Biochemical and Biophysical Characterization of a Prokaryotic Mg²⁺ Ion Channel: Implications for Cost-effective Purification of Membrane Proteins

Abstract

Although magnesium is the second most abundant cation present in the cell, the transport mechanism of Mg²⁺ across membranes is poorly understood. Importantly, the prokaryotic MgtE Mg²⁺ channel is related to mammalian SLC41A1 transporters and, therefore, biochemical and biophysical characterization of MgtE and its orthologs assumes significance. To date, the purification and structure determination of MgtE from Thermus thermophilus has been carried out using the widely used nonionic detergent, n-dodecyl-β-D-maltopyranoside (DDM). However, DDM is an expensive detergent and alternative methods to produce high-quality proteins in stable and functional form will be practically advantageous to carry out structural studies in a cost-effective manner. In this work, we have utilized ‘dual-detergent strategy’ to successfully purify MgtE channel in a stable and functional form by employing relatively inexpensive detergents (Triton X-100 and Anzergent 3-14) for membrane solubilization and subsequently changed to DDM during purification. Our results show that Triton X-100 and Anzergent 3-14 extract MgtE well and the quality of purified protein is comparable to DDM-extracted MgtE. Interestingly, addition of high concentration of salt and glycerol during solubilization does not significantly affect the quantity and quality of MgtE. Importantly, limited proteolysis assay, circular dichroism spectroscopy and ensemble tryptophan fluorescence strongly support the use of Triton X-100, in particular, as an inexpensive, alternative detergent for the purification of MgtE without compromising the structural integrity of the channel and Mg²⁺-induced gating-related conformational dynamics. Overall, these results are relevant for the cost-effective purification of stable and functional membrane proteins in general, and magnesium channels, in particular.

1. Introduction

Membrane proteins perform many important biological processes such as ion transport, electrical excitability, cell communication, signal transduction and protein secretion, and are associated with diseases like heart disease, cancer, neurodegenerative diseases etc. Importantly, membrane proteins constitute ~30% of proteins produced by genomes of lower and higher organisms [1,2], and ~60% of approved drugs target membrane proteins, of which G-protein coupled receptors and ion channels constitute the largest groups [3-6]. This highlights the importance of understanding the mechanism of membrane proteins’ function, which is very critical for biomedical research and drug discovery. Despite recent successes in determining the highresolution structures of membrane proteins, the study of membrane proteins at atomic level is still quite challenging due to poor expression and extraction, low yield of functional protein, the complexity and heterogeneity of source membranes [7] and the low success rate of forming well-ordered 3D crystals [8].

In particular, extraction of membrane proteins from the native membrane source remains the biggest obstacle in obtaining large quantities of pure, stable and functional proteins that are needed for characterizing the functionally-relevant structural dynamics utilizing sophisticated biophysical approaches. In this context, detergents serve as an indispensable tool for the membrane solubilization and extraction of membrane proteins [9-14]. However, the membrane protein purification process in general, and detergent-mediated membrane solubilization in particular, is very expensive. Any improvement in the extraction of pure, stable and functional membrane proteins in a cost-effective manner is, therefore, expected to facilitate the structure-guided approaches that are in need for the drug discovery pipeline.

In this paper, we have tested the effect of relatively inexpensive detergents such as Triton X-100 and Anzergent 3-14 and a widely used expensive detergent, DDM (n-dodecyl-β-D-maltopyranoside) in successful extraction of the magnesium ion channel MgtE from E. coli membranes with structural and functional integrity. MgtE is a homodimeric Mg²⁺ channel of 450 amino acids with a monomeric molecular mass of ~ 50 KDa having a transmembrane C-terminal domain and a cytosolic N-terminal domain. The putative closed state three-dimensional X-ray structure of MgtE homodimer has been determined at high-resolution in detergent micelles [15,16]. The structure reveals that the cytosolic domain is divided into two sub domains: The N-terminal N-domain and the C-terminal cystathione-β synthase (CBS) domain. The N-domain of MgtE has six Mg²⁺ binding sites possibly making it functions primarily as a Mg²⁺ sensor. Electrostatic repulsion amongst several acidic residues of the N-domain of MgtE keep the channel open in the absence of Mg²⁺, whereas neutralization of these charges helps to close the channel [16,17]. In this regard, Mg²⁺ acts not only as a permeating ion, but also the regulatory one that modulates the gating behavior of MgtE. Despite the availability of a crystallographic snapshot in one functional state, namely the closed state, the gating and permeation mechanisms in membranes are poorly understood partly due to the unavailability of structural information in other functional states.

Our results show that the relatively inexpensive detergents (Triton X-100 and Anzergent 3-14) can faithfully be used to solubilize and extract MgtE in a properly folded and stable conformation. Further, the quantity and quality of MgtE extracted using these alternative detergents are comparable to procedures with DDM. Importantly, our results demonstrate that the secondary and tertiary structural features, along with the gating-related dynamics, are preserved suggesting that these detergents do not disrupt the functional and structural integrity of the protein during extraction.

2. Materials and methods

2.1. Materials

E. coli C41 (DE3) strain were purchased from Agilent (Santa Clara, CA). n-dodecyl-β-D-maltopyranoside (DDM), Triton X-100 and Anzergent 3-14 were obtained from Anatrace (Maumee, OH). Mouse anti-6X His tag antibody and goat anti-mouse IgG H&L (HRP) was from Abcam (Cambridge, MA). Proteases were obtained from GoldBio (St. Louis, MO). Pre-stained molecular weight markers were obtained from Bio-Rad (Richmond, CA). All other chemicals used were of the highest purity available from Merck (Kenilworth, NJ) or Amresco (Radnor, PA).

