Steady State Kinetics of Mannitol Phosphorylation Catalyzed by Enzyme IImtl of the Escherichia coli Phosphoenolpyruvate-dependent Phosphotransferase System

Phosphorylation Catalyzed by

enzyme IImtl can be adequately described by the following.
Mannitol,, + P-HPr +mannitol-Pi, + HPr (Eq. 1) With respect to the substrate mannitol,,, the enzyme catalyzes two reactions, (i) transport across the membrane and (ii) phosphorylation.The emphasis in kinetic studies of this enzyme has been on the phosphorylation reaction catalyzed by the enzyme solubilized in detergent, where the discrimination between "in" and "ou~'' is obscure (3-6).The most straightforward interpretation of the phosphorylation reaction in a noncompartmentalized system would be that exactly the same reaction is catalyzed as the one denoted in Equation 1, i.e. phosphorylation measures transport as well.However, since one of the two reactions of mannitol is invisible in this experimental setup this interpretation is not necessarily true.In fact, several lines of evidence indicate that the mechanism underlying the phosphorylation reaction is much more complex and a model showing the relation between the phosphorylation activity and the transport activity of the enzyme has been proposed (7).In this model two pathways for mannitol phosphorylation may be recognized, (i) binding to the periplasmically oriented binding site followed by translocation and phosphorylation and (ii) direct binding to the cytoplasmically oriented binding site followed by phosphorylation.
Here we describe the kinetics of mannitol phosphorylation catalyzed by enzyme IImtl under a number of different physical conditions that, potentially, assay these different kinetic pathways.
The data set generated in this paper will be used in the accompanying paper (8) to test a hypothetical kinetic scheme for enzyme IImt'.

Methods
Growth Conditions-E.coli strain ML308-225 was grown at 37 "C in medium 63 (11) containing 0.5% mannitol as the carbon source.Cells were grown in 5-liter flasks, filled with 2 liters of medium, and aerated by continuous shaking.Cells were harvested at A650 of 1.0.
Membrane Vesicles-These were prepared essentially as described (12).The vesicles were washed once with 25 mM Tris, pH 7.6, 1 mM dithiothreitol, 1 mM NaN3.Aliquots of 50 pl containing 1 mg/ml membrane protein (13) were stored in liquid nitrogen.Samples were thawed rapidly before use.Each sample was used only one time.
Activity Measurements-All experiments were performed at 30 "C.The activity of enzyme IIm" was measured by following the formation of [3H]mannitol-P or ["Clmannitol-P in time in a total volume of 100 pl.The buffer contained 25 mM Tris, pH 7.5,5 mM dithiothreitol, and 5 mM MgCl,.Four samples of 20 p1 each were withdrawn at consecutive times and analyzed for labeled mannitol-P as described (5).A fifth sample of 10 pl was used to relate the labeled mannitol concentration to the total radioactivity in the sample.A 20-pl sample of 0.8 p~ [3H]mannitol or 100 p~ ["C]mannitol contained 134,000 or 94,100 cpm, respectively.The error in these numbers is typically 1-2%.The procedure results in a time-independent background of 0.26 and 0.38% for [3H]mannitol and [14C]mannitol, respectively.The amount of mannitol-P formed was usually less than 10% of the initial mannitol concentration to assure initial rate measurements.P-enolpyruvate-dependent phosphorylation was initiated by addition of labeled mannitol to the enzyme suspension incubated with 5 mM Penolpyruvate, 0.22 p~ enzyme I, and the appropriate concentration of HPr.In the overall reaction the phosphoryl group donor for enzyme IImtl is P-HPr, which is rephosphorylated by P-enolpyruvate in a reaction catalyzed by enzyme I.The concentrations of enzyme I and P-enolpyruvate were chosen such that essentially all HPr was in the phosphorylated state under all turnover conditions.Therefore, the concentration of the substrate P-HPr equaled the concentration of HPr added initially to the mixture.
Membranes were solubilized by suspending I S 0 membranes in a buffer containing 25 mM Tris, pH 7.5, 5 mM dithiothreitol, and 0.2% decylPEG at a membrane protein concentration that was 20 times higher than used in the activity measurement.Subsequently, these solubilized membranes were diluted 20 times into the assay mixture, with or without 0.25% decylPEG for assays above and below the cmc of the detergent, respectively.
