Isolation and Characterization of IIAChb, a Soluble Protein of the Enzyme II Complex Required for the Transport/Phosphorylation ofN,N′-Diacetylchitobiose in Escherichia coli *

N,N′-Diacetylchitobiose is transported/phosphorylated in Escherichia coli by the (GlcNAc)2-specific Enzyme II permease of the phosphoenolpyruvate:glycose phosphotransferase system. IIAChb, one protein of the Enzyme II complex, was cloned and purified to homogeneity. IIAChb and phospho-IIAChb form stable homodimers (3). Phospho-IIAChb behaves as a typical ε2-N (i.e. N-3) phospho-His protein. However, the rate constants for hydrolysis of phospho-IIAChb at pH 8.0 unexpectedly increased 7-fold between 25 and 37 °C and increased ∼ 4-fold with decreasing protein concentration at 37 °C (but not 25 °C). The data were explained by thermal denaturation studies using CD spectroscopy. IIAChb and phospho-IIAChb exhibit virtually identical spectra at 25 °C (∼80% α-helix), but phospho-IIAChb loses about 30% of its helicity at 37 °C, whereas IIAChb shows only a slight change. Furthermore, the T m for thermal denaturation of IIAChb was 54 °C, only slightly affected by concentration, whereas the T m for phospho-IIAChb was much lower, ranging from 40 to 46 °C, depending on concentration. In addition, divalent cations (Mg2+, Cu2+, and Ni2+) have a dramatic and differential effect on the structure, depending on the state of phosphorylation of the protein. Thus, phosphorylation destabilizes IIAChb at 37 °C, potentially affecting the monomer/dimer transition, which correlates with its chemical instability at this temperature. The physiological consequences of this phenomenon are briefly considered.

In Escherichia coli, the operon that encodes the proteins required for catabolism of the chitin disaccharide N,NЈ-diacetylchitobiose, (GlcNAc) 2 , 1 was previously thought to be a "cryptic" cellobiose operon (6,7). The cryptic cellobiose operon is, in fact, a normally inducible catabolic operon required for the utilization of (GlcNAc) 2 . We have proposed that the nomenclature for the proteins encoded by the operon be changed, from Cel (cellobiose) to Chb (di-N-acetyl-chitobiose).
We report (1, 5) 2 that (GlcNAc) 2 uptake is mediated by the phosphoenolpyruvate:glycose phosphotransferase system (PTS). Transport requires the PTS general proteins, Enzyme I and HPr, as well as two soluble sugar-specific proteins, IIA Chb and IIB Chb , and a membrane protein IIC Chb . The overall reaction sequence leading to cytoplasmic sugar phosphate is summarized in Fig. 1 of one of the accompanying papers (2).
The structures of the Enzyme II complexes, i.e. the sugarspecific components or permeases of the PTS, are currently under intense investigation. One suggested nomenclature of the complexes derives from amino acid sequence and functional homologies (8,9). They can be divided into distinct domains, designated IIA, IIB, and IIC. For some sugars the IIA, IIB, and IIC domains are encoded by a single (membrane bound) polypeptide, for others IIBC forms a single membrane protein component with a separate soluble IIA protein. The (GlcNAc) 2 permease offers many advantages for studying the structure/ function of an Enzyme II complex because the IIABC domains are encoded by three separate genes, one each for the IIA, IIB, and IIC. But more importantly, IIB Chb is cytoplasmic, which offers a rare opportunity to conveniently study its interactions with its partners, IIA Chb and IIC Chb .
Based on sequence homologies IIA Chb has been assigned to the lactose family of PTS transporters (8,10), but there is virtually no published information on IIA Chb nor, of course, on phospho-IIA Chb . Although both the crystal and solution structures of an active site mutant protein of IIB Chb have been reported (11), there is little published information on IIB Chb itself and nothing on phospho-IIB Chb , the subjects of the accompanying papers.
In the present report, homogeneous IIA Chb and phospho-IIA Chb are described along with some important properties of the proteins. Analytical sedimentation studies of the proteins are presented in an accompanying paper (3).

Materials
Buffers and reagents were of the highest purity commercially available. [ 32 P]PEP was a kind gift from Dr. N. Meadow (Johns Hopkins University). Cellulose nitrate filters (diameter, 25 mm; pore size, 0.45 m) were obtained from Sartorius. E. coli strain BL21(DE3):⌬EI, containing a kanamycin cartridge in Enzyme I was a kind gift from Dr. F. Chauvin.

Molecular Analysis and Sequencing of DNA
Preparations, analyses, restriction enzyme digests, ligations, and transformations were performed using standard techniques (12,13). Double-stranded DNA was prepared from recombinant clones and sequenced by the dideoxy method using a U.S. Biochemical Corp. Sequenase version 2.0 sequencing kit or alternately by the Genetics Core Facility (Johns Hopkins Medical School) using an ABI-373 automated sequencer.