2.2. Cloning and expression test of MgtE

The gene encoding MgtE from Thermus thermophilus genomic DNA from strain HB8 (ATCC 27634) was PCR-amplified and cloned into a pET28a vector (N-terminal His-tag) between Nde1 and Xho1 restriction sites. The MgtE construct was transformed into chemically competent E. coli C41 (DE3) cells by the heat shock method and grown overnight at 37 °C in Luria-Bertani (LB) broth containing 50 μg/ml of kanamycin. The following day, 1% of the overnight pre-culture was used to start a mini-culture using LB broth containing 50 μg/ml of kanamycin and grown at 37 °C till the cells reached optical density (OD) at 600 nm of 0.5. Protein expression was started by inducing cells with 0.5 mM isopropylthiogalactoside (IPTG) at 0.5 OD. After induction, cells were incubated under constant agitation (250 rpm) at 37 °C for 4 hours, 30 °C for 4 hours or 20 °C for 20 hours. Pre- and post-induction cells (normalized to OD 0.5) were collected, harvested and lysed using Laemmli buffer and electrophoresed in a 10% SDS-PAGE. The gel was electro-transferred to a polyvinylidenedifluoride (PVDF) membrane for 2 hours. Membranes were blocked for 1 hour at room temperature with 3% Bovine serum albumin (BSA) in phosphate buffered saline (PBS), pH 7.4 followed by overnight incubation with mouse anti-histidine antibody (1:5000 dilution) at 4 °C. After three washes with PBS containing 0.02% Tween-20 (PBST), the membranes were incubated with goat anti-mouse antibody (1:10000 dilution) for 2 hours. After three washes with PBST, the MgtE bands were visualized using a GE Healthcare Life Sciences ImageQuant LAS 500 system.

2.3. Detergent-mediated solubilization of MgtE

Small-scale solubilization tests were performed with different detergents as follows: Wild type MgtE was expressed in 100 ml Luria-Bertani (LB) broth containing 50 μg/ml of Kanamycin and was induced at 0.5 OD with 0.5 mM of IPTG. The cells were grown overnight at 20 °C postinduction and then harvested by centrifuging at 5000 rpm and resuspended in 8 ml of 20 mM HEPES buffer. The cells were then sonicated in the presence of protease inhibitors: 1 μg/ml aprotinin, 10 μM leupeptin and 1 μM pepstatin A, and the membrane fraction was separated by ultracentrifugation at 100,000 g. The resuspended membrane fractions were then subjected to detergent solubilization by using 10 mM of DDM, Triton X-100 or Anzergent 3-14, ~50X critical micelle concentration (CMC), for 2 hours at 4 °C in different NaCl concentrations (150 to 1200 mM). The combined effect of the chemical osmolyte (10% glycerol) with high concentration of NaCl (1200 mM) during solubilization was also assessed. The detergent-soluble fraction (supernatant) was then separated by centrifugation at 30,000 g for 45 minutes at 4 °C, and was electrophoresed in a 10% SDS-PAGE followed by Western blotting as described above.

2.4. Large scale expression and purification of MgtE

The wild type MgtE was expressed in 1 L cultures and induced at 37 °C, 30 °C and 20 °C, respectively. The harvested cells were disrupted and the membrane fraction was solubilized using different detergents as described above. The supernatant containing the detergent-solubilized membrane was incubated with Ni²⁺ resin for 15 min. followed by passing it through a column. The resin was washed with 20 mM HEPES, 150 mM NaCl, 1 mM DDM, 50 mM imidazole, pH 7.0 buffer and MgtE was eluted with buffer containing 300 mM imidazole and 1 mM DDM. The eluate was then concentrated in Amicon Ultra 30K filter (Merck Millipore). To analyze whether the channel is properly folded, the purified protein was applied onto a Superdex 75 10/300 column (GE Healthcare) size-exclusion column equilibrated with 20 mM HEPES, 150 mM NaCl, 1 mM DDM, pH 7.0 buffer. The main peak of the gel filtration profiles was collected to measure the concentration of the purified MgtE (normalized to per litre of culture) using the molecular weight (~50 KDa) and molar extinction coefficient (53860 M^-1cm^-1) of wild-type MgtE in a DS-11+ microvolume spectrophotometer (DeNovix).

2.5. Circular dichroism (CD) measurements

CD measurements were carried out at room temperature in a Jasco J-815 spectropolarimeter purged with a nitrogen flow of 15 L/min. MgtE, extracted using different detergents, were measured at a concentration of 2 μM in 20 mM HEPES, 150 mM NaCl, 1 mM DDM, pH 7.0 buffer to obtain a good signal-to-noise ratio. The spectra were scanned with a quartz optical cuvette with a pathlength of 0.1 cm. All spectra were recorded with a bandwidth of 1 nm and integration time of 0.5 s with a scan rate of 50 nm/min. Each spectrum is the average of 10 scans. All spectra were appropriately blank subtracted and smoothed so as to ensure that the overall shape of the spectra remains unaltered. The ellipticity data obtained in millidegree was converted to molar ellipticity ([θ]) by using the following equation:

$$[\theta ]={\theta }_{obs}/(10Cl)$$ (1)

where θ_obs is the observed ellipticity in millidegree, C is the concentration in mol/L and l is the pathlength in cm.

