, and HPr at concentrations ranging from 0 to 90 µM, in a total volume of 10 µl at 10°C. The reactions were stopped by addition of 10 µl of 2× SDS sample buffer.

5

1

View larger version (17K): [in this window] [in a new window]   FIG. 6.   Phosphoryl transfer from IIALacS-6H~P to HPr from S. thermophilus. The dephosphorylation reaction was started by adding 11 µM purified IIALacS-6H~P to 50 mM Tris acetate (pH 7.5) containing 1 mM DTT, 2 mM MgCl2, and HPr at concentrations of 10 µM (), 25 µM (), and 50 µM (). The reaction volume was 10 µl, and the temperature was 10°C. The inset shows the dephosphorylation of the IIALacS-6H in the absence of HPr.

Complementation of crr strains in trans. Since E. coli HPr~P could phosphorylate IIA^LacS, we next addressed the question of whether IIA^LacS could substitute for IIA^Glc in the phosphoryl transfer catalyzed by the glucose PTS. The phosphoryl transfer activity of IIA^LacS was studied in E. coli LM1 (crr manA), a strain defective in glucose transport due to a lack of IIA^Glc as well as a functional mannose PTS. Since the N-terminal residues of IIA^LacS could be important for the interaction with the membrane IICB^Glc protein, we tested IIA^LacS[181] and IIA^LacS[173] as well as IIA^LacS-6H(I548E/G556D) for the ability to restore glucose transport in E. coli LM1 (data not shown). The results were all negative, suggesting that these IIA^LacS proteins are unable to transfer the phosphoryl group rapidly enough to IICB^Glc even though they can be phosphorylated by HPr~P.

View larger version (20K): [in this window] [in a new window]   FIG. 7.   Inhibition of glycerol kinase activity by IIALacS. S. typhimurium PP2178 (crr::Tn10 nagE) was transformed with pKK223-3 (control) (A), pKKELE (IIALacS) (B), pSKIIAm2his [IIALacS-6H(I548E/G556D)] (C), and pSKIIAm1his [IIALacS-6H(H552R)] (D). Cells were grown overnight on a minimal salts medium supplemented with 0.4% DL-lactate. After being washed with minimal salts medium, the cells were diluted to an A660 of 0.35 in minimal salts medium containing 54 mM glycerol and incubated for 30 min at 37°C; glucose was added to a final concentration of 10 mM; the cells were incubated for 1 h at 37°C and then washed with minimal salts medium. Uptake assays with 0.5 mM [14C]glycerol were performed in the presence (open symbols) and absence (filled symbols) of 10 mM 2-deoxy-D-glucose; the cells were equilibrated in the presence or absence of 2-deoxy-D-glucose for 5 min prior to the initiation of uptake.

View larger version (19K): [in this window] [in a new window]   FIG. 8.   Relationship between the activity of glycerol kinase and the extent of inhibition of glycerol uptake by IIALacS () and IIALacS-6H(I548E/G556D) (). Cells were grown as described in the legend to Fig. 7. The data are from several independent experiments that reflect partial induction by glycerol for 15, 20, or 30 min. The glycerol kinase activity was measured as [14C]glycerol uptake rate in the absence of 2-deoxyglucose, and the inhibition of [14C]glycerol uptake was measured from the ratio of the rates in the presence and absence of 10 mM 2-deoxyglucose.

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

In this paper, we describe the functional expression and purification to near homogeneity of the IIA domain of the lactose transport protein of S. thermophilus. On the basis of the similarities in the primary sequences of IIA^LacS, IIA^Glc, and other PTS IIA protein domains (12, 20), we anticipated that IIA^LacS would be able to carry out one or more functions associated with IIA^Glc, i.e., complement crr strains in PTS-mediated glucose uptake, inhibit glycerol kinase, and/or activate adenylate cyclase.

