Thermostability assays provide a generic and versatile tool for studying the functional and structural properties of membrane proteins in detergents
Steven P.D. Harborne1, Martin S. King2 and Edmund R.S. Kunji2
1School of Biomedical Sciences and Astbury Centre for Structural Molecular Biology, University of Leeds, Leeds, LS2 9JT, UK
2Medical Research Council, Mitochondrial Biology Unit, University of Cambridge, Wellcome Trust / MRCKeith Peters Building, Cambridge Biomedical Campus, Hills Road, Cambridge CB2 0XY, UK
Author to whom correspondence should be addressed: E.R.S. Kunji
Tel.: + 44 1223 252850, Fax.: + 44 1223 252875, e-mail: ek@mrc-mbu.cam.ac.uk
†This research was supported by the Medical Research Council (grant MC_UU_00015/1).
Running Head
The versatile thermostability assay
Abbreviations:
AAC: mitochondrial ADP/ATP carrier; APC: mitochondrial ATP-Mg/Pi carrier; ATR: atractyloside; BKA: bongkrekic acid; CATR: carboxyatractyloside; CPM: -[4-(7-diethylamino-4-methyl-3-coumarinyl)phenyl]; DM: decyl maltoside; DDM: dodecyl maltoside; DMNG: decyl maltose neopentyl glycol; LMNG: lauryl maltose neopentyl glycol; PC: phosphocholine; Tm: apparent melting temperature; TOCL/CL: tetraoleoyl cardiolipin (18:1); UDM: undecyl maltoside
Abstract
There are very few generic methods to assess the stability and functional properties of membrane proteins solubilized in detergent. For this purpose, a thiol-reactive fluorochrome N-[4-(7-diethylamino-4-methyl-3-coumarinyl)phenyl]maleimide (CPM) can be used. An unfolding profile is obtained when the fluorochrome becomes fluorescent on reaction with cysteine residues that have been exposed during thermal denaturation of the protein population. The method was initially developed to optimize the stability of membrane proteins for crystallization studies, but in the course of our work we found many other applications. First, the assay can be used to study the binding of inhibitors, substrates, lipids and other effectors to membrane proteins. Second, the assay can be used to understand the dynamics of proteins, allowing states to be defined by changes in accessibility of cysteine residues or by changes in specific amino acid interactions. Finally, the assay can be used to study state-dependent domain interactions, for example, as part of regulatory mechanisms. The CPM thermostability assay represents a broadly applicable and versatile tool for a wide range of applications in the functional and structural analysis of membrane proteins.
Introduction
A pre-requisite for structural and functional studies of purified membrane proteins is stable, homogeneous protein that has retained its native properties in detergent solution (1). However, many membrane proteins, particularly -helical membrane proteins, are inherently unstable in detergent solution. For this reason, the detergent, lipid and ligand additions during protein preparation must be optimized to ensure that the protein is preserved in a biologically-relevant state. Several different methods have been described in the literature to study the structural integrity of membrane proteins. Among them are generic methods, such as analytical size exclusion chromatography (2), but also assays that exploit the functional properties of membrane proteins, such as radioactive ligand binding coupled to temperature challenges (3). As an alternative, Alexandrov et al. (4) introduced a fluorescence-based procedure, which uses a thiol-reactive dye, called CPM, which stands for N-[4-(7-diethylamino-4-methyl-3-coumarinyl)phenyl] (Figure 1A). In the assay, the temperature of the protein sample is increased gradually from 25 to 90 °C, which thermally denatures the protein. CPM reacts with protein thiols to form a blue fluorescent adduct when they become solvent-accessible, allowing the proportion of the protein population that has unfolded to be monitored. The inflection point of the unfolding curve is taken as the apparent Tm and is used as a relative indicator of stability (Figure 1B). Furthermore, it is often useful to compare the difference in relative stability between two conditions (thermostability shift), for example, with and without the addition of a ligand or wild-type and mutant protein. Since cysteine residues are relatively hydrophobic, folded proteins often contain buried cysteine residues. In the absence of native cysteine residues, they can be added by site-directed mutagenesis. We have made this procedure suitable for high-throughput and previously reportedhave shown that unfolding can be monitored reliably with a few micrograms of protein (5,6).
The mitochondrial ADP/ATP and ATP-Mg/Pi carriers as subjects of study
To illustrate some applications of thermostability assays we will use the mitochondrial ADP/ATP (AAC) and ATP-Mg/Pi (APC) carriers as examples. These transport proteins of the belong to the mitochondrial carrier family (SLC25) (7) and catalyze the exchange oftransport adenosine nucleotides across the mitochondrial inner membrane ADDIN EN.CITE ADDIN EN.CITE.DATA ( HYPERLINK \l "_ENREF_8" \o "Harborne, 2018 #10213" 8-10)by a mechanism that is not fully resolved (8-13). The carriers are monomeric (13-19) and have a three-fold pseudo-symmetric structure (14), reflecting their three homologous sequence repeats (20). The structure consists of three similar domains, each comprising two transmembrane α-helices, connected by a short α-helix on the matrix side (21). When inhibited by carboxyatractyloside (CATR), which locks the carrierAAC in an abortive cytoplasmic state (22) ADDIN EN.CITE ADDIN EN.CITE.DATA ( HYPERLINK \l "_ENREF_8" \o "Kunji, 2016 #9464" 8, HYPERLINK \l "_ENREF_22" \o "Klingenberg, 2008 #734" 22), the six transmembrane α-helices form a barrel around a central water-filled cavity, which is accessible from the mitochondrial intermembrane space and cytosol (9,21,23). When inhibited by bongkrekic acid (22), AAC is locked in a state in which the substrate binding site is open to the mitochondrial matrix (9,10). APC is largely homologous to AAC, but has two additional domains; an N-terminal regulatory domain containing two pairs of EF-hands and a linker loop domain with an amphipathic α-helix, both of which are involved in the calcium regulatory mechanism of APC (11,24).