2.6. Limited protease protection assay

The limited protease protection assay with purified wild type MgtE, extracted using DDM, Triton X-100 or Anzergent 3-14, was carried out as follows: 5 μl of purified MgtE (2 mg/ml) was equilibrated with 0 to 32 mM MgCl₂ for 30 minutes, and 2 μl of trypsin (15.6 μg/ml) was added to each reaction mix and incubated at 4 °C for 17 hrs. Wild type MgtE with no added Mg²⁺ and trypsin served as a control. After the addition of 4X laemmli buffer, the samples were run on a 10% SDS-PAGE.

2.7. Steady-state fluorescence measurements

Ensemble tryptophan fluorescence measurements of MgtE, extracted using different detergents, were carried out with a Hitachi F-7000 spectroflourometer using 1 cm pathlength quartz cuvette. MgtE in 20 mM HEPES buffer containing 150 mM NaCl, 1 mM DDM, pH 7.0 represents the open state. The MgtE channel was closed by incubating MgtE in presence of 20 mM MgCl₂ for 30-40 minutes. Excitation and emission slits with a nominal bandpass of 2.5 nm were used to obtain a good signal-to-noise ratio. Background intensities were omitted from each sample to cancel out any contribution due to the solvent Raman peak and other scattering artifacts. Fluorescence polarization measurements were done using a Hitachi polarization accessory. Polarization values were calculated from the equation [18]:

$$P=\frac{I_{VV}-G{I}_{VH}}{I_{VV}+G{I}_{VH}}$$ (2)

where I_VV and I_VH are the measured fluorescence intensities (after appropriate background subtraction) with the excitation polarizer vertically oriented and emission polarizer vertically and horizontally oriented, respectively. G is the grating correction factor and is equal to I_HV/I_HH. All experiments were done with multiple sets of samples and average values of polarization are shown in Table 1. The spectral shifts obtained with different sets of samples were identical in most cases.

Table 1. Fluorescence emission characteristics of MgtE.

Solubilization detergent	Emission maximum (nm)		Fluorescence polarization#
	Open	Closed	Open	Closed
DDM	334	337	0.142 ± 0.008	0.164 ± 0.009
Triton X-100	334	337	0.127 ± 0.006	0.141 ± 0.007
Anzergent 3-14	334	336	0.130 ± 0.001	0.152 ± 0.009

^#Calculated using Eq. 2. The polarization value represents mean ± SE of three independent measurements. The excitation wavelength was 295 nm and the emission was monitored at respective emission maximum. The concentration of MgtE used was 1.6 μM in all cases.

2.8. Acrylamide quenching measurements

Acrylamide quenching experiments were carried out by measurement of fluorescence intensity of MgtE in 20 mM HEPES, 150 mM NaCl, 1 mM DDM pH 7.0 buffer by sequential addition of freshly prepared stock solution (1 M) of acrylamide in water to each sample. After addition of acrylamide the sample was incubated in dark for about 5 minutes before taking the measurement. The excitation wavelength used was 295 nm and emission was monitored at 334 nm. After dilution correction, the inner filter effect correction was made using the following equation [18]:

$$F=F_{obs}\text{antilog}\left[\left(A_{ex}+A_{em}\right)/2\right]$$ (3)

where F is the corrected fluorescence intensity and F_obs is the background subtracted fluorescence intensity of the sample, A_ex and A_em are the measured absorbance at the excitation and emission wavelengths. The absorbance of the samples was measured using a Jasco V-650 ultraviolet (UV)-visible spectrophotometer. Quenching data were analysed by fitting to the Stern-Volmer equation [19]:

$$F_0/F=1+K_{SV}[Q]$$ (4)

where F ₀ and F are the fluorescence intensities in the absence and presence of the quencher, respectively, [Q] is the molar quencher concentration and K_SV is the Stern-Volmer quenching constant.

2.9. Time-resolved fluorescence measurements

Fluorescence lifetimes were calculated from time-resolved fluorescence intensity decays using a picosecond pulsed LED-based TCSPC fluorescence spectrometer and MCP-PMT as a detector. Lamp profiles were measured at the excitation wavelength (295 nm) using Ludox (colloidal silica) as the scatterer. To optimize the signal-to-noise ratio, 10000 photon counts were collected in the peak channel. All experiments were performed with a bandpass of 8 nm. Fluorescence intensity decay curves obtained were deconvoluted with the instrument response function and analysed as a sum of exponential terms:

$$F(t)={\displaystyle {\sum }_i{\alpha }_i\exp \left(-\frac{t}{{\tau }_i}\right)}$$ (5)

where F(t) is the fluorescence intensity at a time t and α_i is a preexponential factor representing the fractional contribution to the time-resolved decay of the component with a lifetime τ_i such that Σ_iα_i = 1. Mean (average) lifetimes <τ> for triexponential decays of fluorescence were calculated from the decay times and pre-exponential factors using the following equation [18]:

$$<\tau >=\frac{{\alpha }_1{\tau }_1^2+{\alpha }_2{\tau }_2^2+{\alpha }_3{\tau }_3^2}{{\alpha }_1{\tau }_1+{\alpha }_2{\tau }_2+{\alpha }_3{\tau }_3}$$ (6)

The mean fluorescence lifetime, tH, can be directly calculated using a model-independent approach from the histogram of photons obtained during fluorescence lifetime measurements using the following equation [20]:

$${\tau }_H=\frac{{\displaystyle {\sum }_{i=p}^n\left(N_i-noise\right)t_i}}{{\displaystyle {\sum }_{i=p}^n\left(N_i-noise\right)}}-t_p$$ (7)

where N_i and t_i denote the number of detected photons in the i-th channel and the corresponding value on the time axis, respectively, n is the total number of channels in the histogram, p is the channel with the highest number of detected photons (peak of the decay) and t_p is the corresponding time.