The in vitro phosphorylation assays indicate that IIA^LacS-6H can be phosphorylated by HPr~P from S. thermophilus, B. subtilis, and E. coli, and the results suggest that the phosphorylation site in IIA^LacS is His-552. This residue corresponds to His-90 of IIA^Glc in E. coli, which has been shown to be the phosphoryl-accepting site (4). Although IIA^LacS can be phosphorylated by HPr~P, the phosphorylation rates are lower than for phosphorylation of IIA proteins involved in PTS-mediated transport (14). The rate constant for the phosphoryl transfer between S. thermophilus HPr~P and IIA^LacS at a temperature of 10°C was k₁ = 4.3 × 10² M¹ s¹; the rate constant for the reverse reaction (IIA^LacS~P to HPr) was k₁ = 1.1 × 10³ M¹ s¹. Although these phosphoryl transfer rates increased 10-fold at 37°C, it is evident that phosphorylation of IIA^LacS by HPr~P from S. thermophilus is 4 orders of magnitude slower than phosphorylation of IIA^Glc by HPr~P from E. coli (14) (k₁ = 6.1 × 10⁷ M¹ s¹). With HPr~P from B. subtilis and E. coli, the differences are even much larger, as one might expect, when interacting proteins from different sources are compared (Fig. 5). The phosphorylation rate of IIA^LacS by HPr~P from S. thermophilus increased 1 order of magnitude when Ile-548 and Gly-556 were substituted by Glu and Asp, respectively, and increased 3 and 4 orders of magnitude when IIA^LacS-6H(I548E/G556D) was phosphorylated by HPr~P from E. coli and B. subtilis, respectively (Fig. 6). The corresponding acidic residues (Glu-86 and Asp-94 of IIA^Glc in E. coli) are conserved in all PTS IIA protein domains but are not present in the non-PTS IIA protein domains (Fig. 1). The crystal structure of IIA^Glc from E. coli shows that Asp-94 participates in phosphoryl transfer, as the backbone amide nitrogen of this residue is positioned to form an H bond to the phosphoryl group (16). It has been suggested that this H bond stabilizes the transition state (trigonal bipyramid) form of the P atom. The conserved Glu-86 is more exposed to the surface of the molecule and is proposed to serve as the recognition site for one of the interacting PTS proteins (12). Substitution in the -glucoside IICBA of E. coli of Asp-551, the equivalent of Asp-94 in IIA^Glc, by an alanine residue decreased the phosphorylation rate relative to the wild-type protein (30). The higher rate of phosphorylation of IIA^LacS-6H(I548E/G556D) than of IIA^LacS-6H is consistent with the predictions one could make from the three-dimensional structure of IIA^Glc of E. coli and substantiates the critical role of these residues in the phosphorylation of IIA. Importantly, the rates of (de)phosphorylation of IIA^LacS-6H(I548E/G556D) by HPr~P from E. coli and B. subtilis are similar to those of S. thermophilus, which allowed us to assess some of the functional properties of IIA^LacS in a heterologous system (see below).

The inability of IIA^LacS to substitute for IIA^Glc in phosphorylation of E. coli IIB, as determined by assays of PTS-mediated glucose uptake, is likely due to the low phosphorylation rate of IIA^LacS by E. coli HPr~P. Additionally, the inability of IIA^LacS to complement the lack of IIA^Glc in PTS-mediated glucose uptake could also be due to an incompatible amino terminus. In this respect, it is worth noting that cell extracts of E. coli contain two electrophoretically distinguishable forms of IIA^Glc, a slow form and fast form. The fast form [IIA^Glc(fast)] is the product of an endopeptidase that cleaves the N-terminal heptapeptide from the mature form. IIA^Glc(fast) is fully active in accepting the phosphoryl group from HPr~P, but it has only 3% of the phosphodonor activity of the intact protein (13) and has a smaller effect on inhibition of methyl--D-thiogalactopyranoside uptake (inducer exclusion) than IIA^Glc(slow) (15). This finding suggests that the N-terminal region of IIA^Glc could participate in the interaction with its partner molecules, such as IIB^Glc and the lactose permease. The IIA^LacS protein that was used in the majority of the experiments is 173 amino acids long and corresponds in length to IIA^Glc(slow). Addition of another eight amino acids (part of putative Q-linker region in LacS) did not affect the (de)phosphorylation activity. Shortening of IIA^LacS[173] by 19 residues abolished the ability of the protein to accept the phosphoryl group from HPr. Future experiments are required to clarify which amino acid residues are involved in the interaction between IIA^Glc and IIB^Glc to facilitate phosphoryl transfer to glucose.