The carriers have been proposed to function according to a “single binding centre gated pore mechanism”, which is an alternating access mechanism (23,25), which is supported by structural analysis (9,10) . They have a single central substrate binding site (26-28) and two salt bridge networks that form part of the gates at the cytoplasmic and matrix side of the carriers, which regulate access to the binding site in an alternating manner. The matrix salt bridge network is formed by the charged residues of the three-fold pseudo-symmetrical PX[DE]XX[KR] motif on transmembrane -helices H1, H3 and H5, forming ionic interactions when the carrier is in the cytoplasmic state (21,29), which can be braced by glutamine residues (23). The cytoplasmic salt bridge network is formed by the charged residues of the three-fold pseudo-symmetrical [FY][DE]XX[RK] motif on transmembrane -helix H2, H4 and H6 (28), which form ionic interactions in the transport cycle (9,10,23,25,30), which are braced by tyrosine residues (9,10).
Applications of the thermostability assays
We have used thermostability analyses to provide experimental evidence for different aspects of the transport and regulatory mechanism of mitochondrial carriers. For example, (i) to guide structural work (10), (ii) to identify detergents and lipids that stabilize or destabilize mitochondrial carriers (5,31,32), (iii) to characterize the interactions of inhibitors and cardiolipin (33) (iiiiv) to analyze and extended the known substrates of carriers and other transport proteins (6) ADDIN EN.CITE Majd201810091(29)100911009117Majd, H.King, M. S.Palmer, S. M.Smith, A. C.Elbourne, L. D. H.Paulsen, I. T.Sharples, D.Henderson, P. J. F.Kunji, E. R. S.Screening of candidate substrates and coupling ions of transporters by thermostability shift assaysBioRXiv 367805 BioRXiv 3678052018 https://doi.org/10.1101/367805 ( HYPERLINK \l "_ENREF_29" \o "Majd, 2018 #10091" 29), (iv) to study the state-dependent accessibility of residues directly (25), (vi) to provide evidence that specific residues are interacting in a state-dependent way (25) and (vii) to study domain-domain interactions as part of a regulatory mechanism of APC (24). The main focuses of our studies have been the ADP/ATP carriers Aac2p and Aac3p from Saccharomyces cerevisiae (5,31,32), AAC from Thermothelomyces thermophila (10,34), which has two cysteine residues; C65 in matrix helix h12, and C229 in transmembrane helix H5 (Figure 2A), and APC from Homo sapiens (SCaMC1/SLC25A24), which has four cysteine residues; C15 at the N-terminus of the regulatory domain (see Note 1), C330 in matrix helix h34, and C391 and C398 in transmembrane helix H5 (Figure 2B) (12).
In a number of the above examples the thermostability assay provided unique insights into protein function. For example, our recent findings show that thermostability shifts occur in the presence of substrates but not in the presence of closely related compounds (6) ADDIN EN.CITE Majd201810091(29)100911009117Majd, H.King, M. S.Palmer, S. M.Smith, A. C.Elbourne, L. D. H.Paulsen, I. T.Sharples, D.Henderson, P. J. F.Kunji, E. R. S.Screening of candidate substrates and coupling ions of transporters by thermostability shift assaysBioRXiv 367805 BioRXiv 3678052018 https://doi.org/10.1101/367805 ( HYPERLINK \l "_ENREF_29" \o "Majd, 2018 #10091" 29). This result was not intuitive as transported substrates, in contrast to inhibitors, bind only transiently and relatively weakly, leading to a conformational change that in turn causes the release of the substrate on the other side of the membrane. This principle allowed us to develop a high-throughput screening method for the identification of transport proteins, where potential substrates were identified from a large set of chemically related compounds, including stereo-isoforms (6) ADDIN EN.CITE Majd201810091(29)100911009117Majd, H.King, M. S.Palmer, S. M.Smith, A. C.Elbourne, L. D. H.Paulsen, I. T.Sharples, D.Henderson, P. J. F.Kunji, E. R. S.Screening of candidate substrates and coupling ions of transporters by thermostability shift assaysBioRXiv 367805 BioRXiv 3678052018 https://doi.org/10.1101/367805 ( HYPERLINK \l "_ENREF_29" \o "Majd, 2018 #10091" 29).
We have also used the thermostability assay to study state-dependent accessibility of residues (25) and we found that residue C65 on matrix helix h12 was occluded when AAC was locked in the cytoplasmic state by CATR, whereas it became accessible to the water phase when AAC was locked in the matrix state by another state-dependent inhibitor, bongkrekic acid (BKA). This effect manifests itself by a raised baseline of fluorescence at the start of the thermostability experiment, as CPM reacts directly with C65 without a requirement for unfolding. In principle, this approach could be used to probe water-accessible residues of proteins in a state-dependent manner. We were also able to provide evidence that specific residues interact in a state-dependent manner. The central question was whether charged residues of the cytoplasmic network interact when AAC is locked in the BKA-bound abortive matrix state (25). The rationale was that mutation of residues involved should eliminate these interactions, leading to a decrease in thermostability. Indeed, there was a clear correlation between the interaction energy of the cytoplasmic network and the thermostability of the carrier when modulated by mutations, but only in the BKA-bound state and not in the CATR-bound state (25). Thus, it is possible to use thermostability assays to demonstrate the formation of state-dependent amino acid interactions. The conclusions of these studies have been confirmed by the structure of the matrix state (10).