3. Results

3.1. Temperature-dependent expression of MgtE

Previous excellent studies by Nureki and co-workers [15,16] on MgtE from T. thermophilus have used the post-induction temperature of 20 ºC to express MgtE in C41(DE3) cells. Since the objective of these studies is to obtain the crystal structure of MgtE, it is not very clear whether the post-induction temperature of 20 ºC is really needed for proper purification of the protein or to obtain better crystals. We have systematically checked the expression of MgtE in C41(DE3) cells at different post-induction temperature. Our expression test results show that there is no leaky expression (pre-induction lane) and the expression of MgtE is considerably increased as a function of decreasing temperature and the maximum expression has been observed at post-induction temperature of 20 ºC (Fig. 1a). Large scale purification of MgtE shows that the protein is pure irrespective of the post-induction temperature used as determined by SDS-PAGE (Fig. 1c). In addition, size exclusion chromatography profiles do not contain void volume peak, suggesting that non-specific aggregation of MgtE is not present irrespective of the temperature used for postinduction. However, the gel filtration profiles of the purified MgtE when expressed at 30 and 37 ºC are slightly broader and not particularly homogeneous compared to when expressed at 20 ºC. The purified MgtE elutes at ~10.6 ml with a homogenous peak when expressed at 20 ºC indicating the good quality of the purified protein (Fig. 1b), as shown previously [15,16]. Importantly, significant increase (~3 fold) in the overall yield of protein per litre of culture when expressed at 20 ºC, compared to when expressed at higher post-induction temperature (Fig. 1d). Based on these results, we have used the post-induction temperature of 20 ºC to express MgtE for subsequent studies.

Fig. 1. Expression of MgtE at different temperature.

(a) Western blot of MgtE expressed in C41 E. coli cells before (pre-induction) and after IPTG induction at different post-induction temperature as indicated. MgtE migrates as ~50 KDa band. (b) Size exclusion chromatography (SEC) of purified MgtE, expressed at different temperature, in DDM micelles, and (c) SDS-PAGE of MgtE peak fraction. (d) Yield of purified MgtE obtained per litre of culture grown. See Materials and methods for details.

3.2. Extraction of MgtE from membranes by detergents

The molecular structures of detergents used in this study are shown in Fig. 2. While DDM is a popular, nonionic detergent for both purification and crystallization of many membrane proteins [14], Anzergent 3-14 is a promising zwitterionic detergent that has been used for the crystallization of a voltage-sensing domain [21]. Triton X-100 is probably the oldest classical nonionic detergent that has been widely used in membrane biology [19,22,23] but significantly underrepresented in the use of membrane protein extraction and purification. We have chosen these relatively mild detergents in such a way that the properties such as the CMC, aggregation number and micellar size are comparable [14,24,25].

Fig. 2. Molecular structure of detergents used to extract MgtE.

DDM, n-dodecyl-β-D-maltopyranoside; Triton X-100, α-[4-(1,1,3,3-tetramethylbutyl)phenyl]-ω-hydroxy-poly(oxy-1,2-ethanediyl), where n = 9-10 units of ethylene oxide; Anzergent 3-14, n-tetradecyl-N,N-dimethyl-3-ammonio-1-propanesulfonate. N_agg and CMC represent the aggregation number and critical micelle concentration, respectively.

Our results on solubilization screening of MgtE show that DDM extracts MgtE well and the protein migrates as a ~50 KDa band (Fig. 3a) – expected for a MgtE monomer with 450 amino acids, which is in excellent agreement with the previously published reports [15,16]. Importantly, Triton X-100 (Fig. 3b) and Anzergent 3-14 (Fig. 3c) detergents could faithfully extract MgtE as efficiently as the more expensive DDM when all detergents being used at ~ 50X CMC for solubilization. Considering the fact that these detergents, Triton-X-100 in particular, are inexpensive, our results show a great promise to purify magnesium channels in a cost-effective manner.

Fig. 3. Effect of salt and glycerol on detergent solubilization of MgtE.

Western blot analysis of MgtE extracted using (a) DDM, (b) Triton X-100 and (c) Anzergent 3-14 detergents during solubilization as a function of increasing NaCl concentration, as indicated. (d) Anzergent 3-14 and Triton X-100 extracted MgtE in the absence (lanes 1 and 4) and presence of 10% glycerol (lanes 2 and 5); and a combination of 10% glycerol and 1.2 M NaCl (lanes 3 and 6). Similar results were obtained for DDM as well (not shown). MgtE migrates as ~50 KDa band. See Materials and methods for other details.

