Lactose Transport System of Streptococcus thermophilus. The Role of Histidine Residues

,

the phosphoeno1pyruvate:sugar phosphotransferase system (Poolman et al., 1989; Liau et al., 1991).In earlier reports, the IIA protein and IIA protein domains were designated Enzyme I11 and Enzyme I11 domains, respectively.The IIA protein domain of Lacs contains 2 histidine residues (His-537 and His-552) that are conserved in the homologous PTS' proteins (Fig. 1) (Poolman et aL, 1989).His-552 of Lacs corresponds to His-91 of the IIAGLC protein of E. coli, which has been shown to be phosphorylated by HPr-P (Dorschug et al., 1984).The lactose transport protein (Lacs) is a secondary active transport system that catalyzes the uptake of various a-and 8galactosides in symport with a proton as well as exchange of galactosides.'The heterologous exchange of lactose for galactose most likely occurs during metabolism of lactose by S. thermophilus and Luctobucillus bulgaricus (Poolman, 1990).
Although the overall similarity between the carrier domain of Lacs and the lactose carrier (Lacy) of E. coli is not significant (Poolman et al., 1989), a region (H376/E379) in Lacs can be identified that is similar to the H322/E325 region in Lacy (Fig. 1).Evidence has been presented indicating that His-322, Glu-325 (putative helix X), and Arg-302 (putative helix IX) of Lacy are in close contact with each other and that these residues could form a "charge relay" system that participates in the coupled transmembrane movement of protons and galactosides (Puttner et al., 1986(Puttner et al., , 1989;;Carrasco et al., 1986Carrasco et al., ,1989;;Menick et al., 1987;Lee et al., 1989).Recently, the requirement of an ionizable histidine residue at position 322 in Lacy for galactoside/proton symport has been questioned (King and Wilson, 1989a, 1989b, 1990;Franco and Brooker, 1991).It has been shown that some His-322 mutants of Lacy, which do not build up a galactoside concentration gradient, catalyze galactoside-dependent proton transport (King and Wilson, 1989a, 1989b, 1990).Moreover, the H322N mutant of Lacy still accumulates lactose against a concentration gradient, although the levels of accumulation are low compared to those of the wild-type strain (Franco and Brooker, 1991).
In this study, we assess the role of the histidine residues in the carrier and IIA protein domains of the lactose transport protein of S. thermophilus.The results indicate that His-552 is phosphorylated by HPr(His-P) from Bacillus subtilis, but that the conserved histidine residues in the IIA protein domain of Lacs are not essential for transport.His-376 in the carrier domain of Lacs could serve a role similar to that of His-322 in Lacy.The similarities and differences between the proposed active-site residues of Lacs and Lacy are discussed.
C .Foucaud and B. Poolman, submitted for publication.

EXPERIMENTAL PROCEDURES
Bacterid Strains and Plasmids Bacterial strains are described in Table I.The cells were grown aerobically in Luria broth at 37 'C (Sambrook et al., 1989) unless indicated otherwise.Plasmid pEKS8 containing a 4.1-kilobase EcoRI chromosomal DNA fragment of S. thermophilus, encoding the lactose transport protein (Lacs), has been described (Poolman et al., 1989(Poolman et al., , 1990)).For the construction of pSKE8, the 4.1-kilobase EcoRI fragment of pEKS8 was ligated in the EcoRI site of pBluescript IISK+ (Stratagene).PlasmidpAVLl is pACYC177 that carries a 9.5-kilobase HindIII chromosomal fragment of Klebsiella pneumoniae KAY2026, carrying the nag operon, including nagE, which codes for IICBA-Nacetylglucosamine (Vogler et al., 1988).