Finally, the thermostability assay allowed us to study domain-domain interactions that are part of a calcium regulatory mechanism in APC (24). We showed that removal of the regulatory domain by truncation led to a much more stable carrier domain, indicating that the amphipathic α-helix bound strongly to the carrier domain, locking it in a stable conformation. By mutating individual residues of the amphipathic α-helix, it could be shown that the N-terminal part of the helix was bound to the regulatory domain in the calcium-bound state, whereas the C-terminal part bound to the carrier domain in the absence of calcium. These helix interactions are key to our proposed locking-pin mechanism of calcium regulation of APC (24).
In this paper, we will demonstrate the basic protocol for the thermostability assay, highlighted by three specific examples. First, we show the use of thermostability assays to evaluate the effects of detergent, lipids and inhibitors on the stability of AAC in detergent solution. Second, we study the effect of buffer compostion on the stability of APC, and finally, we show how these assays can be used to guide crystallization trials.
A note on the equipment used for the thermostability assays with CPM
The requirements for thermostability assays with CPM are the ability to perform a rapid and tightly controlled temperature ramp, to excite the CPM fluorophore at or close to its optimum of 387 nm and to read the fluorescence intensity at or close to the emission optimum of 463 nm. Therefore, most commercially available qPCR machines are suitable for these measurements. However, the Rotor-Gene Q 2plex HRM cycler (Qiagen) is our instrument of choice, as it has a rotor to run either 36 or 72 samples (5). The rotary function of the machine ensures a uniform temperature across all samples and counteracts problematic foaming associated with membrane protein solubilized in detergent. We have previously reported that the excitation and emission wavelengths of 440–480 nm and 505–515 nm are not ideal, but the measurements are so sensitive that unfolding can nonetheless be monitored accurately (5). Furthermore, the accompanying software, designed for double-stranded DNA melt analysis, provides a rapid and convenient tool to estimate apparent melting temperatures of protein populations from the derivative of the unfolding curve.
Materials
Strains and plasmids
1. S. cerevisiae strain WB.12 (MATα ade2-1 trp1-1 ura3-1 can1-100 aac1::LEU2 aac2::HIS3), which lacks functional Aac1p and Aac2p carriers (35), is used for the expression of Thermothelomyces thermophila AAC in a functional form (see Note 2).
2. S. cerevisiae strain W303.1B (MATα leu2-3,112 trp1-1 can1-100 ura3-1 ade2-1 his3-11,15) is used for the expression of Homo sapiens APC isoform-1 (SCaMC1/SLC25A24) in a functional form (see Note 2).
3. S. cerevisiae expression vectors containing wild-type or mutant AAC (pYES-pMIR2-AAC2) and wild-type or mutant APC (pYES-pMIR2-APC1).
Expression and purification reagents
1. YPG medium (1% (w/v) yeast extract, 2% (w/v) tryptone, 2% (v/v) glycerol).
2. YPD medium (1% (w/v) yeast extract, 2% (w/v) tryptone, 2% (v/v) D-glucose).
3. Synthetic-complete tryptophan-dropout medium (SC-Trp; Formedium, order number: DSCK1008).
4. Tetraoleoyl cardiolipin (18:1) (Avanti Polar Lipids, order number: 710335C).
5. L--phosphocholine (Avanti Polar Lipids, order number: 840051C).
6. EDTA-free complete protease inhibitor tablet (Roche Diagnostics Ltd, order number: 11873580001).
7. Dodecyl maltoside (Glycon Biochemicals GmbH, order number: D97002).
8. Undecyl maltoside (Glycon Biochemicals GmbH, order number: D99012).
9. Decyl maltoside (Glycon Biochemicals GmbH, order number: D99003).
10. Lauryl maltose neopentyl glycol (Anatrace, order number: NG310).
11. Decyl maltose neopentyl glycol (Anatrace, order number: NG322).
12. Breaking buffer (0.65 M sorbitol, 100 mM Tris-hydrochloride, pH 8.0, 0.2 % bovine serum albumin, 5 mM EDTA, 5 mM amino hexanoic acid, 5 mM benzamdine hydrochlorate).
13. Wash buffer (0.65 M sorbitol, 100 mM Tris-HCl, pH 7.4, 5 mM amino hexanoic acid, 5 mM benzamdine hydrochlorate).
14. Tris-buffered glycerol (TBG) (100 mM Tris-HCl, pH 8.0, 10 % glycerol).
15. Solubilization buffer (20 mM imidazole, 150 mM NaCl, 20 mM HEPES-NaOH, pH 8.0, and an EDTA-free complete protease inhibitor tablet).
16. Buffer A (20 mM HEPES-NaOH pH 8.0, 150 mM NaCl, 20 mM imidazole, 0.1% dodecyl maltoside (AAC) or lauryl maltose neopentyl glycol (APC) and 0.1 mg ml−1 tetraoleoyl cardiolipin (18:1)).
17. Buffer B (20 mM HEPES-NaOH pH 8.0, 50 mM NaCl, 0.1% dodecyl maltoside (AAC) or lauryl maltose neopentyl glycol (APC) and 0.1 mg ml−1 tetraoleoyl cardiolipin (18:1)).
18. Factor Xa protease (New England BioLabs, order number: P8010L).
19. Ni-Sepharose high performance (Amersham Biosciences, order number: 17-5268-01).
20. Empty micro bio-spin chromatography columns (Bio rad, catalog number: 7326204).
21. BCA assay kit (Thermo scientific, order number: 23225).
Reagents for thermostability assays using CPM
1. Assay buffer (25 mM HEPES-NaOH pH 7.5, 50 mM NaCl, 0.1% detergent, 0.1 mg ml-1 lipid (see Note 3)).
2. DMSO, anhydrous (Life technologies, order number: D12345).
3. CPM (Sigma, order number: C1484).
4. PCR tubes (Axygen, order number: PCR-02-C).
5. Hampton Additive Screen (Hampton research, order number: HR2-138).
6. Silver Bullets Screen (Hampton research, order number: HR2-096).
Equipment
1. Rotary qPCR machine (Rotor-Gene Q 2plex HRM cycler).
2. For protein purification, we use an ÄKTAprime (GE Healthcare).
3. For yeast fermentation, we use an Applikon 140 Pilot System with an eZ controller.
4. For small scale expression of protein in yeast, we used 2.5 liter full-baffle TunAir® shake flasks (Sigma-Aldrich, order number: Z710822).