It has been shown that high concentration of salts added during the detergent-mediated solubilization step increases the extraction efficiency of membrane proteins [26,27]. Extraction of membrane proteins from their native membranes to membrane-mimetic environments like detergent micelles subjects the protein to chemical and mechanical stress which might compromise the stability of the extracted protein. Polyhydric alcohols like glycerol are, therefore, used as chemical chaperones to stabilize proteins, i.e., protect the protein from unfolding [28,29]. We have used a range of sodium chloride concentrations (150 to 1200 mM) and the chemical chaperone, glycerol, at the time of solubilization, to see their individual or combined effect on the efficiency of detergent-mediated extraction of MgtE. It is interesting to note that there is no appreciable change in the extraction efficiency of MgtE regardless of concentration of NaCl used during membrane solubilization (Fig. 3a-c). Even addition of 10% glycerol in the presence of high concentration of NaCl (1.2 M) does not significantly affect the extraction efficiency (Fig. 3d) irrespective of the detergent used during solubilization. This suggests that detergents, and not the tested additives, play an important role in MgtE solubilization.

3.3. MgtE extracted by alternative detergents is stably folded and a-helical

Although MgtE is extracted well by Triton X-100 and Anzergent 3-14 detergents, it migrates as a monomer (a single band at 50 KDa) in presence of a harsh anionic detergent, sodium dodecylsulfate (SDS), commonly used in SDS-PAGE. Since SDS is known to destabilize oligomers, and MgtE channel functions as a dimer in physiological conditions [16,30,31], we have therefore tested the structural integrity of MgtE using size exclusion chromatography (SEC) and CD spectroscopy. We have adopted ‘dual-detergent strategy’ by employing inexpensive detergents (Triton X-100 and Anzergent 3-14) for membrane solubilization and subsequently changed to DDM during purification. This is particularly important for Triton X-100 because of the presence of the aromatic ring (see Fig. 2) that absorbs strongly in the UV region and interferes with protein quantification and size exclusion chromatography [14]. Further, this gives an opportunity to directly compare the effects of inexpensive detergents in preserving the structural and functional integrity to that of previously used expensive detergent, DDM.

SEC analysis is used to assess the stability of a protein since it separates proteins based on their native hydrodynamic radius without perturbing the tertiary or quaternary structure of the protein. We observe that there is no protein in the void volume (~6-8 ml of elution) suggesting that there is no non-specific aggregation of MgtE (Fig. 4a). It has previously been shown that MgtE in DDM buffer migrates as a dimer in SEC analysis [15,16]. Our results show that MgtE, which is extracted by Triton X-100 and Anzergent 3-14, displays predominantly a single, homogenous peak that is consistent with DDM-extracted MgtE gel filtration profile, suggesting that even the cheapest detergent (Triton X-100) preserved the dimeric conformation of the channel (Fig. 4a).

Fig. 4. Stability of MgtE extracted using various detergents.

(a) SEC of MgtE extracted with either DDM, Triton X-100 or Anzergent 3-14 during solubilization are shown. (b) Representative far-UV CD spectra of MgtE extracted using various detergents during solubilisation. CD spectra were recorded using 2 μM MgtE in DDM micelles. See Materials and methods and Eq. 1 for other details.

It is well established that the surface charge of micelles play an important role in stabilizing the secondary structure of membrane interacting proteins [32]. To investigate whether MgtE extracted with nonionic Triton X-100 and zwitterionic Anzergent 3-14 has similar secondary structure, we carried out far-UV CD spectroscopy in the wavelength range of 200 to 250 nm. The CD spectra of MgtE shows a characteristic a-helical conformation indicating that the secondary structure of MgtE, which is extracted by Triton X-100 and Anzergent 3-14, is well preserved and similar to that of DDM-extracted MgtE (Fig. 3b). This suggests that Triton X-100 and Anzergent 3-14 detergents are not only extracting MgtE well from membranes, but also extract in a proper, stable form during solubilization.

3.4. Mg²⁺-induced protease protection of MgtE

Limited proteolysis assay is a commonly used sensitive technique to monitor protein dynamics to determine conformational changes and structural reorganization induced by ligands [33,34]. Trypsin is a serine protease that recognizes and hydrolyzes at exposed basic amino acids (arginine and lysine) on a protein. When the ligand-induced conformational changes in a protein involve shielding these basic amino acids, the protein gets protection from the action of trypsin, which is directly indicative of the ligand-mediated function of protein [35]. The N-domain of MgtE has many such basic amino acids which might be exposed when it shuttles from closed to open state. Therefore, in the open state of MgtE, i.e., in the absence of Mg²⁺, the N-domain gets chopped off into several fragments generating a truncated MgtE (ΔN). Since Mg²⁺ is not only a permeating ion, but also a gating regulator [16,17] and N-domain is proposed to be the Mg²⁺ sensor due to the presence of several Mg²⁺-binding sites [16], MgtE undergoes a transition from open to closed state in the presence of Mg²⁺. This results in an increased protection of MgtE from protease as a function of increasing Mg²⁺ concentration as shown previously for DDM-extracted wild type MgtE [17,36]. Figure 5 shows that MgtE is susceptible to proteolysis in the absence of Mg²⁺ (open state) as expected in case of DDM, Triton X-100, Anzergernt 3-14-extracted MgtE. Upon increasing the Mg²⁺ concentration (0-32 mM), MgtE gets more pronounced protection from trypsin, and interestingly, with the appearance of the full-length MgtE at more than 8 mM Mg²⁺. It should be noted that the appearance of full-length MgtE in trypsin protection assay has been observed in the literature only in the presence of ATP [36], whereas the protection of full-length MgtE is apparent at concentrations of 16 mM Mg²⁺ or more even in the absence of ATP (Fig. 5). Considering that MgtE fully closes at a concentration of 10 mM Mg²⁺ and above [16], our results capture the ligand-induced conformational changes associated with gating. This is true whether DDM, Triton X-100 or Anzergent 3-14 is used to extract MgtE during solubilization (Fig. 5), demonstrating that the alternative, inexpensive detergents like Triton X-100 and Anzergent 3-14 extracts MgtE probably in a stable and functional form.