Construction of IIA Chb Overexpression Vector
The open reading frame corresponding to the chbA gene was cloned into the pET21a (Novagen, Madison, WI) overexpression vector using polymerase chain reaction and primers specific to the ends of the gene. The primers were designed with unique restriction sites at each end to facilitate the cloning procedure. The polymerase chain reaction generated fragments were first cloned into the pNoTA/T7 shuttle vector (Prime PCR Cloner Cloning System; 5 Prime 3 3 Prime, Inc., Boulder, CO) and then subcloned into pET21a (according to the manufacturer's recommendations and procedures). Primers were designed (at the 5Ј end of the gene) such that the start site (ATG) codon would be ligated directly to the NdeI/start site of pET21a. The nucleotide sequences of the primers are given below. The engineered restriction sites are underlined, and the start site of the gene is in bold. The primers used were as follows. 5Ј-GAGGAACGACATATGATGGATCTC-3Ј (NdeI site) and 5Ј-TCACTGGCTGGATCCTCGACTCC-3Ј (BamHI). The isolated subclones in pET21a were confirmed by sequencing the entire insert.

Purification of IIA Chb
When the protein was overexpressed from pET21a and crude extracts were subjected to SDS-PAGE, a new major band was detected that migrated at the expected molecular mass (12.75 kDa). SDS-PAGE was therefore used to detect the protein throughout the purification procedure.
Step 1: Crude Extract-Two liters each of LB media supplemented with 100 g/ml ampicillin in two 6-liter flasks were inoculated with 80 ml of an overnight culture of E. coli strain BL21(DE3):⌬EI harboring the plasmid pET-chbA. The culture was shaken vigorously at 37°C until A 600 was about 1.0 (2-3 h) before being induced by the addition of 1 mM (final concentration) isopropyl-1-thio-␤-D-galactopyranoside or 3 mM melibiose. Cells were then allowed to grow for an additional 2 h and harvested by centrifugation at 4000 ϫ g for 10 min at 4°C. The following steps in the purification were conducted at 0 -4°C unless otherwise stated. The cell pellet was washed twice with 500 ml of 50 mM Tris-HCl buffer, pH 8.0, containing 1 mM EDTA and resuspended in the same buffer using 4.0 ml of buffer/g (wet weight) of cells. The cells were disrupted by two passages through a Wabash French Press. Cell debris was removed by centrifugation at 12,000 ϫ g for 15 min.
Step 2: Streptomycin Sulfate Precipitation-Nulceic acids were precipitated with a solution of streptomycin sulfate (160 l of 10% stock/ml crude extract) added dropwise with stirring. The white precipitate was removed by centrifugation at 200,000 ϫ g for 30 min and discarded.
Step 3: DEAE Chromatography-The supernatant from Step 2 was transferred to a 50-ml DEAE-Sepharose CL-6B column equilibrated in the same buffer. The column was washed with 100 ml of 50 mM Tris-HCl buffer, pH 8.0, followed by 100 ml of buffer containing 0.1 M NaCl. A gradient (1 liter) from 0.1 to 0.5 M NaCl in the buffer was applied, and 9.0-ml fractions were collected. The IIA Chb protein eluted at around 0.25 M NaCl. Fractions were pooled based on estimated protein amount and purity as determined by SDS-PAGE.
Step 4: Metal (Cu 2ϩ ) Chelate Affinity Chromatography-Pooled fractions from Step 3 were dialyzed against 50 mM sodium phosphate buffer, pH 7.5, containing 0.1 M NaCl and transferred to a Chelating Sepharose Fast Flow column (Amersham Pharmacia Biotech). The chelating gel (30 ml) was first charged with CuCl 2 (60 ml, 2 mg/ml) before being equilibrated in the starting buffer 50 mM sodium phosphate, 0.1 M NaCl, pH 7.5.
After loading the sample, the column was washed with two column volumes each (60 ml) of the following solutions: 50 mM sodium phosphate buffer, pH 6.5, containing 0.1 M NaCl; 50 mM sodium phosphate buffer, pH 5.5, containing 0.1 M NaCl; and 50 mM sodium phosphate buffer, pH 7.5, containing 1 M NH 4 Cl. The column was re-equilibrated in the starting buffer before being eluted with a gradient (300 ml) of 0 -20 mM imidazole in 50 mM sodium phosphate buffer, pH 7.5, containing 0.1 M NaCl.
The protein IIA Chb eluted at around 7-10 mM imidazole. Peak fractions were pooled and concentrated by pH precipitation. Precipitation was carried out as follows (at 4°C); the pH of the pooled fractions was gradually lowered by the dropwise addition of 500 mM monobasic sodium phosphate, pH ϳ4.3, until the pH was around 5-5.5. The solution was stirred for 1 h before the precipitate was collected by centrifugation (10,000 ϫ g, 30 min). The pellet was resuspended in 100 mM dibasic sodium phosphate buffer, pH ϳ9.2. The final pH was 8.0, and the solution was dialyzed against 25 mM MOPS, pH 7.0.
Step 5: Mono-Q FPLC-The FPLC Mono-Q purification step was performed at room temperature. The fraction from Step 4 was transferred to an FPLC Mono-Q HR10/10 (Amersham Pharmacia Biotech) column equilibrated in 25 mM MOPS, pH 7.0. The column was washed with two column volumes of buffer (16 ml) before being eluted with a linear gradient (160 ml) from 0 to 0.5 M NaCl in 25 mM Tris-HCl buffer, pH 8.0. The protein eluted between 0.3 and 0.35 M NaCl. The IIA Chb containing fractions were pooled and concentrated by pH precipitation as described above.
Step 6: Gel Filtration Chromatography-An FPLC HiPrep 26/60 Sephacryl S-100 column (Amersham Pharmacia Biotech) was equilibrated in 25 mM sodium phosphate, pH 8.0. The pooled concentrated fractions (4 -8 ml) from Step 5 were transferred to the column and eluted with the same buffer. Purified IIA Chb was dialyzed against 5 mM sodium phosphate buffer, pH 8.0, and lyophilized for long term storage.