5. For the disruption of yeast cells, we used mechanical breaking with glass beads of 0.5 – 0.75 mm diameter in a Dyno-Mill (Dyno-Mill, Multi-Lab).
Software
1. Prism (GraphPad; www.graphpad.com).
2. Gnuplot version 5.0 (Thomas Williams, Colin Kelley et al.; www.gnuplot.info).
3. RotorGene Q software (Qiagen).
Methods
Large-scale expression of wild-type and mutant AAC in yeast
1. For each wild-type and mutant AAC, a single colony of S. cerevisiae containing plasmid was used to start 50-100 ml primary pre-cultures in YPG + 0.1% glucose medium, with incubation at 30 °C overnight (>18 h) and shaking at 200 rpm (see Note 2).
2. Primary pre-cultures were used to inoculate five liters of YPG + 0.1% glucose medium secondary pre-cultures to a starting OD600nm of 0.05. Cultures were incubated at 30 °C overnight (>18 h) and shaking at 225 rpm.
3. The five-liter secondary pre-cultures were used to inoculate 100 liter of YPG medium in the fermenter, where they were grown at 30 °C for 24 h.
Small-scale expression of wild-type and mutant APC in yeast
1. For each wild-type and mutant human APC1, a single colony of S. cerevisiae containing the correct plasmid was used to start 5-10 ml primary pre-cultures in SC-Trp + 2% glucose medium, with incubation at 30 °C overnight (>18 h) and shaking at 200 rpm (see Note 2).
2. Primary pre-cultures were used to inoculate 500 ml of SC-Trp + 2% glucose medium secondary pre-cultures to a starting OD600nm of 0.05. Cultures were incubated at 30 °C overnight (>18 h) and shaking at 200 rpm.
3. The 500 ml secondary pre-cultures were used to inoculate 10 liter of YPD medium. 1 liter of total culture was used in each 2.5-liter full-baffle TunAir® shake flasks, which were incubated at 30 °C for 24 h with shaking at 225 rpm.
4. Cells were harvest by centrifugation (4,000 x g, 20 mins, 4 °C).
Isolation of yeast mitochondria
1. Yeast cell pellets were re-suspended in 1 liter of breaking buffer per 500 g of cells.
2. Cells were lysed by one pass through a Dyno-Mill (see Note 4).
3. Whole cells and debris were removed by two rounds of centrifugation (3,000 x g, 15 minutes, 4 °C).
4. Mitochondria were isolated by centrifugation (30,000 x g, 1 hour, 4 °C).
5. Mitochondria were re-suspended in washing buffer and harvested by centrifugation (30,000 x g, 1 hour, 4°C).
6. Mitochondria were re-suspended in TBG and harvested by centrifugation (30,000 x g, 1 hour, 4 °C).
7. Total mitochondrial protein concentration was determined by BCA assay and mitochondria were re-suspended in TBG to a final total protein concentration of 20 mg ml-1 (see Note 5).
8. Mitochondria were flash frozen in liquid nitrogen, and stored at -80 °C.
Preparation of lipid for protein purification and addition in stability assays
1. Tetraoleoyl cardiolipin (18:1) and L--phosphocholine were supplied dissolved in chloroform.
2. Typically, 100 mg of lipid was dispensed into a glass vial and the lipid was dried under a stream of nitrogen to remove the chloroform.
3. For thorough removal of chloroform, lipids were re-suspended in the same volume of diethyl ether, vortexed and once again dried under a stream of nitrogen.
4. Lipids were solubilized in 10% (w/v) detergent (either dodecyl maltoside or lauryl maltoside neopentyl glycol) by vortexing for 4 h at room temperature to give 10 mg ml−1 lipid in a 10% detergent stock. The stocks were snap-frozen and stored in liquid nitrogen until use.
Purification of wild-type and mutant AAC and APC for stability assays
1. Isolated yeast mitochondria (1 g total protein) were solubilized in 2% detergent (dodecyl maltoside (AAC) or lauryl maltose neopentyl glycol (APC)) by inversion mixing with solubilization buffer at 4 °C for one hour.
2. Particulate material was removed by ultracentrifugation (140,000 x g, 45 min, 4 °C).
3. The soluble fraction was loaded onto a Ni-Sepharose high performance column at 1 ml min−1 on an ÄKTAprime.
4. The column was washed with 40 column volumes of buffer A.
5. The column material was washed with a further 20 column volumes of buffer B.
6. For each preparation, the column material was re-suspended with 400 μl buffer B and transferred to a vial containing 5 mM CaCl2 and 10 μg Factor Xa (AAC) or 75 μg Factor Xa (APC), vortexed thoroughly, and incubated at 10 °C overnight for AAC or 4 °C overnight for APC.
7. The cleaved protein was separated from the resin using centrifugal filtration through micro bio-spin columns, the protein concentration was determined, and the sample was snap-frozen and stored in liquid nitrogen.
Basic thermostability assay
1. A 5 mg ml−1 stock of CPM dissolved in DMSO was diluted 50-fold into assay buffer containing 20 mM HEPES-NaOH pH 8.0, 150 mM NaCl, 0.1% dodecyl maltoside for AAC or lauryl maltose neopentyl glycol for APC as well as 0.1 mg ml-1 lipid (see Note 3)
2. This working stock was vortexed and allowed to equilibrate in the dark at room temperature for 10 min (see Note 6)
3. One to four micrograms of protein were added into a final volume of 45 μl of assay buffer in 200 μl thin-walled PCR tubes, and 5 μl CPM working solution was added.