Eualuution of the Data-Rates were estimated from best fits of a straight line through the four consecutive time points and the background.The kinetic data was subjected to nonlinear fit procedures and the residual errors were analyzed graphically (14, 15).A residual is defined as the difference between the experimental and the calculated value divided by the calculated value d = (yewycale)/ycslc.The mean residual error of n data points is defined by Z(ABS &)/n.Values for kinetic parameters are averages from at least two independent experiments.

RESULTS
Kinetics of Mannitol Phosphorylation Catalyzed by Enzyme Z P " Solubilized in DecylPEG-We routinely purify enzyme IIm" in the presence of decylPEG, a detergent with a relatively low cmc of 0.035% (v/v).The enzyme is stable for prolonged periods of time in the presence of decylPEG concentrations above the cmc.Activity measurements indicate that the turnover rate of enzyme IImt' is independent of the decylPEG concentration over a large range of concentrations above the cmc (not shown).A concentration of 0.25% decylPEG was chosen in the first series of experiments.The buffer contained 25 mM Tris, pH 7.5, 5 mM MgC12, and 5 mM dithiothreitol.In this buffer the enzyme is in the dimeric state over a wide range of turnover conditions (6).
The rate of mannitol phosphorylation by P-HPr catalyzed by cytoplasmic membranes solubilized with decylPEG was measured over a range of mannitol concentrations from 1.25 to 640 p M in two independent experiments.The P-HPr concentration was 24 p M which is close to saturating (see below).The data from the two experiments were separately fitted to a single saturation curve resulting in mean residual errors of 34 and 29%.Analysis of the residuals shows a similar nonrandom distribution for both sets of data, indicating that the experimental data is not well described by a single saturation curve (Fig. 1, inset A ) .A double-reciprocal plot of the data clearly shows biphasic kinetics (Fig. 1, main plot).In the concentration range of 1.5 pM to about 5 p~ the data fit within experimental error to a single saturation curve (the

FIG. 1. Kinetics of mannitol phosphorylation catalyzed by cytoplasmic membranes solubilized in decylPEG.
The rate of mannitol phosphorylation ( u ) was measured at mannitol concentrations of 1. 25, 1.67, 2.5, 5, 10, 20, 26.7, 40, 80, 160, 320, and 640 p ~.The P-HPr concentration was 24 WM.The membrane protein concentration and the interval between the time points were adjusted according to the expected rate, ranging from 83.8 to 250 ng/ml and 2.5 to 20 min, respectively.The main plot shows the data plotted according to Lineweaver-Burk (0 and 0, duplicate experiments).Inset A, residual analysis after fitting the data to a single saturation curve.The data points are numbered as listed above.Inset E , residual analysis after fitting the data to the sum of two saturation curves.
Znset C, analysis of the low affinity phase.The difference of the actual rate and the extrapolated rate from the high affinity regime (Au) was plotted in a Lineweaver-Burk plot.The markers in the insets (0 and 0 ) correlate with those in the main plot.mean residual error is 0.9%) resulting in an affinity constant for mannitol KGt' = 2.45 pM and a maximal rate of 644 nmol/ min.mg membrane protein.We will refer to mannitol concentrations below and above 5 p~ as the high and low affinity regime, respectively.Subsequently, the data from the two experiments was fitted separately to the sum of two saturation curves resulting in mean residual errors of 3.4 and 3.2%.In spite of these small errors, analysis of the residuals (Fig. 1, inset B ) shows that the residuals are not randomly distributed around zero.Instead, the residuals of the two experiments cluster similarly around the x axis suggesting that the low affinity regime is not described adequately by a saturation curve.Znset C in Fig. 1 shows the low affinity regime when analyzed as the difference between the measured rates and the rates calculated with the parameters from the high affinity regime.The low affinity regime manifests itself most strongly in the mannitol concentration range from 20 to 100 p ~. Phenomenologically, the mannitol concentration giving half the maximal rate is about 60 p ~.
The affinity for P-HPr at a saturating mannitol concentration of 1 mM was estimated from the rates measured between The rate of mannitol phosphorylation ( u ) was measured at concentrations of mannitol of 0.2, 0.27, 0.4, 0.5, 0.67, 0.8, 1.0, and 2.0 p ~.The P-HPr concentrations were 3 (O), 0.6 (0), 0.3 (a), and 0.1 (0) p ~.The membrane protein concentration in the experiments were either 25 or 50 ng/ml.The left plot shows the data plotted according to Lineweaver-Burk.The right plots show the residuals of the data at the four P-HPr concentrations after fitting the entire data set to the rate equation pertinent to a ping-pong mechanism.The data points are numbered as listed above.
