Effect of detergents on galactoside binding by melibiose permeases

Abstract

The effect of various detergents on the stability and function of melibiose permeases of Escherichia coli (MelB_Ec) or Salmonella typhimurium (MelB_St) were studied. In n-dodecyl-β-d-maltoside (DDM) or n-undecyl-β-d-maltoside (UDM), WT MelB_St binds melibiose with an affinity similar to that in the membrane. However, with WT MelB_Ec or MelB_St mutants (Arg141→Cys, Arg295→Cys or Arg363→Cys), galactoside binding is not detected in these detergents, but binding to the phosphotransferase protein IIA^Glc is maintained. In the amphiphiles lauryl maltose neopentyl glycol (MNG-3) or glyco-diosgenin (GDN), galactoside binding with all the MelB proteins is observed, with slightly reduced affinities. MelB_St is more thermostable than MelB_Ec, and the thermostability of either MelB is largely increased in MNG-3 or GDN. Therefore, the functional defect with DDM or UDM likely results from relative instability of the sensitive MelB proteins, and stability, as well as galactoside binding, is retained in MNG-3 or GDN. Furthermore, isothermal titration calorimetry of melibiose binding with MelB_St shows that the favorable entropic contribution to the binding free energy is decreased in MNG-3, indicating that the conformational dynamics of MelB is restricted in this detergent.

Membrane transporters, receptors, and channels play crucial roles in cellular functions by moving molecules across cell membranes. Detergents are essential tools for studying the structure and function of membrane proteins; however, selection of a detergent that retains activity and conformational stability is challenging because knowledge about detergents is still limited. The mild, nonionic detergent n-dodecyl-β-d-maltoside (DDM) is probably the most commonly used detergent for structure determination and functional analysis. The novel amphiphiles lauryl maltose neopentyl glycol (MNG-3) and glyco-diosgenin (GDN) have been shown to be superior to DDM or n-undecyl-β-d-maltoside (UDM) in maintaining solubility of several membrane proteins, including the melibiose permease of Escherichia coli (MelB_Ec) or Salmonella typhimurium (MelB_St), and G protein-coupled receptors.^{1, 2} However, their effects on substrate-binding affinity and binding thermodynamics to MelB have not been tested yet.

Both MelB_Ec and MelB_St catalyze symport of galactosides with H⁺, Na⁺, or Li^+3-7 and are well-characterized members of the glycoside-pentoside-hexuronide/cation subfamily^{8, 9} of the major facilitator superfamily of membrane transport proteins.^10-13 The X-ray crystal structure of MelB_St shows that MelB is composed of N- and C-terminal domains containing six irregular transmembrane helices surrounding a deep aqueous cavity open to the periplasmic side.¹² This overall fold is similar to other major facilitator superfamily of transport proteins, such as the lactose permease (LacY).^{14, 15}

MelB_St is effectively extracted from membranes with DDM or UDM, and the purified protein in UDM is monodisperse on gel filtration chromatography¹² and does not precipitate when stored at 0 ºC for months. MelB_St solubilized in UDM binds melibiose,^{12, 16} nitrophenyl-α-galactoside (α-NPG),¹⁶ and the fluorescent sugar 2’-(N-dansyl)aminoalkyl-1-thio-β-d-galactopyranoside (D²G).¹² Melibiose binding affinity with right-side-out (RSO) membrane vesicles containing WT MelB_St or with the purified proteins in UDM is comparable, with a K_d of ~1 mM.^{6, 12} Surprisingly with UDM-solubilized MelB_Ec, melibiose reversal in Trp→D²G fluorescence resonance energy transfer (D²G FRET) experiments is not detected. Thus far, all Trp→D²G FRET data with MelB_Ec were derived solely from studies with reconstituted proteoliposomes or membrane vesicles.^17-20 In this communication, the effect of UDM, DDM, MNG-3, and GDN on ligand binding by MelB_Ec or MelB_St is characterized by isothermal titration calorimetry and/or D²G FRET assays. The results indicate that the functional defect with DDM or UDM likely results from relative instability of the sensitive MelB proteins, and that MNG-3 or GDN maintains the stability and galactoside binding with either MelB.