Gel Electrophoresis
For denaturing gels, unless otherwise noted, protein samples were heated at 100°C for 3-5 min in loading buffer (65 mM Tris-HCl buffer, pH 6.8, with 0.3% SDS, 5% glycerol, 5 mM DTT, 0.1 mg/ml bromphenol blue) prior to being electrophoresed in a vertical 12-18% polyacrylamide slab gel, pH 8.0, with a 6% stacking gel, pH 6.8. Gels were stained with Coomassie Blue G-250. Native gel electrophoresis was carried out essentially as described for denaturing gels except SDS was omitted from all buffers and samples were not boiled prior to loading. The Novex gel electrophoresis system and precast gels (Novex, San Diego, CA) were also employed under both native and denaturing conditions. Stained protein bands on the gels were quantitated by densitometric scanning using the Eagle Eye Video System (Stratagene, La Jolla, CA).

N-terminal Amino Acid Sequencing
The purified proteins were subjected to N-terminal amino acid sequencing using an Applied Biosystems 475A protein sequencer (Department of Biological Chemistry, Johns Hopkins School of Medicine).

Filter Binding Assay
The assay was performed essentially as described (14). For measurement of IIA Chb phosphorylation the assay reaction mixture (20 l) contained 50 mM Tris-HCl buffer, pH 8.0, 10 mM MgCl 2 , 1 mM DTT, 5 mM NaF, 2-5 pmol of purified Enzyme I, 5-10 pmol of purified HPr, and 100 -1000 pmol of purified IIA Chb . The reaction was started by the addition of 2 mol of [ 32 P]PEP (10 -20 cpm/pmol). Aliquots were taken over the time course, the reaction stopped by dilution with 1.0 ml of ice-cold buffer (10 mM Tris-HCl buffer, pH 8.0, 150 mM NaCl) and rapidly filtered through cellulose nitrate filters (Sartorius). The filters were washed twice with 1 ml each of ice-cold buffer, immersed in 4 ml of Packard Ultima-Gold XR liquid scintillation counter mixture, and counted in a Packard Liquid Scintillation Spectrometer.

Phosphorylation of IIA Chb
Phosphorylation reactions for preparative isolation of phospho-IIA Chb were performed on a large scale as described above. Typically, 1.0-ml reaction mixtures contained 50 mM Tris-HCl buffer, pH 8.0, 10 mM MgCl 2 , 1 mM DTT, 0.1-0.3 nmol of purified Enzyme I, 0.25-0.5 nmol of purified HPr, and 35-100 nmol of purified IIA Chb . The mixture was incubated for 45-60 min at 37°C.
The phosphoprotein was purified by native gel electrophoresis using 16% acrylamide gels. Stacking and resolving gels and buffer were identical to the SDS-PAGE procedure except that SDS was omitted from all buffers. Protein samples were maintained at 4°C prior to loading onto the gel, and the sample loading buffer was modified to contain the following (final concentrations): 0.1 M Tris-HCl buffer, pH 9.3, 5% glycerol, 0.1% (w/v) bromphenol blue, 2 mM DTT. To identify the position of the phosphoprotein in the gel after electrophoresis, lanes on both sides of the gel were cut and either briefly stained with Coomassie Blue or (where proteins were labeled with 32 P) further cut into slices, immersed in 3.0 ml of scintillation fluid, and counted as described above. The remaining portion of the gel was then sliced and the desired (phospho)-protein product recovered by electroelution at 250 V for 2 h at 4°C in an Amicon Centrilutor. Electroelution buffer was either 12.5 mM Tris borate, pH 8.7, or 15 mM Tris-CAPS, pH 9.3. The eluate was concentrated to 0.5 ml by ultrafiltration in a Centricon-3 and stored in aliquots at Ϫ80°C until used.

Alternate Procedure for Phosphorylation and Purification of Phosphoproteins
The phosphorylation reaction was performed as described above or using HPr coupled to Affi-Gel-15-agarose beads beads (see below). The reaction was terminated by sedimenting the HPr beads (10,000 ϫ g, 5 min), and the supernatant was applied to a Superdex 75 HR 10/30 column (10 mM ϫ 300 mm; Amersham Pharmacia Biotech) equilibrated and eluted with 25 mM sodium phosphate buffer, pH 8.0, using a Amersham Pharmacia Biotech FPLC system. Protein in the eluate was monitored by absorbance at 280 nm. Fractions corresponding to phospho-IIA Chb were either used immediately or stored at Ϫ70°C; the phosphoprotein was stable for at least 3 days under these conditions.