Apart from its function in the uptake and phosphorylation of sugars, IIA^Glc interacts with several non-PTS enzymes such as glycerol kinase, the MalK component of the maltose transport system, and the melibiose and lactose transporters of E. coli, resulting in inactive complexes (23). In a previous study (36), it was shown that wild-type LacS protein, in which the IIA domain is fused to the membrane-bound carrier domain, does not inhibit glycerol kinase. This could point to the inability of membrane-bound IIA^LacS to interact functionally with glycerol kinase, but more likely the expression of LacS in S. typhimurium was too low for significant inhibition to be observed. It is worth noting that membrane-bound IIA^Nag, as part of the IICBA^Nag complex, is able to inhibit glycerol kinase (37), indicating that inducer exclusion is not exclusively mediated by cytosolic IIA^Glc. Since we could express IIA^LacS separate from the carrier domain of LacS, we also studied the ability of the various IIA^LacS proteins to inhibit glycerol uptake. Although IIA^LacS~P and IIA^LacS-6H (I548E/G556D)~P were not able to rapidly transfer the phosphoryl group to IICB^Glc, the uptake of glycerol in S. typhimurium PP2178 was inhibited by the addition of 2-deoxyglucose, most likely because IIA^LacS~P or IIA^LacS-6H (I548E/G556D)~P was dephosphorylated through the redirection of phosphoryl groups to HPr and subsequently to the mannose PTS. 2-Deoxyglucose is taken up by the mannose PTS and thus indirectly influences the phosphorylation state of IIA^LacS. The extent of inhibition of glycerol uptake was dependent on the actual uptake rate of glycerol (amount of glycerol kinase) (Fig. 8), suggesting that IIA^LacS inhibits glycerol kinase by forming a stoichiometric complex with the enzyme as observed for IIA^Glc and glycerol kinase (37). Hurley et al. (7) identified the sites of interactions between glycerol kinase and E. coli IIA^Glc, which mainly involve hydrophobic and electrostatic interactions and a Zn(II) binding site. The Zn(II) binding site is made up of the two active-site histidines of IIA^Glc (His-75 and His-90), Glu-478 of glycerol kinase, and an H₂O molecule. In the absence of Zn(II), IIA^Glc binds to glycerol kinase primarily via the hydrophobic patch and without participation of the His residues. Except for the histidines that coordinate the Zn atom, the positions are poorly conserved in the non-PTS IIA protein domain(s). Nevertheless, IIA^LacS was able to inhibit glycerol kinase. Our results also indicate that substitution of His-552 for Arg (His-552 in LacS corresponds to His-90 in E. coli IIA^Glc) abolished the putative interaction of IIA^LacS with glycerol kinase. Possibly, the binding constant of the mutant to glycerol kinase is decreased as suggested for IIA^Glc(H90Q) (16). Whether the Zn atom plays an important role in the interaction of IIA proteins to glycerol kinase is unclear, as IIA^Glc(H75Q) seems as effective an inhibitor as the wild-type protein (cited in reference 16).

In gram-negative enteric bacteria the phosphorylated form of IIA^Glc stimulates adenylate cyclase, whereby the expression of many catabolic enzymes is regulated through changes in cAMP levels, e.g., the expression of genes for succinate and citrate catabolism (1, 23). Reddy et al. (25) suggested that the acquisition of a negative charge at His-90~P of E. coli IIA^Glc is responsible for the interaction with adenylate cyclase. Although IIA^LacS and IIA^LacS(I548E/G556D) could be phosphorylated by HPr from E. coli both in vitro and in vivo, these IIA^LacS proteins were not able to stimulate adenylate cyclase sufficiently, if at all, to restore growth of crr mutants on succinate and citrate. The in vivo phosphorylation is suggested by the identification of two species of IIA^LacS (IIA^LacS and IIA^LacS~P) in cell extracts of E. coli cells (unpublished data); it also follows from the glycerol kinase inhibition experiments.

The equilibrium constant (K_eq = k₁/K₁) for the phosphoryl transfer reactions between IIA^LacS and HPr from S. thermophilus is 0.5, which is approximately threefold lower than the corresponding value for HPr and IIA^Glc from E. coli (K_eq = 1.4 ± 0.5 [14]). This implies that at a given phosphorylation potential, the phosphorylation state of IIA^LacS (and we assume LacS as well) in S. thermophilus will be lower than that of IIA^Glc in E. coli. Since phosphorylation of LacS inhibited the lactose-H⁺ symport reaction (17), this condition will be met in vivo only when the [HPr(His~P)]/[HPr] ratio is relatively high. Vadeboncoeur and coworkers (34) reported that in stationary-phase cells of S. salivarius, the [HPr(His~P)]/[HPr] ratio is approximately 2.5 for S. salivarius. On the other hand, actively growing cells of S. mutans and S. salivarius contained mainly HPr(Ser-P) and HPr(His~P)(Ser-P), with very little HPr(His~P) and free HPr. Thus, the degree of phosphorylation of LacS will be low in exponentially growing cells of S. thermophilus, which implies that the system will have full activity. In resting cells of S. thermophilus, a partial inhibition of LacS-mediated methyl--D-thiogalactopyranoside transport was observed (17).