Materials and Methods

Materials

D²G was kindly provided by H. Ronald Kaback and Gérard Leblanc. Synthesis of MNG-3 and GDN was described previously.^{1, 2} DDM, UDM, and DM were purchased from Anatrace. All other materials were reagent grade and obtained from commercial sources.

Plasmids

The expression plasmid pK95ΔAH/WT MelB_Ec was from Gérard Leblanc. ²¹ The pK95ΔAH-based plasmids were used for overexpressing WT MelB_St⁶ and MelB_St mutants R141C, R295C, or R363C.¹³

Preparation of RSO vesicles

RSO membrane vesicles were prepared from E. coli DW2 cells by osmotic lysis,^{6, 22, 23} resuspended with 100 mM KP_i (pH 7.5), and stored at −80 °C.

Trp→D²G FRET

RSO membrane vesicles or detergent-solubilized samples were used for Trp→D²G FRET measurements with an Amico-Bowman Series 2 (AB2) Spectrofluorometer. Trp residues were excited at 290 nm, and emission spectra were recorded between 430 and 550 nm. In time traces, emission was recorded at 465 nm for MelB_Ec or 490 nm for MelB_St. Successive addition of 10 μM D²G, 20 mM NaCl, and excess melibiose was done for all D²G FRET measurements.

Determination of IC₅₀ of melibiose for the half-maximal displacement of bound D²G (10 μM) was carried out as described.^{6, 13, 24} Briefly, stepwise addition of melibiose was performed during D²G FRET until no further change in the FRET signal was observed. The IC₅₀ was determined by hyperbolic fitting (OriginPro).

MelB overexpression and membrane preparation

Overexpression of MelB and membrane preparation were carried out according to previous protocols.¹²

Thermostability test

Thermostability assay of MelB in DDM or GDN was carried out as described^{1, 2, 25, 26} except for the use of 2% detergent concentration. Briefly, membranes containing MelB_St or MelB_Ec at 10 mg/mL in 20 mM Tris-HCl (pH, 7.5), 200 mM NaCl, 10% glycerol, and 20 mM melibiose were incubated with 2% (w/v) of DDM or GDN at 0 ºC for 10 min, and subsequently placed at a given temperature (0, 45, 55, and 65 ºC) for 90 min. Samples were ultracentrifuged at 355,590 g for 45 min at 4 ºC. Equal-volume solutions were analyzed by SDS-PAGE and immunoblotted with Penta-His-HRP antibody. The thermostability data of WT MelB_St²⁷ and MelB_Ec¹ in MNG-3 have been published.

MelB purification in different detergents

Purification of MelB has been reported.¹² Briefly, E. coli DW2 cell membranes at 14 mg/ml were extracted with 1.5% UDM or 1.2% MNG-3, and purified MelB in 20 mM Tris-HCl (pH 7.5), 100 mM NaCl, 0.035% UDM or 0.01% MNG-3, and 10% glycerol was flash-frozen in liquid nitrogen, and stored at −80 °C.

Expression and purification of phosphotransferase IIA^Glc

Expression and purification of IIA^Glc of E. coli were performed as described.^{16, 28} Purified IIA^Glc in 20 mM Tris-HCl (pH 7.5), 100 mM NaCl, and 10% glycerol was stored at −80 °C.

Protein concentration

Protein concentration was determined with the Micro BCA Protein Assay (Pierce Biotechnology, Inc).