Site-directed Mutagenesis
The mutagenic primers are listed in Table 11.Site-directed mutagenesis was carried out by the method of Kunkel et al. (1987).
Mutations in pSKE8-Single-stranded uracil-containing DNA of pSKE8 was isolated from E. coli CJ236 (dut-, ung-)/pSKE8 after infection with helper phage M13KQ7 (Sambrook et al., 1989).Closedcircular heteroduplex DNA with the desired mutations was synthesized in vitro as described (Kunkel et al., 1987) and transformed to E. coli JMlOl (ung+).Plasmid DNA was isolated from a number of transformants, and plasmids bearing the desired mutation(s) were identified by digestion with restriction enzymes for which a site was lost or created by the mutagenesis (see Table 11).Subsequently, mutations were verified by nucleotide sequencing of double-stranded DNA using the dideoxy chain termination method (Sanger et al., 1977) and a set of primers complementary to a region of lacs located 50-100 bases downstream or upstream of the mutation site.Each mutant was isolated independently at least twice.
Mutations in pEKS8"The 648-base pair KpnI-EcoRI fragment of pEKS8, comprising the carboxyl-terminal 323-base pair region of lacs, was ligated in the replicative form of M13mp18/M13mp19.Recombinant phage DNA was used to transfect E. coli CJ236; and single-stranded uracil containing M13mp18 or M13mp19 DNA, including the KpnI-EcoRI fragment, was isolated.Synthesis of the complementary strand and screening of mutants were carried out as described above.The entire region of lacs was sequenced.Subsequently, the KpnI-EcoRI fragment of pEKS8 was exchanged for the corresponding fragment containing the desired mutation that was isolated from the replicative form of the MI3 DNA.

Preparation of Cell Suspensions
Overnight cultures or exponentially grown cells (Am = 0.4-0.8)were harvested by centrifugation, washed twice, and resuspended to a final protein concentration of 20-80 mg/ml in 50 mM potassium phosphate (pH 6.5), 5 mM magnesium sulfate (KPM buffer).Measurements of the membrane potential and analysis of the effects of ionophores (valinomycin, nigericin) on transport were performed with in S. thermophilus 9151 E. coli cells that were treated with EDTA essentially as described previously (Sarkar et al., 1988).Concentrated cell suspensions were stored on ice until use.

Transport Assays
Transport experiments were performed at 20, 30, or 37 "C as specified in the figure legends.
Active Tramport-Cells were diluted to a final protein concentration of 0.6-1.2mg/ml in KPM buffer containing 10 mM D-lithium lactate as the electron donor.After 2 min of pre-energization in the presence of oxygen, radiolabeled substrate was added; and at appropriate time intervals, the uptake reaction was stopped by addition of 2 ml of ice-cold 100 mM LiCl.Cells were collected by filtration on a 0.45-pm cellulose nitrate filter (Millipore Corp.) and washed with 2 ml of ice-cold LiCl.
Efflux and Exchange-Preloading of cells with sugars was achieved by incubation (overnight at 4 "C) of the concentrated washed cell suspensions with the appropriate concentration of radiolabeled galactoside in the presence of deoxyribonuclease I (20 pglml).The next day, potassium azide and carbonyl cyanide m-chlorophenylhydrazone were added to final concentrations of 30 mM and 50 pM, respectively; and the cells were incubated for another 2 h at room temperature essentially as described (King and Wilson, 1990).For efflux and exchange, 1-p1 aliquots of concentrated cell suspension (50-80 mg/ ml) were diluted into 500 pl of KPM buffer containing no substrates and unlabeled galactosides, respectively.The transport reaction was stopped by rapid filtration as described above.

Protein Purification
Enzyme I and HPr of B. subtilis and E. coli were purified as described previously (Reizer et al., 1989, 1992).

I1
Mutations in the lactose transport gene of S. thermophilus The nucleotides changed by the site-directed mutagenesis are underlined.The nucleotide changes resulting in the desired amino acid substitution are indicated in boldface type.The last column indicates whether a new restriction was created or a restriction site was lost as a result of the mutagenic event(s).

Miscellaneous
["PIP-enolpyruvate was prepared as described previously (Reizer et al., 1984).For the calculation of intracellular concentrations, a specific internal volume of 3 pl/mg of cell protein was used.Protein was measured by the method of Lowry et al. (1951) with bovine serum albumin as a standard.Plasmid DNA was isolated by the alkaline lysis method (Birnboim and Doly, 1979).For sequencing of doublestranded DNA, plasmid DNA was further purified by one or two polyethylene glycol precipitation steps and additional phenol/chloroform extraction, followed by ethanol precipitation and denaturation by alkali treatment (Sambrook et al., 1989).(Bassilana et al., 1987).