The markers in the plots on the right correlate with those in the left plot.

Kinetic parameters for P-enolpyruuate-dependent mannitol phosphorylation catalyzed by enzyme X P ' under different physical conditions
The kinetic parameters of purified enzyme P t ' , enzyme IImCl solubilized from IS0 vesicles, and assayed above the cmc of decylPEG (solubilized), enzyme IImCl embedded in the membrane of IS0 vesicles (IS0 vesicles), and enzyme 11"" solubilized from IS0 membranes and assayed below the cmc of decylPEG are presented.The parameters are described in the text.The affinity constants are in p ~, the maximal rates in min" and nmol/min.mgfor the purified enzyme and the other conditions, respectively. (3) e Number of experiments.
Single low affinity phase.
3 and 24 p~ P-HPr.The rate increased according to a single saturation curve with an affinity constant for P-HPr, KgHPr, of 4.7 f 0.6 p~ (Table I).
The results thus far suggest that the kinetic behavior of enzyme IImt' may be described by four parameters.In the high affinity regime, the affinity constant for mannitol, KEt'(HA), and the maximal rate, V,.,(HA), and in the low affinity regime, the affinity constant for HPr, KgHPr(LA), and the overall maximal rate, Vm..(TOT).The four parameters may be estimated from the saturation behavior of the rate between 1 and 5 p~ mannitol at saturating concentrations of P-HPr (KEt'(HA) and Vm.,(HA)) and between 3 and 24 p~ P-HPr number of data point at saturating concentrations of mannitol (Table I).These four parameters were determined again after purification of enzyme II"*' from the membranes.The results were qualitatively the same as observed with the solubilized membranes (Table I).The presence of the low affinity regime causes the turnover number of the enzyme to be a factor of 2-3 higher than reported before (4).
In an earlier paper from our laboratory (4), the rate of mannitol phosphorylation catalyzed by purified enzyme 11"" solubilized in the detergent Lubrol PXlOO was measured in the mannitol concentration range from 1 to 10 p ~, which in part extends into the low affinity regime described above.Different concentrations of P-HPr resulted in a set of parallel lines when analyzed in Lineweaver-Burk plots in accord with the expected "ping-pong" type of mechanism (4).We have repeated these experiments in the mannitol concentration range from 0.2 to 2 p~ in which there is no interference from the low affinity regime, and with an extended range of P-HPr concentrations.The whole data set was fitted to the rate equation pertinent to the ping-pong mechanism resulting in a mean residual error of 7%.Graphical analysis of the residuals revealed a nonrandom deviation of the experimental data from the rate equation for a ping-pong mechanism (Fig. 2, right).A double-reciprocal plot of the data shows that with 3 p~ P-HPr the data can adequately be described by a single saturation curve (0).This line parallels the line in the high affinity regime at 24 p~ P-HPr shown in Fig. 1.However, lowering the P-HPr concentration below 3 p M seems to affect the rate at the higher mannitol concentrations relatively more than at the lower concentrations, resulting in nonlinear relationships.This abnormal behavior was confirmed by measuring the rate in the P-HPr concentration range from 0.25 to 24 p~ at a fixed mannitol concentration of 5 pM (Fig. 3).Analysis of the residuals after fitting the data to a single saturation curve reveals a nonrandom distribution (insert, 0 ) .The double-reciprocal plot shows that the data fit reasonably well to a single saturation curve between 1 and 24 p M (mean residual error of 2.5%).The affinity constant for P-HPr equals The rate of mannitol phos- phorylation catalyzed by IS0 membranes (0) and IS0 membranes solubilized in 0.25% decylPEG (0) was measured at P-HPr concentrations of 0.25, 0.33, 0.5, 1, 2, 6, 12, and 24 PM, The mannitol concentration was 5 p ~. Main plot, the data plotted according to Lineweaver-Burk.Inset, residual analysis after fitting the data to a single saturation curve.The data points are numbered as listed above.
The markers correlate with those in the main plot.