Coupling of HPr to Affi-Gel 15 Beads
Purified HPr (kind gift of Dr. R. Mattoo) was coupled to Affi-Gel 15 (Bio-Rad) according to the manufacturer's recommendations. Briefly, 3.6 mg of purified HPr in 2 ml of 10 mM MOPS buffer, pH 6.5, were added to 1.5 ml of washed Affi-Gel 15 resin and reacted at 4°C for 4 h on a rotating shaker. About 50% of the protein remained in the solutions and was removed. Unreacted groups on the resin were blocked with ethanolamine as recommended by the manufacturer.

Kinetics of Hydrolysis of Phosphoprotein as a Function of pH
The rate of hydrolysis of [ 32 P]phospho-IIA Chb was determined as a function of pH. The following buffers were used: McIlvaine's sodium phosphate-citric acid broad range buffer from pH 2.0 to 8.8, Bates' and Bowers' boric acid-KCl buffer (pH 8.0 -10), sodium phosphate, sodium acetate, sodium borate, Tris-HCl buffer, MOPS-HCl, and TAPS-HCl. Three separate procedures were used to measure the remaining 32 Plabeled protein and the product [ 32 P]P i : DEAE-Paper Chromatography-Phosphoproteins were separated from [ 32 P]P i by chromatography on Whatman DEAE-paper (DE-81) using conditions (15) where the phosphoproteins remained at the origin, whereas phosphate migrated with the solvent front. Briefly, aliquots of the reaction mixtures (25-50 l) were spotted on strips of DE-81 (2 ϫ 10 cm) and developed by ascending chromatography using a solvent containing 35% ethanol, 10 mM CAPS, pH 10.5, 1 mM EDTA, and 50 mM KCl, for 1-2 h at room temperature. The strips were dried, cut in half, immersed in 3.0 ml of scintillation mixture, and counted as described above.
Filter Binding Assay-The filter binding assay was as described above.

Protein Determination
Protein concentrations and extinction coefficients were calculated for the purified proteins by quantitation of nitrogen content by the Kjeldahl method as modified by Jaenicke (18). For routine protein measurements, the Bio-Rad dye binding assay was employed using bovine serum albumin as the standard.

CD Spectroscopy
CD experiments were conducted using a Jasco-J710 spectropolarimeter with a Peltier type cell holder (Jasco model PTC-348W). Each wavelength spectrum scan was obtained by averaging 4 -5 spectra using 1-nm intervals at 50 nm/min and a 1-mm rectangular cell. Buffer scans (25 mM sodium phosphate, pH 8.0) were accumulated and subtracted from the sample scans, and the mean residue ellipticity was calculated. CD temperature scans were performed by varying the temperature over the indicated range at a rate of 1°C/min, and the mean ellipticity was measured at 222 nm. Experiments measuring the reversability of thermal denaturation were conducted at a rate of 2°C/min.

Effect of Divalent Cations on CD Spectra and Thermal
Denaturation of IIA Chb and Phospho-IIA Chb Because IIA Chb binds to a Cu 2ϩ chelate column, it was of interest to study the effects of divalent cations on the CD spectra and on thermal denaturation. Three metals were tested, Cu 2ϩ , Mg 2ϩ , and Ni 2ϩ . Final cation concentrations (as their chlorides) were 1 mM, whereas the proteins were used at 20 -200 M concentrations.