RESULTS
Histidine Mutagenesis-The lactose transport protein (Lacs) of S. therrnophilw contains 11 histidine residues (circled in Fig. lA), of which 5 are present in the carrier domain and 6 in the IIA protein domain.Each of the histidine residues was replaced by site-directed mutagenesis with glutamine or arginine (Table 11).Glutamine and arginine were chosen as replacements on the basis of amino acid similarity coefficients that were obtained from the main chain torsion angle distributions of highly resolved protein structures (Niefind and Schomburg, 1991).Plasmids with the histidine mutations were used to transform E. coli HB101, and transformants were streaked on lactose MacConkey agar plates.The lac phenotype of each of the mutants appeared to be indistinguishable from that of the wild-type strain, i.e.HBlOl/pSKES or HBlOl/pEKS8.
Lactose Transport by Histidine Mutants-Each of the plasmids bearing a mutation resulting in the replacement of a single histidine residue was used to transform E. coli DW2 (AlacZY).Lactose transport by DW2/pSKE8 (wild-type and histidine mutants) is shown in Fig. 2. With the exception of the H376Q (and to some extent, H155Q) mutant protein(s), the initial rates of uptake and the steady-state levels of lactose accumulation of the histidine mutants were similar to those of the wild-type.Similar results were obtained when transport To ensure that secondary mutations did not contribute to the observed phenotype, the transport activity of each mutant was determined with two independent isolates.The data from a second set of mutants did not differ significantly from those presented in Fig. 2. To analyze the transport properties of each of the mutants in more detail, the affinity constant (KT) and maximal rate of uptake ( VmaX) for lactose were determined (Table 111).The data show that the H376Q mutant had a lower KT and Vmax for lactose, whereas the kinetic properties of the other mutants were not significantly affected.
Since Lacs transports lactose as well as melibiose (a- galactoside), TMG, galactose, and other galactosides: the effects of the H376Q mutation on the uptake of these carbohydrates were compared (Fig. 3).To avoid metabolism of lactose and melibiose, transport studies were carried out with E. coli DW1 (AlacZY, ArnelAB)/pSKE8(wild type) and DWl/ pSKE8(H376Q).The kinetic parameters for lactose uptake by E. coli DW1 bearing the wild-type or H376Q mutant proteins were similar to those obtained with E. coli DW2 bearing the correspOnding transport protein (Table 11).The data presented in Fig. 3 show that uptake of melibiose and TMG was markedly more affected by the H376Q mutation than uptake of lactose.In fact, the uptake rates were too low to determine accurately the kinetic parameters of the H376Q mutant (Fig. 3, B and C, insets).The KT values for uptake of lactose, melibiose, and TMG by LacS(wi1d type) were 0.7, 0.53, and 0.27 mM, respectively.The Vmax values for uptake of lactose, melibiose, and TMG were 270, 140, and 60 nmoll min/mg of protein, respectively.
In contrast to transport of lactose (Fig. 3A), TMG was not accumulated against a concentration gradient by the H376Q mutant (Fig. 3C).To establish whether this difference indeed reflected differences in the ability to accumulate various sugars or whether the apparent failure to accumulate TMG was due to higher efflux by passive and/or facilitated diffusion, a number of experiments were carried out.First, efflux of TMG was analyzed by diluting [14C]TMG-loaded cells into buffer without TMG.No significant differences were observed between the wild-type and H376Q mutants (Fig. 4, 0 ) .Second, the level of accumulation of TMG by the wild-type and H376Q proteins was determined over a wide range of concentrations (5-165 PM) and at different pH values (5.5-8.0).If passive efflux of TMG, in combination with a reduced rate of uptake, affects the accumulation level, one might observe a significant :' B. Poolman, unpublished data.