Isothermal titration calorimetry

ITC measurements with a Nano Isothermal Titration Calorimeter ( TA Instruments) and data process using the NanoAnalyze version 2.3.6 software^{29, 30} were performed as described.^{16, 28} MelB in 20 mM Tris-HCl buffer (pH 7.5) containing 100 mM NaCl, 10% glycerol, and a given detergent was placed into the sample cell, and melibiose and IIA^Glc were prepared in the same buffer used for the permease. Data fitting using one-site independent binding model³¹ yields the association constant (K_a) and enthalpy change ΔH values. ΔG = - RT ln K_a. The dissociation constant (K_d) = 1/K_a. Entropy change (−TΔS) is calculated from equation ΔG = ΔH - TΔS.

Results

Trp→D²G FRET

D²G is a fluorescent sugar analogue for MelB and LacY.^{6, 17, 18, 20, 32} As reported,^{6, 17} with RSO membrane vesicles containing WT MelB_Ec, Trp→D²G FRET is observed upon addition of D²G (Fig. 1, upper left panel, red curve), which is increased by Na⁺ (green curve) and decreased by consecutive addition of melibiose (blue curve), to a level slightly higher than the background (black curve). The binding affinity for Na⁺ or melibiose is reflected by Na⁺ stimulation and melibiose displacement of bound D²G, respectively. With RSO vesicles containing WT MelB_St, the D²G FRET is also obtained,⁶ but the FRET signal is less intense (Fig. 1, lower left panel). RSO vesicles containing MelB_Ec or MelB_St bind D²G in the presence of Na⁺ with a K_d of approximately 3 or 10 μM, respectively, and bind melibiose with a K_d of approximately 0.5 or 1 mM, respectively.⁶

Fig. 1. Trp→D²G FRET.

RSO vesicles prepared from DW2 cells expressing WT MelB_Ec (upper panel) and WT MelB_St (lower panel) at a protein concentration of 1.0 mg/ml, or solubilized with 1% of an indicated detergent, were excited at 290 nm. Emission spectra were recorded between 430 and 550 nm in the absence of (black or curve 0) or presence of 10 μM D²G (red or curve 1), with successive addition of 20 mM NaCl (green or curve 2) and 120 mM melibiose (blue or curve 3). A.U., arbitrary unit.

When RSO vesicles containing WT MelB_St were solubilized with detergent DDM, UDM or DM without purification of the protein, a specific D²G FRET signal was also detected (Fig. 1). Surprisingly, MelB_Ec solubilized with DDM, UDM or DM does not exhibit Na⁺ stimulation or melibiose reversal of fluorescent intensity, which are not improved by changing the DDM or UDM concentration from 0.5 to 2.0% (data not shown). Strikingly, with detergent MNG-3 or GDN, Na⁺ stimulation and melibiose reversal of D²G FRET are obtained with both MelB_Ec and MelB_St (Fig. 1). Notably, the intensity of D²G FRET is protein- and detergent-dependent, in addition to the fraction of bound D²G. The IC₅₀ for melibiose displacement of bound D²G was determined to analyze galactoside-binding affinity. MelB_Ec in MNG-3 or GDN and MelB_St in each detergent exhibit slightly increased IC₅₀ values than observed with RSO membrane vesicles (Table 1)

Table 1.

IC₅₀ for melibiose displacement of bound D²G^a (mM)

Permease	RSO membrane vesicles	Detergent solubilization
		DDM	UDM	MNG-3	GDN
MelBEc	0.66 ± 0.16b	NDc	ND	1.20 ± 0.35	1.82 ± 0.18
MelBSt	2.42 ± 0.22	5.65 ± 1.75	4.79 ± 0.71	4.22 ± 0.82	3.19 ± 0.41
R295C MelBSt	4.04 ± 0.42d	/e	/	5.49 ± 0.83	/

^aMelibiose concentration for the half-maximal displacement of bound D²G (IC₅₀) in the presence of 20 mM NaCl.

^bSEM, (n = 2).

^cNo detectable signal.

^dData from reference #12;

^eNo measurement.