RESULTS
Purification of IIA Chb -The gene encoding IIA Chb was cloned into the pET21a overexpression vector using a polymerase chain reaction approach as described under "Experimental Procedures." The protein was purified from E. coli BL21(DE3):⌬EI harboring pET:IIA Chb . A deletion of Enzyme I was used to ensure that IIA Chb was isolated in its unphosphorylated form. During purification of IIA Chb , it eluted as a blue solution, i.e. complexed to Cu 2ϩ , from the Cu 2ϩ chelating column. The Cu 2ϩ ion was removed by dialysis against buffer (typically, 20 mM sodium phosphate, pH 8.0) containing 1 mM EDTA and 0.1 mM DTT.
SDS-PAGE of the purified protein is shown in Fig. 1A. The protein migrates with an apparent molecular mass of 12-14 kDa, which agrees with a predicated molecular mass of 12.75 kDa from the gene sequence (7,19). As we report elsewhere (3), the native protein is a homodimer.
The N-terminal amino acid sequence of the purified protein was obtained as described under "Experimental Procedures" and agreed with that predicted from the coding sequence, indicating that the protein is not processed during its expression (data not shown).
Phosphorylation of IIA Chb -In the accompanying report (1) phosphorylation of (GlcNAc) 2 and its analogue Me-TCB was reconstituted in vitro using a mixture of (purified) Enzyme I, HPr, IIA Chb (for which IIA Glc could not substitute), and membranes containing IIC Chb .
Direct measurement of the phosphorylation of IIA Chb (and IIB Chb ) could be monitored using [ 32 P]PEP and autoradiography after SDS-PAGE (phospho-IIA Chb migrates with IIA Chb under these conditions). Native, nondenaturing gel electrophoresis was also used. These gels offered several advantages over SDS-PAGE. Fig. 1B shows that IIA Chb is phosphorylated by HPr in the presence of Enzyme I and PEP. (The Coomassie stain shows a light band corresponding to Enzyme I in each lane.) Furthermore, Fig. 1B also shows that a shift in the gel mobility of IIA Chb occurs when it is phosphorylated and that the phosphoproteins were stable in the native gels and were easily separated from the unphosphorylated proteins. Use of [ 32 P]PEP and autoradiography confirmed the identities of the phosphoproteins (data not shown). Omission of Enzyme I, HPr, or PEP yielded no phosphorylation of IIA Chb .
The kinetics of phosphorylation of IIA Chb are shown in Fig.  1C. The Chb proteins can be phosphorylated using either catalytic or stoichiometric quantities of Enzyme I and HPr. The degree of phosphorylation was assayed by determining the relative ratios of the differentially migrating forms of the stained proteins in the native gels (scanning densitometry).
Stability of Phospho-IIA Chb -To systematically study the stability of the phosphoprotein, it was purified using Superdex-75 gel filtration chromatography as described under "Experimental Procedures." Gel filtration chromatography is capable of separating Enzyme I and HPr from IIA Chb but cannot separate IIA Chb from P-IIA Chb . The phosphoprotein preparations were analyzed for the degree of phosphorylation using native gel electrophoresis. For the studies described below, preparations of phospho-IIA Chb were used only where there was no detectable unphosphorylated protein (i.e., Ͻ5%). As reported in the accompanying paper, both IIA Chb and phospho-IIA Chb are dimers (3). The corresponding monomers were not detected in the sedimentation experiments. Phospho-IIA Chb can therefore contain (on the average) from 1 to 2 mol of phosphate/dimer. When it is phosphorylated under the conditions described above, 1.5-2.0 mols of phosphate were incorporated per dimer based on protein concentration determined by a micro-Kjeldahl procedure and the specific activity of the [ 32 P]PEP. From the amino acid sequence similarity/identity of IIA Chb and the IIA Lac proteins (discussed below), phospho-IIA Chb should be a phosphohistidinyl protein, with the phos-phoryl group linked to the His 89 moiety (8,10).
A characteristic feature of the phosphohistidinyl linkage in phosphorylated PTS proteins is the sensitivity of this bond to acid, hydroxylamine, and pyridine and its stability to alkali (20). Hydrolysis of [ 32 P]phospho-IIA Chb was therefore followed at 25°C, and Fig. 2 gives the rate constants for hydrolysis (as t1 ⁄2 values) as a function of pH. The protein precipitated below pH 5, prohibiting accurate assessment at lower pH values. Fig.  2 shows that the phosphoryl linkage is stable at high pH. The rate was also determined at 37°C at pH 8.0 in the presence and absence of hydroxylamine and pyridine (Fig. 3). The phosphoryl linkage is sensitive to hydroxylamine and pyridine, characteristic of a phosphoramidate linkage.
Effect of Temperature and Concentration on Hydrolysis Rates of Phospho-IIA Chb at pH 8.0 -For physico-chemical studies, it was important to determine the conditions of maximum stability of the phosphoprotein at or near physiological pH. Gel electrophoresis was therefore used to analyze the hydrolysis process for 24 h at different temperatures and at pH 8.0 and gave the following results (as a percentage of phosphoprotein remaining): 37°C, 30 -40%; 25°C, 75%; 16°C, Ͼ90%; 10°C, stable. Unexpected results were obtained, however, when the experiments were conducted kinetically to obtain rate constants.
First order hydrolysis rate constants were determined at two temperatures, 25 and 37°C, and the results are given in Table  I. The rates, expressed as t1 ⁄2 were 1800 and 250 min, respectively. This 7-fold change seemed inordinately large for a 12°C change in temperature.