FIG. 1.
A , secondary structure model of lactose transport protein of S. thermophilus based on hydropathy plot of deduced amino acid sequence (Poolman et al., 1989)  trations only.Some accumulation (2-3-fold) of TMG was observed a t concentrations <50 PM and at pH 5.5-6.0 (data not shown); the level of TMG accumulation by the wild-type strain also decreased with increasing TMG concentration.In conclusion, the data suggest that the H376Q mutation affects sugar recognition by Lacs as well as the coupling between galactoside and proton transport.
To assess the sugar recognition properties of the wild-type and H376Q proteins in more detail, the effect of an 85-fold excess of unlabeled sugar on uptake of lactose was determined.The data presented in Table IV show that the galactosides lactose, melibiose, TMG, P-D-galactopyranosyl-1-thio-P-Dgalactopyanoside, isopropyl-1-thio-8-D-galactopyranoside, and galactose inhibited uptake of ['4C]lactose (47 p~ final concentration), whereas maltose (a-glucoside) and sucrose (fructofuranoside) did not.
The pattern of inhibition of lactose uptake by the unlabeled sugars does not indicate major differences in sugar recognition of the wild-type and H376Q B. Poolman, S. Yoast, and B. F. Schmidt, unpublished data.transport proteins, although raffinose (trisaccharide/galactoside) was a more effective inhibitor of LacS(H376Q) compared with LacS(wi1d type).
The data presented above describe the effect of H376Q substitution on uphill (Ap-driven) transport by the lactose carrier protein.Under these conditions, a complete translocation cycle includes binding and release of H' and galactoside as well as reorientation of the loaded and unloaded binding sites.Since Lacs also catalyzes an exchange reaction involving only binding and release of galactoside and reorientation of loaded binding sites, nonequilibrium exchange of TMG and lactose by the wild-type and H376Q transport proteins were compared (Figs. 4 and 5).For the exchange reaction, cells were preloaded with [I4C]TMG or ["'C]lactose and, following treatment with azide/carbonyl cyanide m-chlorophenylhydrazone (see "Experimental Procedures"), diluted 500-fold into buffers containing unlabeled TMG (or lactose) at concentrations ranging from 0 to 20 mM.Since the rate of net galactoside transport was negligible in these "energy-poisoned cells (inferred from uptake by unloaded cells) (data not shown), 2. Lactose uptake by E. coli DW2/pSKE8 wild-type and histidine Lacs mutants.Cells were suspended in KPM buffer (pH 6.5) supplemented with 10 mM D-lithium lactate to a final protein concentration of 0.8 mg/ml.After 2 min of pre-energization, ["C] lactose was added to a final concentration of 10 PM, and uptake was assayed for different time intervals.Cells harvested in the stationary phase of growth were used the assay temperature was 30 "C.HxQ, histidine mutants except H155Q and H376Q.0, wild-type ( W T ) Lacs; 0, H155Q mutant; 0, H376Q mutants.The shaded area denotes the results obtained with all other histidine mutants of Lacs.

Kinetic parameters of lactose transport by wild-type and histidine
Lacs mutants Experimental conditions were similar to those described in the legend to -Fig. 3,except that lactose concentrations ranged from 10 p M to 1.6 mM. the observed isotope exit reaction reflects the real exchange of galactoside.Fig. 4 shows that initially the release of ["C] TMG from the cell was monoexponential with a slope that depends upon the external TMG concentration.The exchange reaction catalyzed by the H376Q mutant was somewhat slower than that of the wild-type strain (the KT values for external TMG were very similar, i.e. between 0.6 and 1.0 mM).This can be seen more readily from the kinetic analysis of lactosei,/ lactoseout exchange (Fig. 5).At an internal lactose concentration of 4.5 mM (see legend to Fig. 5), the exchange kinetics yielded KT values for external lactose of -10 mM, which is about an order of magnitude higher than that of Ap-driven lactose uptake (Fig. 3 and Table 111).