4. The solution was vortexed and allowed to equilibrate in the dark for a further 10 min (see Note 6).
5. PCR tubes containing the reaction mixture were transferred to the 36-position rotor and loaded into the RotorGene Q.
6. An initial pre-incubation step of 90 s was set to allow the temperature to equilibrate to 25 °C.
7. Measurements were made in 1 °C intervals from 25 to 90 °C with a ‘wait between reading’ set to 4 s, which equated to a ramp rate of 5.6 °C min−1.
8. Data were analyzed and melting temperatures (the inflection point of the melting curve) were determined with the software supplied with the instrument.
9. Raw and analyzed data were exported from the software and Prism (Graphpad; www.graphpad.com) was used to interpret the results graphically.
Applications
Effects of detergents, lipids and inhibitors on stability of the mitochondrial ADP/ATP carrier
The alkyl maltoside series of detergents are commonly used for the solubilization and purification of membrane proteins. Using the CPM assay, we determined the apparent stability of AAC in each detergent in the presence or absence of the AAC-specific inhibitor CATR (Figure 3A). The structures of the bovine (21) and yeast (23) enzymes inhibited with CATR reveal an extensive network of polar and hydrophobic interactions between protein and inhibitor, stabilizing AAC and increasing the apparent melting temperature by 25–30 °C relative to the apo-state. As shown previously, the detergents with larger micelles, and smaller critical micelle concentrations, are more stabilizing (15,18,36,37). Furthermore, we demonstrated that the neopentyl glycol version of decyl maltoside, decyl maltose neopentyl glycol (DMNG (38)) is more stabilizing in both the presence and absence of CATR (Figure 3B). We also investigated the stabilizing effect of lipid. Cardiolipin, which is known to associate tightly with AAC (10,23,39,40), had a significant stabilizing effect compared to phosphocholine (PC) (Figure 3C).
1. The protocol for the basic thermostability assay was followed with a few modifications as detailed below.
2. AAC was purified in dodecyl maltoside. Typically, two micrograms of protein were diluted twenty-fold into a final volume of 45 μl of assay buffer containing 1% detergent and, when required, 50 μM ADP and 20 μM CATR. ADP is required to allow the protein to cycle between conformations, thus access both the cytoplasmic and matrix states. Lipid, when required, was added at a concentration of 0.1 g/g of detergent.
3. The assay mixture was allowed to equilibrate in the dark for 10 min at room temperature for the compounds to take effect. In the absence of inhibitor, the assay mixture was incubated on ice in the dark for 10 min, as these proteins are highly unstable.
Effects of buffer composition on stability of the mitochondrial ATP-Mg/Pi carrier
The thermostability assay can be used to screen the optimum stabilizing buffer conditions rapidly for the purification and storage of membrane proteins. We found that the thermostability of the human mitochondrial ATP-Mg/Pi carrier isoform 1 (APC1) decreases significantly above pH 8 or below pH 6.5 (Figure 4A). Furthermore, sodium chloride concentrations above 200 mM had a detrimental effect on APC1 stability, particularly at suboptimal pH conditions (Figure 4A). Therefore, during preparation of APC1, buffer conditions of 50 mM NaCl, and 25 mM HEPES-NaOH pH 7.5 were chosen for ongoing studies.
1. The protocol for the basic thermostability assay was followed with a few modifications as detailed below.
2. Three micrograms of protein were added into a final volume of 45 μl of assay buffer made by adding 5 µl of concentrated stock condition as detailed below to 40 µl of diluted protein.
3. All conditions contained final concentrations of 0.1% lauryl maltose neopentyl glycol and 0.1 mg ml-1 lipid (as per basic protocol).
4. A deep-well block containing 10-fold concentrated conditions was prepared, where columns 1-8 contained increasing sodium chloride concentrations (250, 500 mM and 1, 2, 3 and 5 M), and rows A-H contained different buffers (A; 250 mM Tris-HCl pH 8.5, B; 250 mM Tris-HCl pH 8.0, C; 250 mM Tris-HCl pH 7.5, D; 250 mM Tris-HCl pH 7.0, E; 250 mM MES-NaOH pH 6.5, F; 250 mM MES-NaOH pH 6.0, G; 250 mM sodium-citrate pH 5.0, H; 250 mM sodium-citrate pH 4.0).
5. Once diluted, the final conditions were 25 mM buffer and differing concentrations of sodium chloride (between 25 and 500 mM).
6. Once collected, melting temperatures from first derivative analysis were compared to one another using a heatmap (Figure 4A). Simple heatmaps can be produced in Microsoft Excel, or more complex heatmaps plotted in dedicated graphing software such as gnuplot (www.gnuplot.info) or Prism (Graphpad; www.graphpad.com).