Fig. 5. Protease protection assay of MgtE in DDM micelles.

SDS-PAGE shows trypsin-treated MgtE (10 μg) extracted by (a) DDM, (b) Triton X-100 and (c) Anzergent 3-14, during solubilisation, in the absence and presence of trypsin (16 μg/ml) with increasing concentration of Mg²⁺ as indicated. The ΔN (~37 KDa band) represents MgtE lacking N-domain. Control samples do not contain trypsin and MgCl₂. See Materials and methods and text for details.

3.5. Organization and dynamics of MgtE

Aromatic residues like tryptophan are known to localize at the protein-membrane interface in a membrane protein [37-39]. They form a so-called “aromatic belt” around the protein and is known to stabilize the protein at the membrane interface. Hence, the orientations of these residues can act as a reporter of the conformational homogeneity of a membrane protein. We used steady-state fluorescence approaches to monitor the organization and dynamics of MgtE during gating by utilizing the intrinsic tryptophan fluorescence. MgtE has six tryptophan residues per monomer of which four are at the membrane interface and the remaining two are in the N-domain. As mentioned earlier, MgtE can be stabilized in open and closed states in the absence and presence of Mg²⁺, respectively.

The fluorescence emission maximum¹ and polarization values of open and closed states of MgtE that has been extracted with different detergents are shown in Table 1. The ‘average’ tryptophan emission maximum of MgtE in open state is 334 nm irrespective of the detergents used for membrane solubilization and MgtE extraction. Interestingly, the emission maximum of tryptophan is red shifted by ~3 nm upon closing the channel in the presence of 20 mM Mg²⁺. The observed tryptophan fluorescence emission maximum (334 to 337 nm) is indicative of the interfacial localization of tryptophan residues as shown previously for ion channel-forming peptides in membrane-mimetic systems [32,40-42]. Considering four of the six tryptophan residues of MgtE per monomer is expected to be localized at the membrane interface based on the crystal structure of closed state MgtE in detergent micelles [16], our results indicate that the positions of tryptophans at the membrane interface might not change considerable during gating. Apart from red shifted emission maximum, the closed state is also accompanied by ~15% increase in fluorescence intensity (Fig. 6). This suggests that the polarity of the microenvironment around the tryptophan residues is altered, which could be due to ligand-induced conformational changes (see Fig. 5). Further, we have carried out fluorescence polarization measurements, which has recently been shown to be a powerful approach for monitoring the dynamic behavior of ion channel gating [43]. The polarization values of MgtE, shown in Table 1, generally indicate that the average rotational mobility of tryptophan is representative of motionally restricted environments. Interestingly, the polarization values significantly increase in the closed state reflecting a reduced conformational flexibility of tryptophan residues when the channel transitions from open to closed state (Table 1). In other words, this indicates that, on average, the rotational dynamics of tryptophan residues gets more restricted during gating. We therefore propose that the higher dynamics of tryptophan residues is a characteristic feature of the open state of the MgtE channel. Importantly, all these gating-related dynamic changes triggered by Mg²⁺in tryptophan fluorescence are consistent among MgtE extracted using different detergents (Fig. 6 and Table 1).

Fig. 6. Steady-state tryptophan fluorescence of MgtE.

Intrinsic tryptophan fluorescence emission spectra of MgtE extracted using DDM, Triton X-100 and Anzergent 3-14 detergents during solubilization. The fluorescence spectrum corresponding to open (no Mg²⁺) and closed (20 mM Mg²⁺) states of MgtE is shown in black and red, respectively. The excitation wavelength used was 295 nm and emission was monitored at 334 nm. Purified MgtE (1.6 μM) in 20 mM HEPES, 150 mM NaCl and 1 mM DDM pH 7.0 buffer was used in all cases. See Materials and methods for other details.

In general, fluorescence lifetime of tryptophan is well known to be sensitive to solvent, temperature and excited state interactions [44,45]. The fluorescence lifetimes of MgtE in DDM micelles are shown in Table 2. As seen from the table, all fluorescence decays were fitted with a triexponential function and the mean fluorescence lifetimes were calculated using Eq. 6 and are shown in Table 2. The high values of lifetimes indicate that the tryptophan residues are not completely exposed to aqueous environments, a phenomenon which is fairly common for most membrane proteins because they preferentially localize at the membrane-mimetic interface. Apart from conventional calculation of mean fluorescence lifetimes, we have also calculated the tryptophan fluorescence lifetimes (τ_H), obtained from the histogram of photons counted during the measurement using Eq. 7 (Table 2), using a recently developed model-independent approach [20]. Obviously, model-dependent and model-independent analyses of lifetimes yield slightly different values of lifetimes. However, tryptophan lifetimes obtained using both these approaches for MgtE extracted using Triton X-100 and Anzergent 3-14 detergents are very much comparable to the values obtained for MgtE extracted using DDM. Since fluorescence lifetime serves as a faithful indicator of the local environment in which given fluorophore is placed [45,46], the remarkable similarity of tryptophan fluorescence lifetimes of MgtE extracted using different detergents clearly suggest that the structural properties (tertiary structure in this case) of MgtE is preserved even upon extraction using inexpensive detergents. This is in excellent agreement with our previously discussed CD results (see Fig. 4b).