The second unexpected finding was obtained when the hydrolysis rates were determined at different concentrations of phospho-IIA Chb . The results are also presented in Table I. As can be seen, at pH 8.0 and 37°C, the rates of hydrolysis increase 4-fold as the phosphoprotein is diluted. The effects of temperature and concentration did not readily fit a pseudo first order reaction (the activity of water remains constant), and an explanation was sought.
Fluorescence Studies-The protein IIA Chb contains no Trp and only one tyrosine per monomer. The fluorescence excitation and emission spectra of the Tyr in phospho-IIA Chb and IIA Chb were therefore compared but showed no detectable dif- 32 P-Labeled phospho-protein was purified by electroelution from a native gel or by gel filtration on a Superdex-75 FPLC column as described under "Experimental Procedures." Assays for hydrolysis were conducted at 25°C by the DEAE-paper method. Rate constants were determined from logarithmic plots (phosphoprotein remaining versus time) as a function of pH in the following buffers: ࡗ, sodium acetate; , Tris-HCl buffer; q, McIlvaine's and Bates'/Bowers'. ference (data not shown). The results suggest that the environment around the Tyr does not change significantly when the protein is phosphorylated.
Thermal Denaturation of IIA Chb and Phospho-IIA Chb -Thermal denaturation of IIA Chb and phospho-IIA Chb were studied using circular dichroism. Based on the amino acid sequence similarity to analogous PTS proteins, IIA Chb was predicted to have a high helical content. This prediction was confirmed by CD spectroscopy (Fig. 4A). There was a slight decrease in the normalized mean ellipticity of the protein upon phosphorylation, but the protein still exhibited Ͼ80% helical content at 25°C.
CD spectroscopy was then used to study thermal denatur-ation of the two proteins. Here, in contrast to the data discussed above, a dramatic difference was observed. The phosphoprotein was much more sensitive to thermal denaturation. Two types of experiments were conducted. In the first (Fig. 4), the CD spectra are shown from 190 to 250 nm for IIA Chb and phospho-IIA Chb respectively. In the second, discussed below, thermal denaturation was followed at 220 nm. Fig. 4 shows that the helical content of both proteins decreased as they unfolded, but phospho-IIA Chb is much less stable than IIA Chb . For example, at 45°C there is only a slight decrease in the signal for IIA Chb (Fig. 4B), whereas more than 50% of the helicity is lost from phospho-IIA Chb (Fig. 4C). More pertinent to the hydrolysis results reported above, at 37°C phospho-IIA Chb exhibits about 35% less helicity than at 25°C, whereas IIA Chb was essentially unaffected . Fig. 5 shows the effect of temperature on ellipticity at 220 nm. The temperature of a solution containing either IIA Chb or phospho-IIA Chb was slowly raised over the range 10 -90°C at pH 8.0, 25 mM phosphate buffer. Both IIA Chb and phospho-IIA Chb showed at least two transitions (Fig. 5A). The thermal denaturation of IIA Chb was reversible. Likewise, a second temperature scan of phospho-IIA Chb was identical to the first if the phosphoprotein was heated to the end point of the first transition (65°C) but not higher. Less than 10% of the protein was dephosphorylated during the thermal denaturation (data not shown). Continued repeat scans of phospho-IIA Chb resulted in ellipticity curves intermediate between the original phosphoprotein and the native protein, eventually resulting in a curve superimposable on the nonphosphorylated protein (data not shown). Native gel electrophoretic analysis revealed a gradual increase in the fraction of dephosphoprotein until eventually the protein was completely dephosphorylated. These results indicate that the thermal denaturation resulting over the course of the first transition is reversible, both for IIA Chb and phospho-IIA Chb .
The thermodynamic relationships between the two proteins cannot be accurately assessed by changes in the spectra (which presumably are related to structural effects). These relationships can be determined, however, by differential scanning calorimetry. Preliminary differential scanning calorimetry experiments gave results consistent with the CD data, and more extensive studies are in progress. For the remainder of this report, we focus on the first, or major, transition that is detected by CD (Fig. 5B). The observed T m values were 56°C for IIA Chb and 40.8°C for phospho-IIA Chb . A 15°C change in T m shows that phosphorylation of IIA Chb results in a major destabilizing effect on the structure of IIA Chb . It is important to note (Fig. 5A) that a significant fraction of the phosphorylated molecules are destabilized at both 25 and 37°C but that the effect (relative to IIA Chb ) is much more apparent as the temperature is increased beyond 20 -25°C.
Effect of Protein Concentration on Thermal Denaturation of IIA Chb and Phospho-IIA Chb -The results described above showed that the t1 ⁄2 of hydrolysis of the phosphoprotein changed with protein concentration. As a consequence, the effects of protein concentration on the thermal denaturation of IIA Chb and phospho-IIA Chb were examined, and the results are presented in Fig. 6. As the concentration decreases from 150 to 10 M the T m of IIA Chb changes slightly, from about 56 to 54°C. By contrast, there is a large change in the T m of phospho-IIA Chb , from about 45 to 34°C. These results are discussed below.
Effect of Divalent Cations on Thermal Denaturation of IIA Chb and Phospho-IIA Chb -One step in the purification of IIA Chb was to bind it to a Cu 2ϩ chelating column. As explained under "Experimental Procedures," the protein was eluted from the  a Rate constants were obtained from the progress curves for the pseudo first order hydrolysis reactions, and are presented as t1 ⁄2 values in minutes.