0.59 p ~.However, below 1 p~ the rate decreases more rapidly than expected.Kinetics of Mannitol Phosphorylation Catalyzed by Enzyme IFt' Embedded in the Membrane of I S 0 Vesicles-In the high affinity regime, the rates of mannitol phosphorylation by enzyme IIm" embedded in the membrane of IS0 vesicles is much lower than observed with the solubilized enzyme (Fig. 4, 0 and X, respectively).The data up to 5 p~ mannitol fits to a single saturation curve with a mean residual error of 1.6%.The affinity constant for mannitol K$'(HA) equals 9.3 p~ and the maximal rate was 234 nmol/min.mg.The low affinity regime shows a pronounced second phase, resulting in much smaller differences between the rates catalyzed by the IS0 vesicles and the solubilized enzyme.Moreover, analysis of the saturation behavior of the rate between 3 and 24 pM P-HPr at a saturating mannitol concentration of 1 mM yielded an affinity constant for P-HPr of KEt'(LA) = 15 f 3 pM for the IS0 vesicles uersus 4.7 f 0.6 ptM for the solubilized enzyme.As a consequence, the maximal rate in the two conditions of the enzyme (V,,,(TOT)) is about equal (Table I).
The affinity constant for mannitol in the high affinity regime is significantly different with the solubilized enzyme and the enzyme in IS0 membranes, KG'(HA) = 2.3 0.3 and 9.3 & 0.5 p ~, respectively (Table I).In contrast, the behavior with respect to P-HPr was quite similar for the two conditions of the enzyme (Fig. 3, 0 and 0 ) even though the deviation from a single saturation curve with the IS0 membranes was less pronounced.In the concentration range of 1 to 24 p~ P-HPr the data for the membrane bound enzyme (0) fitted a saturation curve with a mean residual error of 0.8% and an  0), after solubilization in the presence of 0.25% decylPEG (X), and after solubilization in the presence of 0.01% decylPEG (0).The P-HPr concentration was 24 PM.The rates of the experiments with the IS0 vesicles (0) and in 0.25% decylPEG ( X ) are averages of two independent experiments.affinity constant of 0.84 p~ compared to 0.59 pM for the solubilized membranes.Below this concentration range the rate fell more rapidly than expected.The most important difference between the kinetics seems to be that, over the range of P-HPr concentrations tested, the rates with the solubilized enzyme were a factor of 5 higher than observed with the membrane bound enzyme.
Kinetics of Mannitol Phosphorylation Catalyzed by Enzyme IPtl Diluted Under the CMC of DecylPEG-Solubilization of cytoplasmic membranes with 0.25% decylPEG followed by dilution to 0.01% decylPEG which is below the cmc of the detergent, renders enzyme 11"" in a meta-stable state.The rate of mannitol phosphorylation catalyzed by the enzyme under these conditions was stable for about 1 h after which the rate reduced gradually, probably because of precipitation of the enzyme.Addition of decylPEG during this first hour resulted in an enzyme with the same activity as observed before the dilution step indicating complete reversibility of the treatment.
In the range from 1.25 to 640 pM mannitol and at 24 p M P-HPr the kinetics of enzyme IImt' under this physical condition is fitted by a single saturation curve with a residual error of 1.9% and an affinity constant for mannitol of 91 pM (Fig. 4,0).The maximal rate, inferred from a P-HPr titration at saturating mannitol concentration, is twice as high as observed with the enzyme solubilized and measured above the cmc or with the IS0 vesicles (Table I).Lowering of the P-HPr concentrations results in decreasing fits to a single saturation curve (Fig. 5, right).Lowering the P-HPr concentration from 12 (0) to 0.5 pM (0) did not have a significant effect on the rate in the range from 2 to 5 p~ mannitol but it lowered the rate dramatically at the higher mannitol concentrations.At the lower mannitol concentration limit, the kinetics followed a similar saturation curve but as the P-HPr concentration was lowered, it seemed to set a limit to the maximal rate at infinitely high mannitol concentrations.