Table 2. Fluorescence lifetimes of MgtE in open state.

Solubilization detergent	α1	τ1 (ns)	α2	τ2 (ns)	α3	τ3 (ns)	<τ>$ (ns)	τH % (ns)
DDM	0.28	2.26	0.63	6.24	0.09	0.67	5.62	4.93
Triton X-100	0.26	2.69	0.61	6.37	0.13	0.91	5.68	4.99
Anzergent3-14	0.32	3.21	0.53	6.60	0.15	0.91	5.69	4.99

The concentration of MgtE was 1.6 μM. The excitation wavelength was 295 nm and the emission was monitored at 334 nm. See Materials and methods for other details.

^$Calculated using Eq. 6

^%Calculated using Eq. 7.

3.6. Acrylamide quenching of MgtE tryptophan fluorescence

The above results show that the tryptophan residues in the transmembrane region of MgtE reside at the membrane-mimetic interface. To examine the accessibility and location of MgtE in DDM micelles, fluorescence quenching experiments were performed with acrylamide, which is a widely used neutral aqueous quencher of tryptophan fluorescence [47]. Figure 7a shows representative Stern-Volmer plots of acrylamide quenching of tryptophans of MgtE extracted using different detergents. The slope of such a plot (K_SV) is related to the degree of exposure (accessibility) of MgtE tryptophans to aqueous phase. In general, the higher the slope, the greater the degree of exposure, assuming that there is not a large difference in fluorescence lifetime. In fact, the bimolecular quenching constant (k_q) is a more accurate measure of the degree of exposure since it takes into account the differences in fluorescence lifetime. Since the tryptophan lifetimes of DDM, Triton X-100 and Anzergent 3-14-extracted MgtE are near identical in our case (see Table 2), the quenching results and interpretation due to K_SV values do not suffer from lifetime-induced artifacts. The calculation of k_q is therefore not considered in our case.

Fig. 7. Acrylamide quenching of MgtE tryptophan fluorescence.

(a) Representative data for Stern-Volmer analysis of acrylamide quenching of MgtE tryptophan fluorescence in DDM micelles. F ₀ is the fluorescence in the absence of quencher, F is the corrected fluorescence in the presence of quencher. The excitation wavelength was 295 nm and emission was monitored at 334 nm. The concentration of MgtE used was 1.6 μM. (b) Degree of tryptophan fluorescence quenching by acrylamide (at 130 mM) is shown. See Materials and methods and text for other details.

It is well documented that the K_SV value for a completely exposed tryptophan is ~18 M^-1, whereas the corresponding value for tryptophan localized at the micellar interfacial region is in the range of ~ 4-7 M^-1 depending on the surface charge of micelles [32]. We have obtained the K_SV values of 4.5 M^-1, 4.9 M^-1 and 5.4 M^-1 for MgtE extracted using DDM, Triton X-100 and Anzergent 3-14, respectively. These values are well within the range observed for tryptophan localized at the membrane interface. This conclusively shows that the MgtE tryptophans, especially the ones present in transmembrane region, resides at the micellar interface as inferred (see Fig. 6 and Table 1). In addition, the magnitude of K_SV values also suggests that two of the cytoplasmic tryptophans might also be shielded from the aqueous phase in the open state of MgtE. This view is supported by ~35-40% overall quenching observed in the presence of 130 mM of acrylamide (Fig. 7b). Taken together, these results suggest that the solvent accessibility of the MgtE tryptophans are similar in all cases and reflect that the conformational homogeneity of MgtE remained constant regardless of the detergents used to extract. Overall, the inexpensive detergents, particularly Triton X-100, extracts MgtE in a stable and functional form without compromising the structural integrity.

4. Discussion

Membrane proteins are potential drug targets. However, they represent less than 2% of the reported crystal structures in the Protein Data Bank (PDB) because the structural characterization of membrane proteins using sophisticated biophysical approaches is remarkably challenging and quite expensive. The structural studies of any membrane protein require the protein to be extracted from its native membrane into a membrane-mimetic stable environment [13]. The membrane solubilization and subsequent membrane protein extraction is achieved by using detergents, which are soluble amphiphiles and above a critical concentration (strictly speaking, a narrow concentration range), known as critical micelle concentration (CMC), self-associate to form thermodynamically stable, noncovalent organized molecular aggregates called micelles [9,10,19,48]. The solubilization step of membrane protein purification typically requires the detergent concentration to be above the CMC. As a result, a higher concentration of specialized nonionic detergents (~5 to more than 100X CMC) is needed to extract the membrane protein from the source membrane to create water-soluble protein-detergent complexes [10,12], compared to concentration of detergents needed in the downstream steps of purification.

Although detergent solubilization of membrane proteins is arguably the ‘rate-limiting’ step for a successful membrane protein purification, the detailed mechanisms of the action of detergents on membrane components have not been fully understood [10,12]. In addition, despite the commercial availability of large number of detergents, there is no ‘universal detergent’ that can be used to extract the different classes of membrane proteins. This is obvious from the fact that the extraction efficiency of expressed membrane proteins can vary dramatically with different detergents [13], and clearly highlights the importance of detergent screening to identify the best detergent for determining the extraction efficiency and stability of a target protein. The membrane protein purification process is therefore very expensive and the bulk of the purification cost lies in the choice of detergents suitable for extraction, purification and crystallization.