b Phospho-HPr is the only phosphoprotein shown where the phosphoryl group is linked to the ␦1-N position of the His residue.
c The molar concentrations of the phosphoprotein were calculated based on the monomer molecular mass, 12,850 g/mol. Monomeric species were not detected in the equilibrium sedimentation experiments (3) with IIA Chh and phospho-IIA Chb . However, it should be stressed that all of these experiments were conducted at temperatures Յ25°C. Temperatures such as 37°C could not be used because of hydrolysis of the phosphoprotein over the prolonged times required for equilibrium sedimentation experiments. column with an imidazole gradient, but the solution was blue, indicating that IIA Chb binds Cu 2ϩ with high affinity; the copper was removed by dialysis against EDTA. The effects of the divalent cations Mg 2ϩ , Cu 2ϩ , and Ni 2ϩ on the thermal denaturation of IIA Chb and phospho-IIA Chb was studied because of the affinity of the protein for Cu 2ϩ and because Mg 2ϩ has been found in the crystal structure of a similar protein, IIA Lac . The Mg 2ϩ was bound to Asp 81 in the core of the protein between its subunits (10). Control experiments (data not shown) established that 2 mM EDTA had no effect on the thermal denaturation curves of IIA Chb and phospho-IIA Chb and that EDTA completely reversed the divalent cation effects discussed below including the marked effects on thermal denaturation of the proteins.
The results are presented in Figs. 7 (IIA Chb ) and 8 (phospho-IIA Chb ), containing three panels in each. Panel A is the wave length scan from 190 to 250 nm. Panel B is the thermal denaturation experiment, followed at 220 nm. Panel C consists of the data in Panel B normalized for the first transition.  Table II show that there is a noticeable increase in stability of 20 M IIA Chb in the presence of 1 mM Mg 2ϩ , but the effects of 1 mM Cu 2ϩ and Ni 2ϩ were entirely unexpected. The T m values increased as follows (1 mM chloride salts): Mg 2ϩ , 2.75°C; Cu 2ϩ , 15.75°C; Ni 2ϩ , 29.25°C. Mg 2ϩ is considered in detail below. Cu 2ϩ and Ni 2ϩ were also tested over the concentration range 20 M to 1 mM using 20 M IIA Chb . An effect was observed on the CD spectrum at the lowest concentration, and the full effect was found at 100 M concentrations of the ions (data not shown).
Thermal unfolding experiments (Fig. 8, B and C) were also conducted with phospho-IIA Chb in the presence of the ions. In view of the effects of protein concentration on stability of the phosphoprotein, the ion studies were conducted at 22 and 150 M phospho-IIA Chb . Mg 2ϩ (Fig. 8 and Table II) showed a greater effect with phospho-IIA Chb than with IIA Chb . With II-A Chb , the T m increased about 2°C in the presence of Mg 2ϩ . By contrast, at each phospho-IIA Chb concentration tested, Mg 2ϩ increased the T m by 9°C. Interestingly, at the high phospho- IIA Chb concentration, Mg 2ϩ brought the T m to 55°C, about the same value as obtained with IIA Chb Ϯ Mg 2ϩ . Cu 2ϩ and Ni 2ϩ gave complex results: (a) ions appeared to decrease the helicity of the protein at room temperature; (b) the thermal unfolding process appears to proceed via several intermediates with two or three temperature transitions observed. DISCUSSION (GlcNAc) 2 is a PTS sugar in E. coli (1). The chb or chitobiose operon contains three genes originally thought to be required for the uptake and phosphorylation of cellobiose but now known to encode the three (GlcNAc) 2 -specific PTS transport proteins, IIA Chb , IIB Chb , and IIC Chb , respectively (7). There appear to be no published reports on the properties of IIA Chb nor on the respective phosphoprotein.
The phosphoryl group in phospho-IIA Chb exhibits properties of a typical ⑀2-N (i.e. N-3) phospho-His derivative with repect to akali stability, acid lability, and sensitivity to hydroxylamine and pyridine (Figs. 2 and 3). A plot of the first order rate constants as a function of pH also indicate that the phosphoryl group is linked to ⑀2-N (i.e. N-3) of the imidazole His ring rather than to the ␦1-N (i.e. N-1), as in phospho-HPr. Table I compares the rate constants for the hydrolysis of some phospho-PTS proteins with phospho-IIA Chb at pH 8.0, 37°C. The most interesting comparison is between E. coli phospho-IIA Chb and Staphylococcus aureus phospho-IIA Lac because of their sequence and structural similarities. The t1 ⁄2 for phospho-IIA Lac (called III Lac in Ref. 20) is 58 min, whereas the corresponding values for phospho-IIA Chb varied from 250 to 1080 min, depending upon protein concentration. It was this concentration effect that led to the experiments discussed below.
IIA Chb (previously IIA Cel ) is a member of the lactose/cellobiose PTS Enzyme II permeases. There is considerable amino acid sequence identity/similarity within this group of proteins (8,10) For example, IIA Chb shows 33% identity in its sequence to the S. aureus IIA Lac protein, and 35% to the homologous protein from Lactococcus lactis. The latter has been crystallized, and its structure has been determined (10). The structures of the protein monomers also show similarity. L. lactis IIA Lac is 83% ␣ helix, whereas the CD spectrum indicates that IIA Chb is from 75-85% helix.
There are also marked differences between E. coli IIA Chb and the IIA Lac proteins of S. aureus and L. lactis. For one, both IIA Lac proteins form stable trimers (10,20), whereas IIA Chb forms a very stable dimer. Second, phosphorylation of S. aureus IIA Lac trimer is thought to result in dissociation to phospho-IIA Lac monomers (21), but no monomer of either IIA Chb or phospho-IIA Chb was detected in the analytical sedimentation experiments. Third, phosphorylation has only a small effect on the CD spectrum of IIA Chb at pH 8.0 and 25°C (Fig. 4A), whereas there was a major change in the structure of phospho-IIA Lac as judged by relative reactivity to antibodies.