Effect of detergents on MelB_St mutants

R141, R295, or R363 residue forms multiple interactions between the two domains of MelB_St¹³ for stabilizing the outward conformation.¹² Replacement of R141, R295, or R363 of MelB_St with Cys significantly inhibits melibiose uptake with little effect on melibiose or Na⁺ binding as shown by melibiose transport with intact cells and the D²G FRET assay with RSO vesicles.¹³ When R141 or R363 are replaced with Lys, the conservative replacement mutants catalyze melibiose uptake. With RSO vesicles containing mutants R141K or R363K MelB_St, Na⁺-stimulation and melibiose reversal of fluorescent intensity are detected after solubilization in UDM or MNG-3 (Fig. 2a). Strikingly, there is no change upon addition of Na⁺ or melibiose with RSO vesicles containing MelB_St mutants (R141C, R295C, or R363C) after solubilization with UDM (Fig. 2b, middle column). When using MNG-3 (Fig. 2b, right column), Na⁺-stimulation and melibiose reversal of D²G FRET are observed with all three mutants, although intensity is weaker than that with RSO membrane vesicles (Fig. 2b, left column). With mutant R295C MelB_St in MNG-3, the IC₅₀ for melibiose displacement of bound D²G is 5.49 ± 0.83 mM (Table 1).

Fig. 2. Time trace of Trp→D²G FRET.

With excitation wavelength of 290 nm, emission was recorded at 465 nm for MelB_Ec or 490 nm for MelB_St. 10 μM D²G, 20 mM NaCl, melibiose (Mel, black) or water (red) were successively added at a 60-sec intervals. Left column, recorded with RSO vesicles; middle and right columns, recorded with RSO vesicles after solubilization with 1% UDM or 1% MNG-3, respectively. (a) R141K and R363K MelB_St mutants; (b) R141C, R295C, and R363C MelB_St mutants.

Sugar binding by ITC

To determine the K_d value for melibiose binding with MelB purified after solubilization with UDM or MNG-3, ITC measurements were carried out. As shown previously, titration of WT MelB_St in UDM with melibiose yields a K_d of 0.97 ± 0.02 mM and ΔG of −17.20 ± 0.07 kJ/mol in the presence of Na⁺ (Fig. 3, Table 2).¹⁶ Energetically, melibiose binding is driven by both favorable enthalpy (ΔH of −10.33 ± 0.36 kJ/mol) and entropy (−TΔS of −6.87 ± 0.28 kJ/mol) (Fig. 4b). Remarkably, titration of the WT MelB_Ec in UDM with melibiose (10 mM), even at an increased concentration (100 mM), yields a similar heat change to that observed by injection of melibiose into buffer in the absence of protein (Fig. 3, right column). Similarly, titration of mutant R141C MelB_St with melibiose at 100 mM also exhibits no binding isotherm (Fig. 3). These data clearly indicate that WT MelB_Ec and mutant R141C MelB_St in UDM do not bind melibiose, which correlate well with the D²G FRET results.

Fig. 3. ITC measurement of melibiose binding with MelB in UDM.

MelB was extracted with UDM and purified in buffer containing 0.035% UDM. Thermograms (with y-axis on the left side) were recorded at 25 °C during the titration of WT MelB_St, mutant R141C MelB_St, or WT MelB_Ec (80 μM) with melibiose at 10 mM (black curves) or 100 mM (pink curves). Injection of melibiose to the buffer in the absence of a protein is used for the control. Melibiose binding to the WT MelB_St has been reported.^{12, 16} Accumulated heat change (ΔQ, with y-axis on the right side) of each injection was plotted against the melibiose/MelB molar ratio (on the top), and fitted to a one-site independent binding model.

Table 2.