The most commonly used detergents for membrane protein purification and crystallization are alkyl maltosides (n-dodecyl-β-D-maltopyranoside, DDM and n-decyl-β-D-maltopyranoside, DM) and glycosides (n-octyl-β-D-glucopyranoside, OG and n-nonyl-β-D-glucopyranoside, NG) [8,14]. These nonionic detergents are very expensive and alternative, inexpensive detergents are therefore required to reduce the purification cost, yet extracting stable and functional protein, to facilitate the structural studies of membrane proteins. Triton X-100 is one of the oldest, classical nonionic (uncharged) detergent containing polyoxyethylene chains as hydrophilic moieties characterized by a low CMC of ~ 0.3 mM [19,48]. It has been extensively used to explore the detergent resistant (insolubility) membranes and functional membrane domains in cell biology [22,23,49]. In general, the extraction efficiency of a target membrane protein by a detergent is directly related to the hydrophobicity of the micellar core, which is significantly higher for Triton X-100 compared to DDM, DM and OG [50]. Although it has been used to solubilize several membrane proteins in the functional state before the explosion of membrane protein structural biology [51], use of Triton X-100 in membrane protein purification is significantly underrepresented (only ~ 2% compared to ~ 40% in case of DDM) for a few practical reasons [14]. These include its chemical heterogeneity and the interference with the protein quantitation due to its strong absorption in the UV region of the electromagnetic spectrum [12,14,50].

Interestingly, Triton X-100-mediated extraction of stable and functional membrane proteins has recently made significant strides [27,52]. It has been shown that, for a well-characterized potassium channel KcsA [43,53-55], inexpensive detergents like Triton X-100 can be used to extract large amounts of stable and functional protein in the presence of high concentration of salt and glycerol [27]. In addition, Triton X-100 is also shown to increase the quality and quantity of the pentameric ligand-gated ion channels (pLGIC) of prokaryotic origin [52]. In this work, we have tested the potential applicability of such inexpensive detergents (Triton X-100 and Anzergent 3-14) along with additives such as salt and glycerol in the extraction of a Mg²⁺-selective ion channel MgtE. Further, we have utilized ‘dual-detergent strategy’, CD spectroscopy, steady-state and time-resolved tryptophan fluorescence to demonstrate the feasibility of these detergents to be used to produce large amounts of stable and functional protein for structural characterization in a cost-effective manner.

Our current knowledge on molecular basis of Mg²⁺ transport comes from the functional and structural studies on prokaryotic CorA and MgtE Mg²⁺ channels [30,31]. It is now known that MgtE, like CorA, is the primary Mg²⁺ transport system in ~50% of all prokaryotic species. In addition, unlike K⁺ and Na⁺ channels, the architecture of the Mg²⁺ transporters is not conserved as is evident by the recent crystal structures of prokaryotic Mg²⁺ channels. For instance, the functional unit of CorA Mg²⁺ channel is a pentamer [56-58] whereas MgtE is a dimer and only the structural snapshot of the closed state of MgtE in detergent micelles is available [15,16]. It is intriguing that these two evolutionarily distinct Mg²⁺ transporters seem to have conceptually similar Mg²⁺-dependent gating mechanisms with different architectures, which makes them unique. Since the molecular mechanism of Mg²⁺ gating and transport across membranes remains obscure due to paucity of structural information, our study related to biochemical and biophysical characterization of MgtE assumes significance.

Our results demonstrate that the post-induction temperature is crucial for a significant expression of MgtE. Importantly, we show that inexpensive detergents, particularly Triton X-100, could potentially be used in the detergent solubilization step to extract MgtE in a stable and functional form as has been shown for Triton X-100-mediated extraction of K⁺ and ligand-gated ion channels [27,52]. However, in our case, the extraction efficiency is not affected by the presence of high concentration of salt and glycerol and this is true irrespective of the detergents used during solubilization. This is in agreement with a recent study in which it has been shown that the effect of salt on Triton X-100-mediated extraction of pLGIC is negligible [52]. This is not surprising since it is well-known that the optimal extraction conditions for one membrane protein may not be applicable to other members of the same class or different classes of membrane proteins [13,50]. Taken together, our results strongly support the use of Triton X-100 as an inexpensive, alternative detergent for the purification of MgtE without compromising the structural integrity of the channel and Mg²⁺-induced gating-related conformational dynamics. Considering that the cost of Triton X-100 is ~200 fold cheaper than the widely used expensive DDM, it will be worthwhile to include Triton X-100 for routine detergent screening for membrane protein extraction. Overall, our results will have potential implications for the cost-effective purification of membrane proteins in general, and MgtE family of magnesium channels, in particular.

Abbreviations

BSA

bovine serum albumin

CBS

cystathione-β-synthase

CD

circular dichroism

CMC

critical micelle concentration

DDM

n-dodecyl-β-D-maltopyranoside

DM

n-decyl-β-D-maltopyranoside

HEPES

hydroxyethylpiperazine ethane sulfonic acid

IPTG

isopropylthiogalactoside

NG

n-nonyl-β-D-glucopyranoside

OG

n-octyl-β-D-glucopyranoside

PBS

phosphate buffered saline

PVDF

polyvinylidenedifluoride

SDS-PAGE

sodium dodecyl sulfate-polyacrylamide gel electrophoresis

SEC

size exclusion chromatography

TCSPC

time-correlated single photon counting