Effects of crystallization additives on stability of the mitochondrial ATP-Mg/Pi carrier
The thermostability assay can be used to screen commonly used crystallization additives that stabilize the protein. There are a number of commercially available screens; here, we used the Hampton additive screen and the Silver Bullets screen. Five stabilizing hits were observed in the Hampton additive screen: B11; sodium-citrate, D3; spermine, D4; hexammine cobalt (III) chloride, D10; ATP and E1; EDTA (Figure 4B). Four stabilizing conditions were observed in the Silver Bullets screen (see Note 7): D3; MES, PIPES, hexamminecobalt (III) chloride and HEPES pH 6.8, D4; gadolinium (III) chloride hexahydrate, samarium (III) chloride hexahydrate, benzamidine hydrochloride and salicin, E4; protamine sulfate and H4; 1,4-diaminobutane, 1,8-diaminooctane, cadaverin, spermidine and spermine (Figure 4B). These hits provide possible stabilizing additives to add to APC1 in crystallization trials, and for some hits, physiological relevance can be drawn from the types of compound that stabilize. EDTA has previously been shown to stabilize APC as it chelates calcium ions, which leads to inactivation of APC, making it more stable (24). Furthermore, sodium-citrate also chelates calcium ions at pH 7 and could be stabilizing APC in the same way as EDTA. ATP is a substrate of the protein and stabilizes the carrier by interacting with the residues of the substrate binding site, rescuing parts of the population from thermal denaturation. Hexamminecobalt (III) chloride was identified in two hits and is possibly an analogue for a fully hydrated magnesium ion for which there is a binding site in APC. There were several hits with polyamines, such as spermine, likely to be providing a generic stabilizing effect, as also observed for other carriers.
1. The protocol for the basic thermostability assay was followed with a few modifications as detailed below.
2. Three micrograms of protein were added into a final volume of 45 μl of assay buffer with compound. The compounds were added by a 10-fold dilution of the stock compound from either the Hampton Additive or Silver Bullets screens using a multi-channel pipette (5 µl into the assay mixture for a final reaction volume of 50 µl).
3. All conditions also contained a final concentrations of 0.1% lauryl maltose neopentyl glycol and 0.1 mg ml-1 lipid (as per the basic thermostability assay protocol).
4. Once data were collected, results were analyzed by calculating the thermostability shift (∆Tm; Tm of the protein in the presence of compound minus the Tm of the protein without compound). Compounds were considered to stabilize if they had a positive ∆Tm.
Notes
Note 1. We have previously shown that APC1 residue C15 is solvent exposed in the natively-folded state (24); therefore, thermal melting of APC1 with CPM gives a stability for the carrier domain only and not the regulatory or amphipathic helix domains.
Note 2. For the work outlined here, we expressed AAC and APC using an S. cerevisiae expression system followed by purification using nickel affinity and an on-column cleavage approach. However, it should be noted that these details are not important for carrying out the CPM thermostability assay. The only prerequisites are that the protein is reasonably pure (>85%), and that buried cysteine residues are present in the protein.
Note 3. The choice of detergent and lipid is protein-dependent and needs to be empirically determined. For AAC, the default detergent and lipid were 0.1% dodecyl maltoside and 0.1 mg ml-1 tetraoleoyl cardiolipin (18:1). For APC1, the default detergent and lipid were 0.1% lauryl maltose neopentyl glycol and 0.1 mg ml-1 tetraoleoyl cardiolipin (18:1).
Note 4. We have found previously that mechanical lysis of yeast cells using glass beads is an effective method to break the cells and milder than methods that rely on pressure and shear force to break open the cells. We get better quality protein preparations with less contaminants.
Note 5. We typically recovered 30-50 mg of total mitochondrial protein per liter of yeast culture.
Note 6. We have found that pre-incubation of CPM with detergent and protein:detergent:lipid micelles is an essential step to avoid significant drift in baseline fluorescent signal during melting. Presumably CPM differentially partitions into the micelles, which needs to proceed to equilibrium, otherwise changes in fluorescence during the run are observed due to equilibration of the dye rather than thermal denaturation of the protein.
Note 7. Silver bullets screen contains mixtures of compounds in each condition.
References
1. Bill, R. M., Henderson, P. J., Iwata, S., Kunji, E. R., Michel, H., Neutze, R., Newstead, S., Poolman, B., Tate, C. G., and Vogel, H. (2011) Overcoming barriers to membrane protein structure determination. Nature biotechnology 29, 335-340
2. Newstead, S., Kim, H., von Heijne, G., Iwata, S., and Drew, D. (2007) High-throughput fluorescent-based optimization of eukaryotic membrane protein overexpression and purification in Saccharomyces cerevisiae. Proc Natl Acad Sci U S A 104, 13936-13941
3. Magnani, F., Serrano-Vega, M. J., Shibata, Y., Abdul-Hussein, S., Lebon, G., Miller-Gallacher, J., Singhal, A., Strege, A., Thomas, J. A., and Tate, C. G. (2016) A mutagenesis and screening strategy to generate optimally thermostabilized membrane proteins for structural studies. Nature protocols 11, 1554-1571
4. Alexandrov, A. I., Mileni, M., Chien, E. Y., Hanson, M. A., and Stevens, R. C. (2008) Microscale fluorescent thermal stability assay for membrane proteins. Structure 16, 351-359