Detergent effect on energetics of melibiose binding to MelB^a

Permease	Ka (/mol)	Kd (mM)	ΔG (kJ/mol)	ΔH (kJ/mol)	-TΔS (kJ/mol)
MelBSt in UDM	1030 (27.50)	0.97 (0.02)	−17.20 (0.07)	−10.33 (0.36)	−6.87 (0.28)
MelBSt in MNG-3	402.18 (20.60)	2.51 (0.13)	−14.68 (0.13)	−13.65 (0.36)	−1.21 (0.40)
MelBEc in UDM	NDc	ND	ND	ND	ND
MelBEc in MNG-3	782.05 (38.65)	1.28 (0.06)	−16.51 (0.12)	−9.33 (0.04)	−7.18 (0.16)

^aAll measurements were carried out at 25 °C. The data are presented in Figs. 3 and 4.

^b SEM, number of tests = 2 - 4.

^cNo detectable signal.

Fig. 4. Thermodynamics of melibiose binding with MelB in MNG-3.

WT MelB_St or MelB_Ec was solubilized with MNG-3 and purified in buffers containing 0.01% MNG-3. (a) Thermograms (with y-axis on the left side) were recorded at 25 °C during the titration of MelB (95 μM) with melibiose at 30 mM (MelB_St, black curve) or 10 mM (MelB_Ec, black curve). Injection of melibiose at 30 mM or 10 mM to the buffer in the absence of a protein is used for the control (magenta curves). ΔQ (with y-axis on the right side) was plotted against the melibiose/MelB molar ratio (on the top), and fitted to a one-site independent binding model. (b) Energetics of melibiose binding with MelB in UDM or MNG-3. Free energy change (ΔG), enthalpy change (ΔH), and entropy change (−TΔS) are obtained from curve fitting in Figs. 3 and 4 as described in the materials and methods. Errors, SEM, number of tests = 2 - 4.

Both WT MelB_Ec and MelB_St were purified after solubilization with MNG-3. Titration of MelB_St in MNG-3 with melibiose exhibits slightly reduced binding affinity with a K_d of 2.51 ± 0.13 mM (Fig. 4a, Table 2), which is 2.5-fold higher than that in UDM (Table 2). Interestingly, the favorable entropic contribution to the total free energy (ΔG) is largely reduced from 40% in UDM to less than 10% in MNG-3 (Fig. 4b, red bars), yielding a −TΔΔS_{MNG-3 - UDM} of 5.66 kJ/mol. As a partial compensation, the favorable enthalpic change (ΔH) is increased and contributes to ΔG at greater than 90%. Thus, the decrease in melibiose binding affinity with MelB_St in MNG-3 is due solely to a loss of entropy.

While no sugar binding is observed with MelB_Ec in UDM, titration of MelB_Ec in MNG-3 with melibiose reveals a typical binding isotherm with a K_d of 1.28 ± 0.06 mM (Fig. 4a), which is about 2.5-fold higher than that measured with RSO by D²G FRET⁶ but similar to the K_d obtained with proteoliposomes by flow-dialysis assay using [³H]nitrophenyl-α-d-galactopyranoside.³³ Energetically, both enthalpy and entropy contribute favorably to the free energy, which is similar to melibiose binding with MelB_St in UDM (Fig. 4b).

IIA^Glc binding by ITC

Previous ITC measurements show that the phosphotransferase IIA^Glc, a regulatory protein, binds to MelB_St,¹⁶ MelB_Ec,¹⁶ or LacY²⁸ in UDM, yielding K_d of ca. 3, 25, or 5 μM, respectively. When melibiose is pre-incubated with MelB_St, IIA^Glc affinity is 3-fold decreased, and the binding rate is faster.¹⁶ However, melibiose has no effect on IIA^Glc binding to MelB_Ec (Fig. 5, also see ref. ¹⁶), which is now recognized as lack of melibiose binding. When titration of mutant R141C MelB_St in UDM with IIA^Glc, a binding curve similar to WT MelB_St is obtained at a K_d of 2 μM (Fig. 5). Again, melibiose shows no effect, which is consistent with a lack of melibiose binding.