DISCUSSION
The main kinetic characteristics of enzyme IImt' described in this paper are, (i) the biphasic mannitol-dependent kinetics FIG. 5. Kinetics of mannitol phosphorylation catalyzed by solubilized membranes under the cmc of the detergent.Cytoplasmic membranes were solubilized in 0.25% decylPEG and subsequently diluted to 0.01% decylPEG in the assay mixture.Final membrane protein concentration: 250 ng/ml.The rate of mannitol phosphorylation was measured at mannitol concentrations of 2, 2.67,4,8, 16, 32, 64, and 128 PM.The P-HPr concentrations were 0.5 (U), 1 (W), 3 (O), and 12 (0) p ~. Left plot, the data plotted according to Lineweaver-Burk.Rightplots, residual analysis after fitting the data at each P-HPr concentration separately to a single saturation curve.The data points are numbered as listed above.The markers correlate with those in the main plot.l/[rnannitoll (1/pM) number of data point of both the solubilized enzyme and the IS0 vesicles, (ii) the stimulation of the rate in the high affinity regime following solubilization of IS0 vesicles, (iii) the similar maximal rates for solubilized enzyme and IS0 vesicles, (iv) the deviation from ping-pong kinetics at limiting concentrations of P-HPr, (v) the monophasic kinetics of the solubilized enzyme diluted under the cmc at saturating P-HPr, (vi) the deviation from saturation kinetics at P-HPr limitation under the same condition, and (vii) the 2-fold increase of the maximal rate under this condition.In the accompanying paper (8) a hypothetical kinetic model for enzyme IImt' will be presented and the present data will be used to test the kinetic performance of the model.In this "Discussion" section only experimental aspects of the results and the interpretation of the different physical conditions will be discussed.Cytoplasmic Membranes os a Source of Enzyme ZP'-We have determined the kinetics of mannitol phosphorylation catalyzed by enzyme IImt' using cytoplasmic membranes derived from E. coli cells induced for the mannitol transport system as the source of the enzyme.This has the advantage that the kinetic behavior of the enzyme can be compared directly under different physical conditions.Nevertheless, care should be taken that enzyme IImt' is solely responsible for the kinetics.Therefore, a number of the experiments carried out on solubilized membranes were repeated with purified enzyme IImt' in the detergent decylPEG.Table I shows that the biphasicity of the mannitol-dependent kinetics pertinent to the solubilized cytoplasmic membranes is retained with the purified enzyme, indicating that it is a property of enzyme IImt' and not caused by some other component present in the membrane preparation.Other experiments with the purified enzyme showed that all the reported properties could be ascribed to enzyme IImt' (not shown).Apparently, enzyme 11"" behaves like an independent entity after solubilization from cytoplasmic membranes, a physical condition comparable with the purified enzyme kept soluble in detergent.
Homogeneity of the Enzyme Population-Enzyme IImtl is the only mannitol-specific transport protein known to be expressed during growth of E. coli on mannitol.Nevertheless, inhomogeneity in the enzyme 11"" population may arise during the isolation procedures of the membranes or the enzyme.Biphasic kinetics might be indicative for such inhomogeneity since the resulting kinetics would be a summation of the kinetics of the different populations of enzyme molecules.
Residual analysis suggests that the mannitol-dependent kinetics (Fig. 1) is a property of a single population rather than a summation of two independent populations.Moreover, the contribution of the high affinity phase to the overall maximal rate ( Vm,.(HA)/Vm.,(TOT),Table I) is not significantly different for solubilized IS0 membranes and for enzyme IImtl purified from these membranes.No shift between two hypothetical populations can be detected during the purification procedure which is more in line with a homogeneous enzyme population.
Different Physical Conditions for the Enzyme-The kinetic pathways of an enzyme may be better understood by monitoring the kinetic properties under different physical conditions.In the case of a transport protein where a substrate molecule is transported from one compartment to another, the presence or absence of substrate in either one or both of the two compartments is likely to have its effect on the kinetic behavior.Although the two compartments have disappeared after solubilization of cytoplasmic membranes, enzyme IImt' will still have its "cytoplasmic" and "periplasmic" side.However, both sides of the protein see exactly the same concentration of mannitol.On the other hand, with enzyme IImt' embedded in the membrane of IS0 vesicles the two sides of the protein face two different compartments, the internal and external volume.The external concentration of mannitol can be controlled experimentally.The internal mannitol concentration is determined by the activity of the enzyme, assuming that passive diffusion of mannitol over the membrane is negligible (16, 17) and is likely to be different from the external concentration.