TABLE IV
Substrate specificity of LacS(wild type) and LacS(H376QJ Lactose uptake was assayed at different time intervals between 0 and 90 s essentially as described in the legend to Fig. 2. Cells harvested in the stationary phase of growth were used; the assay temperature was 30 "C; the final [13C]lactose concentration was 47 phi.The sugars indicated were added to a final concentration of 4 mM (85-fold excess).
LacS(wi1d type) Table 111, on the other hand, indicate that the H552Q mutation does not affect transport activity.This suggests that phosphorylation of His-552 does not occur or that phosphorylation of the protein is not essential for transport activity.To assess the functional activities of the IIA protein domain of Lacs, several approaches were undertaken.First, Ap-driven uptake and exchange of lactose by DW2/pEKS8 and PPA209/ pEKS8 (ptsI-) were compared.The results obtained with the and [I4C]TMG were added to final concentrations of 45, 44, and 43 pM, respectively; and the assay temperature was 37 "C.For the kinetics of galactoside uptake (insets), cells harvested in the stationary phase of growth were used.Initial rates of uptake were estimated from uptake after 10 s (in triplicate) at galactoside concentrations varying from 10 p~ to 1 mM.The assay temperature was 37 "c; the final protein concentrations were between 0.8 and 1.2 mg/ml.WT, wild type.FIG. 5. Nonequilibrium exchange of lactose by E. coli DW2/ pSKE8 LacS(wi1d type) and LacS(H376Q).Cells were loaded in the presence of 2.9 mM ['4C]lactose as described under "Experimental Procedures."From the amount of ["Cllactose taken up (expressed as nanomoles/milligram of protein) and using a specific internal volume estimated to be 4.5 and 4.8 mM for E. coli DW2/pSKE8 LacS(wi1d of 3 pl/mg of protein, the final internal lactose concentrations were type) and LacS(H376Q), respectively.Exit of ["Cllactose was determined following 500-fold dilution of a concentrated cell suspension (-80 mg of protein/ml) into KPM buffer (pH 6.5) containing no lactose (efflux) or 0.1, 0.2, 0.4, 1.0, 2.0, 4.0, 10, or 20 mM lactose (nonequilibrium exchange).The release of ['4C]lactose from the cell was monoexponential, and first-order rate constants ( k = In 2/P) could be determined.The initial rates ( V ) of exit were obtained by multiplying k with the internal lactose concentration ( V = k X Si).The assay temperature was 30 "C. ptsI mutant were similar to those of the wild-type mutant (data not shown).Second, manipulation of the level of PTS phosphoprotein intermediates by addition of nonmetabolizable sugar analog methyl-a-D-glucopyranoside and 2-deoxyglucose (Nelson et al., 1986) did not affect the lactose transport activities.Third, pEKS8 (lacs+) was unable to complement E. coli LM1 (crr-, nugE) with respect to transport of amethylglucoside.In accordance with previous observations (Vogler et al., 1988), pAVL1 containing nugE was able to restore the uptake of a-methylglucoside in E. coli LM1 to -20% of the wild-type activity (strain LR2-167).Fourth, by examining the sequences of the IIA protein(s) (domains) of the lactose transport proteins of S. thermophilus and L. bulgaricus and the PTS proteins indicated in Fig. IC, two positions were identified at which the lactose transport proteins differed significantly from the PTS proteins, i.e.Ile-548 and Gly-556.Three residues are close to proposed phosphorylation site residue (His-552); and the crystal structure of the IIAG'" protein domain of the glucose PTS of B. subtilis indicates that the equivalent residues, i.e.Glu-79 and Asp-87, can assume a role in recognition of the interactive PTS proteins (the IIB protein domain or HPr) and facilitate the phosphoryl transfer reaction (Liau et al., 1991).The double mutant LacS(I548E/G556D) was constructed, but the I548E/G556D mutations did not affect lactose transport activity (data not shown).

Sugar
Phosphorylation of Lactose Transport Protein-To establish whether Lacs could be phosphorylated by HPr(His-P), inside-out membrane vesicles of E. coli T184/pKK223-3 (lacs-, vector control) and T184/pEKS8 (lacs+) and various mutants (see "Experimental Procedures") were isolated.The results shown in Fig. 6 indicate that the wild-type mutant as well as the 1548E/G556D mutant were readily phosphorylated by ["'PIP-enolpyruvate in the presence of purified Enzyme I and HPr isolated from B. subtilis.By contrast, no phosphorylation of the mutant H552Q was detected.The phosphorylation of the proteins was not significantly affected by methyl-a-Dglucopyranoside. Surprisingly, but in accordance with the data presented above, phosphorylation of LacS could not be detected with Enzyme I and HPr isolated from E. coli (data not shown).Altogether, these results indicate that PTS-catalyzed phosphorylation of His-552 of Lacs can occur, but that it does not serve an essential functional role in the transport of lactose under the conditions employed in this study.