Fig. 5. Thermogram of IIA^Glc binding.

Titration of WT MelB_Ec (30 μM, at 25 °C) or R141C MelB_St (50 μM, at 20 °C) with IIA^Glc (400 or 455 μM) in the absence or presence of 10 mM melibiose was recorded, respectively. Both MelB were purified in buffer containing 0.035% UDM after extracted with UDM. Injection of IIA^Glc to the buffer in the absence of a protein is used as the control. The K_d for IIA^Glc binding to WT MelB_Ec has been reported.¹⁶

Comparison of thermostability between MelB_St and MelB_Ec

Thermostability studies with MelB_St and MelB_Ec in DDM, MNG-3, or GDN have been reported,^{1, 2, 27} but the previous focus was not on direct comparison between the two MelB proteins. Data on DDM or GDN (Fig. 6) were obtained from similar studies but at higher concentration; the MNG-3 data are from two publications.^{1, 27} All the tests were done in the presence of melibiose and NaCl.

Fig. 6. Thermostability.

Equal volume of MelB samples after solubilization with 2% DDM or GDN, as well as 1.5% MNG-3 in the presence of 20 mM melibiose and 200 mM NaCl without (−) or with (+) ultracentrifugation was loaded onto SDS-12% PAGE. MelB proteins were detected by Western blotting using anti–His tag antibody. 10 μg membrane proteins were used for test under each condition. *, SDS-16% PAGE. The data of MNG-3 have been reported^{1, 27}.

DDM quantitatively solubilizes either MelB_St or MelB_Ec at 0 °C (Fig. 6, lanes 1-2). After incubation at elevated temperatures (45, 55, or 65 °C) for 90 min, MelB_Ec exhibits strong aggregations on the western blot (Fig. 6, lanes 3-7). With MelB_St, only slight aggregation is observed at 45 °C; at 55 °C, the soluble fraction of MelB_St is greater than that of MelB_Ec, as shown with the samples after ultracentrifugation (lane 6); at 65 °C, both MelB proteins disappear from the solutions (lane 8).

MNG-3 or GDN maintains either MelB in soluble fraction completely after incubation at 55 °C; GDN also keeps all MelB in solution even after incubation at 65 °C for 90 min (Fig. 6).² The data indicate that either MelB exhibits an increased thermostability in MNG-3 or GDN, by approximately 10 °C or 20 °C, respectively. In addition, as observed in DDM, MelB_St in both MNG-3 and GDN shows less aggregation, clearly indicating that MelB_St is more thermostable than MelB_Ec (Fig. 6).

Discussion

MelB_St and MelB_Ec share more than 85% sequence identity, and the residues necessary for cation and sugar binding are highly conserved.¹⁰ The common detergents DDM and UDM completely extract either permease from the membrane,^{1, 2} and the purified proteins exhibit no aggregation on ice for weeks to months, but the thermostability of WT MelB_St is clearly better than WT MelB_Ec (Fig. 6). Interestingly, DDM and UDM, which work well for MelB_St, do not support galactoside binding by MelB_Ec, or by some MelB_St mutants (Figs. 1-3). Thus, DDM or UDM causes abnormal conformations of these sensitive proteins, but not MelB_St. It is likely that these effects are due to relatively poor stability in these detergents. The DDM and UDM effects are subtle and reversible without denaturation/aggregation (Fig. 6) because the same protein samples bind the regulatory protein IIA^Glc (Fig. 5). In addition, WT MelB_Ec after reconstitution into proteoliposomes binds Na⁺ or Li⁺ and galactosides.^{17, 18, 20, 34} When using MNG-3 or GDN, galactosides binding with WT MelB_Ec and the MelB_St mutants are obtained (Figs. 1, 2, 4; Tables 1 and 2). ITC measurement with WT MelB_Ec in MNG-3 yields a K_d value of 1.2 mM for melibiose in the presence of NaCl.