The results indicate a remarkable difference between the kinetics of solubilized enzyme IImtL when determined above or below the cmc of the detergent.Since, under both conditions, there is no compartmentalization, the different kinetics must reflect differences within the enzyme itself, caused by the different surrounding of the hydrophobic part of the enzyme.The microenvironment of the hydrophobic membrane domain provided by the detergent micelle or the phospholipid bilayer may be comparable, allowing the same mobility in terms of conformational changes.Breakdown of the micelle surrounding the enzyme by lowering the detergent concentration below the cmc may reduce this mobility to such an extent that transitions of the hydrophobic domain are inhibited.More specifically, conformational changes involved in the translocation of mannitol may be blocked.This tentative interpretation is supported by the notion that the kinetics at saturating concentration of P-HPr shows two phases above the cmc and one below the cmc (Fig. 4, x and 0, respectively) suggesting that part of the kinetic scheme is inaccessible below the cmc.Homogeneity of the I S 0 Vesicle Preparation-Membrane vesicles prepared by passing cells through a French press have predominantly an inside-out orientation, but a small fraction of vesicles with a right-side-out orientation and open structures will be present (18, 19).Enzyme IImt' molecules in the right-side-out vesicle fraction will not contribute to the rate of phosphorylation of mannitol since the internal volume is not accessible to P-enolpyruvate, enzyme I, and HPr.The physical condition of enzyme IImtl in membranes that do not form closed vesicles is identical to the solubilized enzyme with respect to the substrate mannitol (see above).Therefore, the kinetics of mannitol phosphorylation catalyzed by the IS0 membrane preparation is likely to be a summation of the kinetics of I S 0 vesicles and the kinetics of open membrane sheets that will be similar to the kinetics of the solubilized enzyme (20).The high affinity regime in the mannitol-dependent kinetics catalyzed by the I S 0 membrane preparation could arise from the open membrane structures (Fig. 4,O and x).Then, the real kinetics of the IS0 vesicles would be described by the low affinity regime.In the accompanying paper (8) it will be demonstrated that this interpretation is consistent with the proposed kinetic model.Subtraction of the high affinity regime from the overall kinetics of the IS0 membrane preparation would leave the kinetics of the IS0 vesicles, but, unfortunately, it also contains the low affinity contribution of the open membrane structures.The different affinity constants for mannitol in the high affinity regime observed with the solubilized membranes and the open structures in the IS0 membrane preparation (KgL(HA) = 2.4 and 9.3 PM, respectively) may reflect the slightly different microenvironment of enzyme IImtl under the two conditions.The drawback of this is that the contribution of the open structures to the low affinity regime of the kinetics of the I S 0 membrane preparation cannot be quantified in any reliable way.It can only be concluded that the affinity for mannitol is relatively low at saturating concentrations of P-HPr and that the maximal rate is similar to that observed with the solubilized membranes.
The similar P-HPr concentration-dependent kinetics catalyzed by the solubilized membranes and the open membrane structures (Fig. 3) in the high affinity regime seems to be consistent with the notion that the different environments most likely are restricted to the membrane-bound part of enzyme IT"", which is the interaction site for mannitol.This would leave the environment of the cytoplasmic domains of enzyme IImtl, which constitute the phosphorylation pathway, identical under these conditions.
FIG.2.Kinetics of mannitol phosphorylation catalyzed by solubilized membranes in the high affinity domain.The rate of mannitol phosphorylation ( u ) was measured at concentrations of mannitol of 0.2, 0.27, 0.4, 0.5, 0.67, 0.8, 1.0, and 2.0 p ~.The P-HPr concentrations were 3 (O), 0.6 (0),0.3 FIG. 3. P-HPr-dependent kinetics of mannitol phosphoryl-ation in the high affinity domain.The rate of mannitol phos- phorylation catalyzed by IS0 membranes (0) and IS0 membranes solubilized in 0.25% decylPEG (0) was measured at P-HPr concentrations of 0.25, 0.33, 0.5, 1, 2, 6, 12, and 24 PM, The mannitol concentration was 5 p ~. Main plot, the data plotted according to Lineweaver-Burk.Inset, residual analysis after fitting the data to a single saturation curve.The data points are numbered as listed above.The markers correlate with those in the main plot.
mannitol] (1IpM) FIG. 4. Comparison of the kinetics of mannitol phosphorylation catalyzed by enzyme IIm" under different physical conditions.IS0 membrane vesicles were assayed in the absence of decylPEG (