DISCUSSION
On the basis of hydropathy analysis and a comparison of the primary structures of Lacs and MelB, a secondary structure model of the lactose transport protein is proposed (Fig. 1A).The polypeptide is organized into 12 hydrophobic transmembrane a-helical domains that are connected alternately by periplasmic (OUT) and cytoplasmic ( I N ) hydrophilic charged segments.The topology of the periplasmic loops that connect helixes I and 11, I11 and IV, and V and VI as well as the cytoplasmic loop that connects helixes VI-VI1 has been confirmed by analyzing lacs-alkaline phosphatase gene (phoA) fusions (Manoil and Beckwith, 1986): The IIA protein domain (boxed sequence in Fig. lA) is proposed to be on the cytoplasmic side of the membrane.Lacs contains 11 histidine residues (circled in Fig. lA), of which 5 are present in the carrier domain and 6 in the IIA protein domain.None of the histidine residues of Lacs are conserved in the homologous melibiose carrier protein (MelB).Although Lacs and the lactose carrier protein (Lacy) of E. coli are not homologous (Poolman et al., 1989), a region similar to the proposed catalytic site of Lacy was identified in Lacs (Fig. 1B).The conserved charged residues in the regions flanking His-376 of Lacs and His-322 of Lacy (indicated by arrows) are also conserved in the Lacs protein of L. bulgaricus, the Lacy protein of K. pneumoniae, and the raffinose transport protein (Ram) of E. coli (Fig. 1B).His-322 (arrow 3) and Glu-325 to form a H+ relay system that could function as the chemical pathway for H+ movement through the lactose carrier (Carrasco et al., 1986(Carrasco et al., , 1989;;Puttner et al., 1986Puttner et al., , 1989;;Menick et al., 1987;Lee et al., 1989).Although evidence against this proposal has been presented (King andWilson, 1989a, 1989b;Franco and Brooker, 1991), it is evident that this region is crucial for energy transduction by the carrier protein.Mutations at position 322 in Lacy also affect the sugar recognition properties of Lacy (Franco et al., 1989;Collins et al., 1989).Indeed, some properties of the LacY(H322) mutants resemble those of LacS(H376Q).For instance, active accumulation of lactose, melibiose, and TMG is highly compromised in cells bearing the H376Q protein.Furthermore, the affinity for galactosides (lactose) is altered by the H376Q substitution.The histidine residue corresponding to His-376 in Lacs and His-322 in Lacy is not conserved in MelB.Recent experiments in which the histidine residues of MelB were replaced with arginine residues indicated that only His-94 is important for transport activity (Pourcher et al., 1990a).The H94R mutation caused complete loss of sugar binding and transport, whereas melB was expressed normally.
Interestingly, the residue corresponding to Glu-325 of Lacy (Fig. lB, arrow 4 ) is conserved in the lactose transport proteins of E. coli, K. pneumoniae, s. thermophilus, and L. bulgaricus; the raffinose transport protein of E. coli; and the melibiose carrier of E. coli.Substitution of Glu-325 in Lacy leads to a transport protein that does not catalyze lactose/H+ symport or lactose efflux, but catalyzes exchange and counterflow at normal rates (Carrasco et al., 1986).Similarly, substitution of the corresponding residue in MelB (Glu-361) for glycine, aspartate, or alanine affects sugar and cation translocation, but not recognition of the substrates (Pourcher et al., 1990b).
Mutations of the conserved lysine (Fig. lB, arrow 2) have been isolated for Lacy after selection for enhanced recognition of maltose (a-glucoside) and resistance for TDG (Collins  et al., 1989).The LacY(K319N) mutant protein exhibits diminished recognition of P-galactosides and @-glucosides.The conserved acidic residue corresponding to arrow 1 (Fig. 1B) has not been investigated in any of the transport proteins.Overall, comparison of the sequences presented in Fig. 1B and the data obtained from mutagenesis experiments suggests that the region connecting helixes X and XI of Lacs is important for substrate recognition.
A difference worth noting between the Lacs and Lacy regions is the higher polarity of the residues around His-376 (LacS) in comparison to the region around His-322 (Lacy).As a consequence, hydropathy analysis predicts the His-322 region of Lacy to be in the membrane, whereas the His-376 region of Lacs is predicted to be in a cytoplasmic loop.Nevertheless, the local environment of the important residues in both protein segments might be similar if the regions are located near the head groups of the lipid bilayer.
The effect of the H376Q substitution on galactoside transport by Lacs has been analyzed by measuring Ap-driven uptake and exchange under nonequilibrium conditions.Assuming that the V,,, of the exchange reaction reflects the level of expression of the lactose transport protein, the effect of the H376Q mutation on the kcat of Ap-driven uptake is -2fold smaller than indicated by the V,,, values for uptake of lactose, melibiose, and TMG (Fig. 3).The observation that Ap-driven transport of galactosides is more severely affected by the His-376 mutation than the exchange reaction suggests that the histidine at position 376 is important for energy transduction, i.e. coupled movement of galactosides and protons.The results also indicate that His-376 is not obligatorily required for the active accumulation of galactosides.
The in vitro phosphorylation assays indicated that Lacs can be phosphorylated by HPr(His-P) of B. subtilis and that His-552 in the IIA protein domain is most likely the phosphorylation site.Nevertheless, the in vivo transport experiments that were performed in E. coli bearing Lacs did not provide an indication regarding the functional and/or regulatory role(s) of the phosphorylation reaction.Although B. subtilis IIAG1" can readily replace E. coli IIAG1" with respect to sugar transport and regulation (Reizer et al., 1992), it is possible that the IIA protein domain of Lacs is a poor phosphoryl acceptor and/or phosphoryl donor of the E. coli HPr(His-P) and/or the IIBGLc domain, respectively.Consequently, discerning the role of Lacs phosphorylation may require the use of PTS constituents derived from homologous systems (or other Gram-positive bacteria) rather than the heterologous system used in the in vivo studies described here.
In addition, we cannot exclude the possibility that phosphorylation of the IIA protein domain of Lacs does not play a direct role in the translocation of lactose across the membrane, but rather functions in P-enolpyruvate-dependent and PTS-catalyzed phosphorylation of the intracellularly formed glucose following hydrolysis of the incoming lactose.Guided by these suppositions, a study is now underway to examine the role of Lacs phosphorylation in S. thermophilus.