5. Crichton, P. G., Lee, Y., Ruprecht, J. J., Cerson, E., Thangaratnarajah, C., King, M. S., and Kunji, E. R. (2015) Trends in thermostability provide information on the nature of substrate, inhibitor, and lipid interactions with mitochondrial carriers. J Biol Chem 290, 8206-8217
6. Majd, H., King, M. S., Palmer, S. M., Smith, A. C., Elbourne, L. D., Paulsen, I. T., Sharples, D., Henderson, P. J., and Kunji, E. R. S. (2018) Screening of candidate substrates and coupling ions of transporters by thermostability shift assays. eLife 7, e38821
7. Ruprecht, J. J., and Kunji, E. R. S. (2019) The SLC25 Mitochondrial Carrier Family: Structure and Mechanism. Trends in biochemical sciences
8. Kunji, E. R. S., Aleksandrova, A., King, M. S., Majd, H., Ashton, V. L., Cerson, E., Springett, R., Kibalchenko, M., Tavoulari, S., Crichton, P. G., and Ruprecht, J. J. (2016) The transport mechanism of the mitochondrial ADP/ATP carrier. Biochim. Biophys. Acta 1863, 2379-2393
9. Ruprecht, J. J., and Kunji, E. R. S. (2019) Structural changes in the transport cycle of the mitochondrial ADP/ATP carrier. Current opinion in structural biology 57, 135-144
10. Ruprecht, J. J., King, M. S., Zogg, T., Aleksandrova, A. A., Pardon, E., Crichton, P. G., Steyaert, J., and Kunji, E. R. S. (2019) The molecular mechanism of transport by the mitochondrial ADP/ATP carrier. Cell 176, 435–447
11. Harborne, S. P. D., and Kunji, E. R. S. (2018) Calcium-regulated mitochondrial ATP-Mg/Pi carriers evolved from a fusion of an EF-hand regulatory domain with a mitochondrial ADP/ATP carrier-like domain. IUBMB life 70, 1222-1232
12. Harborne, S. P., King, M. S., Crichton, P. G., and Kunji, E. R. S. (2017) Calcium regulation of the human mitochondrial ATP-Mg/Pi carrier SLC25A24 uses a locking pin mechanism. Sci. Rep. 7, 45383
13. Harborne, S. P., Ruprecht, J. J., and Kunji, E. R. S. (2015) Calcium-induced conformational changes in the regulatory domain of the human mitochondrial ATP-Mg/Pi carrier. Biochim. Biophys. Acta 1847, 1245-1253
14. Kunji, E. R. S., and Harding, M. (2003) Projection structure of the atractyloside-inhibited mitochondrial ADP/ATP carrier of Saccharomyces cerevisiae. J. Biol. Chem. 278, 36985-36988
15. Bamber, L., Harding, M., Butler, P. J. G., and Kunji, E. R. S. (2006) Yeast mitochondrial ADP/ATP carriers are monomeric in detergents. Proc. Natl. Acad. Sci. U.S.A. 103, 16224-16229
16. Bamber, L., Harding, M., Monné, M., Slotboom, D. J., and Kunji, E. R. S. (2007) The yeast mitochondrial ADP/ATP carrier functions as a monomer in mitochondrial membranes. Proc. Natl. Acad. Sci. U.S.A. 104, 10830-10834
17. Bamber, L., Slotboom, D. J., and Kunji, E. R. S. (2007) Yeast mitochondrial ADP/ATP carriers are monomeric in detergents as demonstrated by differential affinity purification. J. Mol. Biol. 371, 388-395
18. Kunji, E. R. S., Harding, M., Butler, P. J. G., and Akamine, P. (2008) Determination of the molecular mass and dimensions of membrane proteins by size exclusion chromatography. Methods 46, 62-72
19. Crichton, P. G., Harding, M., Ruprecht, J. J., Lee, Y., and Kunji, E. R. S. (2013) Lipid, detergent, and Coomassie Blue G-250 affect the migration of small membrane proteins in blue native gels; mitochondrial carriers migrate as monomers not dimers. J. Biol. Chem. 288, 22163-22173
20. Saraste, M., and Walker, J. E. (1982) Internal sequence repeats and the path of polypeptide in mitochondrial ADP/ATP translocase. FEBS letters 144, 250-254
21. Pebay-Peyroula, E., Dahout-Gonzalez, C., Kahn, R., Trezeguet, V., Lauquin, G. J., and Brandolin, G. (2003) Structure of mitochondrial ADP/ATP carrier in complex with carboxyatractyloside. Nature 426, 39-44
22. Klingenberg, M. (2008) The ADP and ATP transport in mitochondria and its carrier. Biochimica et biophysica acta 1778, 1978-2021
23. Ruprecht, J. J., Hellawell, A. M., Harding, M., Crichton, P. G., Mccoy, A. J., and Kunji, E. R. S. (2014) Structures of yeast mitochondrial ADP/ATP carriers support a domain-based alternating-access transport mechanism. Proc Natl Acad Sci U S A 111, E426-E434
24. Harborne, S. P. D., King, M. S., Crichton, P. G., and Kunji, E. R. S. (2017) Calcium regulation of the human mitochondrial ATP-Mg/Pi carrier SLC25A24 uses a locking pin mechanism. Sci Rep-Uk 7
25. King, M. S., Kerr, M., Crichton, P. G., Springett, R., and Kunji, E. R. (2016) Formation of a cytoplasmic salt bridge network in the matrix state is a fundamental step in the transport mechanism of the mitochondrial ADP/ATP carrier. Biochimica et biophysica acta 1857, 14-22
26. Kunji, E. R. S., and Robinson, A. J. (2006) The conserved substrate binding site of mitochondrial carriers. Biochim. Biophys. Acta 1757, 1237-1248
27. Robinson, A. J., and Kunji, E. R. S. (2006) Mitochondrial carriers in the cytoplasmic state have a common substrate binding site. Proc. Natl. Acad. Sci. U.S.A. 103, 2617-2622
28. Robinson, A. J., Overy, C., and Kunji, E. R. (2008) The mechanism of transport by mitochondrial carriers based on analysis of symmetry. Proc Natl Acad Sci U S A 105, 17766-17771
29. Nelson, D. R., Felix, C. M., and Swanson, J. M. (1998) Highly conserved charge-pair networks in the mitochondrial carrier family. J Mol Biol 277, 285-308
30. Springett, R., King, M. S., Crichton, P. G., and Kunji, E. R. S. (2017) Modelling the free energy profile of the mitochondrial ADP/ATP carrier. Biochim. Biophys. Acta 1858, 906-914