Nevertheless, phosphorylation of IIA Chb has a profound effect on the structure of the protein. The most obvious change was observed on thermal denaturation, where phospho-IIA Chb was found to have a T m 15°C lower than the unphosphorylated protein. Thus, phosphorylation significantly destabilizes II-A Chb . Below 25°C, at pH 8.0, the CD spectra of IIA Chb and phospho-IIA Chb are very similar. At 37°C, there is a minimal change in the spectrum of IIA Chb , but there is a striking change in the spectrum of phospho-IIA Chb . At this temperature, there is a loss of close to 35% of the helicity of the phosphoprotein. At the same time and under the same conditions, there is a the large change in the rate constant for the hydrolysis of phospho-IIA Chb (Table I). The t1 ⁄2 is 1800 min at 25°C and 250 min at 37°C, a substantially greater effect than expected for a 12°C change in temperature.
A second set of data required explanation. The rate constants for the hydrolysis of phospho-IIA Chb varied with protein concentration at 37°C (Table I). The same result was obtained in the thermal unfolding experiments, where the T m of phospho-IIA Chb changed significantly with concentration, whereas there was only a slight change with IIA Chb (Fig. 6 and Table II).
We interpret these results as follows: (a) At pH 8.0, in the temperature interval 25-37°C, there is little change in the structure of IIA Chb but a profound change in the structure of phospho-IIA Chb , leading to a loss of stability of the phosphoryl linkage in the phosphoprotein. (b) The effects of protein concentration on both the T m and the t1 ⁄2 of hydrolysis of phospho-IIA Chb reflect a change in the monomer/dimer transition of the phosphoprotein relative to IIA Chb . In monomer/dimer transitions of this type, the fraction of monomer/dimer increases significantly as the absolute concentration is decreased, and the data are explained if the phosphomonomer is less stable than the phosphodimer. The monomer/dimer transition has a K dissoc Յ 10 Ϫ7 M at 25°C, as deduced from the sedimentation experiments (3). Using this K dissoc value, one can calculate that  Table II. C, the spectra of IIA Chb at 25°C in the presence of divalent metal ions.
FIG. 8. Effect of divalent metal ions and temperature on the ellipticity of phospho-IIA Chb . The procedures described in Fig. 7 for measuring the ellipticity of IIA Chb were applied to the phosphoprotein. A, mean residue ellipticity at 220 nm of purified phospho-IIA Chb (22 M). B, data in A normalized for the first transition. The mid-point temperatures (T m ) derived from the figure are presented in Table II. C, The spectra of phospho-IIA Chb at 25°C in the presence of divalent metal ions. at 1 mM total protein, the monomer would constitute 1% of the total, whereas at 1 M protein, the monomer would comprise 27% of the mixture.
Finally, the structure of IIA Chb was surprisingly affected by certain divalent cations (Figs. 7 and 8 and Table II). In the presence of 1 mM divalent cation, the T m increased 3°C in Mg 2ϩ , about 16°C in Cu 2ϩ , and 29°C in Ni 2ϩ . Mg 2ϩ increased the T m of phospho-IIA Chb by 9°C. We do not know whether these cations have a stabilizing effect on the monomers or on the dimers or both nor the mechanism for these large effects. But it is important to note that Mg 2ϩ , for example, increases the stability of phospho-IIA Chb to where it is similar to that of the dephosphoprotein in the absence of divalent cation. Whether the metal ion effects are physiologically relevant remains to be determined, but as noted above, the crystal structure of IIA Lac revealed the presence of Mg 2ϩ or Mn 2ϩ between subunits. If IIA Chb has a similar structure, it may explain the increased stability of the protein in the presence of the metal ions to thermal denaturation.
The change in structure of phospho-IIA Chb between Ͻ25°C and 37°C may be physiologically important. Because the temperature change markedly affects the rate of transfer of the phosphoryl group to water, it may significantly increase the rate of transfer to IIB Chb , and thence to (GlcNAc) 2 as it is translocated by the membrane sugar receptor, IIC Chb . It is conceivable, for instance, that phosphoryltransfer from phospho-IIA Chb to IIB Chb to the (GlcNAc) 2 /IIC Chb complex proceeds as a quaterary complex of the three proteins and the sugar and not as usually visualized as a separate sequence of bimolecular reactions. If this is true, then it may be important to destabilize phospho-IIA Chb to accelerate its interaction with or insertion into the membrane. Appropriate kinetic studies are in progress to test these ideas. The effect of Mg 2ϩ on the kinetics will also be measured because Mg 2ϩ appears to stabilize phospho-IIA Chb .
Is destabilization of the active site domain a prerequisite for rapid phosphotransfer between proteins? Both the crystal and NMR structures of an analogous PTS protein, IIA Glc (also called III Glc ), are well established, as is the NMR structure of phospho-IIA Glc (22)(23)(24)(25)(26)(27)(28). NMR studies were conducted at 36.5°C, and some chemical shifts were found in phospho-IIA-Glc , but only in four dipeptide segments around the active site His. There were no significant structural changes in large segments of the protein as is reported here for phospho-IIA Chb or as has been reported for phospho-IIA Lac . It remains to be seen whether the structural effects observed upon phosphorylation of the monomeric IIA Glc or the dimeric IIA Chb (and trimeric IIA Lac ) best exemplifies the IIA proteins as a group.