FIG. 4 .
FIG.3.Uptake of lactose, melibiose, and TMG by E. coli DWl/pSKE8 LacS(wi1d type) and LacS(H376Q).Experimental conditions were the same as described in the legend to Fig.2, except P-enolpyruvate-dependent phosphorylation of His-552 of the ~ lactose transport protein.The data presented in Fig. 2 and exponential phase of growth were used ["C]lactose, [3H]melibiose, that for the time course of galactoside uptake, Eells harvested in the

FIG. 6 .
FIG. 6.In vitro phosphorylation of wild type and mutantsLacS proteins.Phosphorylation of Lacs was achieved as described under "Experimental Procedures."The minus and plus signs refer to the absence and presence of 1 mM 2-deoxyglucose in the phosphorylation assay.The positions of phosphorylated Enzyme I (E,) and Lacs are indicated.The first lane shows the results of a control phosphorylation reaction using membranes that do not bear LacS.WT, wild type.

TABLE I Bacterial strains
a P. W. Postma, unpublished data.

thermophilus OUT 9153 A Conserved residues in carrier domain of
. Histidine residues are circled, and His-376 is also indicated by number.B, conserved residues in carrier domain of Lacs and lactose, melibiose, and raffinose transport proteins of E. coli.The region corresponds to the loop that connects helixes X and XI of LacS.Sources of sequences are as follows: E. coli (ec) Ram, Aslanidis et al. (1989); K. pneurnoniae (kp), Lacy, Mcmorrow et al. (1988); E. coli Lacy, Buchel et al. (1980); S. therrnophilus, ( s t ) LacS, Poolman et al. (1989); L. bugaricus (lb) LacS, Poolman et al.' and E. coli MelB, Yazyu et al. (1984).C, Conserved residues in IIA protein domain of LacS.The region shown corresponds to the proposed phosphorylation site of LacS.Sources of sequences are as follows (see also B ) : Crr, Nelson et al. (1983); NagE, Rogers et al. (1988); BglS (also indicated as BglF), Schnetz et al. (1987); Streptococcus rnutans (srn) ScrA, Sato et al. (1989); and B. subtilis ( b s ) PtsG, Gonzy-Tr6boul et al. (1989) and Sutrina et al. (1990).Residues corresponding to His-537 and His-552 in the lactose transport protein of S. t h rmophilus are indicated by arrows 1 and 2, respectively.Shaded arrows indicate residues that are conserved in the PTS proteins, but not in LacS.L a & Lactose Transport in S.