31. Chipot, C., Dehez, F., Schnell, J. R., Zitzmann, N., Pebay-Peyroula, E., Catoire, L. J., Miroux, B., Kunji, E. R. S., Veglia, G., Cross, T. A., and Schanda, P. (2018) Perturbations of native membrane protein structure in alkyl phosphocholine detergents: a critical assessment of NMR and biophysical studies. Chem. Rev. 118, 3559-3607
32. Kurauskas, V., Hessel, A., Ma, P., Lunetti, P., Weinhaupl, K., Imbert, L., Brutscher, B., King, M. S., Sounier, R., Dolce, V., Kunji, E. R. S., Capobianco, L., Chipot, C., Dehez, F., Bersch, B., and Schanda, P. (2018) How detergent impacts membrane proteins: atomic-level views of mitochondrial carriers in dodecylphosphocholine. J Phys Chem Lett 9, 933-938
33. Crichton, P. G., Lee, Y., Ruprecht, J. J., Cerson, E., Thangaratnarajah, C., King, M. S., and Kunji, E. R. S. (2015) Trends in thermostability provide information on the nature of substrate, inhibitor, and lipid interactions with mitochondrial carriers. J. Biol. Chem. 290, 8206-8217
34. King, M. S., Kerr, M., Crichton, P. G., Springett, R., and Kunji, E. R. S. (2016) Formation of a cytoplasmic salt bridge network in the matrix state is a fundamental step in the transport mechanism of the mitochondrial ADP/ATP carrier. Biochim. Biophys. Acta 1857, 14-22
35. Hashimoto, M., Shinohara, Y., Majima, E., Hatanaka, T., Yamazaki, N., and Terada, H. (1999) Expression of the bovine heart mitochondrial ADP/ATP carrier in yeast mitochondria: significantly enhanced expression by replacement of the N-terminal region of the bovine carrier by the corresponding regions of the yeast carriers. Biochim Biophys Acta-Bioenergetics 1409, 113-124
36. Prive, G. G. (2007) Detergents for the stabilization and crystallization of membrane proteins. Methods 41, 388-397
37. le Maire, M., Champeil, P., and Moller, J. V. (2000) Interaction of membrane proteins and lipids with solubilizing detergents. Biochimica et biophysica acta 1508, 86-111
38. Chae, P. S., Rasmussen, S. G., Rana, R. R., Gotfryd, K., Chandra, R., Goren, M. A., Kruse, A. C., Nurva, S., Loland, C. J., Pierre, Y., Drew, D., Popot, J. L., Picot, D., Fox, B. G., Guan, L., Gether, U., Byrne, B., Kobilka, B., and Gellman, S. H. (2010) Maltose-neopentyl glycol (MNG) amphiphiles for solubilization, stabilization and crystallization of membrane proteins. Nat Methods 7, 1003-1008
39. Klingenberg, M. (2009) Cardiolipin and mitochondrial carriers. Biochimica et biophysica acta 1788, 2048-2058
40. Duncan, A. L., Ruprecht, J. J., Kunji, E. R. S., and Robinson, A. J. (2018) Cardiolipin dynamics and binding to conserved residues in the mitochondrial ADP/ATP carrier. Biochim. Biophys. Acta 1860, 1035-1045
41. Arnold, K., Bordoli, L., Kopp, J., and Schwede, T. (2006) The SWISS-MODEL workspace: a web-based environment for protein structure homology modelling. Bioinformatics 22, 195-201
Legends to figures
Figure 1. The thermostability assay using CPM. A) Chemical structure of N-[4-(7-diethylamino-4-methyl-3-coumarinyl)phenyl] (CPM). B) Illustration of the thermostability assay. The protein population is denatured by a thermal melting ramp between 25 and 90 °C. CPM reacts with thiol groups of cysteine residues as they become exposed, forming a fluorescent adduct with an excitation/emission optima of 387/463 nm. The inflection point of the melting curve is taken as the apparent melting temperature (Tm) and used as a relative marker of protein stability.
Figure 2. Positions of cysteine residues in AAC and APC. A) Membrane view of the comparative homology model of AAC of Thermothelomyces thermophila generated with SwissModel (41), based on the related structure of Aac2p of Saccharomyces cerevisiae (PDB: 4C9G) (23). B) Model of human APC from protein model database: PM0080481 (24). Cysteines are shown as red spheres.
Figure 3. Effects of detergents, lipids and inhibitor on stability of the mitochondrial ADP/ATP carrier. AAC was purified in dodecyl maltoside and was diluted twenty-fold into buffer containing detergent/lipid. The thermostability data are obtained from reference (5). A) The unfolding profiles of purified AAC diluted into buffer containing dodecyl maltoside (DDM, black line), undecyl maltoside (UDM, red line) or decyl maltoside (DM, blue line) with (dashed line) or without (solid line) 20 μM CATR and 50 μM ADP. The graphs on the right are the corresponding derivative curves. The apparent melting temperatures are indicated. B) as A), except AAC diluted into decyl maltose neopentyl glycol (DMNG, red line). C) as A) except AAC diluted into buffer containing decyl maltoside (DM, blue line), decyl maltoside in the presence of phosphocholine (DM+PC, red line), or decyl maltoside in the presence of tetraoleoyl cardiolipin (DM+CL, black line), in the presence (dashed line) or absence (solid line) of 20 μM CATR and 50 μM ADP.
Figure 4. Effects of buffer composition and crystallization additives on stability of the human mitochondrial ATP-Mg/Pi carrier APC1. A) The apparent melting temperature of APC1 has been plotted on a three-dimensional graph (heatmap) against the sodium chloride concentrations and the pH values of the buffer. The Tm of APC1 is represented by color intensity as described in the key. The Tm represents the average from three independent thermostability curves. B) The apparent melting temperature of APC1 in the presence of each compound was plotted as its difference from the apparent melting temperature of APC1 without added compound (Tm). Stabilizing conditions from the Hampton Additive Screen (E1, D4, D10 and B11) or Silver bullets screen (D3, D4, E4 and H4) are annotated. The Tm values are in each case derived from a single thermal profile.
1