The effect of DDM and MNG-3 on a LacY mutant with Cys154→Gly has also been studied,³⁵ The C154G LacY, which likely favors an intermediate periplasmic-open conformation in the membrane, collapses to a lower-energy, periplasmic-closed conformation in DDM. Notably, MNG-3 stabilizes C514G LacY in the membrane-embedded form.³⁵ It has been also reported that the muscarinic acetylcholine receptor requires cholesterol or its derivatives for protein stability and that MNG-3 stabilizes the receptor without cholesterol derivatives.³⁶ In addition, the off-rate of MNG-3 from the G protein-coupled beta(2)-adrenoreceptor has been shown to be four orders of magnitude lower than that of DDM,³⁷ clearly demonstrating that MNG-3 binds the membrane protein tightly.

Structurally, both MNG-3 and GDN have a branched dimaltoside hydrophilic headgroup,^{1, 2} which plays an important role in membrane protein stabilization. The two alkyl-chains in MNG-3 could have a multivalent effect, binding to membrane proteins tighter than DDM or UDM. GDN has a flat panel-like structure with lipophilic groups, and thus likely forms self-assemblies with strong intermolecular interactions around membrane proteins. Both MNG-3 and GDN also have critical micelle concentration (CMC) values (0.001%¹ and 0.002%², respectively) lower than DDM (0.008%). While the CMC is only one of many factors determining membrane protein stability, a detergent with a low CMC value has high tendency to form stable micelles, which could contribute to formation of stable protein-detergent complexes. Thus, the unique structural and biophysical features of the novel amphiphiles and resulting tight interactions with membrane proteins could be responsible for the increased stability and for retaining the galactoside binding in MelB. The tight binding of MNG-3 is supported by the thermodynamic characterization using ITC. The results clearly show that the favorable entropy change contributing to melibiose binding with MelB_St in MNG-3 is reduced with a −TΔΔS_{MNG-3 - UDM} of 5.66 kJ/mol. Unfortunately, this parameter for MelB_Ec is not available because sugar binding is not observed with MelB_Ec in UDM. Loss of entropy implies that the conformational dynamics of MelB_St is restricted, in all probability because MNG-3 interacts tightly with the protein.

Interestingly, melibiose binding with MelB_St in MNG-3 exhibits a 2.5-fold increase in K_d value. The decrease in galactoside binding affinity likely results from the restricted conformational dynamics. The galactoside binding with MelB is proposed to use an induced-fit mechanism,^{16, 28} which is similar to the sugar binding in LacY.^{38, 39} Based on this notion, a galactoside induces conformational change of MelB from an open state to an occluded state to form a completely liganded binding site. It has been proposed that IIA^Glc inhibits such a process by restricting conformational entropy of MelB, resulting a largely reduced melibiose binding.¹⁶ It seems that MNG-3 also hinders the induced-fit process in MelB_St. Thus, for an optimal galactoside binding of each MelB, a balance between structural stability and conformational dynamics is needed, which is largely affected by detergent micelles.

Abbreviations and Textual Footnotes

MelB_St

melibiose permease of Salmonella typhimurium

MelB_Ec

melibiose permease of Escherichia coli

E. coli

Escherichia coli

MNG-3

lauryl maltose neopentyl glycol

GDN

glyco-diosgenin

DDM

n-dodecyl-β-d-maltoside

UDM

n-undecyl-β-d-maltoside

DM

n-decyl-β-d-maltoside

D²G

2’-(N-dansyl)aminoalkyl-1-thio-β-d-galactopyranoside

CMC

critical micelle concentration

D²G FRET

Trp→dansyl-galactoside fluorescence resonance energy transfer

ITC

isothermal titration calorimetry

K_d

dissociation constant

RSO

right-side-out membrane vesicles

ΔQ

accumulated heat change

ΔG

free energy change

ΔH

enthalpy change

−TΔS

entropy change